First Legal Test of the GPL

Trepidity writes "In stark contrast to the plethora of false alarms recently, there's a pretty clear-cut case that Vidomi, a DVD ripping product by SloMedia, is composed of a great deal of code from VirtuaDub, a GPLd product. As SloMedia have refused all requests to either release their source or stop using the code, the developer is planning to file suit with the aid of the Free Software Foundation, in what could be the first legal test of the GPL's enforceability."

Comments

Re:Get a life

[Anonymous Coward]

Why do a hundred different people insist on submitting this (completely incorrect) argument?

This (entirely bogus) argument has been posted several times already. If you like this (totally counterfeit) argument, then you can read the previous posts without having to crapflood Slashdot with additional copies of it.

There are also a lot of replies to this (utterly absurd) argument which explain how the GPL (and copyright law in general) actually works.

Re:Huh?

[Anonymous Coward]

Yes, but remember that linking to a dll occurs at runtime not at compile time, so the people linking to the dll are the end users, not Vidomi. I don't see how Vidomi can be held accountable for actions that end users might take with their program.

Look, it doesn't matter whether you have a warm fuzzy feeling inside that this isn't a GPL violation. What matters is copyright law.

The fact that the linking occurs at runtime is a completely moot point. It may give you a warm fuzzy feeling inside that this isn't a GPL violation, but copyright law does not care about warm fuzzy feelings or runtime loading of DLL's. All that matters under copyright law is that Vidomi is a derived work with respect to VirtuaDub. Period. You simply can't create and distribute a derived work without the permission of the original copyright holder.

Re:No Credit for Nuke Either

[Anonymous Coward]

I'd say that's credit enough. I can't stand these web pages with "powered by.." plastered all over them. It's so tacky looking.

Besides, anybody who gives a rat's ass about what a website uses will check the source. Anyone else is just there for the content and that's all they should see.

Re:Huh?

[Anonymous Coward]

So if you want to include someone else's library in your proprietary program without the library creator's consent, you simply have your program pop up a dialog box when it starts, saying:

[X] I want this program to work.
[ ] I don't want this program to work.

If the user checks the first checkbox, the program loads the stolen library and runs fine. If the user checks the second checkbox, the program doesn't load the stolen library, and does something useless (I guess this is up to you -- maybe you could draw pictures of little goats on the screen). According to your logic, the program now "runs" fine without the library.

Try doing this with a library from a large proprietary software corporation and see how far you get.

Re:Huh?

[Anonymous Coward]

Is the following program a derived work of libc?

main() { printf("Hello world!\n"); exit(0); }

Legally, yes. If it weren't, there would be no reason for the existence of the LGPL (which glibc is licensed under) in addition to the GPL. Richard Stallman may be -- er -- eccentric, but he takes the law seriously. He drafts these licenses with professional legal advice based soundly on copyright law.

This is not to say that copyright law is perfect, flawless, handed down from the Almighty, etc. Far from it. But the law is all that really matters once a case comes to court. (Well, it also matters who can afford to hire the best lawyers and/or bribe the judge ... but that's a topic for another discussion.)

Not So Clear (was Re:Grey area)

[Anonymous Coward]

> Actually dynamic linking to GPL code does make
> the linking program based on that code. There
> is no grey area when is come to the GPL.

Not So clear: tell me what happen if I a write a GPL winamp plugin?

The plugin is a DLL. When I use this plugin winamp will be dynamically linked with my GPLed DLL.
Does Winamp have to be GPLed ? No, clearly not!
Does the GPL allow this? Does this mena the user is in this case breaking the GPL? Or does the plugin is itself violating its license, that is t say: you cannot enforce GPL on such a plugin?

Then next step. Let's say your plugin is very very cool. Winamps people says "hey let's bundle this plugin with every winamp, with the source and the credits". Are they viloating the GPL?
I would say yes - the letter of the GPL doesn't allow this, but is it the spirit? Is it not just "mere aggregation" (like RedHat bundling commercial software with a lot of GPLed things)?

Another possibility

[Anonymous Coward]

Others have commented already on what happens if the FSF wins or loses this suit. What happens if they go ahead with it when there's really no point to it (as has been pointed out many times already)? Does the FSF risk making the whole OSS movement look like a bunch of sue-happy idiots that no one will want to work with?

Re:Huh?

[Anonymous Coward]

Yeah, but let's be honest - this is one of the more questionable parts of the GPL. The GPL itself doesn't say anything specific [IIRC], but RMS has decided that dynamic linking in a DLL is not OK, but dynamic linking to non-GPL OS functions [at load time] is OK, which is a completely arbitrary distinction based more on intent than on actual technical differences.

It would be easy to play games with this. What if a non-GPL program needs a GPL program to function, but communicates using pipes [and the GPL program modified to use pipes is released]? What if they communicate using method X? It's all dynamic linking at the end of the day, just in various forms of indirection. Once SOAP is widespread, can a non-GPL program make RPC calls against a GPL program? How far will RMS push this? Because at the boundary this kind of "abuse" can and will take place.

I look forward to seeing the results of this case after the expert witnesses have been called to testify.

In this DMCA age...

[Anonymous Coward]

You don't need to waste time renaming functions and variables. All you have to do is bury a clause in the click-license legalese that says:

42. You agree that you will not diff the binary to check for similarities with GPLd code.

Re:In this DMCA age...

[etrnl]

And then someone hears about it, and inspects the files on the CD without installing it and clicking to aggree to it.

Or you put their asses in court and subpeona it to investigate it... using the DMCA to shut them down until the court case is settled.

--etrnl--

Re:In this DMCA age...

[borgquite]

And Avery Lee could just put a little clause in addition to the GPL for VirtualDub to say:

x. No developer called Vidomi may download, use or link to any portion of this code.

(just like those websites that say government agents 'may not access this site')

Sorted! ;)

--

' Ore stabit fortis a fine placet ore stat '

I'm confused on a couple of points.

[Anonymous Coward]

I'd like someone in the know to please clarify what is and is not a violation of the GPL. The part I'm getting stuck on is the separate/independent clause.

Here's the gist. If I design a system or solution which runs on Linux, and uses a few GPL'ed tools, do I have to GPL the *system*? Here are some facts about this situation:

• No GPL'ed code is modified. We're talking your basic './configure && make && make install' here.
• I provide the GPL'ed tools and their source when I distribute/publish my "system"
• The system is written such that it needs the GPL'ed tools to run. There's no easy method for making the system work without the GPL'ed tools.
• It is quite possible (nay, simple) to distribute my proprietary un-GPL'ed "system" completely independent of the GPL'ed stuff. In fact, we may do that anyways, have a two-stage install, first stage is to install all the GNU/linux stuff and tools, stage 2 is to add in our "value add". But my "system" will not work without being able to use the GPL'ed tools. In other words, it is trivial to distribute and publish the work separately, but the "system" will not function properly or at all without the availability of the GPL'ed tools.

Opinions and thoughts anyone?

Re:I'm confused on a couple of points.

[spitzak]

Sorry, the reason your non-GPL program can call fopen() is that fopen() is under the LGPL, not because other systems have an fopen() call.

Re:I'm confused on a couple of points.

[RGRistroph]

I believe the matter hinges on how your code uses the GPL'ed tools.

If it uses them by dynamically linking them into the executable at run time (i.e., it links with a GPL'ed .so file), then the FSF's intention is that you should have to GPL your code. If you don't like it, consider that some libraries are under a separate license, the LGPL, for that reason -- the LGPL doesn't restrict dynamic load and link.

Another factor will be if there are other libraries that your code will link with just fine, other than the GPL'ed one. If there are, then you are not specifically using the GLP'ed library, but some standard interface or semi-standard. In that case you are probably ok. So if it is a library for which their exists a proprietary equivalent on Solaris, for example, then just try to set up your code so it will compile with either one without trouble.

If you are using these tools just by running them, then you are in the clear for sure. I.e., if your tool is a script and you run /bin/bash, no problem.

In summary: Linking GPL stuff means you are GPL if you distribute it, with certain exceptions; running GPL stuff means nothing.

I suggest reading the GPL and then reading some of the various discussions and FAQs on it. It is pretty clear, and short, compared to commercial licenses.

Re:I'm confused on a couple of points.

[RGRistroph]

You can write code that calls fopen() and not release it under the GPL, inspite of the fact that there is a GPL'd version of fopen() in the GNU/linux system.

On the other hand, if you are calling a function that is only available in the GPLd library, then you are obviously "using" the GPLd code, not merely some standard call.

That is all I meant. I didn't mean that you could link with a GPL'd library and then duck all the GPL restrictions by just also linking with a non-GPL'd library.

Re:I'm confused on a couple of points.

[RGRistroph]

yes, fopen() was a bad example. So pick one that is under the GPL then.

Re:I'm confused on a couple of points.

[harlows_monkeys]

Another factor will be if there are other libraries that your code will link with just fine, other than the GPL'ed one

That's not a factor at all. Work X either is or is not a derivative work of work Y. The existence or lack thereof of another work, Z, that can be used in place of Y is irrelevant.

The combination of X and Y is a derivative of Y, of course, but X is not.

Re:A true test of the GPL

[Anonymous Coward]

For EVERYONE's sake, let's hope that Vidomi is correct and that dynamic linking does NOT make something a derivative work.

I would disagree.

Consider that libraries are CREATED with the intent that someone else will use the API and header files.

Libraries created with the intent that someone else will use the API and library files are licensed under the LGPL. That's what it's for.

I have never agreed with RMS (and Trolltech's) assumptions that dynamic linking makes something a derived work.

You are entitled to your opinion.

Imagine if someday the kernel developers decided that all of a sudden ALL works ever developed under linux were actually GPL'd

Then suddenly a lot of people would stop using Linux. What's the problem? If you don't like a license agreement, you don't use the product. You don't argue that it means something it doesn't.

Fortunately, RMS is wrong.

*chuckle* I may not agree with RMS on some things, but I'm not stupid enough to bet against him.

Re:Update from Vidomi

[Anonymous Coward]

>We recognize that Avery?s opinion is not binding
>on anyone

Bzzt. Wrong. He's the copyright holder! He has specifically stated that what you're doing is not okay. Therefore, you do not have the right to modify, distribute, or copy the code.

>We are attempting to deal with this GPL issue
>openly and with integrity.

You are not respecting the author of code which you rely upon for functionality.

Using your argument, I could create a wrapper that loads your entire application as a DLL. I could then freely distribute your copyrighted code with it because, hey, it's not a derivative work.

Re:easy:

[Anonymous Coward]

Hint: what you did in college may not work in a copyright infringement lawsuit.

Re:Vidomi's position

[Yarn]

4) Contact the copyright owner and negotiate a different license agreement.

Re:A true test of the GPL

[Mark J Tilford]

Well, I believe that it was originally Library GPL; although now RMS prefers to call it the Lesser GPL.
-----------

FUD, FUD, FUD

[oGMo]

While IANAL, it's obvious you're going on mostly hearsay, popular opinion, and misunderstandings (whether intentional or not). Let's see:

Consider that libraries are CREATED with the intent that someone else will use the API and header files.

Yes, they are, and it's not too unreasonable that the person who creates the library may wish it to only be used for Free Software purposes. Like, say, TrollTech. This fits perfectly with their intent. It's their code, if you want your own nice widget set that's closed, write it yourself, or pay TrollTech their due. Next up:

I have never agreed with RMS (and Trolltech's) assumptions that dynamic linking makes something a derived work.

Your opinions on the matter don't affect the matter, of course. Save this thought for later. Going on:

It would certainly make ALL third party software vendors beholden to OS vendors, for example.

Hmm, better not give software vendors any bright ideas here.

Every program ever written under linux would be a derivative work of the kernel, which is GPL. Imagine if someday the kernel developers decided that all of a sudden ALL works ever developed under linux were actually GPL'd (not that this is a likely scenario, but it is a possible one if RMS is right).

OK, this is complete tripe. The standard C/C++ library (glibc) is under the LGPL, Linus has made special provision for allowing use of the kernel interface however you'd like (allowing binary drivers for instance). Let me reiterate: this is just pure FUD.

Fortunately, RMS is wrong. GPL libraries do not make all dynamically linked programs derivative works, and I expect this lawsuit to uphold that point (and sanity in general).

Um, and RMS is wrong why? Remember, your opinions don't make it so. You certainly haven't shown it legally to not be the case. The only real reasoning you've shown is FUD. This assertion, therefore, doesn't seem to have a whole lot of weight behind it.

Remember, RMS has quite a legal team. He very well understands what the license means, and it's been analyzed by many other legal teams (IBM, who could find no holes; MS, who is running scared). Let's see what the real lawyers with informed opinions have to say.

Re:VirtualDub is GPL, not LGPL

[Chris Johnson]

Why not? Why would IE be _entitled_ to someone's third party plugins just for existing? I see no problem with the position that if they want to play proprietary software, it should be possible to make a license for stuff they _cannot_ take advantage of, and therefore they must reverse-engineer and/or separately develop anything from scratch.

What is so wrong with that scenario? It's not like _we_ are allowed to use _IE_ code anytime we want. GPL ain't public domain in intentions. GPL is a _protected_ _area_. It makes perfect sense to exclude some. If you want to be more inclusive, use LGPL- that is what it is FOR. If it bugs you that GPLed stuff is out there teasing you but you can't have it without giving up your old ways- aw poor baby :P

Re:Vidomi's position

[The Man]

There are no "Linux system libraries." Linux is a kernel. It exports functionality via system calls, not via any libraries. You are probably thinking of glibc, which has nothing whatever to do with Linux, and happens to be under the LGPL license which specifically permits this sort of activity. There's a reason RMS calls it "GNU/Linux" - beyond the fact that he's a nutjob.

Re:GPL is NOT LGPL

[Ian Bicking]

Something basic like Qt or GTK is likely enough to be on the computer already that you don't need to distribute it with your binaries, so binaries that link against it are not considered to be derived works.

This isn't the case. GTK is LGPL, so it doesn't matter. Qt most definately is not LGPL -- of you are writing a proprietary program, you cannot link it against Qt without purchasing a seperate license. This is irregardless of the ubiquity of the Qt libraries.

Re:Huh?

[Ian Bicking]

Another important non-technical interpretation of this is intention. It seems quite clear that SloMedia turned the software into a DLL with the intention of circumventing the GPL. I think this will be more important than the spirit of the GPL, because spirit is vague and open to much debate, and no one at SloMedia was obliged to read RMS's writings to understand what he was trying to do with the GPL. But it is possible to pretty much prove what SloMedia's intention was (and hopefully the FSF will be able to do that).

Re:Vidomi's position

[Ian Bicking]

There's also a possible:
4) Pay Avery Lee to license his work to SloMedia (assuming he holds sole copyright).

It's pretty common that Free Software programmers write proprietary code for a living. And many of them would be willing to provide non-GPL licenses for money. I believe this has been a major source of revenue for Cygnus (don't know what it is now that Redhat owns them). And it's only fair.

Re:Vidomi's position

[Danse]

You are correct. Oversight on my part. Since we were talking about corporate use primarily, I assumed that the program would be distributed and neglected to mention that you weren't obligated to release anything if you don't distribute it. Thanks.

Re:Vidomi's position

[Danse]

Which proves again,GPL is completely unusable in commerical settings.

Depends on what you want to do with the software. If it's something that you don't plan to sell anyway (and IBM and Microsoft both have a lot of software that would qualify) then there's no real problem with using GPLed software.

If you want real and usefull code exchange between profesionals with no social or political strings attached go for BSD style license.

Seems to me that corporate types should be even more averse to BSD style licenses than to the GPL. After all, GPL software authors are asking for something in return for their work, it just isn't money. BSD licensed software authors are just giving their code away. Which one is closer to selling as the corporations want to do?

It really pisses me off that these corporations think they can jam all sorts of crap into their EULAs to take away our rights to do anything with their software except exactly what they want us to do with it, but then they turn around and balk at the GPL. Seems like a bunch of hypocrites.

Re:Vidomi's position

[Danse]

Nope, you got it wrong. It is GPL proponents who are outraged at corporation for their EULAs and yet they are using the very same methods.

Not even close. You can do whatever you want with a GPL'd program as long as it doesn't violate copyright law. You don't even have to agree to the license. However, if you'd like to use the GPL'd code in a program of your own, then you can do so by agreeing to the GPL which states that you must release the source code to the resulting derivative program as well.

Contrast that with standard industry EULAs which routinely take away all sorts of rights such as publishing benchmarks, using the program in a way that criticizes the maker, making backup copies, running the software on more than one machine, etc. It becomes quite ridiculous. There is no real comparison between the GPL and EULAs. At least with GPL'd software you can use it without having to agree to such a ridiculous licensing agreement. If these corporations put this many restrictions on simply USING their software, imagine what kinds of restrictions they would put on using their source code!

Re:If this claim stands, then "Bye-bye Be!"

[demon]

I believe the Gnome base libraries are LGPL'd, actually, so that wouldn't affect commercial software (they'd only have to make public any changes to the library sources that they made, if any).
_____

Sam: "That was needlessly cryptic."

Re:GPL, Distribution, Intent and Spirit

[bhurt]

This makes writting proprietary code for Linux *very* tricky. How do you know you're not including some file which includes some file from either the /usr/include/linux or /usr/include/asm directories?

For example, you include to handle POSIX errnos- a very common occurance. /usr/include/errno.h includes /usr/include/bits/errno.h, which in turn includes /usr/include/linux/errno.h But /usr/include/linux is a symlink to /usr/src/linux/include/linux, and the file /usr/src/linux/include/linux/errno.h is part of the GPL'd, not LGPL'd, linux kernel! So every program which includes has to be GPL'd!

Why Not Money?

[Omega]

The FSF isn't adverse to taking money. In fact they are a non-profit charity and they accept donations. It says so on the GNU website. While the Actual Damages from such a lawsuit might not come to any amount of money, the Punitive Damages could run into the millions.

Specifically when you're dealing with a corporation, the only right and wrong they understand is money. So in order to send a message that abusing the GPL is wrong, I think the FSF should seek huge punitive damages.

Re:Why Not Money?

[Dastardly]

IANAL. But, I think some actual damages would be involved since they are using a piece of software in a manner they were not licensed to use. And, since the VirtuaDub author could have licensed the code to Vidomi under some other license that included money he should get paid actual damages as well.

Re:Why Not Money?

[mamba-mamba]

IANAL either, but I feel sure you are right. In general it seems as though copyright holders can sue copyright infringers for royalties. It might not be considered damages per se, but it would still be money. Then the copyright holder could donate some or all of the money to the FSF, if desired.

On the other hand, suing *only* for cease and desist (asking for no money) gives an almost overwhelming moral high ground to FSF/VirtuaDub. When Sun *first* sued MS over java logo useage, numerous legal observers felt that because they were only asking MS to cease and desist, they had a tremendous moral advantage in the case. Later, I think Sun asked for royalties or damages, too.

Something to think about.

MM

GPL vs. L-GPL

[Omega]

I think many people here are confusing the GPL and the LGPL.

GPL'ed code CANNOT BE USED AS A LIBRARY.

LGPL'ed code can. The author chose the GPL. Vidomi has to live with that decision and either GPL their code or use a different library.

Think of the GPL as King Midas. Anything code that touches it turns GPL.

But the solution for Vidomi is simple. Create a back-end, GPL'ed program using the library and just make execv calls to it from the non-GPL program/frontend.

Re:Vidomi depends on GPL'ed libraries

[jd]

I've not seen the libraries, yet, but are you sure they're GPLed and not LGPLed? It would make a huge difference, in all of this.

Re:What's with the GNU icon?

[jaraxle]

It's a gnu, but personally I see a penis with two red balls holding a blanket with a hat on... Freud would have a field day :) jaraxle

The same company violates MS's copyright as well

[Andy Tai]

What's interesting about this GPL violation is that the product in question also includes illegal copies of Microsoft software (DivX, a hacked version of Microsoft MPEG4 version 3 codec DLL). Thus Microsoft can also sue this company for copyright violation.

Of course you cannot have GPLed and Microsoft's proprietary software in the same program. So this product cannot be distributed legally at all.

Re:This is just what the FSF has been waiting for.

[Jason Earl]

Vidomi's software is quite obviously a derivative work of VirtualDub, in that without VirtualDub's software Vidomi's doesn't do any of the things that it is advertised as being capable of.

An appropriate metaphor would be if I wrote down my time-honored recipe for toasting bread and then created a cookbook containing this recipe and 400 other recipes straight out of the Betty Crocker cookbook. I could use this cookbook all I wanted to in the confines of my own house (fair use), but if I tried to sell copies of this cookbook it would rightfully be described as a derivative work.

The GPL states that any time that you link (dynamically or statically) a piece of GPLed software you are creating a derivative work. That is why BSD advocates get so excited about GPLed libraries like GNU readline. They know that GNU readline is cool, but linking your software with it forces you to release your software under the GPL because it is a derivative work.

This is also why most GNU system libraries are licensed under the LGPL (a completely different license). Here's a link from the newly created GPL FAQ

This is just what the FSF has been waiting for.

[Jason Earl]

The FSF has been waiting patiently for a violation of the GPL by a company small enough that they have a good chance of actually winning the case. Unfortunately they aren't the owners of the copyrighted software, but apparently the author of VirtualDub is willing to let the FSF do most of the heavy lifting. Once they win a case then they have precedent supporting their view of the GPL.

Hopefully this actually goes to court, and hopefully the FSF get the win that they have been waiting for. Otherwise we are in for some interesting times on the Free Software front.

Re:This is just what the FSF has been waiting for.

[aozilla]

From the GPL: "Activities other than copying, distribution and modification are not covered by this License"

Indeed, the GPL does not even need to be mentioned in this case, because it does not apply. The act in question is the creation of a derivative work. It is illegal to create a derivative work without permission. The GPL does not cover the creation of a derivative work, if that work is not a modification, which in this case, it is not.

Re:This is just what the FSF has been waiting for.

[aozilla]

What the GPL states about what is a derivitive work is meaningless. I could write a license that says that everything ever made is a derivitive work, but that doesn't make it so. It would be quite a precedent set to say that Vidomi's software is not a derivitive work (and in my opinion a positive one), but that doesn't mean that it's impossible.

Ahem...

[slothbait]

> RMS (the gun-nut)

I believe that you are confusing RMS and ESR. To my knowledge, RMS is uninterested in firearms. And that's a rather small difference compared to their relative beliefs in software licensing.

Be careful with your TLA's...

--Lenny

Re:Ahem...

[sreeram]

Maybe it was just a typo. He really meant RMS but didn't type gnu-nut properly :)

Re:Vidomi's position

[cpt kangarooski]

This is true, but there is a big caveat. (and some smaller ones)

The one most germane to this discussion is that should you legally acquire software, you do NOT have to agree to the conditions of any license in order to make any copies necessary for using it or for making backups.

Since, due to the pecularities of most computers software must be copied from some initial medium like CD to some other medium like RAM, that's why there was ever a problem at all. Had you a machine that could run programs w/o ever making copies in whole or in part, copyright would really never even enter the picture.

Now a precondition of the sale might be to agree, but the very very meagre case law on this issue indicates (last time I checked, IANAL) that things like shrinkwrap licenses don't count. Thus MS would actually have to ist down with you and hash out a contract prior to money leaving your wallet and product being held in your hand to enforce their EULAs.

(Other types of exceptions exist, e.g. manditory licensing under Congressionally determined terms)

Of course, in this particular case, what you're saying does apply, but do beware of making such blanket statements. I would add however, that also if you don't like their terms, write your Congressmen in support of copyright reforms that would prohbit such terms from being set at all! Copyright's hardly immutable.

Re:Don't go to court!

[cpt kangarooski]

Tough noogies.

The courts can't say that odd contract provisions are only permissible when you're a commercial concern. If they're valid at all - and MS's EULA rests on the same foundation - the GPL must also be valid. Either way it's a win for the FSF.

Contracts can get pretty strange before they really cross the threshold into unenforcability. Requiring that someone kill themselves would not be valid, requiring that they wear a duck on their head or have to license derivative software under certain terms is.

Re:Removing vs. Opening

[spitzak]

Most likely outcome is that they will get the authors of the original code to agree to sell/donate it under a different license to them. This will look like a win for everybody.

They could also be pains in the ass and work around the GPL this way: My understanding is that this is a plugin that plays movie files. Their big hole is that they are advertising features of this plugin, and that there is no other plugin. They could remove the advertising and say something like "Plays many movie formats" and then also provide some dummy plugin (perhaps also under the gpl) that does something useless like play .gif movies. That way the program "works" without the gpl plugin and they can legitimatly say it is not a derived work. This may make some gpl purists mad, but I don't see this as being too bad an outcome either.

Re:GPL can probably be circumvented

[spitzak]

I think the LGPL can cover your "interfaces" problem just fine.

Most of the your problems and absurdities are due to your assumption that people will put stuff under the GPL when it makes absolutely no sense to do so (at least assumming they want their stuff to be used).

In for your printer driver example, it is true that a GPL printer driver can not legally be used by a non-GPL program. Contrary to what MicroSoft is trying to claim, this does not mean that MSWord is suddenly forced to be GPL. It means you cannot use MSWord with this printer driver. This probably makes this printer driver pretty useless for the majority of users and thus the author was really stupid for putting it under the GPL rather than the LGPL or another license.

Re:Grey area

[Dastardly]

Actually dynamic linking to GPL code does make the linking program based on that code. There is no grey area when is come to the GPL. Linking in pretty much any form is a violation fo the GPL. Consider this... Let's say I take some GPL software break it down into various components, compile those as dynamically linked libraries, then write my own proprietary application that links to those libraries I created from GPL code. This has to be a GPL violation otherwise there is no GPL. It is possible that a court might decide this, but I would not consider that a good thing. Secondly, the existence of the LGPL implies that dynamic linking is a violation of the GPL, otherwise why would the LGPL exist. Dastardly

Funny thing about that

[crisco]

Is that Avery complied with Microsoft's request. He might not have agreed with it but he did remove the offending code and binaries and moved on. Now he's on the opposite side of a similar issue and he can't get any kind of respect from the offender.

I think the people at Vidomi are a little jealous that one guy can code up a program that makes their product look like a toy. VirtualDub is like a Swiss Army Knife of video, it might be a little clumsy but it is small and it does everything you need it to.

Chris Cothrun
Curator of Chaos

Re:VirtualDub is GPL, not LGPL

[johnnyb]

The problem with that argument is that you have to use the actual headers for the link. Thus the license breakage. If you were able to create your own headers independently that allowed linking, you would probably be okay for that reason. However, if you use their headers to create the link, you ARE including their program into yours.

Re:VirtualDub is GPL, not LGPL

[johnnyb]

If Microsoft was distributing it as part of IE, then yes, it would. In fact, you can't create a GPL plug-in for IE (well, at least you can't distribute the binaries). This was the same problem KDE had a while back before QT released their code under the GPL.

Re:OT: Slashdot Submission process...

[Johann]

Because /. editors do not act in concert. I personally have submitted 3 stories that were rejected only to be submitted by another /.er and posted. IMHO - This is the most irrating aspect of /.
--
"In the land of the brave and the free, we defend our freedom with the GNU GPL."

More info...

[Uruk]

You can find more information about this at Advogato where one of the guys involved in this posts about his experiences.

It also contains some technical evidence as to which functions were lifted and how they know.

Ha ha ha!

[jimdesu]

Good morning boys & girls, can you say CLASS ACTION?

These guys are harming the whole community.

Re:Ha ha ha!

[QuantumG]

yer, they FSF is only got a few dozen lawyers who donate their time to it. 99% of costs of a court case is in the lawyers.

Legal vs. GPL?

[TBone]

No, what the have done is legal according to the LGPL, not the GPL. GPL2 and LGPL1 were released together to address exactly this issue.

Re:No harm intended/done?

[TBone]

The SUN JVM in an interpreter/compiler, and is licensed such that works produced by the output of the compiler are not subject the the licensing of that compiler. This issue is about one program being included inside another.

Lawsuit? Probably not.

[John Fulmer]

I really dislike extremely misleading articles like this. Avery Lee said that he is asking for HELP to enforce the GPL with Vidomi, not that he's filing suit. There are other forms of help that do not necessarily involve litigation.

I've been a VirtualDub supporter/contributor (financially) for a couple of years. It's a great program, and Avery, what I know of him, is a pretty quality guy. Is he thinking of suing Vidomi? I don't know. And neither does anyone else, from the references posted in this article.

I, and I assume others, submitted a MUCH less inflamitory, more accurate entry. I wonder why THEY weren't used instead? It's this sensationalism in article selection and lack of even basic fact checking that really makes me wonder about value of Slashdot as a 'news' portal.

jf

Re:Lawsuit? Probably not.

[dkixk]

Unfortunately, posting unchecked, sensationalistic articles puts /. in "good company", e.g. the recent treatment of the scandal of the White House vandals.

Re:Lawsuit? Probably not.

[QuantumG]

sigh. I can understand your frustration, but unfortunately in this case litigation is exactly what Avery is intending (this is from personal email with him that I will not quote for privacy reasons).

Re:a GPL DLL issue I once pondered

[ZxCv]

You can release your code any way you want to. It matters to a plugin developer who wants to release his plugin for your app under the GPL. If the plugin works in such a way that it does extensive communication with your app (ie, not just fork,exec or pipes, etc) then it cannot be licensed under the GPL because your app is not licensed under the GPL.

I had a (somewhat) similar issue - it gets murky

[Booker]

I wrote a Type 1 MICR font (GnuMICR) by hand, as far as I know, the only freely available MICR font out there, and licensed it under the GPL. (Postscript is a language, my font is a "program", IMHO).

I had a request from a guy who wanted to do a closed-source check printing program, and wanted to use my font. I told him that I would license it to him for a fee, but that he could not use it under the GPL if he hard-coded the font into his application. i.e., his application WOULD NOT WORK without my font. If he had made a font selection dialog, and happened to distribute my font with his program as one option, then that would probably have been fine by me.

For kicks, though, I asked the GPL newsgroup, and none other than Linus responded that he thought they were in the clear if they wanted to bundle my font with their app. I didn't really pursue it much further...

In any case, this DLL linking is probably more clear cut, and I think it's most likely a violation. But whenever you get interactions between GPL and non-GPL'd code, it gets tricky.

Re:Vidomi's position

[Zagadka]

Your analogy doesn't work.

If you were to make a shell for MS Windows, you could distribute your shell without any problems. You could even distribute it along with a copy of Windows, provided you had a license to distribute copies of Windows.

That same is true if you replace Windows with a GPLed piece of software, except that the precise terms of the license to distribute GPLed software is different than for Windows.

Here's an interesting thought experiment: if they didn't distribute the GPLed code in question, but instead only distributed their own binary, and recommended a list of mirrors where the GPLed software was available, would they be violating the GPL? Keep in mind that if you were to write a shell for MS Windows and distribute it without Windows but instead point users at microsfot.com (or even just say "requires MS Windows XX") it would be perfectly legal.

Unless Vidomi's code actually statically links with the GPL'ed code, then I think the FSF will either have to concede that it doesn't violate the GPL, or they'll have to admit that the GPL is not strictly less restrictive than plain old copyright.

Re:Vidomi's position

[Zagadka]

well, no, my analogy doesn't work when you misread and/or change it. I specifically said "distribute a copy of Windows under my own license while ignoring MS's license". I thought that that made it clear that "..provided you had a license to distribute copies of windows" was NOT included.

My point is that your anaogy is invalid because you made the a priori assumption that Vidomi is violating the GPL by shipping the DLLs. If you think about it a bit, you'll see that this assumption is incorrect.

Vidomi is shipping two things:

• a modified version of the VirtuaDub DLLs
• their own proprietary software that dynamically links with the VirtuaDub DLLs

Ignore the proprietary software for a moment. Aren't they allowed to distribute the modified DLLs? Yes, provided they allow access to the source (which they do). So the DLLs are being distributed while still abiding by the GPL. In other words, if you want to make an anology including Microsoft Windows, you have to assume that the company is shipping licensed copies of Windows, because the VirtuaDub DLLs they're shipping do conform to the GPL.

The problem everyone has is with the proprietary code that Vidomi wrote. That's why I'm asking how you'd feel about it if they weren't shipping the modified DLLs at all. Because the DLLs are GPL compliant, so shipping those DLLs isn't the real issue here.

In your thought experiment, they would have had to compile at least a header into the original program in order to dynamically link later. Whether this was the original header or a reverse engineered header, its still a derivative of the original work. The header would be GPLd, compiled in -> entire work GPLd. You need the LGPL for that to work.

Unless the VirtuaDub headers contain inline code, that argument doesn't hold up at all, especially not if Vidomi was using reverse engineered headers. Using your logic, WINE is completely illegal, because it contains "reverse engineered headers" from Windows. Typical headers are really just interfaces, and it's long been argued that duplicating that form of interface is fair use.

Programs that use GPL'd DLLs/libs/modules/etc

[Vic]

What would happen in a similar case.... Let's say I do this:

Write a BASH script that completely relies on other GNU utilities, such as find, grep, wc, sed, gawk..... Would this mean that I'd have to release the script under the GPL also?

I wish the best for the GPL, and the author of the original code in this story. I hope it manages to strengthen the GPL, as there still seem to be some ucertainties.

-Vic

Re:Programs that use GPL'd DLLs/libs/modules/etc

[hattig]

Nope, as the script would not be including these programs into its own memory space (as a linked library would do), but in their own, doing their own thing.

Also, they are arguably core OS utilities as well. If you took a non core OS GPL utility (say an application that did X), and wrote your closed source program that called X for part (or all) of its functionality, then provided that program as part of the distribution, then what is the situation? The program isn't linked in to your own program in any way, however your program is arguably a derived work. The same goes for pipes and CORBA, etc.

The GPL has to sort these issues out now - GPL software distribution terms and conditions, like:

You may only distribute this code/application as part of a GPL'd application unless the code does not provide functionality to said application, but rather provides an additional facility. (e.g., supplying emacs with a ray tracer, or something).

Re:Programs that use GPL'd DLLs/libs/modules/etc

[QuantumG]

stupid stupid stupid. His little script _would_ have to be GPL if he distributed it with the GPL tools.

Re:Oh look! More /. GPL alarmism

[bstadil]

You are wrong. Look at their "interpretation" of the GPL. What they say is that using function calls to GPL'ed is OK as it "only" interacts with the software.

Re:Grey area

[mandolin]

Secondly, the existence of the LGPL implies that dynamic linking is a violation of the GPL, otherwise why would the LGPL exist. Dastardly

1) I'm not totally sure but I'm pretty sure that the LGPL and GPL licenses predate widely-used dynamically-linking technology. Which would explain why the GPL leaves it so open to 'interpretation'. This is exactly the same damn reason (well, one of) trolltech wouldn't release under the GPL at first.

2) the LGPL's existance (or lack thereof) has no effect on whether dynamic linking is a violation of the GPL. (..and I'm suprised a post ranked as a 3 would miss this..) What would have an effect would be a judge's interpretation of the GPL.

Re:Vidomi's position

[Just Another Perl Ha]

In their quote from the GPL, they left out one very important clause. Here's the whole section (with their edited portion highlighted):

If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.

Thus, it is not the intent of this section to claim rights or contest your rights to work written entirely by you; rather, the intent is to exercise the right to control the distribution of derivative or collective works based on the Program.

They don't seriously think that creative editing can get them out of this, do they?

Huh?

[mcelrath]

Looks like (relevant parts of) the source code is available right here. Could be a short case.

--Bob

Re:Huh?

[jms]

Suppose you buy a piece of software and after installing and using it for a while, you decide to read the shrinkwrap agreement. You discover to your horror, that by running the software you agree to give your first-born son to the software company. They sue to get your son. Would the judge hand him over?

That would depend on whether or not you live in a state that has passed the UCITA legislation.

Re:Huh?

[grub]

Well, yes, the program can run, but it can't perofrm the functions it is intended to perform, so it can't do the things it advertises, and the things users obtain it for, so practically, and legaly, it can NOT ber run with out it.

Nonsense. I could write "Hello, World" with some funky twist so that it will only work on Linux, yet release my code as BSD licensed.

What you are suggesting is tantamount to "Your car won't run without gas, so your car is the property of Exxon."

Hell, the programmers of GPL code need food and water, do they by extension own the food on the shelves of the supermarkets?

grub

Re:Huh?

[rhavyn]

You're confusing software licenses and copyright licenses. A copyright license explicitly states how an author for something will allow someone to distribute it. IANAL, but I don't think there is anything you can't put in a distribution license. If someone doesn't like it, they can't distribute it. No one has any rights under the law to distribute something that is copyrighted without permission.

This has nothing to do with using the software, it has to do with the fact that the GPL says that if you *distribute* a derived program (AKA, any program that is distributed with the GPL code and needs it to perform it's inteded function), you need to release the code to your program to. And it's tough to argue that Vidomi works the way it's advertised without the GPL code that they've stolen.

Re:Huh?

[rhavyn]

This has been brought up dozens of times in this thread alone. GLibc is under the LGPL. The LGPL allows linking without requiring you to release your code. The GPL does not allow you to make a derivative program without releasing the source. What is a derivative program? Anything that depends on the GPL code to perform it's intended function and/or is packaged with the GPL code.

In this case, the software would not have all the features they are advertising without the GPL'd code. So according to how the GPL is written, it's a derived work.

Furthermore, the GPL is a copyright license for distribution. That means that the author can say pretty much whatever he wants. If the GPL is shown to be invalid, then it will be illegal under copyright law to further distribute *any* software under the GPL. Without a license you have no rights whatsoever to distribute the software.

Re:Huh?

[rhavyn]

The GPL explicitly states that if the program requires the GPL to do its intended purpose then the program is derived from the GPL code. Now, go over and look at the website and notice how many features that are advertised as being part of this program won't work without the GPL'd code. Then come back here and tell me that it would fulfill it's intended purpose without that DLL.

Re:Huh?

[Erik Hensema]

When you compile a DLL, the linker generates a .dll file and a .lib file. This .lib file is linked during compile time.

Also, you probably need GPL'ed header files to make use of the library.

Re:Huh?

[Erik Hensema]

Yes, but I don't think they're distributing code statically linked against the lib file. So the fact that the compiler happens to spit one out is irrelevent.

They must distribute code statically linked against the lib file, because when they wouldn't, they couldn't use the DLL.

Windows however has a dlopen()-like interface, so true runtime linking is also possible.

Re:Huh?

[JabberWokky]

Yes, but remember that linking to a dll occurs at runtime not at compile time, so the people linking to the dll are the end users, not Vidomi. I don't see how Vidomi can be held accountable for actions that end users might take with their program.

Ah, but you fail to realize that lawyers are not programmers. Most of the time, this is a bad thing, but this may very well be a good thing in this case.

They won't care if it's Corba, DCOM, or staticly compiled. Hell, they don't even know what that means. All the law cares about is: "Did this action violate the spirit of the license". Sure, they'll try and interpret the license every which way by pulling out each relevant sentence and dissecting the meaning of every word, but it won't be a technological analysis. Technical, yes, but not technological.

They will look, see what the license says:

2.b) You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License.

(later in 2) These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.

I boldfaced the important bits. Are they distributing a piece of software that is a work (as a whole) that is based on the GPLed software? If so, it must be GPLed, even if it's connecting to the code on a seperate server... as a higher level interpretation (which is my own, and not legal), even if the GPL code were to reside on a server: if the app requires the GPLed code as a basic part of its core functionality, then the GPL applies to the work as a whole.

--
Evan YMMV IANAL, but I used to work with them.

Re:Huh?

[Darren.Moffat]

glibc which is what Linux uses for C runtime is under LGPL not GPL. LGPL specifically allows the linking. The virtualdub code being used in the Vidomi case is under GPL not LGPL.

Re:Huh?

[bdowne01]

It looks like the argument that is spurring is not whether they are releasing the open-source pieces of the code or not, it's whether they are allowed to keep their "proprietary" portions closed-source while their application uses the open-source pieces for most of it's functions.

Part of the GPL states that they're allowed to do this only if their application keeps the code completely seperate, but it doesn't appear that they are.

Re:Huh?

[ADRA]

Legally, the "product advertised" does work, even if some of it is open sourced.

The program does not segfault if the dll is missing, so on GPL terms, the code is not essential to the product. The GPL has no bearing on the product's marketing. The product as shiped comes with all the functions as promissed, so why should customers care if some of the program is open source. PS: They don't unless they are lame nerd people like us.

By taking a position of superiority you show how nearsighted you are. Thus Spake ADRA

Re:Huh?

[Pedersen]

Relevant is, in this case, irrelevant. GPL says that if you link your code with mine (which I've released under the GPL), all your code must then be released under the GPL. If you only want to release my code,and not yours, you had better hope I chose the LGPL (or BSD, or...), which will let you do that. Otherwise, you're SOL. All your code is out, or you don't use my code.

TTYL. Gotta reboot a TLA server now (hate these long nights! :)

Re:Huh?

[hattig]

it becomes crystal clear that all the GPL'd code is in a self-contained DLL

Which is entirely besides the fact of the whole matter, which is that the main application knows about and actively makes use of the GPL code in the library, and hence anyone with half a brain (read: half a jury if you are lucky) will conclude that the library is part of the application. Or are you arguing that khtml is not a part of konqueror? Or that the windows HTML library is not a part of IE? (initially, Microsoft would now let you think that it is part of windows, and IE is one tool that uses it)

Dynamic or static does not affect the above. I can't link against unknown library X if I don't know what it does can I? I need to know what it does in order to use it, and make it a part of the application.

Re:Huh?

[hattig]

Some points in response:

1. Their program will not work as advertised though. This they are advertising the GPL code as a feature of the program, and that could be seen legally to be an indicator that the code is part of the software, not an independent program that doesn't affect the closed source software at all.

2.Customer's don't care - of course not, even if they are breaking copyright law by proxy! Now they would care if the law was like that for handling stolen property (whether or not you knew it was stolen, it is taken away from you, and you go to court).

Just my opinions. The GPL has specified what to do about DLLs already - hence the LGPL. It has yet to decide what to do about CORBA/pipes/etc - other methods of possibly using GPL code in closed source products.

Remember above all - the GPL was chosen by the author and copyright holder of the software as his preferred means of distributing the software. He didn't choose the LGPL.

Re:Huh?

[hattig]

Is the following program a derived work of libc? main() { printf("Hello world!\n"); exit(0); }

Yes, hence Linux libc is released under the terms and conditions of the LGPL, not the GPL!

Re:Huh?

[Velox_SwiftFox]

And Exxon could indeed demand that if they owned a patent on the production and use of gasoline.

"Be sure to buy plenty of Motor Mutton. You will have to, you see." - Bruce Mcall Bulgemobile brochure, National Lampoon

Re:Huh?

[ffatTony]

You're right, but from the comments posted on that page it appears as though the argument is whether or not they are obligated to provide source code for the portions they wrote in adition to the gpl'd code. Their work requires Mr. Lee's GPL'd code. Is their code a derivative work? If so, as I understand it must be GPL'd too.

Re:Huh?

[dimator]

commonly distributed

This whole "distribution" thing is what it all hinges on, I think. If this company had TWO links for you to download on their page, one their own proprietary front-end, and one GPL'ed shared lib of GPL'ed code, would this be OK? They are not being distributed togther (meaning in the same installer, zip, whatever) right?

If I can make that claim, then the only problem for any company wanting to use GPL'ed code would be the two-download problem (who wants to download two things when installing crap?) This can easily be circumvented by including something that downloads the GPL'ed library from the proprietary installer, so the user/installer does not have to deal with it. As long as you have the source to the GPL'ed shared lib up somewhere, you'd be OK wouldn't you?


---

Re:Huh?

[scotch]

The code is free BUT YOU CAN"T USE IT unless ...

s/USE/DISTRIBUTE/

s/warmiak/fud-meister/

Re:Huh?

[scotch]

No, that is not my opinion of the political right at large, just of you (as I clearly stated). You might be surprised to learn than I agree with some of the goals and ideals of the republican platform. What I have a problem with is the deluge of caustic rhetoric from over-eager supporters like yourself and certain conservative media outlets. While I understand all conservatives don't share your black and white view of the world, those like yourself can be so vocal as to drive away people who have a more centric view on average of the issues.

I don't hope to convince you of anything; really I am just having a lot of fun baiting you. ;)

Re:Huh?

[scotch]

In another thread, I showed how you used smear tactics while at the same time claiming that conservatives rarely do this (search for "ironic" in the motif story).

"Sometimes things are black and white"

Like what, your sig? That's usually what gets you involved in these off-topic threads. Here's a test to see how much of a party drone you are. Name one major platform issue of the republican party that you disagree with. Can you do it?

Or are you too far gone?

I never said you promoted any particular media outlet. Everthing I've posted directly applies to you and your illogical (more irony) and unsupportable claims.

I can't really waste much more time on this - you just ignore most of the points other raise anyway.

I'm heading off on vacation.

HAND

Re:Huh?

[Saidin]

I think that would really only be true if the proprietary part of the program could run independantly of the GPL dll. If the program requires that DLL to run, and do anything useful (other than displaying an error message :) ) than you really can't claim they are separate, no matter how you distribute them.

Re:Huh?

[aozilla]

The GPL says that any derivitive work is covered by the GPL. Whether or not dynamic linking is considered a derivitive work is not the decision of RMS, it is a decision of the courts. In fact, if dynamic linking is not a derivitive work, there is indeed nothing RMS or anyone can do, because the modifications and distributions (except potentially distributing both together) are perfectly in sync with the GPL.

Assuming the GPL itself stands up, this is a win-win situation for free software. If dynamic linking is allowed, it will be legal for free software to link to proprietary software dlls, as long as the end-user herself is responsible for the actual obtaining of those proprietary dlls. Just think what this could do in terms of WINE.

Personally, I'm in favor of less strict copyright laws, not more. I'd prefer for the GPL to lose this one.

Re:Huh?

[rgmoore]

I think that you're missing the bigger issue. Basically, Vidomi is turning a GPLed library into a dll and linking a proprietary program against it. They're releasing the code for the dll, but not the program that links to it. Since the GPL claims that work that is commonly distributed as an organic whole constitutes a derivative of any packages that make up the whole, that appears to be a GPL violation; the whole thing should be released under the GPL. If Vidomi wins, it will mean that the GPL is substantially weaker than people have generally believed, as it will be possible to make part-GPL/part-proprietary programs by isolating the GPLed part in a dll.

Re:Huh?

[ahrenritter]

When reading their site a few minutes ago, I saw the latest concession they are making to try to become more compliant. They have stated they are going to pull the GPL'ed dlls out of the primary setup and make them an optional "plug-in" with a seperate setup. This seems fair to me, and while IANAL, it seems to hold true to the spirit of the GPL. Now while I didn't see their website before the fit hit the shan (although I did check Google's cache and it was the same), their credits page seems *very* upfront about what they did and to a common bloke like me, seems like they are at least attempting to be GPL compliant. What do y'all think?

Re:Huh?

[DmitriA]

First of all, libc is under LGPL, so that part of your point is moot.

Second of all, you are partially right in that you can right close-source linux kernel drivers and modules that LINK with the GPLed Linux kernel.
HOWEVER, the only reason you are allowed to do that is because Linus speicifically gives permission for people to do that in his license (he can do that, since he holds the copyright to the Linux kernel and, thus, can license it under any license he chooses to - whether it is strict GPL or modified derivative of it)

Re:Huh?

[pcidevel]

The GPL itself doesn't say anything specific [IIRC], but RMS has decided that dynamic linking in a DLL is not OK, but dynamic linking to non-GPL OS functions [at load time] is OK, which is a completely arbitrary distinction based more on intent than on actual technical differences.

Actually this is incorrect. libc is distributed under the LGPL which allows dynamic linking. The GPL explicitly forbids dynamic linking. This guy could have released his code under the LGPL if he wanted to allow this use, he didn't. Read my statement above as to why this is VERY unfortunate for the Free Software and Open Source movements. This will be our most harmfull moment ever.

Re:Huh?

[DarkEdgeX]

I'm not seeing the GPL violation here-- they put the code in a DLL, distribute all of their code for the DLL (AFAIK), and their app only uses the DLL if it's there, but CAN BE RUN WITHOUT IT.

In truth, if they wanted to avoid a GPL issue, they could write a new DLL, or have someone else write a DLL that would work with their application, but really, this sounds like a simple case of plug-ins (ala Photoshop), where the code in the DLL isn't needed for the main program to run.

Re:Huh?

[DarkEdgeX]

If the code in the DLL is nothing more than a plug-in (and the main program can run without it, as has been said I believe on one of the linked-to websites), then how is this a violation of the GPL if the source code for the DLL is free (as is the DLL itself) and downloadable from their website?

Think of apps like Adobe Photoshop, and think of what would be the case if some 3rd party wrote a plug-in for Photoshop using GPL'd code-- would this force Adobe to open source the entire Photoshop product? Of course not. The only difference here is that it's not a 3rd party that's using the GPL'd code. The app still runs without the DLL though.

Re:Huh?

[DarkEdgeX]

If I write a plugin for Photoshop, I'm using Adobe's API for writing Photoshop plugins. I can reverse-engineer that API if I must, but that does not violate Adobe's copyright. If I write an application that uses a GPL'd DLL, I must open that DLL (use it's header in my code) in order to use it, and that act is what violates the GPL.

I'm sorry, I'm not seeing the relevance of the header here. You can re-write the header since you have the GPL'd source for the DLL, so what exactly is the violation? There'd be a point to this rant if Vidomi had made their own DLL with their own custom modifications and then not released the source code to THAT, or if Vidomi had copied/pasted the code from the GPL source directly into their main applications code, but this isn't the case. The application is modular, and could run with or without the DLL, or with different DLL's from different people (GPL or no GPL code). So how again is the header relevant?

As for the differences between a DLL and a plugin, you're part right, part wrong-- now that Vidomi has released their product, there's nothing stopping anyone from writing a new DLL that performs the same (or similar) functions as the existing DLL to the same spec as the first one-- with this completed DLL in hand, one could delete/rename the GPL DLL, and replace it with the non-GPL DLL and run it normally. Literally, this application doesn't require the GPL code to run.

Re:Huh?

[DarkEdgeX]

Shrug, a feature list is a long ways away from a GPL violation. The app would still run fine without the DLL, it just wouldn't be able to perform all of it's tasks. Big deal-- if I delete some plugins from Photoshop, I suddenly can't save JPEG or GIF files, and my happy lens flare filter will disappear; but the marketing says it's all there. Whoops, the marketing didn't mention it was in seperate libraries!

Re:Huh?

[DarkEdgeX]

Their program will not work as advertised though. This they are advertising the GPL code as a feature of the program, and that could be seen legally to be an indicator that the code is part of the software, not an independent program that doesn't affect the closed source software at all.

Yeah, while it might get the case through the front door of the courthouse, once the product is looked at from a technical point of view and we've eschewed the marketing crap, it becomes crystal clear that all the GPL'd code is in a self-contained DLL.

Just my opinions. The GPL has specified what to do about DLLs already - hence the LGPL. It has yet to decide what to do about CORBA/pipes/etc - other methods of possibly using GPL code in closed source products.

I don't see how the LGPL applies to Dynamic Link Libraries though-- the LGPL seemed to me to be more intended for use with STATIC libraries (libraries that must be linked against, for example, the C runtime libraries). With DYNAMIC libraries though, the GPL or LGPL would be enough to give the public the right to access code in the DLL as long as the code IN the DLL remained GPL'd and open source.

Re:Huh?

[DarkEdgeX]

Or that the windows HTML library is not a part of IE? (initially, Microsoft would now let you think that it is part of windows, and IE is one tool that uses it)

Actually the HTML library isn't part of IE. Anyone can use the HTML library (MSHTML), and that's what Microsoft intended. Anybody could also re-create the API and write their own custom HTML library and as long as it re-created all of the interfaces and COM objects that MSHTML did, Internet Explorer would still run.

Again, the app that was written will run fine in the absence of the DLL that contains the GPL'd code-- it just won't perform many (any?) functions. Now if someone else were to re-create a new DLL that performed the functions required, they could include their DLL and the app would expose all of these additional features. Just as a plugin for Photoshop does.

It's more appropriate to realize that this app (and apps like it) are nothing more than a shell or wrapper, and that they don't care what library is present as long as the one that IS present properly recreates the functions it expects.

Re:Huh?

[DarkEdgeX]

They ship the DLL with their product, therefor, according to the GPL, they must GPL the entire thing.

Uh huh, so if he just made the files two seperate downloads this would put him in the clear?

I think it's basically just nit-picking at that point, and a waste of time to try to enforce. If the community wants GPL'd code to flourish and make it into the mass market, attitudes like this aren't going to exactly bolster their position.

Re:Huh?

[HyperbolicParabaloid]

>>CAN BE RUN WITHOUT IT. Well, yes, the program can run, but it can't perofrm the functions it is intended to perform, so it can't do the things it advertises, and the things users obtain it for, so practically, and legaly, it can NOT ber run with out it.


-------------------------

Re:Huh?

[NoOneInParticular]

From the copyright statement of glibc:

The GNU C Library is free software; you can redistribute it and/or modify it under the terms of the GNU Library General Public License as published by the Free Software Foundation; either version 2 of the ...

Notice the 'Library' statement in there. It is LGPL, so linking against it is ok. About the socket and file access... you might be right, though I think the people at FSF might have an answer for that too. If this were the case maybe glibc is guilty of a GPL violation. Now that would be ironic

Re:Huh?

[3.1415926535]

For the last time, this is copyright law and not software license law. As an individual, you have absolutely no rights to distribute except those explicitly granted by a contract with the copyright holder. Under the GPL, the only way you can get the right to redistribute the work is by agreeing to a contract (NOTE: not a license, a contract) with the copyright holder agreeing that you will distribute the source to your derivative work with the GPL'd code included. If you don't agree to this license, you have no right to distribute your derivatve work. End of story.

Re:Huh?

[3.1415926535]

Step 1: Bob downloads Jim's copyrighted GPL'd work. Bob has no right to distribute (or even use) Jim's work.

Step 2: Bob accepts Jim's contract, granting him the right to use and redistribute his product as long as he releases anything he makes with it, INCLUDING PROGRAMS LINKED WITH IT. It shouldn't matter if this falls under some technical definition of "derivative work", it's still an action that requires GPL'ing the new code under the terms of the GPL-contract with Jim.

(By the way, IANAL so I don't know if this how the GPL actually works, but it's how it SHOULD work.)

Re:Huh?

[Rick the Red]

Looks like (relevant parts of) the source code is available right here. Could be a short case.

Actually, your statement (and theirs) would be true if the VirtuaDub code was released under the LGPL, but it's not, it's under the GPL, so removing it to a library (in this case a DLL) isn't good enough. You're right, it should be a short case. They violated the GPL, they should quickly lose.

Re:Huh?

[Rick the Red]

If I write a plugin for Photoshop, I'm using Adobe's API for writing Photoshop plugins. I can reverse-engineer that API if I must, but that does not violate Adobe's copyright. If I write an application that uses a GPL'd DLL, I must open that DLL (use it's header in my code) in order to use it, and that act is what violates the GPL.

Photoshop does not need to read or understand my plugin code in any way to use that plugin. Vidomi must understand VirtuaDub's code in order to use it. That's the difference between a plugin and a DLL, and it's the difference between honoring and violating the GPL.

Re:Huh?

[Ayende Rahien]

You are more or less correct.
You don't have to GPL your own product if it's sufficently seperate from the GPL code (I think that DLL files are enough for that).
/However/, if you ship those DLLs with your programs, then you *must* release the whole thing under the GPL.

Re:Huh?

[Ayende Rahien]

> don't see how the LGPL applies to Dynamic Link Libraries though-- the LGPL seemed to me to be more intended for use with STATIC libraries (libraries that must be linked against, for example, the C runtime libraries). With DYNAMIC libraries though, the GPL or LGPL would be enough to give the public the right to access code in the DLL as long as the code IN the DLL remained GPL'd and open source

They ship the DLL with their product, therefor, according to the GPL, they must GPL the entire thing.

Re:Huh?

[Ayende Rahien]

Nope, if the properity part is sufficently seperated from the GPL code, it doesn't have to be GPL.
You force it to be GPL if you distribue the whole thing together.
I think that his idea about the installer would work.

Re:Huh?

[Ayende Rahien]

Yes, if Photoshop was distribued with GPL plug-ins, then Adobe would've to GPL it.

Re:Huh?

[Ayende Rahien]

> Uh huh, so if he just made the files two seperate downloads this would put him in the clear?

I don't *think* so.
That they ship the whole thing together just makes it easier to say that they violate the GPL.
Shipping it as two seperated downloads will get you into the courts deciding "is dynamic linking is derived work or not?"

Re:Huh?

[BlueTurnip]

They're releasing the code for the dll, but not the program that links to it.

Yes, but remember that linking to a dll occurs at runtime not at compile time, so the people linking to the dll are the end users, not Vidomi. I don't see how Vidomi can be held accountable for actions that end users might take with their program.

There's a larger issue here too. Can the terms of a license agreement require that licensees give up arbitrary rights pertaining to things other than the licensed code? Suppose you buy a piece of software and after installing and using it for a while, you decide to read the shrinkwrap agreement. You discover to your horror, that by running the software you agree to give your first-born son to the software company. They sue to get your son. Would the judge hand him over?

My point is there are limits to what rights must be relinquised as part of a software licensing agreement. I predict, as I have for a long time, that the GPL under this interpretation will be unenforceable.

Re:Huh?

[BlueTurnip]

All that matters under copyright law is that Vidomi is a derived work with respect to VirtuaDub. Period.

Vidomis modified version of the library is certainly a derived work and they've released the full source code for it.

They also release a program which when run happens to link to the dll. Please explain to me exactly how that program is a derived work.

Is the following program a derived work of libc?

main() { printf("Hello world!\n"); exit(0); }

Re:Huh?

[BlueTurnip]

When you compile a DLL, the linker generates a .dll file and a .lib file. This .lib file is linked during compile time.

Yes, but I don't think they're distributing code statically linked against the lib file. So the fact that the compiler happens to spit one out is irrelevent.

Also, you probably need GPL'ed header files to make use of the library.

This is probably the only potentially valid argument for saying the executable which links the dll is a derived work. But since include files generally don't create code, but only define function prototypes, etc., an argument could be made that it isn't sufficient to make the executable code a derived work. It will be interesting to read the judge's opinion.

Re:Huh?

[BlueTurnip]

Yes, I am aware of the existence of the LPGL and the reasons for creating it. But that doesn't make my program a derived work. In fact, that "Hello, world!" example program existed before Gnu libc ever did! So I find it odd that you assert it's a derived work!

Just because someone creates a license like LGPL as a result of a possible interpretation of law, doesn't make that interpretation of law correct.

Re:Huh?

[BlueTurnip]

If it is compiled and linked (statically or dynamically) to libc, then hell yes it is a derived work.

In the statically linked case, I would agree with you, not so for the dynamically linked case.

You cannot perform dynamic linking without embedding derived information from the original library into your executable e.g. routine names or offsets etc.

In my particular example, only the routine name "printf" is embedded, and that name was part of an open standard long before Gnu libc. Now in the case before the courts there may well be routine names that were created by the copyright holder, but unless those routine names are trademarks, I don't see how use of the mere names constitues copyright infringement. It is functioning code that is copyrightable under software copyright law, not interfaces or routine names. That's why the original IBM PC BIOS was legally cloned (the clones certainly used the same offsets at least.)

This makes it a derived work under law, and no matter how much you wish it wasn't, it is.

Well, the law doesn't depend on what either of us wishes. The judge will decide.

Re:Huh?

[BlueTurnip]

But the whole question here is whether or not this program is a derivative work because it dynamically links to a GPL library. Personally, I don't see how it is. Some have argued that because some header files may have been used, that makes it derivative, but unless these header files actually contain code that is included in the executable, I don't see how it is.

Re:Huh?

[BlueTurnip]

It shouldn't matter if this falls under some technical definition of "derivative work"

I think it matters a lot under U.S. copyright law. People are allowed to set the terms of copyright of their own works or works derived from it, but not other works. So it really hinges on whether or not code that links at runtime to a DLL is derived from the code that produced the DLL. Although in some cases it could, I don't think it generally does.

Suppose I write a book. You buy it and notice in the front cover it says that by reading the book, you agree to transfer the copyright of your first novel to me? Does that sound reasonable?

Re:Huh?

[BlueTurnip]

IANAL, but I don't think there is anything you can't put in a distribution license.

Sure, you can put anything you want in the license, that doesn't make in enforceable. The Supreme Court decision which established the First Sale Doctrine (in 1905) I believe, established that fact, when a publisher included in the copyright notice the fact that the book can't be resold. The Supreme Court held that the book could in fact be resold, despite the terms of the copyright notice.

Re:Huh?

[Treylis]

>Irrational love for so called "gun owners rights"

How are gun rights irrational? Better watch what you say around here, or ESR will get ya. ;-)

Re:Huh?

[warmiak]

Dude, you are following me around this place ..

That's not good and possibly bordering on sick.

Re:Huh?

[warmiak]

"I find you to be almost completely full of shit. "

Isn't that your opinion of conservatism and republican party at large ?

Re:Huh?

[warmiak]

"as to drive away people who have a more centric view on average of the issues."
And who are you to say that having "more centric" view is always a good thing ?
Remember, for every case in history of "cool heads" being ultimately right there are counter examples of people who completely fucked up their chance simply by trying to be "objective and cool."
Sometimes things are black or white.

"What I have a problem with is the deluge of caustic rhetoric from over-eager supporters like yourself and certain conservative media outlets."

What in particular? Show me one ultra-conservative and completely out of touch with reality idea I promoted in any of my posts ?
Can you ?

Re:Huh?

[warmiak]

"That's usually what gets you involved in these off-topic threads. Here's a test to see how much of a party drone you are. Name one major platform issue of the Republican Party that you disagree with. Can you do it?"

I am not even a Republican.
I just can't stand Democrats anymore and see them as much bigger danger to our way of life than Reps are. As for Republicans, I completely fail to understand their rabbit and blinded push for war on drugs, their irrational love for so called "gun owners rights" and many other issues as related to social problems.
In fact, most of the reasons I respect Reps for are related to their views on economical issues and not their social agenda (although sometimes both are related.)

Re:Huh?

[Obelisk1010]

If dynamic linking requires the release of source code, then Linux is done with regards to commercial development. The C runtime is used by alot of commercial apps that are not open source. In fact, taken further, the use of basic system services such as sockets or files could be considered GPL violations.

I think Vidomi's in the clear on this one. They have open sourced the DLL and construing dynamic linking as inclusion is ignoring the last decade of software development.

Re:Huh?

[Obelisk1010]

but RMS has decided that dynamic linking in a DLL is not OK, but dynamic linking to non-GPL OS functions [at load time] is OK

That will be it's undoing. As soon as you get arbitrary, a good lawyer has a lot of room to manuver.

It will boil down to the fact that the GPL'd code is not in the same file. It's the only distincion that both techies and lay people both will find clear. The pipes/RPC/SOAP arguments you made will be thrown at whatever experts the FSF uses.

Re:No harm intended/done?

[The Swedish Chef]

Imagine you write an application in Java... the Sun JVM is freely downloadable, so you can expect the user to go get it himself. But for the convenience of the user, some companies bundle the JVM with the product. Not a really big licensing issue.

It's not a big issue because Sun explicitly grants developers permission to redistribute the JRE. Links:

• Runtime FAQ
• Java 2 Runtime Environment License

Re:DEADLY VIRUS!

[ergo98]

you know, no one forces you to use GPL'd code. No one is preventing you from rolling your own. I personally like the nature of the GPL, it means that I can release my code for free (speech) and have assurance that my gift to the community won't be taken by some company and sold as a proprietary piece of software. and above all

Any application that is running on Linux is utilizing the work of GPLd code, therefore should only GPLd code be legal to run on Linux? If you say no then tell me why: I'd love to hear the definition of what is acceptable use and what is not. It seems to me that acceptable use is whatever will get the FSF and the GPL in the door, at which point the rules change. IBM, Sun, Oracle: They're all idiotic to go near Linux or GPLd code with a 200 foot pole.

my gift to the community won't be taken by some company

This is the funniest irony of the GPL community: Somehow if someone uses GPLd code they have "stolen" it (despite the fact that their use of it in no way inhibits its freedom and accessibility), yet GPL fanatics will be the first to tell you that there is no such thing as Intellectual Property and that the etherial nature of code makes it free for the taking.

In other news I am hardly surprized to see my parent post modded to Troll. How very surprizing. This whole GPL (I will not say OSS as the BSD license is respectable and BSD developers are truly selfless) is such a sham. The talking heads on top, raking in the rewards (incl. fame) for the movement, change their tune all the time to insidiously infect the world with their religion. It's a real sucker that doesn't see through it.

Again if I use a function in a library (which itself may be opensource), how does that differ from making a SOAP call to a GPLd web server? Would the SOAP call, and the fact that my code "relies" upon that code, require me to GPL my code? What if NASDAQ put up a stock quote server that was running on a GPLd web server : Would every single application that connects to it need to be released under the GPL under threat from the "free" (but harbouring lots of lawyers) FSF? I have no doubt the GPL hoardes will say "no of course not!" but it's only to make the GPL palatable.

Re:GPL code is proprietary

[ergo98]

If however like me, you prefer to preserve your code as open and the source code available to normal users, you use the GPL

This is classic open source FUD. If someone utilizes some parts of open source code in a proprietary, commercial, $1,000,000 / copy piece of software, in no way does that limit or subvert the original open source code (it's still just as free as it originally was). If I release a BSD-License MPEG compressor and Microsoft uses parts of it in Media Player 9, that doesn't make my code disappear. Indeed, as many open source fanatics as want to can make competitors to the MS product with the same code, and this naturally means that the effective ability to charge in a commercial product is limited by marketing, and by the value-added put in by Microsoft, or whatever vendor happened to integrate open source code.

Re:DEADLY VIRUS!

[ergo98]

Definition : Troll - Anything that puts forth an argument that I or my cadre of cohorts disagree with i.e. Someone advocating that religious persecution is heinous would be a troll in Nazi Germany.

You can utilize GPL'd applications, you just can't directly link to it. This isn't a moral issue, it's a licensing issue, please don't make it more than it is.

Define "directly link". A DLL is dynamically loaded at runtime (it isn't linked at compile time), and while it is loaded and runs in the common memory space of the host application you could just as easily (about a 20 second job for most functional applications that don't used shared memory. A little more if you're using shared memory) separate it to an out-of-process COM server that communicates via COM stubs. You could just as easily stub it out to work via a HTTP server through SOAP and XML.

The FSF and GPL fanatics like to keep their options open so they can "enforce" the GPL only so far that they don't offend people who are a little smarter than accepting the viral conditions and religion of the GPL, but they keep their options open in the future. If they can effectively attack a DLL then there is nothing stopping them from imposing the same conditions on functionality called through a HTTP layer (despite the semantics it's the same bloody thing). I would love to see technical specifics dictating what is and what isn't covered under the GPL.

the rest of your post is so troll-like it's not worth my time. A little common sense would do you some good

Pot-Kettle-Black. The problem is that perhaps I have a little too much common sense so I can see past narrow interpretations that serve a movement at a time. If a DLL is open to attack then there is nothing but myopic people who can't understand parallels stopping them from going after anything that remotely utilizes GPL'd code.

I wouldn't have any problem with a company taking and using my code, it's simply a license issue, if you don't like the license that code comes with, roll your own, don't whine and bitch about how you really want to use that code but the license it comes with is SOOOOO UNFAIR.

No it isn't unfair, however it's viral and this is something that so many in the movement fail to acknowledge. Sure, come on in, enjoy yourself...oh by the way we own your soul. The point is that given the very loose definitions defined by the GPL (note that I've heard that theoretically everything that runs on Linux has to be GPLd but for the fact that Linus has unilaterally stated that that is not so) it is incredibly dangerous for any company to touch anything GPLd with a 200 foot pole.

Re:Microsoft must be drooling

[ergo98]

It's interesting that you call that natural conclusions of the various possible outcomes "FUD" (a word that much like troll refers to whatever is of the opposing view) when in reality they would indeed be the logical conclusions.

If Vivendi is forced to release their code then yes: If you use GPLd software you lose your IP (oh, wait...you can make money on distribution and service...bwahahahahahaahahahahahahahahahahaaha).

Re:DEADLY VIRUS!

[ergo98]

This troll has been answered many times over in this discussion. In short: an application that is running on Linux is utilizing kernel system calls and glibc, neither of which is licensed under the plain vanilla GPL. If you want to know the details, read some of the replies to the other trolls.

Generally the reply is "Uh, don't you worry about it. Your software doesn't have to be under the GPL...for now...". http://www.linux.org/info/gnu.html : Please point out the special wording that says "Oh, but don't worry...it's not really under the GPL...". Whatever.

Re:Vidomi's position

[ansible]

Um, no.

It proves again that the GPL can't be indescriminately combined with proprietary software.

But remember that proprietary != commercial. Though you do have to structure your business differently if you're developing GPL'd software.

Real Men read binaries all the time

[ansible]

You might be able to get away with such a Perl script hack in college, but not in the real world.

Changing variable names won't substantially change the binary code the code compiles into.

Even switching compilers and optimization levels doesn't give you much obfuscation. Besides, how many viable compilers are there for any particular platform? It's not hard to try them all.

To substantially change the generated code (in a hard to detect fashion), you need to change the fundamental structure. And then you need to be careful and not introduce new bugs. And you still have to test it a lot. You might as well write it from scratch.

Try running 'objdump --disassemble /bin/ls' on your Linux system for yucks.

Re:Real Men read binaries all the time

[jonathanclark]

Changing variable names won't substantially change the binary code the code compiles into.

Changing variable names won't change the binary code the code compiles into by a single bit. Compilers don't care what variables are named as long as you follow proper syntax.

Re:Real Men read binaries all the time

[Ayende Rahien]

> Besides, how many viable compilers are there for any particular platform? It's not hard to try them all.

For x86/Windows ?
GCC, LCC (based on GCC), MS-VC, Intel C++ compiler, Borland C++ builder, etc.

That isn't the problem, the GPL code is assembly, meaning that you just debug it to see the "source code".

Re:DLLs and COM

[landley]

>What if I take GPL'ed code and create a COM
>component out of it. I releast the full source.

>3 months later, I decide to start writing a
>closed-source app and I find the component would
>be useful.

>Should my application now be covered under the
>GPL?

You're the author. Your rights to use the code do not originate from a license granting you access to somebody else's copyright. You OWN the copyright.

A license you grant delegating some of your rights to other people doesn't limit YOUR rights to your own work.

Unless, of course, you sign over the copyright. But that's not what a license does. (If your employer paid you to create something, they may own the copyright due to their contract with you. And the FSF wants people to sign their copyrights over to them so they'll have a stronger position if they ever have to sue anybody. But that's a seperate issue.)

Rob

Re:what if....

[DavidTC]

Most of them wouldn't have to rerelease their code, their software is licensed under GPL2 or later.

-David T. C.

Re:what if....

[DavidTC]

No it doesn't. The GPL isn't giving them the right to use the code under the GPL 2.0 or later, they, the authors, are giving that right.

If I say you can use my programs under X, Y, and Z licenses, and X is ruled invalid, you can still use it under Y and Z.

-David T. C.

Re:A true test of the GPL

[Laxitive]

I think you are wrong here, and fall into a common trap - the "technical" trap. You assume that any law is true to it's fullest technical interpretation, and that anything that can be interpreted as meeting those technical considerations, is affected by the law.

But laws are human, and are meant to be interpreted by a human, and not a robot. A human (judge) can take a look at a situation, and make value judgements about the validity of the law or not.

A good example is the Napster case - the judge decided that Napster was illegal because it was aiding in piracy. But all Napster is, is an online opt-in search engine/database. Does that mean that all search engines are now illegal, because they too can be used for piracy? No - because as humans, we can see the difference between Napster and search engines, even though someone can make a claim that Napster == Search engine.

Likewise, it's kind of obvious that this product pretty much relies on GPL code to be useful. Perhaps it offers some token functionality without the GPL code, but, that functionality is pretty irrelavent. A human can take a look at it, and make a value judgement on the situation.

Laws are not interpreted by turing machines (nor are they expressed in lambda calculus) - and that's a very good thing indeed.

-Laxitive

Re:Vidomi's position

[sab39]

"this License, and its terms, do not apply to those sections when you distribute them as separate works?"

Oops.

Re:Crap.

[hawkfan]

It seems to be working out well for Rambus and Apple.

Re:A true test of the GPL

[Omnifarious]

I disagree. I do not consider using the various intel 'trap' instructions to constitute linking to the kernel. The major libraries everybody links against under Linux are all covered under the LGPL which is much more lenient in this regard than the GPL.

The LGPL is specifically designed to address the very concern that you have. If the author had wanted to allow his library to be used in this way, he would've put it under the LGPL.

Re:COM objects?

[IntlHarvester]

COM) It all depends on your interpretation of "derived work". It's true that if significant functionality was derived from a GPL COM object, you'd be in trouble. (I don't know if Virtua Dub actually wrapped the DLLs in COM or not, but it shouldn't matter.)

For example, if I write a GPL media player COM control, does Word and IE have to be open sourced because someone could embed the control there?

DCOM) MS DCOM/COM+ eventually turns into MS-RPC, so your distinction doesn't make any sense.

I can write open source server component (OK - talks to the OS DCOM libraries) and talk to it with a closed source client (talks through the OS DCOM libraries and RPC to get the unrestricted 'output' of the server).
--

Re:A true test of the GPL

[Outland Traveller]

Of course it would help! With the wonders of UTICA and DCMA, you can pretty much put whatever you want in licenses and have it be binding!

Commercial software left mere copyright protections behind long ago. Don't ask me why- I personally do not buy the argument that software needs copyright, patent, AND DMCA monopoly grants in order to foster innovation. If anything, it accomplishes just the opposite. What a travesty!

It's to the FSF's moral credit that the (L)GPL uses the most reasonable protection mechanism for software (copyright) and stops at that.

Re:Oh look! More /. GPL alarmism

[rhavyn]

Actually, it's rather cut and dry if you actually read the license. It explicitly says that the license considers any work with requires the GPL'd code to fulfull it's intended purpose is a derivative work and must fall under the GPL. If you go and read the features that they advertise as being part of their app, it's pretty clear that without those DLL's it won't fulfill it's intended *and advertised* purpose.

Re:GPL, Distribution, Intent and Spirit

[rhavyn]

Linus has explicitly added a clause to the GPL to allow binary modules to link to the Kernel. Under the normal GPL it is not allowed.

Re:A true test of the GPL

[rhavyn]

Perhaps one aught to click on a story before responding. This isn't about linking to a library, this is about taking a GPL'd program, turning it into a library and linking to that. Furthermore, the license of the GPL explicitly states if a program requires the GPL'd code to produce it's intended results then it is a derived work. Without the GPL'd code in Vidomi, it would *not* have all the features it's advertised as having. I think that makes it pretty clear that according to the GPL it's a derived work.

The rest of your arguments are complete crap and have been explained many times in the past. First off, how could you concievable argue that my program is a derivative of the kernel ? Maybe the compiler, linker and libc, but GCC says in it's license that programs compiled with it are *not* considered dirived works and glibc is under the LGPL not the GPL (the LGPL allows linking).

This stuff is simple copyright law. Why can't people seem to understand that?

Re:Inquiry?

[rhavyn]

This is completely legal. It sounds like although your software uses perl and gzip, it doesn't really depend on them (you can always use a different zip tool or scripting language ... your software does it's thing regardless).

The problem here is that the program in question doesn't most of the things it's advertised as doing without the GPL'd code. I personally would consider that a derived work.

Re:VirtualDub and IP

[Samrobb]

The problem is that I am not being told the terms of the contract until after a transaction has taken place. In the US, at least, this is blatantly illegal... except that now, thanks to UCITA, it is specifically allowed in the case of shrink-wrapped software. More significantly, UCITA attempts to explicitly make the "if you are reading this contract, you have already agreed to abide by it" garbage that so many software licenses use legally binding.

I don't know about you, but I consider any law that twists legal principles established over the course of the last millena around 180 degrees to be "insidious".

Re:I had a (somewhat) similar issue - it gets murk

[Samrobb]

Linus was wrong.

Actually, I think he was right. The key point is the first part of the GPL quote:

If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works.

The case he presented was a non-GPL'd program that was not derived from his GPL'd code, but was capable of using his GPL'd code in a generic fashion (via a font selection dialog.) The GPL, in the above passage, seems to explictly recognize this case, as long as the non-GPL'd and the GPL'd products can "reasonably considered independent and separate works&quot.

What's a reasonable way to consider two products independent and seperate works? In the font case, well... if there is nothing in the source code that references the font explicitly, I'd think that would be a good indication that the works are truly independent.

So, if the bloke who wrote the check printing program took the neccesary steps to ensure that it is a "independent and seperate work" from the GPL'd code, it seems that he wouldn't be required to distribute it under the GPL. Which brings up the second point:

...when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License...

If the first assertion is true - that is, the check writing program was an "independent and seperate work" that was not derived from the GPL'd code - then the distribution is not an issue, since the requirement to distribute under the GPL only applies if you distribute the non-GPl'd code "as part of a whole which is a work based on the Program". Since the whole of the work was not based on the GPL'd program, then it would seem that the GPL allows distribution of the GPL'd program in conjunction with the non-GPL'd program, without the requirement that "distribution of the whole" be under the terms of the GPL.

Final note: if the argument that a non-GPL'd program can be forced into being GPL'd by a plug-in or other non-essential component is accepted by a court, I would be very surprised... since it implies that someone other than a copyright holder can alter the license under which a work is distributed. Keep in mind that this is apparently not the situation with VirtuaDub and SloMedia... SloMedia has, by all accounts, failed to produce an "independent and seperate work", which probably puts them four-square in violation of the GPL.

Re:It's funny...

[MSG]

merely because I viewed some GPL source
It's not viewing the GPL'd source that will get you in trouble, it's copying it.

The thing is, if you read someone's code and copy it into your application using your memory as the medium, that's legally the same as just electronically copying the file.

In other words, if you read someone's source code, you're probably bound by the license of said code. Applying their solution to your product, whether you copy it by mind or by wire, is governed by the same licenses.

I understand, BTW, that Microsoft programmers are explicitly forbidden by their management from reading the code to any GPL products to avoid this problem.

Re:It's funny...

[MSG]

He's not talking about copying the code in the sense of "cp gpl.c mynew.c", but he's copying the code all the same.

If you read code to learn from it, you're memorizing someone else's code. Applying that code to your product later may not be copying the code verbatim, but it's copying all the same. If the author of the original code can prove that you've done this (this applies to code under *any* license), then they can pursue the same legal recourse as if you'd used the original code directly.

Re:It's funny...

[MSG]

From the first section of the GPL, describing the intent of the license:
...a "work based on the Program" means either the Program or any derivative work under copyright law: that is to say, a work containing the Program or a portion of it, either verbatim or with modifications and/or translated into another
language

Reading itself is not covered per se, but application of that "algorythm/way of doing it" legally constitutes a derivative work. That means that if you read GPL code to see how a thing is done, the code you write to mimick it must be GPLed itself.

Re:It's funny...

[MSG]

does that mean ... you're bound by joe schmoe license that i invented c under??

Yes. If you design a language, and place a restrictive license on it, then the people who use *your product* are bound by your license. (Although, some people may be legally allowed to reverse engineer the software) Additionally, if your language is an open specification, and you write documentation for the language that includes example code, then you may place a license on those examples. (A *lot* of docs are written this way, unfortunately) People who read your documentation are bound by the license of the example code.

If i read the source code to... diablo 2... can blizzard bust my balls cause i figured out how they did something and put it into my game?
Yes. Yes they can. That's what copyright protects. That's why proprietary software engineers are generally required to stay the hell away from Free Software code, and Free Software engineers don't read proprietary code.

As an example, you won't find the Wine project accepting code from former Microsoft employees, because if those people violate the license used by their former employer, Wine as a whole may be judged to be in violation.

Re:VirtualDub is GPL, not LGPL

[um... Lucas]

So, for a sort of real-world, yet fictuous scenario... Here's the premise:

1 - AOL has license to embed IE in their application. (true)

2 - Mozilla/Gecko has come of age.

3 - AOL hears users asking for a Gecko based browser.

4 - Could it pe possible for AOL to either modify Mozilla in such a way that it pipes it's output to a window in AOL's interface, without them releasing the source (forgetting that they own netscape)? Would there be any way for them to do so, and give users the option of which rendering engine to use?

5 - If 4 is not possilbe, what if a 3rd party developer modified Mozilla in such a way that it operates flawlessly from within AOL. AOL hasn't violated the license. Has the developer? Probably not, if they reverse engineered AOL's interfaces. So would it still be a violation if AOL chose to distribute the 3rd party's library with their application, if they provided source to the linked program?

AGHHHHHHH! The is kind of a brain melting excersize of what ifs, etc...

I can think of MP3 encoders that can use a proprietary or GPL'ed encoder, while themselves being proprietary. The only difference between that and this case (story) is that there are not currently available and comparable DLL's like what they're distributing separately, but with their app. Would it be okay, once there was a second DLL available, so that that one DLL wouldn't be the only way for the program to "function as advertised?"

Re:VirtualDub is GPL, not LGPL

[um... Lucas]

Okay, bad example... but the thing i was aiming for was if mozilla were GPL'ed, and AOL didn't own netscape, what would the outcome of my scenario be?

That AOL can't integrate free code anyhow anyway into their product?

Will the GPL prevent an interested third party developer from extending the functionality of a proprietary program with a GPLed tool?

If those are both no, why can't AOL choose to distribute the program or library with their product, even though they're under two different lcenses? It'd seem that you'd just need to click one okay to one license for the program and another for the library.

Same should go here...

Re:Get a life

[um... Lucas]

For what your aiming at, i think you're off a little bit... For instance, what if the program that's in question was actually a promoted as a photoeditor? True, the photoediting functionality isn't implented yet, but if you include a GPLed library, it will convert digital movies to different formats... Could they get out of this whole thing by adding text editing to their app, and say "hey, it's really a text editor with additional* functionality!"

See they massive gray area?

Re:Vidomi's position

[dap24]

The following language from paragraph 2 of the GPL explicitly concludes that it is possible for a program to in some way interact with GPL code and yet the proprietary independent and separate work not have to be licensed under the GPL:

"If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works... Thus, it is not the intent of this section to claim rights or contest your rights to work written entirely by you; rather, the intent is to exercise the right to control the distribution of derivative or collective works based on the Program."

Vidomi's software MIGHT be considered seperate, depending on how closely you consider dynamic linking to be an interface. But it is most definately NOT independent... by Vidomi's own admittance, it's core functionality is dependent on the GPL'd code!
----
I'm not stoned, I just chugged a pack of fUN dIP!

Re:Vidomi's position

[dap24]

So, why is it okay to build Apache httpd linked to Linux system libraries and release it under the Apache license, but not okay to build a DVD app linked to GPL'd utility code and release it under the app's license?

Because the "Linux system libraries", glibc, is licensed under the LGPL.
----
I'm not stoned, I just chugged a pack of fUN dIP!

Re:GPL can probably be circumvented

[ethereal]

Has anyone ever mentioned to you that using a fixed-width font is very jarring, and makes it look like you're yelling (or at least suffer from voice immodulation syndrome :). You might as well just post in ALL CAPS, except of course the lameness filter won't let you.

Ow, my eyes hurt now...

Caution: contents may be quarrelsome and meticulous!

Re:Not to MLP, but...

[SmittyTheBold]

I find it quite scary I actually got that acronym. For those who didn't, "YSNHTBAL!" = you should not have to be a lawyer!

Derivative Works

[Sloppy]

The header files describing the libraries had to be used to produce the application linked (dynamically) to it and so their application forms a derivative work.

This is a horrible implication for Free Software (and non-monopoly commercial software too), as it could effectively allow someone to use IP laws to protect an interface. This must not be allowed to happen! If judges agree with you on this point, Linux dudes may win the battle but they will lose the war.

If interfacing to another piece of software causes your software to become a derived work of the other, then we're all in for a shitstorm of trouble, and this particular case will be small potatoes by comparison.

---

Re:Implications

[Todd Knarr]

Both of them. Option #1 you state is the LGPL. Option #2 is the GPL. That's why the LGPL was created, to allow option #1. In this case the author considered the LGPL and decided against using specifically to prevent option #1 from being the case.

Re:Vidomi's position

[kinkie]

> However, if you'd like to use the GPL'd code in a program of your own

Wrong. You can do that, as long as you don't distribute the resulting software to anybody, unless also your code is under the GPL.

The GPL is not about use, it's about distribution.

Re:GPL, Distribution, Intent and Spirit

[Royster]

Sure, a 'method of operation' may be a trade secret, but then you have to take measures to ensure it remains secret. Once the secret is out, trade secret protection often ends.

MS realized that they had a losing case against Slashdot and they've done nothing about it since then.

The GPL and Linking

[Royster]

Unfortunately this case, if it ever gets heard, will go up against one of the areas that the GPL is weakest -- dynamic linking.

The question of what constitutes a "derivative work" in the case of software has not ben litigated to my knowledge. The GPL gets its strength from copyright law. Copyright law reserves several rights to the copyright holder among these are the creation of "derivative works".

But Copyright law does not cover anything which is a "method of operation". It could be argued that an API exposed by a DLL or other dynamically linked object is a method of operation and not protected by copyright. Thus merely using a library routine might not be viewed as creating a derivative work.

If this were not the case, we would be in a worse situation. Then MS, for example, could claim that any piece of software using services provided by a MS DLL is a derivative work. Similarly for any platform where routines are exposed through an API and where there isn't license text specifically allowing the use.

Re:The GPL and Linking

[SpinyNorman]

I'd agree that the derivative work part is key. Vidmoi are trying to finesse the GPL by moving the GPL'd code into DLLs, but luckily that packaging reshuffle doesn't change the issue of the degree to which Vidomi's product is derivative of it.

If DLLs defeated the GPL than any GPL program could be trivially modifed to be a DLL and called from a six line commercial application, or more importantly a 50/50 mix like Vidomi's.

Re:The GPL and Linking

[QuantumG]

Vidomi is anything but a 50/50 mix, but I agree with your comment.

Re:GPL, Distribution, Intent and Spirit

[Royster]

The header files describing the libraries had to be used to produce the application linked (dynamically) to it and so their application forms a derivative work.

This is certainly the point of view in the GPL community, but it if far from clear whether the courts will endorse this view. Header files describe an API. An API is arguably a 'method of operation' in the context of copyright law and 'methods of operation' are not protected by copyright.

Is it independent?

[bluestar]

Several reasons the GPL is being violated have been explained. Here's another..

Vidomi themselves quote the GPL:

If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works

How well does their product work without the GPLed parts? If it doesn't, then clearly it's not "independent and separate".

Microsoft must be drooling

[bluestar]

What are the possible results of this?

- Vidomi is forced to release the source to the propietary bits

Microsoft FUD wins, "use GPLed software and lose your IP"

- Vidomi is forced to stop shipping their software

Microsoft FUD wins, "GPLed software is not a sustainable business model"

- Vidomi wins

The GPL is not as strong as we think and Microsoft starts stealing our source code

Re:Microsoft must be drooling

[B.D.Mills]

i hope vidomi is forced to release their software source

The more likely outcome is that Vidomi has to stop using GPL'd source and pay damages for copyright infringement.

Also, Microsoft is relevant to this lawsuit, but not in the way you think.

Look at this comment and footnote from the web site:

Yes, I can dual-license the code, but I have chosen not to in this case [link to footnote]

[Footnote] I have chosen not to because their software is, as far as I can tell, primarily meant to rip DVDs (MPEG-2 + AC3), and also installs the DivX 3.11a codec, a hacked version of Microsoft MPEG-4 V3. I'm not a lawyer, and moral issues aside, it is probably in my interest not to have my code specially licensed in a commercial package which is, with high probability, illegal.

In other words, the software distribution in question may also violate Microsoft's copyright. I would be speaking to Microsoft lawyers about a possible joint lawsuit.

--

Re:Microsoft must be drooling

[Zurk]

micro$oft FUD is irrelevant. have you noticed that the free software community can match it with the best of them ?
i hope vidomi is forced to release their software source -- that would make the GPL bulletproof with a real precendent.

roughly how the GPL would sound in court

[redd]

The GPL is actually pretty solid, so don't worry too much. We'll start with an analogy.. if I sold you some binary software under copyright, then you hex-edited or in some way modified it and gave it to all your mates, you'd be violating copyright law because I didn't give you permission to do that. The GPL is actually a concession of what you CAN do, not what you can't. It's basically a contract that says "you can copy this, but under the following conditions". If you don't obey that contract or you can in some way suggest to the court that the contract is void, then the contract is void, and you are subject to copyright law. Thus, if I write some software and license it to you under the GPL, and you violate that license, I can sue you for copyright infringement. Disclaimer: IANAQLY (q==qualified, y==yet)

Another interesting take on linking

[mattbee]

Of course people are seizing onto the GPL's slippery definition of linking-- to add to the melting pot of opinions, the MySQL folk specify that you must pay them for a special GPL exception license if You have a commercial application that ONLY works with MySQL and ships the application with the MySQL server. This is because we view this as linking even if it is done over the network. e.g. any kind of reliance == linking. I'm not sure whether this makes the case weaker or stronger against dynamic linking as GPL-covered linking.

Re:Another interesting take on linking

[magi]

You have a commercial application that ONLY works with MySQL and ships the application with the MySQL server. This is because we view this as linking even if it is done over the network.

This is getting really fuzzy. Is it a violation to use a proprietary web browser to "link to", i.e., connect through the network with a protocol interface to a GPL-licensed web server? Or a web application such as Slashdot? Wow, we have a case against MS!

Anyhow, GPL is too restrictive to be useful for libraries, which is why there is LGPL. Programs are written to be used and it shouldn't make any difference whether the user is a program or a person. Linking to shared libraries is morally no more stealing than using a GPLed web browser for trading stocks or writing a book with Emacs. That's what even GPLed programs are intended for: for personal benefit (both fun and making your living).

The idea of the GPL is to force users to give back any changes to the tool, not for using the tool. If you want to "get something back" also from the people who just use your GPLed editor for writing their business stuff, give up the GPL, make it proprietary, and charge money for it.

Also, I would believe that gcc wouldn't be as popular as it is, had the libs not been LGPLed. LGPL promotes use, while preserving the central intent of the license.

Um Something is really weird here

[IceFox]

My home page http://www.csh.rit.edu/~benjamin/ has been entirly copied unto virtuadub's site. If they has ssi on you could see all of it. They must have liked the layout of my site that quite a bit. A wierd sensation wakeing up and seeing my website on some other server.

Re:Um Something is really weird here

[IceFox]

nope, I tried with ie, netscape and even wget and I am getting my web site!

Re:Um Something is really weird here

[IceFox]

ooops my bad. Turns out some funky thing is happening to my router... Adding to my todo list to fix....

Re:Um Something is really weird here

[nicodaemos]

You must be using Internet Explorer. I noticed that when I try to access sites that don't respond (virtualdub's site was heavily loaded and wasn't responding to me earlier), IE will instead pop up a site I had previously viewed. Usually this happens to be one of my own sites.

Pretty weird if you ask me, although, I definitely prefer the current behavior over some dark plot by Bill to take me to msn.com -- that would really suck!

BTW, is your page GPL'd or LGPL'd?

Re:scenarios

[Shadowlion]

Actually, this isn't nearly as bad a scenario as it you make it out to be in the worst case.

In the worst case scenario: the GPL is ruled unenforceable. As a result, any code licensed under the GPL reverts to "mere" copyright - which means no corporation can use it _period_ without asking the author or authors for permission. In other words, use of the code will require explicit permission to use rather than the implicit permission granted by the GPL.

Essentially, in the worst-case scenario, GPL'ed software immediately reverts to closed-source software, which is more restrictive than the GPL.


--

Re:A true test of the GPL

[Shadowlion]

First of all, if you really could get around the GPL (say the Linux kernel) by writing a LGPL wrapper (glibc), then it would be pretty easy to just bypass the GPL.

How do you think you have native GNU/Linux applications that are closed-source? The glibc libraries are specifically LGPL-ed so that commercial applications may be developed and linked without requiring that they be GPL-ed themselves.

Second, there's plainly no reference to the LGPL in the text of the GPL.

I don't recall saying that there was.

Third, you haven't read Linux's COPYING file. Linux is not under the GPL -- there's a special Linus modification at the top of the file allowing anyone (even proprietary apps) to link to the kernel.

That's all well and good, and is probably the reason why (as the poster above you stated) people can link to symbols that the kernel exports (select, open, etc.).

However, Linux _is_ under the GPL; Linus, as copyright holder and ultimate licenser, can add as many exceptions onto the licensing as he wants. The GPL gives you certain rights; Linus can give you other rights, or modify the rights that the GPL gives you (but then it no longer qualifies as "true" GPL licensing). In this case, Linus chose to allow people not only the rights afforded by the GPL, but also one that wasn't: linking with non-GPL software.

--

Re:A true test of the GPL

[Shadowlion]

Every program ever written under linux would be a derivative work of the kernel, which is GPL.

*bzzzzt!*

Programs under GNU/Linux are not linked against the kernel. Programs under GNU/Linux are linked against the GNU C library, which is LGPL'ed (Library GPL). The LGPL says that you can link non-GPL'ed software against the C libraries and be OK, which allows for commercial, closed-source applications to be developed for GNU/Linux.

So even if the GPL is held to be enforceable against libraries, commercial apps are safe and can continue to co-exist because the LGPL'ed C library gives them the exception they need.

--

Re:Vidomi's position

[Anomie-ous Cow-ard]

You really don't get it, do you? Hopefully this is an accident, go read http://www.gnu.org/gnu/thegnuproject.html to gain some understanding. You don't have to agree, but you will then argue better if you know the actual goal of your opponent.

But the whole idea of FSF and GPL is to do away with copywrited software and any limitations that come with it.

The whole idea of the FSF it to make software free (libre). The GPL uses copyright law as a means to that end.

-----

The $75,000,000 question

[JabberWokky]

Vidomi can operate without these GPL DLLs can it not?

I thought not, but I don't have a windows box to play with it... I was under the impression that those flask functions in the DLL were crucial to the stated functionality of the program.

Is this true or not?

--
Evan

GPL is NOT LGPL

[Hugonz]

I believe he FSF saw this coming from the very beginning, that's why they came up with the LGPL. Vidori claims that they have released all the changes they've made to the GPLed code itself (the DLL) but those are exactly the term of the LGPL, no the bare GPL.

Re:GPL is NOT LGPL

[maw]

Uhh, on modern systems, the C library can and usually is dynamically linked with an application.

You know, that libc.so nonsense.
--

Re:GPL is NOT LGPL

[rgmoore]

This is incorrect for two reasons. One is that Linus explicitly give permission for non-GPLed programs to use system calls under Linux, so it's perfectly OK to write proprietary software that runs under Linux. More generally, this is not a problem because the GPL considers a work derived if and only if the program is normally distributed as a coherent whole. It's perfectly OK to write non-GPLed code that links to GPLed code if they are distributed separately. A good test of separate distribution is whether one can reasonably expect the GPLed code to exist on the computer already and hence not need to distribute it youself. Something basic like Qt or GTK is likely enough to be on the computer already that you don't need to distribute it with your binaries, so binaries that link against it are not considered to be derived works.

The problem in this case is that the GPLed and non-GPLed parts aren't being distributed separately, and the GPLed code is not something that one could reasonably assume to be present on the system already. The non-GPLed section is effectively crippled when you remove the GPLed stuff, which is a pretty good indication that the composite program is a derived work of the GPLed code. That means that Vidomi should have to follow it's author's wishes and obey the terms of the GPL, which they aren't doing.

Re:GPL is NOT LGPL

[Lozzer]

it's not possible to write software for profit ever again

AFAIK the GPL does not force you to give your software away for no charge. What it does force you to do is dsitribute source (or at least have the source easily obtainable) to people who have bought your code

Only people who have bought your code have the right to use it. The fact that much GPL software is also available for no charge seems to confuse people

Re:GPL is NOT LGPL

[wierdo]

such that it's not possible to write software for profit ever again (atleast not in a world taken over by Linux as the OS of choice).

You seem to fail to recognize the exception that, if the libraries are distributed with your operating system, you may link against them whether or not your application is proprietary, therefore I declare your statement to be misguided and incorrect.

Despite the above, the GPL is still indeed a viral license, but it's more like a common cold virus, rather than the Ebola virus you would make it out to be.

P.S. While I realize that the invocation of the Ebola virus may seem heartless, I submit that it is indeed a good comparison.

-Nathan

Care about freedom?

Re:GPL is NOT LGPL

[pcidevel]

IMHO: The LGPL seems to be meant to be applied to static libraries, ala the C libraries, for example. Dynamic linking is an entirely different beast, and since DLL's are *really* seperate executables, I don't see how the GPL can be applied to a program that USES a DLL with GPL code in it. Now if they had created a library (in C, a Unit in Pascal, etc) and then linked against that, their entire project would need to be open sourced and under the GPL, or else they wouldn't be complying.

Re-read the LGPL.. the LGPL applies ONLY to dynamic linking.. if you statically link the lib you have to release the entire binary under the GPL. Only if you dynamically link the lib under the LGPL can you release your producted as non derived.

I think this case is horribly unfortunate. This is the ammo M$ has been waiting on to PROVE once and for all that the GPL is a horrible viral lisence, and here we go, feeding the FUD machine. This should be dubbed Operation Foot Bullet (stolen from the Operation Clambake pages!), because there is no better way for the FSF to shoot the collective Free Software (and hence Open Source) movement in the foot than to bring this to court. Of course the worst possible outcome is that the case looses. That will still feed the FUD machines (avoid costly legal battles, use MS products) and it will provide a huge hole in the armor that is the GPL (make every GPLed program as a DLL or SO and never have to put your code back in the pool). But even winning this case isn't a win for the GPL or Free Software. If only this could be shuffled under the rug and forgotten, too bad that can never happen...

Re:GPL is NOT LGPL

[DarkEdgeX]

I think this case is horribly unfortunate. This is the ammo M$ has been waiting on to PROVE once and for all that the GPL is a horrible viral lisence, and here we go, feeding the FUD machine.

Well, if everyone (which appears to be the majority in these posts so far) agrees that what this company has done is wrong, then it would appear to be a viral license-- such that it's not possible to write software for profit ever again (atleast not in a world taken over by Linux as the OS of choice).

And the real fact here is, the GPL *IS* a viral license.. saying it's not is a lie.

Re:GPL is NOT LGPL

[DarkEdgeX]

IMHO: The LGPL seems to be meant to be applied to static libraries, ala the C libraries, for example. Dynamic linking is an entirely different beast, and since DLL's are *really* seperate executables, I don't see how the GPL can be applied to a program that USES a DLL with GPL code in it. Now if they had created a library (in C, a Unit in Pascal, etc) and then linked against that, their entire project would need to be open sourced and under the GPL, or else they wouldn't be complying.

Plus as I stated elsewhere, there's nothing stopping someone from writing their own DLL that interfaces with this product that doesn't contain GPL code. This means the product, on it's own, still works without the presence of this GPL-code driven DLL. Just as plug-ins in Photoshop can be missing and the app itself (Photoshop) would still run (just no lens flares, for example).

Did you bother to look at vidomi.com before....

[sweetooth]

posting this? http://www.vidomi.com/index.php?artid=6#news
Also for those don't want to load the page they have three interesting news bits, the first is titled "Credit where Credit is due. They acknowledge the following sites
http://hiroko.ee.ntu.edu.tw/
http://www.freeamp.org/
http://www.flaskmpeg.net
http://www.virtualdub.org/
http://www.mp3dev.org/mp3/

Then they have a link to the source at this url http://www.vidomi.com/download.php?op=mydown&did=3
The next item reports that they will no longer allow download of the GPL'd items with the non GPL'd items, and that they will shortly release a GPL kit.
Where's the problem?

Re:Did you bother to look at vidomi.com before....

[Eric Seppanen]

You ask "Where's the problem?"

The problem would seem to be that Vidomi is a video editing app, and they made it by taking a bunch of GPL code from another video editing app and adding their own stuff, and claiming that they don't have to release source for their own stuff, because they took all the GPLed stuff and wrapped it into a windows DLL.

This kind of stuff is obviously against the spirit of the GPL, and probably against the letter of the GPL as well. That's the problem.
--

Re:Did you bother to look at vidomi.com before....

[Snootch]

This kind of stuff is obviously against the spirit of the GPL, and probably against the letter of the GPL as well. That's the problem.

Yep, it is against ther letter. If you link with a GPL'ed library (in any way - be it dynamic or static), you've gotta GPL the lot.

43rd Law of Computing:

Re:Get a life

[Phill Hugo]

Except that the header files from the shared library are used in the main application making it a derivative and quite subject to GPL.

I think a court would always side with the intent of the developer, especially now they can clearly see that Vidomi has no intent to comply with the GPL or the author's wishes and if they don't protect those, then perhaps all software will become less protected by law.

Re:No harm intended/done?

[Phill Hugo]

You are wrong. Linux applications tend to use the GNU C Libraries which is LGPL - specifically allows closed apps to link against it (even dynamic linking require uses of the header files which are used to form the main application and therefore you have a derivative work).

The Java argument is why Free javalike systems (like kaffe and gnu classpath) were produced, so people can use Java and know that their application can run on a free system.

However, Sun do no require that Java application linking against their classes be subject to their licenses. The GPL does require that.

You cannot link a closed source application to a GPL library becuase it becomes a derivative work, there are parts of the library in the main app even if the bulk is loaded at runtime.

Re:At least it's a Small(er) company

[Phill Hugo]

Of course it is a company like SloMedia. Sony and Cisco's lawyers would no doubt run a mile from the case knowing the likely outcome.

You only have to find out why GCC has a GPL'd Objective C compiler front end to see that larger companies with Real Lawyers tend to consider the GPL clear enough to make them unfavourable in court.

Remember that Vidomi/SloMedia are not asking for the court case, Avery is. They will probably back down if they have any sense, if they do it now, the demands for their application under GPL may die down but I'd be pushing to get it to court just to prove the point.

Its a pretty clear cut case and most courts will side with the intent of the author, especially now Vidomi has published their intent not to play fair on their own web site.

Re:GPL, Distribution, Intent and Spirit

[Phill Hugo]

They've already arguably agreed to the GPL and their users have the right to demand the source for their application under the license propegation.

In the event of court, I think three outcomes are possible...

1) They are deemed not in violotion of the agreement and the GPL as it stands has no safe guard against this (this is UNLIKELY people, judges are looking for FAIR findings and the intent of the GPL here is clear enough. The FSF has compiler and linker programmers to explain this to the judges, they have clear industry set precidents that linking against GPL libs is not permitted and they have a case that most judges will genuinely find interesting to understand.

If you need a precident, courts more often defend the right of the author than the recipient, you've seen that before!

2) They are deemed in violation of a binding agreement and as such must honour it. GPL source for their application to customers.

3) They are deemed to not be bound to the agreement and as such are forbidden from using the GPL code in future unless they properly accept the GPL.

The latter is probably the most likely (the courts will probably side with your view that their code should be protected) and I'd probably give this judgement myself. Both parties walk away relatively happy and neither is massively harmed as a result.

I'd guess a token fine to be paid to the FSF and Avery would be ordered to set a clear winner precident, not enough to warrent appeal but enough to indicate the court is happy with our own precidents.

But that's all assuming Vidomi are stupid enough to do this. They aren't and the GPL will probably never get to court. As soon as you put a decent lawyer in front of it and explain the technology, they just arn't confident of anything other than it being upheld. That's why its never been to court before and a handful of companies have considered it before involving their lawyers.

There's a rumour its because Eben Moglen kicks ass and no one wants to take him on, but I'm not sure if its true ;)

Re:A true test of the GPL

[Phill Hugo]

Yes, Brian allowed XFree to include it under their license but it was originally LGPL.

Re:A true test of the GPL

[Phill Hugo]

All this happened AFTER they GPLd it. NeXT spoke to the FSF and Stallman has this to say about what happened...

NeXT initially wanted to make this front end proprietary; they proposed to release it as .o files, and let users link them with the rest of GCC, thinking this might be a way around the GPL's requirements. But our lawyer said that this would
not evade the requirements, that it was not allowed. And so they made the Objective C front end free software.

Notice that it is the FSF lawyers who said "No" and NeXT accepted this as fair play. This is a precident that a court will consider.

Full text is at...

http://lwn.net/1998/0409/pragmatic.html

GPL, Distribution, Intent and Spirit

[Phill Hugo]

It looks like Vidomi are trying to remove themselves from this while still allowing the code to be used with their closed source product. By decoupling the GPLd code into dynamically shared libraries and distribtuting them seperately, the GPL violation is less obvious but still quite present.

The header files describing the libraries had to be used to produce the application linked (dynamically) to it and so their application forms a derivative work. This is why GPL shared libraries are not linked against elsewhere and is why the LGPL was produced.

Even if Vidomi produced an middle man DLL, that middle man DLL would be subject to the GPL for the same reason and anything linked against it would also be. However, it would be interesting as it would make Vidomi responsible for prosecuting any violation of their proxy DLL. I still think most courts would not favour their case after the situation was clearly explained.

It is really a true shame that Vidomi have put their own interest in front of the Avery's, it shows them in a very bad light, although it would be interesting to see the GPL subject to court-time but I think the outcome is predictable.

I can only suggest they save their time in court, GPL their application and get on with writing applications that people want to use, that is what its all about, right Vidomi?

Re:GPL, Distribution, Intent and Spirit

[csbruce]

I can only suggest they save their time in court, GPL their application and get on with writing applications that people want to use

I'm not sure of the exact type of this company, but for a software company, it's not generally realistic to expect them to give away their crown jewels for free. A more realistic demand would be that they stop using the GPLed code and produce, find, buy, or rent non-GPLed code with the equivalent functionality.

Re:GPL, Distribution, Intent and Spirit

[csbruce]

If a court ruled that the company accepted the terms of the GPL and must release its source code, this outcome would be an unparalleled disaster for the GPL. The Microsoft FUD machine would kick into overdrive and cite the case as proof that to have any involvement with "Open Source" software of any kind is to surrender all of your intellectual-property rights, and the PHBs would be scared senseless.

The best outcome would be an enforceable "cease and desist" with the copyright owner being kind enough not to ask for damages.

Re:GPL, Distribution, Intent and Spirit

[QuantumG]

Avery doesn't want to relicense his code because Vidomi is a dodgy program for ripping DVDs and encoding them with the illegal divx codec stolen from Microsoft. Avery has gone to a lot of trouble to disassociate VirtualDub with "movie pirates" and now Vidomi threatens to bring public perception back to that yet again.

Re:GPL, Distribution, Intent and Spirit

[Amokscience]

As an author who has released GPL'ed software I'd rather see this go to court and finally have some rulings on the GPL. It's one thing to be a company with money to spend of legal fees, it's quite another to be an author making no money off of a product. I'd like to finally see the GPL get some legal rulings to give it more weight.

Re:GPL, Distribution, Intent and Spirit

[naasking]

I think an acceptable compromise/settlement would be if the company paid Avery to distribute the source under a more compatible license(or dual license it), ie. change it from GPL to LGPL so they could distribute it the way they are doing now.

Avery gets some kind of compensation for writing good software and making it free, so it's fair to him, the software remains free and open which is very important, the company avoids legal problems and so saves money and face, and they look good for finding a fair solution by supporting an open source developer and for not trying to sneak around the GPL(though it's kinda too late, they can still get back in good graces).

This is similar to the situation where Mandrake hired the developer behind Bochs and released the source under the GPL, and is now paying him to develop plex86 also under the GPL.

-----
"Goose... Geese... Moose... MOOSE!?!?!"

Re:GPL, Distribution, Intent and Spirit

[Mark Imbriaco]

No, that's not what it's all about. Vidomi is a company. At the end of the day, it's all about making money.

Re:GPL, Distribution, Intent and Spirit

[kinnunen]

It looks like Vidomi are trying to remove themselves from this while still allowing the code to be used with their closed source product. -- The header files describing the libraries had to be used to produce the application linked (dynamically) to it and so their application forms a derivative work. This is why GPL shared libraries are not linked against elsewhere and is why the LGPL was produced.

But isn't this exactly how binary-only (=closed source) kernel modules are created?

--

Re:GPL, Distribution, Intent and Spirit

[Ayende Rahien]

Static lib linking happen at compiling time, by the developer.
Dynamic linking happens on the user's computer.

Re:GPL, Distribution, Intent and Spirit

[slashdot.org]

The header files describing the libraries had to be used to produce the application linked (dynamically) to it and so their application forms a derivative work.

That's an interesting comment.

Let's say I created a header file with functions I would like to see in a DLL. I released it copyleft. Then I took some GPL software and implemented those functions. Obviously those changes now fall under GPL. But what do I care? The header file remains copyleft.

The program now merely uses the DLL. Nowhere is GPL source directly included.

An other way could be to have someone create a DLL (and release GPL source) and reverse-engineer the interfaces in the DLL. Again, no need for a GPL'd header file.

But the question is, how is a DLL different from a static LIB? Everyone seems to understand that a static LIB would obviously violate the GPL (when used by closed source software). What makes a DLL different?

Re:GPL, Distribution, Intent and Spirit

[fsckingdork]

Oracle runs on Linux. That means they must have linked in headers to GPL'd code, which they are therefore loading dynamically.

Now although it's obvious that Vidomi's only reason for their dynamic linking architecture is to avoid having to GPL their code, there is no language in the GPL to explicate the difference between Oracle and Vidomi.

You might say, "It's obvious that Vidomi's dynamic linking architecture exists solely to avoid having to GPL their code," and I think you'd be correct, but that doesn't change the fact that the GPL, as it is written, is no less fulfilled by Vidomi than it is by Oracle.

I'll bet the FSF doesn't even take this case to court, let alone win.

The GPL will not withstand legal scrutiny until it includes explicit language about what forms of software/process interaction are considered to make a work derivative.

- fd

IANAL, TINLA, YMMV, HAND

Re:A true test of the GPL

[Phill Hugo]

> whether dynamic linking to a library makes something a derivative work in the copyright sense.

I think it does since the libraries header files are used in the subsequent work. The courts will always take into account NON COURT precedent in the absence of previous similar findings and there they would run into the NeXT Object C compiler based on GCC (NeXT backed down, its GPL now!), the QT problem, the absence of linking to GPL libs on Linux/UNIX systems, the intent of the LGPL to specifically permit this (this alone should tell you the intent of the GPL is NOT to permit it)

You statement about making 3rd party developers beholded to OS vendors is wrong. They already ARE beholden to them but most sucessful OS's permit free linking to their application without license propegation to the result. The LGPL allows this too and is why glibc and mesa are LGPLd.

The interface between kernel and application is via an LGPL application and Linus specifically permits binary only drivers in Linux (basically making the header files required to use them effecticely LGPL)

You are incorrect on a lot of counts here and you're not alone.

Re:Vidomi's position

[DeathBunny]

>You know, I agree with what you're saying, but >advocacy of the GPL doesn't help if you look >like a raving lunatic as you do it.

>Some hints:
>* Only one exclamation point per sentence >is >ever necessary.

Unless your a porn site! :-)

>* It's 'copyright', not 'copywrite'. Think about >it a second. Copyright deals with rights to >copy, not "writes".
>* It's 'their', not 'thier'. i-before-e rules >don't work, so you should ignore anyone who >tries to teach you one.

I'm at work and do have other things to do, so I'm typing fast without benefit of a spell checker or time to proofread. I've already spent more time on this thread than I should.

So sue me. :-)

Re:Vidomi's position

[DeathBunny]

>Which proves again,GPL is completely unusable in >commerical settings.
>If you want real and usefull code exchange >between profesionals with no social or political >strings attached go for BSD style license.

The GPL has never claimed to be without strings. I don't understand what's so complicated about that. Anytime you use ANY copywrited code (that includes GPL and BSD code) you MUST obey the conditions that code was licensed to you under!!!

The only difference is what conditions the original copywrite holder is asking for. Admittedly, in the case of the BSD license they really aren't asking for anything.

As for the GPL being usable or unusable in commercial settings, it still depends on what you intend to use it *for*. I use lots of GPL'ed code in my job. It works fine for me because my employer isn't planning to re-distribute that code under a more proprietary license. If they *did* wish to release their code under a more proprietary license, then yes, BSD licensed code would then fit our requirements better.

The point is, the original copywrite holder *always* has the right to set the terms on how you can use thier copywrited material. If you don't like thier terms, don't use thier code!!!! Write your own, or find some code that is under a license you *do* like!

Re:Vidomi's position

[DeathBunny]

>Nope, you got it wrong. It is GPL proponents who >are outraged at corporation for their EULAs and >yet they are using the very same methods.At >least commercial entities do not claim to do >have a higher goal like FSF does.
>Who is a bigger hypocrite here ?

Vidomi. Vidomi and the FSF both have the legal right to enforce conditions on the use and redistribution of thier copywrited material.

The FSF doesn't agree with the conditions that proprietary software companies put on thier software, but I haven't seen them try to violate those licenses by, for example, using proprietary code in a free software program.

Vidomi expects us to obey thier license conditions, while at the same time they are attempting to violate and subvert the license conditions of VirtuDub. They seem to believe that thier license matters, other peoples licenses don't.

It doesn't take a brain surgeon to see the hypocracy here.

Re:Vidomi's position

[DeathBunny]

>Agreed. The problem here is that FSF claims to >be on higher moral ground. They do pride >themselves in vilifying other licenses and >sometimes outright accusing others of limiting >freedom ( a very serious charge.)

Where is the problem here? The FSF does believe that proprietary licenses are depriving people of important freedoms. Given that belief, it would be bizarre if they *didn't* accuse proprietary software companies of limiting other peoples freedoms!!

Regardless of whether you agree with Stallman and the FSF or not, I think they've been remarkably consistant and logical in support of thier beliefs. They've talked the talk, and they've walked the walk. They don't just talk about software freedom, they've written hundreds of thousands of lines of code and release them under a license that fits *thier* beliefs in software freedom.

Re:Vidomi's position

[DeathBunny]

>There's also a possible: >4) Pay Avery Lee to license his work to SloMedia >(assuming he holds sole copyright). That is a really good point that I forgot. Thanks!

Re:Vidomi's position

[DeathBunny]

>It might seem tired and overused argument but >GPL really does force people into GPL

That's just pure stupidity. No one is forcing you to use GPL code. As I've said before, anytime you use ANY copywrited material (outside of fair use) in a derived work, you have agreed to abide by the license given to you by the original copywrite holder. If you don't like that license, don't use THAT code.

There are three easy ways Vidomi could have avoided this whole issue:
1) Use VirtuaDub and release thier software under a GPL compatible license.
2) Don't use VirtuaDub, write thier own code.
3) Look for and use code whose license *is* compatible to Vidomi's license, use that.

Seems simple enough.

>license that is not much different than "evil" >they are trying to fight.

That's ridiculous. The GPL prevents you from using GPL code in a proprietary application. That's not evil, that's *the* primary feature of the GPL.

If you don't *want* that feature, don't license your code under the GPL.

And as always, whatever license you use will have to be compatible to the licenses on any other code you have incorporated into your program. That's not specific to the GPL, that applies to everything that isn't Public Domain!

Re:Vidomi's position

[Fluffy the Cat]

It might seem tired and overused argument but GPL really does force people into GPL version of what software freedom is about.

No. Yes, the GPL embodies the concept of "Free" as used by GNU. Why is this a problem? Providing you agree with this definition and see it as a good thing, the GPL is a good choice of license. If you disagree with it, then you're not obliged to use the GPL. The authors of the GPL decided that preventing people from being able to release code released under the GPL in propriatory products was a lesser evil than allowing them to do the same thing, and as a result you can't. I agree with them, and by implication so does everyone that releases software under the GPL. I'm not being forced to do so, and you're not being forced to use my code - nobody is being forced to accept the GNU version of freedom.

Re:Vidomi's position

[Fluffy the Cat]

But the whole idea of FSF and GPL is to do away with copywrited software and any limitations that come with it. Yet you are stubbornly defending GPL using laws this license was designed to destroy.

It's called pragmatism. Idealistically releasing code into the public domain would just result in corporations folding it into propriatory software. Releasing it under the GPL forces them to redistribute the source to the software and encourages them to view this as a viable practice. Over time, people's attitudes are changed.

Re:Vidomi's position

[Fluffy the Cat]

Right. There's reluctance. It's unproven. It's new. But there's awareness. Without the GPL, we wouldn't have had that to the same degree. People would happily take free code and make propriatory apps out of it. The idea is not to get free code used everywhere, the idea is to get free code used everywhere while keeping it free. If that's not what you want, then fine - don't use the GPL. It's not being forced upon you.

EXE versus DLL.

[jonathanclark]

So what's the difference between linking with a DLL and forking a processes that is GPLed - you can achieve the same levels of functionality. In Windows, the DLL and EXE file format are exactly the same. So renaming a .EXE to .DLL would be a violation of GPL? Taking this argument to it's logical extreme, any program that can create a pipe to another process could be in violation of GPL.

Re:EXE versus DLL.

[jonathanclark]

Ohh.. I can't resist a troll...


executables have stripped relocation tables, you can only load them to address 0x400000

That's not true. Most exe's have relocation information. You can strip it out, but it's pretty rare that EXEs have this information removed. Even if they did it just means the process that loads them needs to be loaded at a different base address. There is nothing special about 0x400000, many exes have a base address other than that. Notepad.exe uses 0x1000000 for example (under w2k). Use the program "rebase" that comes with W2k to change the base address to whatever you like. You can load an EXE without relocation information as DLL. Also DLLs don't have to have relocation information assuming they are not going to conflict.

even if the relocation table would be present, the OS doesn't permit loading & dynamically linking against an executable. It was originally possible, but the changed because it can cause security problem.

Try it yourself :

HMODULE m=LoadLibrary("c:\\winnt\\system32\\notepad.exe");

Amazing!! :)

Re:EXE versus DLL.

[tlk nnr]

So what's the difference between linking with a DLL and forking a processes that is GPLed - you can achieve the same levels of functionality.. In Windows, the DLL and EXE file format are ? exactly the same.

Could you try that before posting?

• executables have stripped relocation tables, you can only load them to address 0x400000
• even if the relocation table would be present, the OS doesn't permit loading & dynamically linking against an executable. It was originally possible, but the changed because it can cause security problem.

Obviously you could create a second process and isolate everything with appropriate marshalling. Embed it into a COM server that uses it's own process space, or use stdin/stdout for proprietary marshalling.
But you must DO THAT.
If rip chapters from someone elses book you can't claim that you would have been able to write it youself.
Or at least I doubt that this defense be successful in court.
SloMedia could have avoided the problems by creating a COM [the object model behind Visual Basic and OLE] server, and their app would use that server.
But it seems they didn't do it.

Re:Automatic modification of GPL'd code possible?

[Dr. Manhattan]

Would it be technically possible to write a program that takes source code as input and heavily modifies it [...] so that the result has no similarity with the original?

You mean like this or this or this or this?

All of these are designed so that people can release source code to compile elswhere without revealing "intellectual property", but there are other ways. NVidia ran their "open-source" drivers through the C preprocessor before releasing them, making them nearly useless for development.

Of course, simplistically renaming variables won't change the binary code output much. You need to at least rearrange variables and code, and maybe determine independent statements and reorder them, etc. Perhaps pad arrays with extra space on the end, and so forth.

Re:Inquiry?

[scruffy]

This is a good point. In terms of functionality, there is no difference between a library call and a system call. They both get the job done. Libraries calls are probably more efficient, but even there the difference might not be a lot. If you have to make a lot of calls, you can simply spawn one process and sysin/sysout one line for each call.

So how about this? I turn the GPLed library into a standalone program. In my non-GPLed program, I spawn a process running the library. To make a call, I send one line to the library program and get one line back as the answer. Ok or not?

Almost...

[schon]

link closed source code with a GPL'ed DLL, that's legal.

It's legal depending on the situation

Suppose you write an image viewer library, and released it under GPL... (remember, this is a library, not an app)..

Then someone comes along and takes your library, makes a windows wrapper around it - their program does nothing except provide a user-level interface to your program.. then they release the whole shebang as "Superviewer", selling it for $50.00.. how would you feel?

According to the spirit (and letter) of the GPL, they are in violation.

Re:US Copyright Law 102b

[rking]

It isn't clear to me that you understand the position at all. Vidori don't need to reverse engineer the API to the library; Vidori created the API, they wrote the libary. Specifically, they created it out of the program Virtualdub, for the purpose of using its functionality within their closed source program.

You may feel that the resulting program and library do not between them constitute a derivative work, but reverse engineering really doesn't come into it. Personally I don't think there's any doubt that there is at least a real issue for a court to decide on here.

Re:OT: Slashdot Submission process...

[Covener]

With elegant prose like that how could you have been rejected so much!

Re:VirtualDub and IP

[SpinyNorman]

I thought patents apply to implementations, not ideas? That's why something like the MP3 or MPEG-2 patents are so tricky (impossible?) to get around - because they're so general as to include any possible *way* of implementing the ideas they want to protect.

BTW, who's Vergil Bushnell? No disrespect intended, but should I have heard of him?

Re:GPL v. LGPL

[SpinyNorman]

I don't know the licensing specifics of the kernel/glibc interface, but given that it's a GNU library I'm sure it's been blessed by Stallman as being squeaky clean!

Re:GPL v. LGPL

[SpinyNorman]

The LGPL is the correct choice only if you want to allow your code to be used in commercial programs (such as Ogg Vorbis have chosen), which Avery Lee doesn't. The GPL was the right choice for him, and hopefully the courts will recognise the difference between the GPL and LGPL.

Re:VirtualDub and IP

[SpinyNorman]

Actually, if he had the resources to fight Microsoft he might have won. I'm not sure what law prevents you from reverse engineering a file format!

That's the trouble with the law, and probably what Vidomi were relying on - that the guy with more money often wins regardless of the merits. I guess they wern't expecting the FSF themselves to get involved.

Re:VirtualDub and IP

[SpinyNorman]

FYI that first "UTICA online" link you gave has a "myths" section that claims a lot of Stallman's concerns are invalid.

Among other things it says that UTITA wouldn't change existing reverse engineering law, and says that free software shouldn't be concerned about shrinkwrap licences beacuse a) it's OK to have a non-shrinkwrap licence/warrantee, and b) it's OK to have no warantee at all, like most free software.

I don't know the truth of any of this - just pointing out what the web site says.

Re:A true test of the GPL

[Platinum Dragon]

When I create a binary and use system calls, I am dynamically linking to the kernel in exactly the same sense as I am linking to a dynamic library when I make a call to a function in that library.

From the Linux 2.4 COPYING file:

NOTE! This copyright does *not* cover user programs that use kernel
services by normal system calls - this is merely considered normal use
of the kernel, and does *not* fall under the heading of "derived work".

Just dropping in some useful reference info...

Re:Vidomi's position

[Platinum Dragon]

this merely says that if my program uses/extends gpl code, and also has legitimate non-gpl functionality (menu bar/tool bar/quit button don't qualify as legitimate functionality), then i don't have to release the code of my non gpl using program. i do however HAVE to release the sections of my code that use the gpl code.

Wrongo, boyo.

If you use GPL'd code in your application, and that code gets compiled into a binary or group of binaries that rely on each other to work, the entire code of that app must be GPL'd. This is the "viral" clause that the more virulent anti-GPLers point to as a reason not to use it.

Let's say you have five source files going into a single executable. Two of those files include chunks of GPL'd code from, say, parted (just a shot in the dark). Since identifiable sections of the work are derived from GPL code, the derived work must be distributed under the GPL.

The LGPL might let you get away with that little trick, but not the GPL. The FSF recognized this issue and came up with the LGPL as a less-restrictive alternative.
.

Re:GPL code is proprietary

[Zurk]

so you dont mind anyone taking your code without giving you anything for it ? sorry..not all of us share your opinion and since we wrote the code we license, we have a right to stop a company from doing exactly that... just as you have a right to do anything you like with the code you wrote... even giving it to M$ for use in their products..

A Question of Intent

[ttfkam]

On the contrary, Vidomi's product page advertises the abilities of their software according to the abilities of the GPLed library, not their scaled down, proprietary library.

This is another reason the author is so upset. It's fairly obvious that they made a bare-bones, barely functional implementation of the library for the sole purpose of defrauding the original author's licensing terms (GPL).

This is not the same as a clear cut interface vs. implementation argument. Let's say that someone writes a piece of software that implements a filesystem according to a particular API and this filesystem is under the GPL. We'll call this hypothetical filesystem, XFS. ;) Someone else writes an OS that implements this API. The OS author is not yet in violation as they technically haven't touched any of the GPLed code.

Now let's imagine that the author of the OS writes a barely functional filesystem (like FAT) and offers this as a "default" implementation for the OS. And let's say that the OS author advertises their OS as having journaling, high speed, great reliability, etc. which of course it only has when linked against the XFS code.

That company should not be allowed to simply throw up its hands and say, "Well it really is just an option," when it describes a key piece of functionality for which they advertise. It is quite obvious from their actions that it is not simply an option, it is a fundamental portion of their products abilities.

Back to Vidomi, take away the GPLed code, it does less. THAT makes it a intrinsic part of the overall codebase.

It is NOT the same as interface vs. implementation unless you have an object which only implements half of the interface. If the interface is legit, the half-baked implementation breaks the code! the only way to avoid having the program dump core is to *expect* certain functionality to be missing. This is functionally equivalent to them expecting the GPLed code to be linked, not a coincidence that they happen to implement the same interface.

If Vidomi wrote a fully functional client (a full implementation of the interface) and the GPLed code really was a pluggable option, I might have a little more sympathy for them. Then again, if they had their own library, they wouldn't need the GPLed library, right? The only reason for them to co-op the GPLed code is to get around having to write the complete code within the company. THAT is a violation of the GPL and should be pursued with all vigor.

GPLed headers aren't in the main program

[ttfkam]

They wrapped (thinly) to make the DLL code. The author of VirtualDub noted that the DLL just had a couple of new entry points and function name changes, but the headers from VirtualDub most certainly didn't make it into the main code from Vidomi.

Thus the loophole that they are trying to exploit.

I posted elsewhere why this should not be seen as a simple interface vs. implementation issue however. They clearly violate the license. They violate arguably in the letter and definitely in the spirit.

They either need to own up to it, or get the spanking they so richly deserve.

Expectation of abilities

[ttfkam]

On their main page, they state the abilities of the program. Now, after all of the hubbub, they have an asterix next to a prominent feature which references a footnote stating that the feature requires GPL code.

It isn't that much of a stretch to state that they wrote their program with the intent that a key feature (isn't it? it's mentioned on their main page to intice potential customers) of their program would *require* GPL code. If you were working with motion video clips, wouldn't the "advanced, automated post-processing" be core to your product.

They are trying to take advantage of the codebase (good) without giving anything of substance back (bad). This is precisely what the GPL is meant to prevent. If they just wanted to swap libraries, they should've found something with a different license (LGPL, BSD, W3C) or written it themselves.

What they did instead was write a less-than-minimum implementation so they had something to point to when someone came snooping around.

I agree with the author of VirtualDub. I have no problems with individuals at the company, but the company has definitely done something wrong.

Re:Oh look! More /. GPL alarmism

[ttfkam]

What if Microsoft/Oracle/Sun/Whomever said, "And we're giving credit where credit is due. For efficient routines in the upcoming Windows kernel, special thanks go to the work of Linus Torvalds, Alan Cox, and many other accomplished developers. We would also like to thank Richard Stallman for the excellent command-line utilities to be bundled with the new OS."

And what if the community response was, "So are you releasing the affected code under GPL as well?"

And what if the response back from the company was, "Why? We've acknowledged that it is GPLed and given credit. That should be enough."

No. It's not enough.

I seriously doubt they are bad people. I wouldn't be surprised if they operate under the best of intentions. This, however, is not the point. If someone takes my wallet, thinking it belongs to them, this is not malice. If they do not agree to give it back, then we have a problem. They have acknowledged that it isn't their wallet, but they still want the money in it. Yes! I have a problem with that!

Re:A true test of the GPL

[ttfkam]

From the COPYING file in the Linux kernel source.

NOTE! This copyright does *not* cover user programs that use kernel services by normal system calls - this is merely considered normal use of the kernel, and does *not* fall under the heading of "derived work". Also note that the GPL below is copyrighted by the Free Software Foundation, but the instance of code that it refers to (the Linux kernel) is copyrighted by me and others who actually wrote it.

Linus Torvalds

Programs are linked to the kernel, but are not bound by the GPL. This is due to the above paragraph, not through some loophole or oversight in the GPL. RMS is still right.

Re:It's funny...

[ADRA]

Are you saying that if I read the code for mpeg2dec, then write my own MPEG-2 decoder, then I must release that under the GPL, merely because I viewed some GPL source?

Would you think it ethical if someone had two computers and typed the linux source code line for line on the other and then proprietarily sell the OS as a seperate closed source product? No.

When one is trying to reverse engineer a product legitimately, or create a competing product, don't look at the source or any other NDA information. That is a simple rule of thumb. Live with it and avoid heafty legal bills down the road.

By taking a position of superiority you show how nearsighted you are. Thus Spake ADRA

No Credit for Nuke Either

[hndrcks]

Their website is driven by php-Nuke, a GPL'ed weblog, and all credit has been stripped on that as well... except for the page source:

META NAME="AUTHOR" CONTENT="Vidomi"
META NAME="COPYRIGHT" CONTENT="Copyright (c) 2001 by Vidomi"
META NAME="DESCRIPTION" CONTENT="PC Video"
META NAME="GENERATOR" CONTENT="PHP-Nuke - http://phpnuke.org"

MOD PARENT UP.

[Malcontent]

The first intelligent comment on this thread so far.

Re:Inquiry?

[firewood]

Perl and gzip, on the other hand, are separate programs onto themselves, are called by your code at the command line level, so they run in separate memory spaces.

Not true on MMU-less systems (uclinux, etc.) where everything runs in the same memory space. Even on uniprocessor machines with an MMU, they all usually run in the same physical memory.

Is there really any difference between statically linked libraries and seperate applications on a ROMed embedded MMU-less system?

Re:VirtualDub and IP

[prizog]

No. It was a *patent*. Learn something about the law before contradicting Vergil Bushnell. Patents work like this: One person invents something, and files a patent on it. If the patent is granted, nobody can use that invention without paying her. Even if the reverse engineer it, or independantly invent it.

Re:Inquiry?

[throx]

I'm pretty sure your argument about memory spaces is false. If this were the case, the simple conversion from GPL to LGPL would be to write a GPL'd front end which takes parameters on stdin and sends the results to stdout.

I believe you are thinking too technically about the law. The real test is the intent (I think - IANAL).

I believe the case will be very interesting because if Vidomi actually win then there is a significant hole in the GPL which effectively reduces it to LGPL. That would cause hellish problems with "Free Software" everywhere.

GPL fails when dealing with components?

[throx]

God help us all if this is the case. If RMS believes that separating the processes is enough then in the Win32 world of COM/DCOM things become very complex.

I could start with a simple COM DLL that I place under the GPL (not the LGPL). Now four different proprietary programs invoke this DLL:

i) A program that loads the DLL inproc (Using CLS_INPROC_SERVER).
ii) A program that loads the DLL outproc (Using CLS_LOCAL_SERVER), forcing it to be run in a different process but loaded by rundll32.exe (or possibly svchost.exe)
iii) A program that loads the DLL on a remote machine directly through DCOM.
iv) A program that loads the DLL on a remote machine indirectly through MTS and COM+ (hence the Microsoft middleware is actually loading and executing the methods of my DLL).

In the four cases, (i) above is a violation according to the "same memory space" theory, but all four processes don't actually know or care (thanks to the magic of COM) where the DLL is being run and if it is inproc or not.

Hell!!

Assume for the sake of argument that my DLL has a very application specific interface that would not be easily rewritten without access to the original work in the first place.

As far as I can see in this scenario, the GPL is not specific enough to say what the licensing issues are and how they apply. I don't think RMS took into account the complexities of component software when drafting GPL 2.0, and I'm beginning to think GPL 3.0 is going to be such a monstrosity it will be years before anyone understands it fully.

Inquiry?

[EasyTarget]

What, when it comes to it, is the difference between a Library, and a complete executable called from within another application. Does the GPL make a clear distinction?

I ask since the product I do support for uses perl and gzip as part of it operation. We ship -unmodified- copies of these in binary form and supply the source (simply a copy of the relevent release's source from the Gnu sites) on demand. We call gzip from within some of our code to compress data 'on the fly', and we use perl everywhere, from install scripts through to cron jobs, and a whole bunch of perl utilities we ship that complement the primary product.

My understanding (both from our legal bods, and from stuff I have seen in slashdot discussions) is this is quite legal, and the GPL actually is written to accomodate this sort of use (we use the GNU stuff in it's entirity, without extending it's functionality).

But I fail to see a real difference between calling these utilities as standalone executable, and calling something within a library, surely the net effect is the same?


EZ

Re:Inquiry?

[etrnl]

What you're doing is very different.

You're using exising "applications" en masse. You're not taking the code for perl and the code for gzip, embedding them in your app and selling it. (If you are, you can't do that.)

The GPL has nothing to do with how applications use other applications. It has to do with how you take someone's code and use it in your own.

(Ie, it's not about USE but DISTRIBUTION.)

--etrnl--

Re:Inquiry?

[Courier]

The differences is that perl is a programming language and it's like other language you are allowed to use it to create as you wish.

On the other hand using gzip isn't a problem either. Because you can easily switch to a different program to do the same task or write you own and it would fit in just fine.

The problem as said so many time here is that the "core" fuctionility of this "product" which they get money from people for is from GPLed code.

Re:Inquiry?

[rossjudson]

You know, I thought the same thing initially. But here's an interesting way to think about it: You didn't write a commercial perl interpreter that calls perl to do its interpretation. You aren't writing a gzip program that calls gzip. You are creating substantial, differing functionality; the GPL expressly allows you to call them when you're doing that. I should think that it applies for shells, dynamic, or static linking.

It seems to me that Vidomi has released a video editor that calls VirtualDub, a GPL video editor, to do its work.

It shouldn't stand; it looks like a fairly clear violation to me.

Re:Inquiry?

[NoNeeeed]

This argument could be taken even further. What if I produce a program using GPL code, and GPL it. This then exposes a CORBA interface which other programs can call. Are the IDLs deemed to be GPLed?

Also what if I take a library, wrap it up in a program (GPL it) and use IPC of some form to make the calls to the functionality in the GPLed library. If you created a library that was responsible for sending the calls down the pipes/sockects/queues etc but which appeared as a series of function calls to the user program, should that library be GPLed, it uses no GPL code, merly the protocol used by the GPLed server component (and what if the protocol was licensed before the software was written using a none GPL license and therefore was not a derrived work).

Basically we will see this becomming more and more of a problem as software is increasingly being built out of distributed runtime components (and there is nothing to say you can't use CORBA on a single machine). What is to stop me wrapping libraries in runtime object (programs, activex components, Java RMI objects etc) and then using IPC to access their functionality (distributed or otherwise). If you define the interface/protocol (CORBA IDL, XML DTD) prior to developing the system and release it under no license, and then develope the system around this interface then it can't be derived work.

I'm not saying that this isn't against the GPL, but things start to get pretty messy at this stage.

Re:Inquiry?

[Ayende Rahien]

What about COM components?
They can be run on seperate memory spaces.

Re:Inquiry?

[tb3]

The end result may be the same, but the distinction is the usage. A library (more specifically, a DLL) runs in the same memory space as the calling program and has the same net effect as if it was compiled as part of the program. This is like taking someone else's source code, compiling it into your program, and not releasing the source.

Perl and gzip, on the other hand, are separate programs onto themselves, are called by your code at the command line level, so they run in separate memory spaces.
-----------------

Re:Vidomi's position

[wct]

More clearly, if you actually read the GPL that comes with Linux it has a clause explicitly permitting linking of non-GPL code with the system, which is considered as "normal use".

Re:Get a life

[Pedersen]

Please re-read the GPL. You see, if you use my GPL'ed code in your program, your entire program must then be licensed under the GPL. Kind of a neat bit of legal skullduggery, I think.

Re:GPL v. LGPL

[hattig]

Why couldn't they just make an LGPL library that wrapped the GPL library and link to that instead? How does glibc get to be LGPL, surely it has hooks into the GPLd kernel, which should really make it GPL, right?

The kernel has its own LGPL (Linus' GPL :) ), in that he allows lesser licenses to use its hooks.

You can't make an LGPL wrapper library around a GPL library unless the copyright holder permits it.

Waiting for the Slashdot clock to let me post...

Re:Dangerous argument

[hattig]

To win this, the FSF side has to show that a program that uses an open-source DLL is a "derivative work" based on the DLL.

Nope, the FSF has to prove that the copyright holder did not intend to have his code wrapped up in a library and used by a closed source program, and hence he chose the GPL.

Anyway, maybe all Windows programs that link to kernel32.dll are windows (or kernel32.dll) derivatives - however Windows is not GPL, and this is an entirely different matter! Actually, this stands to reason when you think about it - it is like extending a class in Java. Maybe.

Hot Brain

How do you classify GPL plugins then?

[hattig]

If you had a closed source application, with a plug-in architecture, and somebody made a GPL plugin, then how does the GPL view this? The closed-source app is now making use of GPL code, however the programmers did not intend this obviously.

Now in the Vidomi case, their application replies heavily on the GPL code for a lot of its functionality, whereas in this case the GPL code is only adding some extra user-desired functionality. Now how if the GPL plugin added some great functionality, then I do not think that the plugin could ever be distributed as a part of the application, nor the application advertised as having the GPL'd plugin's functionality, otherwise the GPL would be being broken.

Still, expect Vidomi to argue that it is a plug-in, and that contrary to the application making use of it, it is in fact the reverse - the applications plugin API means that the GPL plugin is in fact making use of the application!

As for the Vidomi case, it is clear cut. They are breaking the GPL. DLL or not, it is providing core functionality to the software. Konqueror works without kHTML (albeit it does nothing!), their software works without the GPL'd DLL (barely, it appears).

The code is not LGPL'd, which would allow the software to link to it, it is GPL'd. Now I am a BSD man myself, but I respect the wishes of people who want to use the GPL to share their work in an environment where other people can use it as long as they don't change it or share their changes with the copyright holder - the author, and follow the GPL in its entirity (excepting special excusions granted by the author).

Vidomi has made the GPL code available, but they haven't followed the GPL in its entirity, and the author has not made any special exclusions. This is the kind of small case that the GPL needs to set a precedent before taking up the larger GPL abusers (allegedly).

It is hard be be clear and to the point when doing law stuff. That is why I am a programmer, not a lawyer. And can't /. make this text area bigger? :)

Grey area

[Hard_Code]

"If [proprietary sections] can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same [proprietary] sections as part of a whole which is a work based on the [GPLed] Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it."

The question then is "does dynamically linking to GPL code automatically make the linking program a work based on that code". I can imagine a case where proprietary standalone tools were developed, each dynamically linking to the same GPLed library for some functions. These tools, or sections of proprietary code, could be "reasonably considered independent and separate works in themselves". Why then is an application (which is nothing but a much larger "section" of proprietary code) automatically infringing on the GPL if it also links against a GPLed library? If *none* of the code within the application is a derivation of GPLed code, then isn't that an "independent and seperate work"? Seems like there is still a grey area here.

Re:Lately?

[QuantumG]

what? you think RMS is gunna go around and sing to them? Enforce == sue, what else is there?

Re:Lately?

[QuantumG]

My discussions with Avery Lee and RMS would indicate otherwise, but I hear ya.

too bad you're not Vidomi

[QuantumG]

you seem to get the fact that Avery Lee didn't want his code used in a dodgy DVD ripping program, it's a shame SloMedia doesn't get it.

Re:What did they do?

[QuantumG]

I can do it, I can do it. I can offer facts without opinion. I'm sure I can do it. yes, they release the source to their modified DLL's. Ahhh.. that wasn't so hard.

Re:OT: Slashdot Submission process...

[QuantumG]

Seriously, this is true. I submitted it to Slashdot, it was rejected. I posted a comment to the AOL crap yesterday, it got modded up to +5, so I resubmitted it, rejected. I posted the story to newsforge and submitted "Newsforge reports VirtualDub GPL violation", rejected. I posted it to LinuxToday, got bored, went to kuro5hin, posted it there. Someone saw my bitch about Slashdot submission process in an editorial comment, submitted it to Slashdot, accepted. It's all about spin. Slashdot editors need to see that headline to trigger their puny little brains into thinking there might be something tabloid here.

Re:(OT)You would like Kuro5hin

[QuantumG]

Yes, where you can have editorial comments which say "hmm, I really think you should rewrite this submission with a bigger summary, because I'm too fucking lazy to click on the link and read the article myself" when all you're trying to do is propogate some links (and yes, it *was* posted under MLP).

Re:Good for the free software foundation....

[QuantumG]

Warezing and selling someone else's stuff are two different things. If you download warez and sell them to people I'd put you in the same boat as SloMedia but at least most people would have the intelligence to not try to do it above board.

it's like a car accident..

[QuantumG]

you look on,
waiting for the head to roll out,
even though you know it will make you sick,
you can't tear your eyes away.

Re:We should thank these ... people

[QuantumG]

WTF are you on about? Microsoft's licenses have been debated in court, are you saying that Microsoft's licenses are weak?

Re:Update from Vidomi

[QuantumG]

I see you're not responding to flames, so I will try to keep it civil. Do you intend to go to court over this matter? If Avery decides to sue that is?

Re:Not to MLP, but...

[Kwil]

From what I've read, the answer is yes in both cases - with a catch.

The catch is that you can not bundle your propietary code with any of the GPL programs that it would use.

You could probably even get away with putting a notice on your software that says your program relies on the GPL'd program "Blah" which can be downloaded from ...

Of course.. I could be just shootin' wind.

Re:GPL can probably be circumvented

[Keeper]

Ok, let's flip the page then.

Let's say I write a windows printer driver and GPL it. Using your logic, MS Word would have to be GPL'd, because word could use my printer driver.

This is, obviously, silly.

I have a feeling that another liscense will have to be created for 'interfaces' -- because what we really have here are 3 things:

* code
* libraries/programs
* how to call the code in the library

The GPL covers code and programs pretty well. The LGPL was meant to address libraries. But nothing has ever mentioned the interface to the library; the assumption that the interface is mated to a specific library and only that library is blatently false.

Something that uses an interface to a library is not a derrative work of the interface; it just uses it. Something that implements an interface would be a derrived work from the interface.

From a different angle (some answers I think would be 'yes', others 'no', however I think you'll find that the question is far more subjective than some make it out to be)....is a car a derrivative work of an engine, or does it just use it? Is an engine derrived from gasoline? Is gasoline a derrivative work of crude oil? Is a tranmission a derrivative work of an engine? Or is it a derrative work of the drivetrain? Or is the drivetrain a derrivative work of the tranmission? Is my word processor a derrivative work of the printer drivers it uses to print on? How about the display drivers? What about the fonts? What about the library which renders the fonts on the screen? Is netscape a derriative work of a webserver? What about a gif library? Or a jpeg library? It certainly uses them, but could the browser exist without gifs or jpegs?

Re:A true test of the GPL

[steveholden]

Well, it was certainly RMS' intention that the GPL should prohibit dynamic linking: in 1999 he explicitly suggested that authors of library routines should GPL them to give free software authors a tactical advantage. It doesn't seem unreasonable that using a DLL makes a program a derivative work, since it would be if it linked the libraries statically.

The more interesting question (to me) is: assuming Vidomi find some way to make a new distribution which does not violate the GPL, and assuming that their original distribution is held to be a violation of the GPL, what remedies does the FSF have for the original violation?

Specifically, will they be able to enforce Vidomi's opening the source of their product, as required by the GPL? If so this would certainly make commercial developers much leerier of GPL violations, probably a good thing for the FSF. Authors do have the right to impose conditions of use, and that's exactly what the GPL does.

Re:A true test of the GPL

[steveholden]

How about if I use C function calls and change the LD_LIBRARY_PATH ? Does my program's licensing change ??

The question at issue has nothing to do with whether Vidomi are using an API. They were distributing GPL'd code with their proprietary code, as a unit, and so now somebody has to decide whether they were in violation of the GPL.

Inasmuch as library functionality can be legally reverse engineered...

Which, of course, it no longer can under the Digital Millennium Copyright Act. Europe appears to have rather more rational laws about such matters, and there it is legal to reverse engineer for the purpose of interoperability.

Re:Update from Vidomi

[Johan Veenstra]

You can ask all the constructive input you want, but the only remaing question is the following:

- Have you distributed the product?

If so, you've released the program under the GPL, wether it says so in your license or not.

If not, you have to get rid of all the GPL'd code before releasing the product, otherwise the above automatically applies.

Why?

- The dll clearly uses GPL'd code, and therefore is GPL'd as well.
- Your program links to a GPL'd library, hence your program is GPL'd as well.

have a nice day,

Johan V.

Re:Get a life

[CaptSwifty]

Actually, they ARE subject to the GPL, as 50 other people pointed out. If you link your non-GPLed program to GPLed code, it is not an indepentant program. Thats the problem the author has. They are linking to GPLed code, which the GPL forbids.

Re:Vidomi's position

[PurpleBob]

You know, I agree with what you're saying, but advocacy of the GPL doesn't help if you look like a raving lunatic as you do it.

Some hints:
* Only one exclamation point per sentence is ever necessary.
* It's 'copyright', not 'copywrite'. Think about it a second. Copyright deals with rights to copy, not "writes".
* It's 'their', not 'thier'. i-before-e rules don't work, so you should ignore anyone who tries to teach you one.


--

Useful insight..

[Convergence]

But, there is another issue to keep in mind:

There is very little precedent for people putting out their source code and allowing other people to make derivative works.

Thus, this issue does not come up in an example like that. Standard proprietary development is done under contractual restrictions.

So, we've got monied interests on one side, and people who want freedom on the other. Personally, I'd say to give it to the good guys.

Re:A true test of the GPL

[sconeu]

Of course it would help! With the wonders of UTICA and DCMA, you can pretty much put whatever you want in licenses and have it be binding!

I know this was tongue in cheek, but does Avery live in MD? It would be interesting to see if we could (ab)use UCITA by enforcing the license under that.

Re:scenarios

[etrnl]

Oh, and who the fuck are you to judge the intent of the author? Especially considering the AUTHOR is the one who's asking the FSF for help SUING the infringing company!!

--etrnl--

At least it's a Small(er) company

[Ted V]

At least it's a company like SloMedia, not Sony or Microsoft. I doubt SloMedia has the same legal resources as Cisco, for example. IANAL;TINLA (I am not a lawyer; This is not legal advice), but it seems like a common legal technique is testing your suits in easy cases to set a tentative precident, which gives you more power in later cases. So in the grand scheme of things, this is probably good for the GPL.

-Ted

Re:At least it's a Small(er) company

[Prior Restraint]

...most courts will side with the intent of the author,...

Actually, in contract disputes where only one side wrote the document, the disputed points are usually construed against the author, unless there's hard evidence to the contrary.

Re:At least it's a Small(er) company

[zhensel]

IANAL;TINLA (I am not a lawyer; This is not legal advice)

If you are going to spell out an acronym, why use it in the first place? Possibly it would be germaine if you used it multiple times in your post, but in this case, the acronym (and the associated description) makes up nearly 15% of the post. Sorry to whine, but this is one of my top 10 Slashdot pet peeves (not that I can list the other 9, but I know them when I see them damn them!) (Oh yeah, one of the other 9 is over use of parenthesis) (another is making posts on a simple point overly long by not ending the damn thing quickly enough).

GPL Question

[RSwan]

I'm not intentionally trying to be a troll so bear with me.

From what I can tell the majority of the pro-GPL arguments seems to be the owner of the copyright chooses the license by which distribution and use of the copyrighted material is controlled. Now the argument in this particular case is the vendor is violating the GPL by only providing the modified GPL source material while keeping the rest of the code propietary. The argument goes since the GPL'ed code is central to the overall program, the rest of the code should fall under the GPL.

Now here goes the potential troll section.

The originator of the GPL'ed code is controlling the distribution of the GPL'ed code by dictating the terms by which the code can be used. If I were to write a license so the material can only be viewed by an approved viewer, according to the main argument expressed in the comments written by pro-GPL commentators, how could a user view the material by an unapproved viewer? The fact the viewer may not be used with the user's favorite tool is immaterial. The user can always get an approved viewer. Fair Use would not necessarily apply because the user has access to an approved viewer, not necessarily a convient one.

For those who haven't figured it out, IANAL and the case I'm talking about involves the MPAA. Personally, I think what the MPAA is doing is stupid and things like the CSS and the regional codes should go away. I don't think I like the viral nature of the GPL because I think it overly restricts the options of the developer. By dictating the terms by which code associated with GPL'ed may be distributed, the GPL is forcing the software to be open not by the superiority of Open Software Process but because of strong arm tactics. Open Software should stand on it's own not be propped up by methods by which Open Software proponents would ordinarily decry.

RSwan

Re:Vidomi's position

[jmccay]

I can see how they would be in violation of the GPL though. The code may be "packaged seperately", but at run time the code becomes a part of the executeable. It gets copied into it's space--at least on windows 32bit systems. (Yes, I know that is a simplification of what goes on.) That is the basic nature of a dll. In this light, the code is a part of the program and they are not distributing the code to the rest of the program.

Under section 2.b.:

"You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License."

I believew this would constitute the requirement of them to provide the source to eveyone. Notice the "in whole or in part" in reference to "any work" it does not say and part of a work (or some resemblance there of).

Further under 2.b.:

"These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it. "

Notice the "apply to the modified work as a whole" with the addition clarification later of "when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License"

I think that should clearly prove that they are in violation with the GPL. I hope the FSF et al. win this one. It will give the GPL some muscle when they go into negotiations with other violators of the GPL.

Removing vs. Opening

[lowLark]

I think the really interesting question here is will the suit seek to force SloMedia to remove their use of GPL'd code, or whether it will attemept to force them to open up their code.

Re:Removing vs. Opening

[Wateshay]

My guess is that the courts will (assuming they rule in favor of the GPL) tell SloMedia that they either have to remove the code immediately or bring themselves into compliance with the license. I seriously doubt they will attempt to force SloMedia to open up their own source regardless of what they do to remedy the problem.

Re:Removing vs. Opening

[ocbwilg]

My guess is that the courts will (assuming they rule in favor of the GPL) tell SloMedia that they either have to remove the code immediately or bring themselves into compliance with the license. I seriously doubt they will attempt to force SloMedia to open up their own source regardless of what they do to remedy the problem.

Dunno. Seems to me that if they cannot guarantee that they can remove from circulation all of the copies of the software/distribution that have GPL'd code in it, then they would need to release the code to those versions. But then I'm not a lawyer or a judge, and it would be an obvious blow to Vidomi's business to have to release the source to their product. Then again, maybe it's not their source anyways...

What did they do?

[scharkalvin]

OK it seems the GPL'ed code is in .DLL form. OK, they can do that, link closed source code with a GPL'ed DLL, that's legal. But did they modify the source of the GPL code to produce a .DLL? In that case they MUST release their modified .DLL code under the GPL. HOWEVER their application that uses the DLL code can STILL be closed source.

SO the question is did they release the source for any modifications they made to the GPL'ed code in the DLLs? If they did not they are in violation of the GPL since their modified DLLs MUST be released UNDER THE GPL. (or the GPLL any way). IMNAL but this is MHO of the thing.

Re:What did they do?

[mdw2]

OK it seems the GPL'ed code is in .DLL form. OK, they can do that, link closed source code with a GPL'ed DLL, that's legal.

Actually, it isn't legal, you can not dynamically link (.so/.dll) to a GPL'd library from a closed source program, an LGPL'd library, yes, but just a GPL'd library, no.

Want some indy electronic (and other) music?

Re:Vidomi depends on GPL'ed libraries

[OmniGeek]

The author of the code in question says it is all GPL, specifically NOT LGPL, as he hadn't built it as libraries; the librariness was a result of post-hoc encapsulation into DLLs by Vidomi. I recommend checking out the links in the header; there's good reading there.

Vidomi depends on GPL'ed libraries

[OmniGeek]

At least according to the VirtualDub fellow's Web site, Vidomi's software is linked to his code as libraries, and is non-functional without it. (They offer a fig-leaf "independent" option that apparently does nothing useful).

Linking to GPL'ed software for its vital program functions means Vidomi's code is NOT an "independent and separate work", and thus is subject to the GPL on the libraries it depends on.

Seems like a smoking gun to me, assuning the VirtualDub fellow is right (and he seems to know what he's talking about).

GPL can probably be circumvented

[Vecna!]

Unfortunately for the principles at stake in this case, I tend to agree with the publisher of the software using the GPL'd code in DLLs that they're not violating the terms of the GPL license.

The GPL is, at heart, a copyright license. It controls the right of a person to use, copy, modify and distribute a work that is copyright. As such, the scope of it's effectiveness is bounded by the scope of the copyright laws.

The crux of the GPL's leverage is it's application in derivative works. If a portion of sourceode licensed using the GPL is combined with other material into a whole, the resulting work will be a derivative work, and the GPL's terms will apply to that work. I seriously doubt that a court will apply any of the computer-science analysis to the code, meaning that terms like "execution thread" and "shared memory" will be ignored. The court will ask "is the GPL'd code combined into a file with the non-GPL'd file", and the answer will be "no". From that point onward, I think the court is going to look very skeptically at the idea that run-time linking can create a derivative work.

In the case of DLLs loaded at runtime, there is never a time when the GPL'd code is combined with the rest of the software to form a derivative work. Interaction with the DLL is in the form of interfaces where the operating system mediates the exchange of data between the calling program and the DLL, and the return of data from the DLL to the calling program. The two pieces of code are never combined into one whole program.

Even if the author of the code in the DLL were to claim that the header files used to construct the DLL calls represented a copyright, and if used in the sourcecode for the calling program represented the creation of a derivative work, the company could simply replace the header files with ordinal values and value types in the description of the external function and call the functions in the DLL without even using the names of the functions or the names of the variables being passed, thus eliminating that source of copyright infringement.

Let me give you another example. Using Windows, printer drivers are stored in DLLs. They are accessed by software at run-time to set printer settings, to rasterize output, and to peform various other kinds of error checking.

If the GPL is held to apply to software that links at runtime to DLLs, it would mean that every Windows printer driver in the world would be required to conform to the GPL in order for GPL'd software to use them.

It has been suggested that the exemption in the GPL for code distributed with the operating system covers printer drivers, but unfortunately many drivers are installed by users and are not a part of the operating system distribution.

Here's an even more extreme example. If the model of "run time calling" is found to create derivative works comprised of the calling software and the target of the call, then anything using an RPC mechanism is also going to have to follow the rules of the GPL. Including the results of RPC calls made using HTTP. Meaning that GPL'd web browsers could only legally connect to GPL'd web servers.

My opinion is that a court will not find the combination of a program using GPL'd code through run-time calls to a GPL'd library to form a derivative work. If it does not form a derivative work, the GPL cannot apply to the calling code. If it does not apply, then distributing a closed-source program that relies on run-time calls to GPL'd DLLs is going to be a legal method of circumventing the copyleft of the GPL.

If the court does hold in that direction, it will essentially mean that the LGPL is a pointless license; anyone will probably be able to encapsulate GPL'd code in DLLs, and as long as they distribute that code in compliance with the GPL, distribute proprietary code that calls those DLLs in the same package. If that happens, the only benefit to using the LGPL will be to allow static linking at compile time to LGPL'd code, something so trivial to replace with a run-time link that the LGPL will become irrelevant.

I am actually glad to see a court test of this issue, because it will help clarify the scope of the GPL and provide the first real road map for how copyleft licenses can be applied to derivative works. It's a valuable lesson to be learned, regardless of the copyleft in question.

Re:GPL can probably be circumvented

[Dirtside]

You said:

Let me give you another example. Using Windows, printer drivers are stored in DLLs. They are accessed by software at run-time to set printer settings, to rasterize output, and to peform various other kinds of error checking. If the GPL is held to apply to software that links at runtime to DLLs, it would mean that every Windows printer driver in the world would be required to conform to the GPL in order for GPL'd software to use them.

You have it backward. GPL software that links to a library doesn't require the library to be GPLed; software that links to a GPL library must be GPLed. At least, that's the contentious issue here.

Let me put it another way. If your printer drivers were written under the GPL, then any software which uses them must also become GPLed (or else violate the terms of the GPL -- if the writer of the software didn't want to GPL his source, he shouldn't have used those drivers). But the example you give is if a GPL program links to drivers. You have it backward.

And as a personal plea, let me request that you not post in monospace fonts; they're hard to read and frustrating. Of course, this is just my opinion; feel free to do as you will.

Re:GPL can probably be circumvented

[clare-ents]

"
Let's say I write a windows printer driver and GPL it. Using your logic, MS Word would have to be GPL'd, because word could use my printer driver.
"

No, false.

The GPL states that a version of Word that *shipped* with your printer driver as part of Word then Word would have to be GPL'd.

Including your printer driver as a part of Word and advertising it as a program that prints to your printer is a violation. The decision to use the driver is made by the software company.

A different case is

Word ships, on the CD is an optional GPL'd program that will install the printer driver for your printer. Then your printer driver is not required for Word's operation and hence Word isn't a derivative product.

Word would be advertised as a program that can print to compatible drivers, drivers for some printers are included on the CD as seperate programs. The decision to use the driver is made by the user.

However, whats going on here is it's been packaged as the second case (after several attempts) but advertised as the infringing first case.

Re:GPL can probably be circumvented

[Whyzzi]

In the case of DLLs loaded at runtime, there is never a time when the GPL'd code is combined with the rest of the software to form a derivative work. Interaction with the DLL is in the form of interfaces where the operating system mediates the exchange of data between the calling program and the DLL, and the return of data from the DLL to the calling program. The two pieces of code are never combined into one whole program. I disagree. The calling program requires the GPL DLL before it can operate, and therefore cannot be considered independant and serperate.

Re:A true test of the GPL

[selectspec]

I am no fan of the GPL but you are completely wrong and the GPL stands on solid footing on this issue. The independent excemption in the GPL applies if the works "can be reasonably considered independent and separate works in themselves" Just because something links to the kernel doesn't mean it is a derivative work, because the code could be "reasonably considered independent." You could compile the code for a different (POSIX compliant) kernel. In the case of kernel modules you are simply utilizing a service (no different from a TCP/IP request to Appache) of the GPL'd application, not creating a derivative work. In this particular case, Vidomi's work is clearly derivative and an extension of the GPL'd work. Their intent of the DLLs was to circumvent the GPL. They will lose if this ever gets to court.

Re:Vidomi's position

[selectspec]

Ah, but RedHat's distribution is not a work "based on the" individual GPL components.

Re:COM objects?

[selectspec]

If you wrote a COM object in a DLL and used the strict GPL with no ammendments (say like Linus' ammendments to the linux kernel's GPL), nobody could use the DLL without beign GPL'd. Now, if you stated in your licence that use of the IDL methods (public API) was excempt, things would be different. The client/server case would not violate the GPL because in the client/server model the client and the server are recognized to be independent programs (your browser is a different program from the web server). However, if your client server model were something a DCOM server and a DCOM client, they would fall under the GPL, because the same portion of the Program (the DCOM object) exists in both Programs.

Re:Vidomi's position

[selectspec]

From the GPL:

These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.

What idiots. Clearly if you write non-derived modules and distribute them individually, such pieces are not under the GPL. Once you package it all up for distribution everything falls under the GPL. What an idiot. Time to get an attorney people or fork over the code!

Re:Vidomi's position

[townmouse]

If Napster advertised 'Master Of Puppets' as their own material, and charged customers for the 'right' to download it, then yes, it would be a Napster-like issue.

Lee's position

[townmouse]

He is refusing to licence his work, mainly on the grounds that he believes Vidomi is illegal.

once it is released that's it right?

[willij3]

i don't claim to be an expert on this stuff but it seems to me that once something is released (under GPL or anything else) that specific instance of that thing (code in this case) is always protected by the license under which it was release.

if this isn't true then no company would ever develop using third party code/software as at any moment the author/copywright owner could come along and demand that the software be removed or a very large fee be paid. any commercial enterprise would be screwed...

in this context if Vidomi violated the GPL then Mr. Avery can command that his code be removed from Vidomi's software. Since Mr. Avery licensed his software under the GPL (for which the established defacto experts are FSF) it makes a lot of sense that Vidomi would want FSF to help them figure out what the word of GPL is. In fact if Vidomi allows Mr. Avery to bend them over and take his code out without consulting the FSF about the validity of his complaint they are allowing him the right to change that license terms of the code that he freely put out on the net some time ago and are opening themselves up to any other vendors/authors whose products/code they have used to come and make demands of them.

A request for clarification regarding library use

[Vicegrip]

Whilst I thought I understood how the GPL works, judging from many of the threads here, apparently I don't.

Some possible cases:

Software loads a KDE lib to do something. Can this software ever be proprietary since KDE is GPLed?

Software links against glibc. Must it then only be GPL?

I was under the impression that a piece of software could use GPL libraries as long as said libraries were not shipped with the product itself: any enhancements made must have the code published under the GPL BUT any GUI or unrelated (to the GPLed code linked to in the library) component of the system can be proprietary. If this is correct, then this presents a special issue for proprietary products since they often link against other 3rd party libraries over which they have little control. Is it possible then for somebody who has accidentally linked against a GPL library to now be liable for making all their code public? I admit this is perhaps more of an issue for Windows programs as they can load libraries at anytime either directly or through COM.

If I wrapped a GPL component with a wrapping layer, publishing that wrapping layer under a "GPL compatible license" would this then satisfy the requirements? Or, as I understand from what this person is claiming, is there no way to publish software that uses GPL code (library format or not/ directly or indirectly) and make it proprietary.

I think it would help A LOT, for people wanting to develop proprietary software for Linux, if there was an authoritative faq somewhere on the does and don'ts, what is expected by the community, and how to publish. I'm interested in writing software for Linux even though all my experience is in Windows. I dont' really expect to ever publish anything that would be worthy enough to sell for money and don't mind sharing my code. But if I ever did come up with a realy nifty idea and wanted to cross platform it, such a faq would help me a lot.

it's copyRIGHT

[kevin805]

What jackass marked legal analysis from someone who doesn't even know what the term is "insightful"? Am I misunderstanding that term? I thought it meant that the posting displayed some innovative thinking, originality, and was interesting in that it pushed thinking in a different direction. I was unaware that the term meant "inaccurately regurgitating arguments made by someone else".

Re:Update from Vidomi

[gss]

Using your argument, I could create a wrapper that loads your entire application as a DLL. I could then freely distribute your copyrighted code with it because, hey, it's not a derivative work

I think then you would be violating Vidomi's license then, ha, ironic isn't it.

Re:US Copyright Law 102b

[RGRistroph]

So ? This is completely off the mark. No one is claiming that because GPL'd code can execute an algorithm, the algorithm is under the GPL. They are just saying that if you distribute a binary that contains GPLd code, you also have to distribute the source. And the point of contention is when the GPLd code in question is in a library. That's it.

Re:Implications

[RGRistroph]

Option #2 is too big. The truth is, some ways of using GPLd code are ok, and some ways are not. You need to make #2 into several little sub-options.

Using code by running it is fine (i.e., a commercial linux distrbution with applications that are not GPL'd, or an application written in Guile that uses the GPL'd interpreter.) Using code without re-distributing it is fine (i.e., MegaCorp has it's programmers do a few kernel or gcc tweaks, but it's only for in-house use and never sees the light of day).

Using code by linking it into the executable, and then distrubuting that code, invoke the viral clause of the GPL and you have to distribute source code with it (well, their might be an exception for when the functions are very generic, so generic that you can find other libraries which will link and run just fine.)

It may help to actually read the GPL and LGPL. It's not cryptic, it is very plain language.

Re:US Copyright Law 102b

[RGRistroph]

You are right that using the API of a library does not necessarily make your code a "derived work."

But sometimes it does. Otherwise, you could use ANY gpl'd code simply by repackaging it as a .so (just remove the main() function and change the link compiler command). The GPL would be almost as weak as the BSD license in that case.

If you link against a library such as glibc.so, and use the fopen() call, no one is going to claim that your code should be GPLd, because you could recompile it on windows or solaris and use their fopen() -- it's obvious that your call to fopen() is more analagous to RUNING someone elses code than it is to MAKING A NEW PROGRAM from it.

But if I include a few elisp configuration files in my package, and make a .so library out of the emacs elisp interpreter, then that should be different matter.

In the Vidori case, they are linking to a library that does some very specific decoding functions.

Also, you say "The binary they distribute requires the API, but not the original source, to work." That is completely orthoganal to the point. Lots of packages out there don't need the source to the libraries to work, in fact that's the usual case. Whether it needs the source to run or not doesn't matter. What matters is if it was developed as a derived work from that library. And since it can't be run using any other interchangable library, they have a weak case.

You can't escape the GPL requirements by packaging up any GPL code you want to use in a DLL and then defining an "API" between your code and the GPL code.

Re:Automatic modification of GPL'd code possible?

[mOdQuArK!]

I think that as long as your "new" source code is merely a straight machine-translation of the original, and still performed exactly the same function, it would be pretty easy to argue in court that the "new" source code is still equivalent to the old. (An aggressive lawyer would go after you for attempting to willfully violate the license terms.)

You'd have to manually add or change functionality to claim creative control over the "new" source code, and given the way the GPL is written, as long as you have any original GPLed code left in the source code (translated or not), you're still in violation.

Of course, in the absense of perfect formal verification tools which can match the functionality of two programs regardless of naming, your "technique" would probably make it difficult for people to easily tell that you were violating the GPL. If it came down to a court case though, I wouldn't count on a simple translation to protect yourself.

Re:Your compiler violates the terms of my code!

[rschwa]

[Darph Bobo]
BRING IT ON!!!
[/Darph Bobo]

Your compiler violates the terms of my code!

[rschwa]

Hey! I released this code under the GPL:

int main (int argc, char **argv)
{
}

Your work is quite plainly derivative of mine, therefore you MUST release it also under the GPL!

Re:Your compiler violates the terms of my code!

[graveyhead]

Don't make me send a bunch of high priced lawyers after your ass, you damn open source hippie freak! Better yet, I'll send Craig Mundie.

Well, your fingers weave quick minarets; Speak in secret alphabets;

Re:VirtualDub and IP

[Ded Mike]

The first link is the web site of the people who are bribing^H^H^H^H^H^H^H...errr...contributing to the campaigns of your freindly neighborhood CongressCritters and State Reptilesentatives: the people who want UCITA to be the law of the land. OF COURSE they're going to say that Stallman's ideas are bunk.

The text of the law they are proposing shows them up for the liars they are. IT WILL outlaw reverse-engineering and IT WILL leave free software authors with unlimited liability. The issue is that this is the software industry codifying the EULAs and FUA that they try to impose on you and I here in the US into law, contrary to the Constitution and their duty to Society. Stallman, the Computer Professionals for Social Responsibility, the Assoiation for Computer Machinery, 2 dozen scholarly organizations, the American Library Association, and over fifty other organizations oppose this nasty law beause it's WRONG and (in the US) UNCONSTITUTIONAL!

BTW, it's UCITA, not UTITA or UTICA.

The original poster asked a question about the outlawing of R-E of file systems and file types being possible, their protection from said R-E and his disbelief that it was POSSIBLE. I pointed out the PROBABILITY of that becoming more widespread thanks to the spread cheeks of your RepublKKKan government.

Re:VirtualDub and IP

[Ded Mike]

RE: I'm not sure what law prevents you from reverse engineering a file format!

It's called UCITA and it's very, very bad . The CPSR, 26 States Attorneys-General, the ACM, and many others oppose it vigorously. You should, too.

Because state reptilesentatives are more worried about their bribes^H^H^H^H^H^H...errr...campaign contributions and the falling tax-base, more than they are about your rights either as consumers or developers, and because the supporters of this unconstitutional law have such deep pockets, UCITA is slowly gaining ground against its opponents.

From Richard Stallman's updated ' Why We MUST Fight UCITA '

UCITA has another indirect consequence that would hamstring free software development in the long term -- it gives proprietary software developers the power to prohibit reverse engineering. This would make it easy for them to establish secret file formats and protocols, which there would be no lawful way for us to figure out.

FIGHT THE POWER!!!!

I'll return that: Huh?

[rakslice]

"Since the GPL claims that work that is commonly distributed as an organic whole constitutes a derivative of any packages that make up the whole, that appears to be a GPL violation;"

I'm not quite sure what you're talking about.
The GPL alludes to the title 17 "derivative work", which isn't defined very well. (It makes sense that the GPL avoids fiddling with broader sets of works, given how the GPL is supposed to function... Remember, a user can always not agree to the GPL.)

As many other people here mention, in dynamic linking, the libraries are never really combined with the calling procedure. Some argue that this doesn't matter; that the library's interfaces are subject to copyright law, and can be the parents of derivative works, dragging the GPL into the fray -- and thus negating all existance. But, current US copyright law has a bunch of exceptions for such things as automated recompiling and modification of interfaces when these are done to allow interoperability.

Given the state of the US judiciary, it will be a barrel of laughs if this case ever lands.

Re:A true test of the GPL

[blakestah]

And if the FSF loses, it won't be that big a deal... GPL 3 would just have to include a clause saying "no dynamic linking with a non-GPL program"

Actually, that would not help at all.

If something is not a derivative work, then the license of some other program is irrelevant,.

Re:A true test of the GPL

[blakestah]

Programs under GNU/Linux are not linked against the kernel.

System calls are not in libc, they are exported directly from the kernel. Things like select, open, ...

Re:A true test of the GPL

[blakestah]

Likewise, it's kind of obvious that this product pretty much relies on GPL code to be useful. Perhaps it offers some token functionality without the GPL code, but, that functionality is pretty irrelavent. A human can take a look at it, and make a value judgement on the situation.

They provide the source code to the GPL libraries. Anyone can take those for at most a copying fee. So the GPL programs do not form part of their intellectual property on which sales could be based.

This case is clear and simple, and it comes down to the question: If I make a program that dynamically links to another, is my program a derivative work of the linking program ?

Think in terms of component models and independence of components. If all linking programs are derivative works, then there is no intrinsic independence of copyright in computer systems at all. That is the statement upon which the defense will call upon human judgment to question the sanity of the FSF lawyers.

Re:"Let's hope Vidomi is correct" ??

[blakestah]

If you put a lot of work and time in your program, and all you want is that the program's source code remains available for everybody, and some company disrespect that wish and steals you code without releasing it, would you he happy about that?

This is not about placing GPL code in proprietary closed works.

This is about whether dynamic linking makes something a derived work.

Vidomi is providing source code to all programs that are GPL'd and is dynamically linking to them with their own code.

Does dynamic linking make something a derivative work in the copyright sense ?? I don't think it does.

Re:A true test of the GPL

[blakestah]

Furthermore, the license of the GPL explicitly states if a program requires the GPL'd code to produce it's intended results then it is a derived work.

If dynamic linking does not make a program a derived work, then it matters for shiit what the GPL says. No license can apply to something that is not a derivative work of the licensed work.

Everyone needs to keep in mind that copyright does not protect functionality - it only protects expression. Copyright does not protect against reverse engineering. If I code against an API and make my binary, is my program a derivative work of the library providing the API ?? How about if there are multiple programs that could provide the API (such as Lesstif, Motif) ?? Is my program a derivative work of Lesstif when I link to its library, and a derivative work of Motif when I link to its library ?? How can that possibly be the case ?

How about if I use C function calls and change the LD_LIBRARY_PATH ? Does my program's licensing change ??

This is not even so difficult conceptually. APIs are NOT protected by copyright intrinsically - only the expression in the source code is. Inasmuch as library functionality can be legally reverse engineered, and the linking program doesn't need to know who wrote the dynamic link library in order to use it, these are independent in the copyright sense.

US Copyright Law 102b

[blakestah]

In no case does copyright protection for an original work of authorship extend to any idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of the form in which it is described, explained, illustrated, or embodied in such work.

Re:US Copyright Law 102b

[blakestah]

The library provides an API. This API can be legally reverse engineered. It is a concept embodied by the source code. If that is true, it is not protected by standard copyright law, and is not protected by the GPL. The API, that is. The source code IS protected, just ideas and concepts embodied in it are not.

That would be a patent level of protection, not a copyright level of protection.

Note that Vidori is distributing a binary that LINKS DYNAMICALLY to a library containing GPLd code. The binary they distribute requires the API, but not the original source, to work.

Re:A true test of the GPL

[blakestah]

US Copyright law provides no protection against ideas, methods, or functionality. It only protects expression. If I have a copy of your copyrighted work, you would need to have an EULA to protect against any usage of its functionality.

Put more simply, you cannot protect functionality with copyright. An additional contract is required, and this is the entire point of an EULA.
The GPL and LGPL and BSD license and artistic license have no contracts. Each grants the user more rights than those intrinsic to copyright, but each takes away NO rights. There are no intrinsic rights in copyright protecting usage. See US Copyright Law 102b.

Re:A true test of the GPL

[blakestah]

Just because something links to the kernel doesn't mean it is a derivative work, because the code could be "reasonably considered independent."

Thanks for making my point. Something that requires an API, but not specific expression used to generate that API, is not a derived work, and thus the license of the program providing the API is irrelevant.

In this particular case, Vidomi's work is clearly derivative and an extension of the GPL'd work. Their intent of the DLLs was to circumvent the GPL.

I think they will argue the intent of the DLL was to make a clear separation between the API provided by the GPLd work and the code that they generated in house to provide additional functionality. Anyone may download the GPl'd code - thus it cannot form a substantive part of Vidomi's intellectual property. The IP is only the in house proprietary portion of the code.

Copyright does not protect ideas and methods embodied in the copyrighted work. In that sense, I do not think library APIs are protected by copyright law.

Now, get yourself an EULA, and everything may be different...

Re:Wrapping GPL code with CORBA

[blakestah]

The GPL cannot be more specific on issues like this - it is beyond its control.

The GPL only applies to copyright issues. If another program interfaces in some way with a GPL program, but is not considered a derivative work, then the interfacing program is not bound by the GPL. Period. There is no re-wording of the GPL that will affect this in any way. GPL is limited by copyright. Copyright can only affect other programs that are derivatives.

The entire question of this lawsuit is whether using a dynamic link library makes something a derivative work. Consider - in many cases you can change the DLL but preserve the API without changing the function of the program. In those cases dynamic linking is clearly NOT making something a derivative work.

In other cases it may be less clear.

A true test of the GPL

[blakestah]

This will be a test of the weakest portion of the GPL that is assumed to be defensible by the FSF - whether dynamic linking to a library makes something a derivative work in the copyright sense.

For EVERYONE's sake, let's hope that Vidomi is correct and that dynamic linking does NOT make something a derivative work. Consider that libraries are CREATED with the intent that someone else will use the API and header files.

I have never agreed with RMS (and Trolltech's) assumptions that dynamic linking makes something a derived work. It would certainly make ALL third party software vendors beholden to OS vendors, for example. Every program ever written under linux would be a derivative work of the kernel, which is GPL. Imagine if someday the kernel developers decided that all of a sudden ALL works ever developed under linux were actually GPL'd (not that this is a likely scenario, but it is a possible one if RMS is right).

Fortunately, RMS is wrong. GPL libraries do not make all dynamically linked programs derivative works, and I expect this lawsuit to uphold that point (and sanity in general).

Re:A true test of the GPL

[blakestah]

> whether dynamic linking to a library makes something a derivative work in the copyright sense.

I think it does since the libraries header files are used in the subsequent work. The courts will always take into account NON COURT precedent in the absence of previous similar findings and there they would run into the NeXT Object C compiler based on GCC (NeXT backed down, its GPL now!), the QT problem, the absence of linking to GPL libs on Linux/UNIX systems, the intent of the LGPL to specifically permit this (this alone should tell you the intent of the GPL is NOT to permit it)

Companies backing down do not set legal precedent. That is the entire point of backing down.

The QT problem has NEVER been addressed in court. They make the same assumptions RMS does - that dynamic linking makes something a derivative. In fact, I think RMS convinced them of it.

The interface between kernel and application is via an LGPL application and Linus specifically permits binary only drivers in Linux (basically making the header files required to use them effecticely LGPL)

When I create a binary and use system calls, I am dynamically linking to the kernel in exactly the same sense as I am linking to a dynamic library when I make a call to a function in that library. I have no necessity to use gcc to make the binary - there are other compilers available, and I can write the program in assembly and in theory create the same executable bit for bit. I don't NEED to include the header file - gcc may require it, but I can certainly code around it.

This has never made sense to me - that ANY dynamic link library makes ALL linking programs derivative works. It seems to me that the entire purpose of creating a dynamic link library is exporting a PUBLIC API and header file that allows someone to create reasonably separate programs that can access a function.

Functions, as even a cursory glance at copyright law will tell you, are not copyrightable. Only the source is. I have no need to use the source to use the API, with the exception of the header file.

If the FSF wins, it would be an almost unprecedented increase in copyright coverage.

Re:A true test of the GPL

[jmv]

Every program ever written under linux would be a derivative work of the kernel, which is GPL.

No. Linux explicitly states (in the Linux license) that. Using Linux system calls is not considered a derived work. Also, note that glibc is LGPL and not GPL to that you can link any program with it.

Re:A true test of the GPL

[naasking]

Or Lesser GPL.

-----
"Goose... Geese... Moose... MOOSE!?!?!"

Re:A true test of the GPL

[Satai]

Whoops, LGPL does mean Library GPL. My bad.

Re:A true test of the GPL

[Satai]

Every program ever written under linux would be a derivative work of the kernel, which is GPL.

Not true. Linus himself has included in the kernel sources a disclaimer allowing for linking to the standard libraries.

In fact, dynamic linking is what the LGPL is all about - and is why it's commonly (mistakenly) called the "Library" GPL.

Re:A true test of the GPL

[Abcd1234]

While I don't entirely disagree with you, there is one particular issue. Imagine company X takes GPLd application Y, swipes a bunch of the code and builds a shared library out of it, and then uses said shared library in a commercial product to provide core functionality. This company then refuses to release the code to the commercial product (although they will release the shared library). Do we want this to be allowable? Under the RMS interpretation, this isn't possible... under yours, it is.

Re:A true test of the GPL

[DmitriA]

Every program ever written under linux would be a derivative work of the kernel, which is GPL.

You are wrong because Linux kernel is NOT licensed under the strict GPL. Linus specifically gives the following exceptions in his COPYING document:

"NOTE! This copyright does *not* cover user programs that use kernelservices by normal system calls - this is merely considered normal use of the kernel, and does *not* fall under the heading of "derived work".

Re:A true test of the GPL

[am 2k]

Somebody who actually believes in mankind... cute.

Likewise, it's kind of obvious that this product pretty much relies on GPL code to be useful.

Yeah, it's obvious to YOU. But is it obvious to somebody that never saw a computer (like most judges)?

Re:A true test of the GPL

[Courageous]

"This will be a test of the weakest portion of the GPL that is assumed to be defensible by the FSF - whether dynamic linking to a library makes something a derivative work in the copyright sense."
----
If this is a "weak" part of the GPL, then the GPL should be revised. Holders of copyrights are permitted under law to dictate whatever terms they please regarding the use of their copyrighted material. That's the strength of the GPL: the original author owns the copyright to the material, while the person using the material agrees to the terms set out in the license. "Permission to copy, distribute, or to excerpt source code from this library is refused in the event that the [insert your own draconian terms here]..." is a perfectly legitimate stipulation. The copyright holder is not under _ANY_ obligation to give up his or her copyright, excepting fair uses clauses, etc.

C//

Re:A true test of the GPL

[Ayende Rahien]

And time, not to mention that there wouldn't be any more lawyers :-D

Re:A true test of the GPL

[mech9t8]

If the FSF wins, it would be an almost unprecedented increase in copyright coverage.

And if the FSF loses, it won't be that big a deal... GPL 3 would just have to include a clause saying "no dynamic linking with a non-GPL program"
--
Convictions are more dangerous enemies of truth than lies.

Re:A true test of the GPL

[mech9t8]

If something is not a derivative work, then the license of some other program is irrelevant,.

Well, I don't think there's an argument over whether the given DLL is a derivative, just whether the front-end program is. Therefore usage of the DLL is governed by the license agreement. Therefore, any restrictions against using that DLL (ie. "don't link to it in a non-GPL program") would be applicable.

I mean, I could release a DLL and say "you must be patting your head when you use this DLL in any way" and if a user is using a program which uses the DLL, they gotta be patting their head or they're in violation.
--
Convictions are more dangerous enemies of truth than lies.

You are so wrong...

[Arker]

I don't know if you are trolling or if you really believe this, but you're completely off-base. A short bit of research at http://www.findlaw.com/ should be sufficient to set you straight. Copyright on source code applies to the code itself - NOT to the underlying algorithms. Just as, for instance, the copyright on "The Hunt for Red October" prevents me from actually copying any lengthy section of that book, but does not prevent me from writing my own novel about a russian submarine captain trying to defect...

If you implement a program using an algorithm you saw somewhere else, but without actually copying any code, you're legally in the clear. Of course, employers may want their employees to do even more than is absolutely legally necessary to avoid the appearance of wrongdoing. There is some guesswork involved in whether or not a court is going to believe that a given work is "derivitive" of another, and if one can make a good case that no one working on project a ever saw the code from project b that could put an end to a suit rather quickly. Otherwise, it might drag out longer and be more expensive to defend. So it does make sense for a microsoft to prohibit their NT coders from even reading linux source for instance. But it's not at all a legal requirement for them to do that.

"That old saw about the early bird just goes to show that the worm should have stayed in bed."

If this claim stands, then "Bye-bye Be!"

[Brett Glass]

Not many people seem to realize it, but BeOS uses GPLed drivers derived from Linux drivers. To avoid being forced to forfeit any reward for writing their very good OS, they make their drivers separate processes; the OS communicates with them via inter-process communications.

If the use mentioned in the story quoted above is contrary to the GPL, then one nasty side effect of this could be to kill Be.

I won't even get into the issue of what happens to programs that call the various GNOME libraries, all of which are GPLed.

--Brett Glass

Re:If this claim stands, then "Bye-bye Be!"

[Brett Glass]

I believe the Gnome base libraries are LGPL'd, actually

No, this is not correct. In a highly controversial edict, Richard Stallman dictated that the GNOME libraries be GPLed. This was shortly after he changed all references to the "Library GPL" to say "Lesser GPL" (shades of Orwell!) on the FSF Web site....

--Brett

Re:GPL v. LGPL

[Error27]

I thought ogg was going to switch to bsd-type license ( http://slashdot.org/articles/01/02/26/1311242.shtm l )

Also the lgpl is slightly more complex than you have explained it. You can't static link non-gpl/lgpl code with lgpl code for example.

it's in assembly

[jon_c]

This would be true, however the guy who made VirtualDub did a good deal of his best work in x86 assembley, so it's very clear from looking at a debugger if they stole his code.

-Jon

VirtualDub?

[Tom7]

I think you must be talking about VirtualDub at virtualdub.sf.net? This is an awesome win32 app for linear editing and video capture (allows you to compress with DivX, for instance). It's one of the most professional pieces of software I've seen...

Re:Answer

[twitter]

It's important that the wishes of the orignial programer are respected. It is right that he recieve credit for his work and that his work is made available. The SloMedia slobs attempted to conceal all of his work as well as their own shoddy hacks. They modified it slightly, wrapped it up and pretended it was their own work.

Plagerism is bad enough, but they are selling his work as well. Others have done the same thing in the past and then used their ill gotten gains to squash the freedoms of the orignal programer with restrictive liscencing and what not. He didn't write his program so SloMedia could enrich themselves by inflicting click through liscences and no copy binary BS on people.

This clearly violates the terms of use for the software. I imagine, from reading the author's site, that he wants either his work to be built on by outside contributors or for money in exchange for the privalige of a closed souce distribution liscence, your standard GPL intent. If software copyright, liscences and IP have any meaning at all, his wishes must be respected just like anyone else's.

Re:Answer

[twitter]

That's EULA crap, and if you want to have an EULA, you should not be using the GPL.

Bull. The GPL is one of the few liscences I'd consider if I ever wanted to share my code. If I were to release without any liscence some creep could scoop up anything of value and lock me and everyone else out. A BSD style liscence could also give me problems like that. The GPL protects my right to share that information in word and deed. As this case demonstrates, the FSF has lawyers.

I'm not sure what you mean by owning all future compatible products. Clearly, someone who wanted to use my work and share it could. Just as clearly, they could make work alikes that I could do nothing about.

As for Windows programing I don't do that anymore if I can avoid it. MS does not have to claim outright ownership of my code. They are free to break it whenever they please and I'm not going to waste my time keeping up with them. Free and beter alternatives are available.

Answer

[twitter]

Other people here have pointed lots of differences out, but I'll make a concrete example.

Your code does something like this:

main()
{my_get_info()
my_code_do_this(info)
my_code_do_that(info)
if (desired) call gzip(info)}

Their code looks more this:
main()
{call restrictive_liscence_scare_consumer()
call cheesy_advert_flash_screen()
call renamed_GPLed_code()
call another_renamed_GPLed_code()
call GPLed_code()}

Your work does something. Theirs is clearly derivative, if not just hijacked, or dare I say "pirated" as most comercial software vendors would? There's a big difference between using a utility like to do something for your program and what they did, which is to rewrap virtua dub so they could sell it.

Rotten tomatoes on you SloMedia!

Re:Answer

[aozilla]

If what they were doing were really that simple, I don't see the big deal. Basically, if they haven't done any real work, why is it so important that we see their source code?

Re:Answer

[aozilla]

First of all, I don't agree that any wishes of the author must be followed. That's EULA crap, and if you want to have an EULA, you should not be using the GPL. From what I've seen, credit was given, and this is not a case of plagiarism.

The main program (not the dlls) clearly violate the GPL. The question is whether or not the GPL applies. Must you get permission from the author of a product to dynamically link to required libraries copyrighted in part by that author. I'm going to ignore the argument that they're not required. They obviously are. But should Microsoft have the right to claim that Windows software is a derivitive work of Windows? Perhaps they should, but to do so would be to abuse their monopoly. In any case, chances are SloMedia will lose, get an injunction, and pay punitive damages. But I don't like this part of the law. Copyright is one thing. Owning the rights to all future compatible products is another. Just because the GPL doesn't work without this "feature" of copyright law, doesn't mean that it is a good thing. The GPL likewise doesn't work for server software (essentially works like a BSD licence, since you don't have to distribute the source). I don't want the law to start enforcing EULAs on usage just to close that hole either.

Re:No harm intended/done?

[BWalint]

I think a better analogy in this case would be running GPL'd software from within Windows. Vidomi 's trying to include the GPL'd code as some sort of plug-in; they've already stated that it runs standalone without the GPL'd code. So, just like how I can run VirtuaDub (a GPL'd "plug-in" to the Windows OS) in Windows, I should be able to have a closed-source Vidomi and have a GPL'd plug-in.

At least, that's how I'm thinking of it, and I'm no GPL expert, but that sounds reasonable to me.

No Credit required.

[SmokeSerpent]

What in the GPL says that if you use GPL'd software, the resulting output must declare or advertise this. If I edit text with emacs, must all my letters end with "This letter written with emacs, a GPL'd text editor."? No. Go away, you're polluting a substantive debate with nonsense.

Irony Meter Explodes

[harlows_monkeys]

One of the comments in their forum was from someone saying basically that it doesn't matter if what they do meets the letter of the GPL or not...they should be respecting the wishes of the author of the GPL code. Yet, the whole purpose of the product is to do something that is against the wishes of the authors of the DVDs that will be ripped with it. My irony meter just exploded.

what if....

[emir]

now hyphoteticaly what happens if judge decides that gpl is invalid? do fsf look for new license ? takes it to higher instance ?

Re:what if....

[emir]

yeah but then amoral companies could continue using their code because its up to them to decide which license they want to use. they are offcourse going to choose gpl v2 and it being invalid allows them to steal code

Doctored Vidomi Logo

[dezwart]

http://froob.net/~dezwart/pics/logo_vidomi_gpl_vio lator.png

Replying to the sig...

[EzInKy]

The way conservatives when national elections is ensuring majority doesn't rule.

Header files

[sigwinch]

Also, you probably need GPL'ed header files to make use of the library.

This is probably the only potentially valid argument for saying the executable which links the dll is a derived work.

Copyright law only cares whether the included work is 1) copyrighted, and 2) included. Meet those criteria, and it's a derived work. A minor infringement is still an infringement.

OTOH, it's pretty easy to write down a list of function names and signatures, and have somebody else independently write new header files. Function names and parameter order are purely functional elements (like numbers in a phone book), and are not subject to copyright. If they actually did this, I think the judge would buy it.

Re:Header files

[sigwinch]

I agree in principle, but it would be a difficult, risky lawsuit. Unless you just enjoy fighing lawsuits, it is better to assume that the binary is a derived work of the headers.

Re:Header files

[BlueTurnip]

I said potentially valid. It would depend, I would think, on whether the header files actually included code, or just function prototypes and variable names. It would be a real stretch to call it a "derived" work, if no code actually got included in the executable.

Purpose of LGPL

[sigwinch]

Is the following program a derived work of libc?

main() { printf("Hello world!\n"); exit(0); }

Legally, yes. If it weren't, there would be no reason for the existence of the LGPL...

Wrong. In the case of dynamic linking, the only difference between the GPL and the LGPL is that the LGPL lets you incorporate header files and other glue in the main program, without having to release the source of the main program. This eliminates the onerous task of writing your own headers from scratch.

IMO, this case comes down to whether SloMedia used VirtualDub header files, or created their own headers from scratch. In the former case, they are screwed, and in the latter case they are not.

Re:Purpose of LGPL

[sigwinch]

To use a DLL, you have to have header files for the exported functions, as well as any symbolic constants used with those functions. For the purposes of copyright law, any binary compiled with those headers is a derivative of them, and therefore affected by the license of the headers.

If those headers come from the original VirtualDub package, they are protected by the GPL, and any binary compiled with them is therefore subject to the GPL.

On the other hand, if SloMedia created their own headers from scratch, those headers are not protected by the GPL, and the binary is therefore not a derived work.

RMS overenthusiastic

[sigwinch]

Amen. The only way RMS will be able to distinguish userland-loaded binaries from kernel-loaded binaries is if the judge doesn't understand him.

Linking as derivation

[sigwinch]

Actually dynamic linking to GPL code does make the linking program based on that code. There is no grey area when is come to the GPL.

True, but so what? This means that when the customer runs the program, they have created a derived work in the memory of their computer. As they do not have the full source used to create that memory image, they must refrain from doing a core dump and distributing copies of the dump.

This has to be a GPL violation otherwise there is no GPL.

Wrong. The GPL would still exist and still be legally enforceable. It just wouldn't have certain nonexistent properties that certain deluded zealots wish it did. If you want a different license that enforces your wishes, write one, but don't be surprised if nobody dares to touch your software.

Re:Vidomi's position

[sigwinch]

They don't seriously think that creative editing can get them out of this, do they?

Who is doing the creative editing? To quote from the GPL In addition, mere aggregation of another work not based on the Program with the Program ... on a volume of a storage or distribution medium does not bring the other work under the scope of this License. Just because the program and DLL are distributed together DOES NOT MAKE THEM A SINGLE WORK FOR THE PURPOSES OF COPYRIGHT.

First, offer a $5 Billion closed source license

[petard]

Before pursuing them in court, the author should offer to license the code for use in a closed-source product for the price of $5 Billion (or a similarly outrageous figure). The company will, quite sensibly IMO, be unwilling to pay the price. At this point, he has actual monetary damages (at least according to the BSA and, apparently, US courts) to sue for. IANAL, but it seems pretty open and shut :-).

easy:

[Argylengineotis]

now's the time to sit down, strip out all the comments, re-arrange a few algorythms and re-name the variables. duh.

Re:VirtualDub is GPL, not LGPL

[BinxBolling]

So, any proprietary software that can be extended by addition of GPLed modules is by definition a derived work? That's more than a little bit scary.

No. What you describe resembles the current case only in certain technical details. You can't call the GPL'd module here a mere "extension" to the closed source program: In addition to advertising itself as providing capabilities that actually come from GPL'd code, the program in question is designed and built around that code. Regardless of the technical details of how the program it uses the GPL'd code, this makes it a derived work. It's quite possible that in another situation, the same technical mechanism could be used without the calling code being a derived work of the called code.

The law doesn't merely care about technical distinctions, such as "static" vs. "dynamic" linking. The law is much more concerned about distinctions (such as intent) that usually have no simple and straightforward technical manifestation.

Re:DEADLY VIRUS!

[mdw2]

you know, no one forces you to use GPL'd code. No one is preventing you from rolling your own. I personally like the nature of the GPL, it means that I can release my code for free (speech) and have assurance that my gift to the community won't be taken by some company and sold as a proprietary piece of software. and above all

NO ONE FORCES YOU TO USE GPL CODE


Want some indy electronic (and other) music?

Re:DEADLY VIRUS!

[mdw2]

okay, I'll bite on your second trolling post of this thread.....

Any application that is running on Linux is utilizing the work of GPLd code, therefore should only GPLd code be legal to run on Linux? If you say no then tell me why: I'd love to hear the definition of what is acceptable use and what is not.

You can utilize GPL'd applications, you just can't directly link to it. This isn't a moral issue, it's a licensing issue, please don't make it more than it is.

This is the funniest irony of the GPL community: Somehow if someone uses GPLd code they have "stolen" it (despite the fact that their use of it in no way inhibits its freedom and accessibility), yet GPL fanatics will be the first to tell you that there is no such thing as Intellectual Property and that the etherial nature of code makes it free for the taking.

...more trolling, I'll bite again. If it wasn't for IP laws, we wouldn't need a GPL. Whether or not intellectual property is legitimate is not what is being discussed, the fact is we have intellectual property and licensing laws. Now, if i released my code under the BSD license, I wouldn't have any problem with a company taking and using my code, it's simply a license issue, if you don't like the license that code comes with, roll your own, don't whine and bitch about how you really want to use that code but the license it comes with is SOOOOO UNFAIR.

the rest of your post is so troll-like it's not worth my time. A little common sense would do you some good.



Want some indy electronic (and other) music?

Re:DEADLY VIRUS!

[mdw2]

I won't argue that the gpl is viral, it's true, however, the gpl can't "infect" anything. Code doesn't get attacked by gpl commandoes and have gpl'd code inserted in it. It's just one open-source license of many, and no one is forcing anyone to start a project and license it under the GPL. If you don't like it, don't use it. It's as simple as that.

Want some indy electronic (and other) music?

Re:Oh look! More /. GPL alarmism

[Wateshay]

You're making the mistake of assuming that RMS's interpretation of the code is correct, and SloMedia's is incorrect. Certainly RMS's interpretation is that which was intended, but it will be up to the courts to decide which one is legally correct based only on the wording of the GPL, and not on RMS's opinion.

(OT)You would like Kuro5hin

[yerricde]

You guys need a system

'K.

where if X percent

Five?

of your readers submit the same story

K5.

you automatically run it

You would love Kuro5hin, where stories are moderated much as comments are on Slashdot.

Disclaimer: Slashdot owner OSDN is a sponsor of Kuro5hin.

Same as a GPL'd plugin for a proprietary program

[yerricde]

They're releasing the code for the dll, but not the program that links to it.

But as the GPL'd library is not necessary for the operation for the main program, the main program could be considered an "operating system" for the library, which would become an "application." Running copylefted applications on a proprietary operating system is common; just look at GNU Emacs for Solaris and Windows and some popular plugins for Winamp. Read More about the operating system loophole in the GPL.

AIM4 license

[yerricde]

Can the terms of a license agreement require that licensees give up arbitrary rights pertaining to things other than the licensed code? Suppose you buy a piece of software and after installing and using it for a while, you decide to read the shrinkwrap agreement. You discover to your horror, that by running the software you agree to give your first-born son to the software company.

Legal. Know the real reason that AOL is blocking Jabber transports from accessing the AIM network? The AIM4 EULA contains language to the effect: "You may not use unauthorized client software to access AOL's servers." Yes, strings are often attached to licenses, and I still use AIM3 (holes and all) on my winbox because its EULA doesn't have such obnoxious terms.

Use on a public web site is Public Performance

[yerricde]

If I have a copy of your copyrighted work, you would need to have an EULA to protect against any usage of its functionality.

Not if you're using it to provide a service to which the public can connect, such as running a server daemon. The courts would probably interpret that as "public performance" of Apache, mod_perl or PHP, MySQL, and whatever nuke/slash/scoop/everything/other weblog engine you're using.

Re:DLLs and COM

[Fester213]

If the original code was GPL'ed, yes, your program using the COM object needs to be too. If the code is LGPL'ed (which would make more sense), then you can create a closed-source program, and are only obligated to release changes to the COM object itself.
-- Fester

OMG

[vex24]

IANAL, but IMHO the GPL and FSF allow this. BTW, HBO ran D.A.R.Y.L last night. LOL.

hahahah clueless company

[small_dick]

1) they don't understand the difference between gpl and lgpl. 2) they think gpl means "open source" (look at their site!) 3) they don't understand that they could solve this by working with the author. must be embarassing to sift through the gpl and construe meaning that isn't there. i think the court would agree that the strongly worded gpl and lgpl definitions far outweigh the "seperate distribution" course they are trying to take. RMS has said many, many times that you absolutely may not treat a GPL lib as an so/dll w/o releasing your linking code. it's in print all over the fsf site.


Treatment, not tyranny. End the drug war and free our American POWs.

Re:Update from Vidomi

[Dahan]

Bzzt. Wrong. He's the copyright holder! He has specifically stated that what you're doing is not okay. Therefore, you do not have the right to modify, distribute, or copy the code.

Bzzt yourself. You're wrong. As the copyright holder, he has released his code under the GPL. As the copyright holder, he can also release his code under the BSD license, or any other license he feels like. But he cannot revoke the GPL from his code just because he doesn't like Vidomi.

You're making like he can say "here's this code which anyone can use, as long as they do A, B, and C," then later go "oh, I don't like Mr. Poopypants, he can't use my code, even though he is doing A, B, and C. Quitit!"

It don't work that way.

Using your argument, I could create a wrapper that loads your entire application as a DLL. I could then freely distribute your copyrighted code with it because, hey, it's not a derivative work.

Wrong once again. Using their argument, Vidomi's code isn't under the GPL. You can't distribute it unless they say you can. It'd be immaterial whether your wrapper was a derived work or not.

Anyways, I'd never release code under the GPL... too many restrictions, too many headaches.

Re:Not to MLP, but...

[Dahan]

Nice and sweet, perhaps, but it misses the point. Vidomi has already "paid" by (arguably) complying with the GPL terms. Now Avery Lee is trying to extract payment above and beyond that.

Too many people are in effect saying what Mr. Lee wants done with his code, he has a right to get. While true morally, it's not the case legally. He's released his code under the GPL. As long as Vidomi is complying with the GPL, Lee's wishes don't mean squat.

Re:Update from Vidomi

[Dahan]

It's not revoking the GPL. They are not complying with the GPL.

Yes, this is what the argument is about--Are they complying with the GPL or not? The AC I was replying to (who may or may not be you :) was saying that because Lee says what the Vidomi folks are doing is not okay, it's not okay. I'm saying that what Lee says no longer matters. It's what the GPL says that matters.

You're missing the point. The reason you cannot distribute it is because of the copyright. They are not respecting the copyright of someone else's work.

Actually, they do have a license to distribute the code. It's called the GPL. Are they complying with the terms of the GPL? That's for the courts to decide, not you, me, the FSF, or Lee, but I think they are.

Re:Not to MLP, but...

[Dahan]

It has already been argued elsewhere here that they in fact did not comply with the GPL.

Yes, well we can argue all we want, but we don't get to decide whether they complied or not :)

So I dispute your premise that Vidomi is legally in the clear here.

I said nothing of the sort. I said that what Lee wants has no bearing on the legality of this. If he wanted to keep control over his code, he shouldn't have open-sourced it. People could talk to him if they wanted to use his code, and he could decide on a case-by-case basis.

If Lee instead decided to sell his software, and Vidomi took it and used it as a reasonably large element in their tool - strike the 'large' part, it's irrelevant - would you not expect that under traditional business law, Lee would be entitled to compensation?

Sure--if he hadn't released his code under the GPL, SloMedia would be infringing on his copyright. But he did GPL his code, so the GPL is what matters now.

Now, Lee entered into the GPL, which, if you read the text of the agreement (or any of a number of posts here), clearly points to exactly where Vidomi was in the wrong.

Not that clear to me :) Anyways, my point is that Lee's wishes are now irrelevant--the question is whether or not SloMedia is complying with the GPL.

Re:Vidomi's position

[aozilla]

You can do whatever you want with a GPL'd program as long as it doesn't violate copyright law. You don't even have to agree to the license.

The question at hand is whether or not you can create a program which requires GPLed code to work. Actually, the other question is whether or not Vidomi's program *does* require GPLed code to work, but let's assume that it does. Let's also assume that Vidomi is not distributing that GPLed code at all, because it would be perfectly possible for them to simply upload the code somewhere and never touch it again.

Now it becomes a napster-like issue. Vidomi isn't doing any direct harm, it is only really the end-users that are. Indeed, this case might work best as a case of contributory copyright infringement.

Re:Update from Vidomi

[aozilla]

Well, it depends if the work was released under a specific version of the GPL, or not. If it was not released under a specific version, the FSF could theoretically change the terms of the license out from under Avery. Which is why you should always specify a specific version when GPLing something.

Re:Vidomi's position

[aozilla]

1) Use VirtuaDub and release thier software under a GPL compatible license.

From the GPL: "Activities other than copying, distribution and modification are not covered by this License"

So in reality, the GPL does not apply at all in this case, unless Vidomi wants to claim that their work is a modification, in which case it is clear that it must be GPLed. The question is a simple one. Is Vidomi's product a derivitive work?

If it is a derivitive work, then Vidomi must get a license to create that derivitive work.

If it is not a derivitive work, then the GPL does not apply.

Re:VirtualDub is GPL, not LGPL

[Mark Imbriaco]

So, any proprietary software that can be extended by addition of GPLed modules is by definition a derived work? That's more than a little bit scary.

If Microsoft were to advertise that certain features were available in Internet Explorer by way of third party plug-ins that just happened to be GPLed, would this force IE to be GPLed itself? I don't think so.

Re:It's funny...

[characterZer0]

He's not talking about copying the code, just reading it and learning from it.

Re:VirtualDub is GPL, not LGPL

[No One]

Mozilla isn't licensed under the GPL; it's MPL/NPL.

Even if it were, AOL owns the copyright, so they can distribute Mozilla in any way they damn well want. The only license terms they'd need to worry about violating are the terms under which MS licensed IE to them.

Not sure about the 3rd party; I don't recall which rights the MPL gives licensees offhand.

--

Re:Vidomi's position

[No One]

Bullshit. The commercial software licenses the FSF is opposed to take away rights that copyright law grants the user. The GPL grants rights to the user that the user would not normally have in law, but only if the user agrees to play by the GPL's rules. If they don't, they can still use the program, modify it, profit by it, do whatever the want with it--except distribute it. Find me one single proprietary software license that gives the user that much freedom even if they don't agree to the license.

The GPL doesn't force you into anything.

--

Re:VirtualDub is GPL, not LGPL

[No One]

If AOL owns the copyright, they can do anything they'd like with it, regardless of the license they release it to others under.

If AOL does not own the source for a GPL'd app, then they have to comply with the GPL. In fact, there's no difference under the law between AOL and some third party. If you're just talking about using the output from one program as the input to another, then I don't believe that not releasing the AOL interface as GPL'd software would be a GPL violation. If they couple the two programs more tightly (for example, as in this case, making Gecko a library AOL required to function correctly), then not GPLing the whole mess probably would be a GPL violation.

If AOL does not own the copyright to GPL'd code, then the answer to your first question is probably yes. The answer to the second question would depend on how the extension was distributed. If it was distributed as a plugin, then I don't believe that would violate the GPL. However, if they were distributing a version of the proprietary app modified by GPL'd code, then they would have to release the whole mess under the GPL.

Vidomi most likely is violating the GPL. At runtime, it's not a matter of piping output from one to the other. They're distributing VirtualDub as a library, so VirtualDub becomes a part of Vidomi's executable code at runtime. And it's not simply a plugin to extend functionality, whatever they claim. That's a legal dodge, and an extremely transparent one. VirtualDub isn't a plugin to Vidomi; if anything, it's the other way around. And that makes Vidomi a derivative work, and thus bound by the terms of the GPL.

--

It isn't about downloading

[elegant7x]

It's about linking. The GPL prohibits linking of code with non-GPL'd code. In that respect the product is in clear violation. There is the LGPL for people who want to allow other people to access their binary code from other products.

Since these dlls aren't part of the OS, and aren't even designed to work with any other programs, it's pretty obvious that they are a 'part' of the program. With out them, the program is nothing. Because of that, the GPL'd software authors have every right to legally require this company to stop distributing the dlls. And sue them for damages (if They filed their copy right with the Library of congress).

They can't force them to open their code, but that might be a reasonable alternative for this company then to simply stop selling it.

Rate me on picture-rate.com

The law is *about* intent

[elegant7x]

Technically, there isn't much difference between someone connecting to a companies server on port 80 and downloading something, and connecting on a port used by MSSQL and grabbing lists of credit cards that are 'publicly available'. But the law really doesn't care.

What makes things crimes are not their technicality, but rather their intent. I would get into far less trouble for running you over accidentally then if I had done it with the pre-meditated intent to kill you. The effect is the same, but the punishment isn't.

Rate me on picture-rate.com

Would that be a bad thing?

[elegant7x]

Taking this argument to it's logical extreme, any program that can create a pipe to another process could be in violation of GPL.

Or network sockets, even. But would it be a bad thing? Remember, it wouldn't have any effect on you, since you can do whatever you want to a GPL'd program in private, as long as you don't distribute it.

Suppose someone wrote a proprietary program that required the distribution of PostgreSQL (or something). The program communicates over Unix Domain Sockets. Now, would that be a GPL violation? Maybe it should be. Really it should be up to the author of the original program how much they want closed source people to be able to profit of their hard work without compensation.

Rate me on picture-rate.com

Hinduism?

[elegant7x]

Hinduism is a very pure and honest religion at heart too, but most people only see the terrorist fundamentalists.

Don't you mean islam?

Rate me on picture-rate.com

Actualy...

[elegant7x]

I think they took the VirtualDub code and wrapped it up in a DLL, then gave away the DLL (but not the main program)

Rate me on picture-rate.com

Mozilla is not GPLed

[elegant7x]

(see subject)

Rate me on picture-rate.com

Huh?

[elegant7x]

The GPL is not a contract. It's a license granting rights that people would not normaly have. Without folowing the GPLs rules, SloMedia is nothing more then a common pirate.

Rate me on picture-rate.com

Right but...

[elegant7x]

This has never made sense to me - that ANY dynamic link library makes ALL linking programs derivative works. It seems to me that the entire purpose of creating a dynamic link library is exporting a PUBLIC API and header file that allows someone to create reasonably separate programs that can access a function.

It was SloMedia, not the virtualDub guy who created the DLL. VirtualDub is just a regular program.

Rate me on picture-rate.com

Re:GPL v. LGPL

[elegant7x]

If the GPL isn't legally binding in this situation, it doesn't matter what the writers meant or what their emotional state was when they wrote it.

Of course it does.

Rate me on picture-rate.com

Virtual Dubya???!!!

[Poligraf]

I didn't know he was GPL-ed ;-)

Qt Implications

[Srin Tuar]

I understand that Trolltech recently agreed to release Qt under the GPL because they became convinced that dynamic linking couldn't be used as a shortcut around the GPL.

If Vidomi wins this case, then Trolltech may decide to reevaluate their position...

This could get interesting. There could be a compromise perhaps- if vidomi can come up with a plugin for their app that can substitute for the GPL library, which would became a separate download...

Re:VirtualDub and IP

[DnA Works]

As I understand it, the *DMCA* prevents unauthorized reverse-engineering of an encrypted file format, which Microsoft could claim fits the ASF (codec as encryption/decryption).

Re:Vidomi's position

[connorbd]

This is a half-assed and noncompliant solution. It's sufficient to purposes if your code is LGPL, but not GPL.

/Brian

Re:Vidomi's position

[connorbd]

They're trying to pretend the code they're using is LGPLed and putting the questionable stuff in DLLs to get around the letter of the license.

/Brian

Re:VirtualDub and IP

[vergil]

Good Point on UCITA.

Even though reverse-engineering (for the most part, DMCA excluded) may be a legally-sanctioned practice in the US, UCITA has a sinister way of blocking it.

Specifically, UCITA gives software publishers wide lattitude to create so-called "contractual use restrictions" -- such as the "no reverse engineering, decompilation, etc" clauses that appear in most EULAs.

Since anyone using a EULA'ed program "agrees" to the terms of the license agreement, reverse engineering becomes not a matter of copyright, but a contractual issue.

Insidious, eh?

Sincerely,
Vergil
Vergil Bushnell

Re:VirtualDub and IP

[vergil]

Here's why I think anti-reverse engineering clauses in shrinkwrapped licenses are insidious.

Theoretically, your assertion about contracts is valid. The legal term "contract" means "meeting of the minds."

However, in the case of mass-market, boilerplate shrinkwrapped EULAs, there really isn't any meeting of the minds. If I download or purchase software as a consumer, I don't sit across the table from the software publisher's lawyers and hammer out a mutually agreeable resolution. EULAs are considered "contracts of adhesion" -- meaning they're unilateral, take-it-or-leave-it "agreements".

Before shooting your mouth off about "communists", why don't you take a look at a few software EULAs (I've put up a few here).

sincerely,
vergil
Vergil Bushnell

VirtualDub and IP

[vergil]

Interesting that VirtualDub is enmeshed in yet another Intellectual Property dispute. In 2000, Avery Lee (the creator of VirtualDub)was forced to remove support for Microsoft's patented ASF (Active Stream Format) file format from his GPL'd program. Apparently, Lee had reverse-engineered the ASF format for compatibility purposes.

Here's a little summary of the VirtualDub/ Microsoft patent dispute with links to more comprehensive articles.

Sincerely,
vergil
Vergil Bushnell

Re:VirtualDub and IP

[Spamalamadingdong]

A contract = I agree to do something and you agree to do something in return. Period. There's nothing insidious about it.

Apparently you've never heard of conscionability limits, statutory rights, and inalienable rights. (IANAL, and even I've heard of these things.)

I know you're just a troll, but I'm sure that even you wouldn't stand for Microsoft adding a term to its Windows licensing agreement (after you're locked into Windows) that you have to pay 10% of your personal/business income to Microsoft for the priviledge of using Windows or Office, and you must continue to pay for the rest of your life. By your logic this would be okay, but courts would probably find Microsoft's terms to be unconscionable and refuse to hold people to them (and possibly make Microsoft pay the legal expenses of the people challenging the terms).
--
Having 50 karma is an itchy feeling; I know I'll get

VirtualDub is GPL, not LGPL

[_|()|\|]

Looks like (relevant parts of) the source code is available

Avery claims that Vidomi is derived from the GPLed VirtualDub. If this is true, then Vidomi source must be released.

Vidomi hopes to create a "user does the link" loophole. This is analagous to the Objective C front end situation described in "Copyleft: Pragmatic Idealism". NeXT gave in. Perhaps Vidomi will, too. Or perhaps, as the headline suggests, this will go to court and set a legal precedent.

Tom Christiansen, a vocal GPL critic, has argued a position similar to Vidomi's. Dynamic linking is an interface. If there is a non-GPLed implementation of the interface, then the program is not necessarily derived from the GPLed library. People have written GPLed replacements for proprietary programs, why not the other way around? However, I suspect that a court would see through the sham replacements in this case. Vidomi advertises functionality that requires GPLed code; therefore, it is a derived work (under copyright law), subject to the license of the original code.

Re:VirtualDub is GPL, not LGPL

[swv3752]

Forsaking Mod points but...

A plugin is in theory a program in and of itself. It could theoretically be done (otherwise you couldn't have any gpl programs on MS Windows.) If the plugin was not embedded but just executed by IE, i.e. ran in a seperate window, it would be legal. I don't see many uses for it, outside of maybe a freeamp plugin, but it could be done.

The point here was that in my example, it is one program executing another program that do not share any functionality. In the vivodi case, you have a program that can only run (or do integral things) if it has this other program (or library if you will).

Re:VirtualDub is GPL, not LGPL

[Ayende Rahien]

No creating your own headers is *very* easy.
You can even write a program that does it.
I don't think that this apply here.

Re:VirtualDub is GPL, not LGPL

[Tech187]

Naw. As you say, non-GPL headers can be thrown together in a jiffy. Or is somebody going to make the claim that 'I have a program that edits text' (basically what a header file is- a description of what a piece of software does) is protected virally too? If so, whomever first said 'I have a program that edits text' owns the rights to every text editor that came after his 'header' described the task.

Re:VirtualDub is GPL, not LGPL

[$hotgun]

If Microsoft were to advertise that certain features were available in Internet Explorer by way of third party plug-ins that just happened to be GPLed, would this force IE to be GPLed itself?

You are correct. That would not be a violation of the GPL.

BUT, if Microsoft distributed the plug-in with IE, AND that program did something critical in the program, such as, um..., displaying web pages, then Microsoft would most definitely be in violation.

The original authors point is that Vidomi is advertising the functionality that is derived from the GPL code as major features of its product. The product will not work as advertised without the code. Furthermore, Vidomi doesn't direct the user on where the DLL containing the GPL code can be obtained and allow the user to go get and install the code. Vidomi has an installer that puts the GPLed code on the user's hard disk, without the user ever knowing that they installed GPL code. For all the end user knows, Vidomi was the author of the entire program.

This program isn't extended by GPLed modules. This program requires GPLed modules. There's a difference. Think of it this way. Your computer is extended by the addition of a text editor, but your computer requires an operating system.

GPL NONSENSE!!

[SPYvSPY]

What is the correct definition of "a whole" under the cited provision? Sold in the same box? Under the same name? It could mean anything.

The GPL is riddled with ambiguity and will most likely support any and all arguments. The only hope is that the judge will have the patience to sort through the encyclopaedic archives relating the drafters' intent.

My guess is that the GPL will sink like a lead balloon.

The lesson: you might hate lawyers, but they draft better contracts than you do.

Re:GPL NONSENSE!!

[SPYvSPY]

Didn't I just say that I challenge anyone to define "a part of the whole" distribution? Have fun.

Re:GPL NONSENSE!!

[SPYvSPY]

It won't hold water. There ought to be a malpractice suit on the horizon, if what you say is true.

Re:GPL NONSENSE!!

[SPYvSPY]

Then why does it suck so bad?

Re:GPL NONSENSE!!

[jdun]

What part of the License do you not understand?

Thankfully, you are wrong...

[SPYvSPY]

It is very wrong to say that the law will always consider intent. Strict liability is the obvious example of no-intent liability. Many other examples exist.

In the world of contracts, it is more accurate to say that assent is the deciding factor. What was the shared intent of the parties in creating this specific instance of the GPL contract? The notion of "shared intent" is often referred to as the "meeting of the minds." This is a very different concept than unilateral intent.

Re:Thankfully, you are wrong...

[odin53]

Actually, she's right, if she's only limiting herself to crimes ("What makes crimes" etc. etc. is what she said). Every crime (and when I say crime, I mean a *crime*, not a violation (e.g., a traffic violation)) requires intent of some kind. Very quickly: a crime is a combination of an act ("actus reus") and a mental state ("mens rea"). You need both to convict someone of a crime. There are no strict liability crimes (aside from low-punishment, low-threshold violations).

Of course, you're definitely right when you point out that the law will not always consider intent. But a contract breach is not a "crime," and one can't say that a company, e.g., "committed a crime" when it's found strictly liable for something. Remember that in these types of cases -- civil cases -- a defendant is found "liable," not "guilty."

Oh look! More /. GPL alarmism

[electricmonk]

Look at the evidence for yourself right here. Please, I implore the editors to at least check their facts (or perhaps it's their morals) before posting blatant flame bait like this. All you serve to do is ruin an innocent company's reputation.

--

Re:Oh look! More /. GPL alarmism

[electricmonk]

And what if the response back from the company was, "Why? We've acknowledged that it is GPLed and given credit. That should be enough."

And perhaps you wouldn't look like such a jackass on Slashdot in the future if you actually clicked the link that I provided in the parent comment, which is NOT simply giving credit. It is also offering the source of the code in question for download, along with whatever modifications that they made to it.

Sheesh, the posters here are getting dumber every time I come back...

--

We should thank these ... people

[surfcow]

... and any like them that come down the road. If the GPL is really so weak that this arguement can even be an debated in court, then the GPL needs to be changed or updated right away. Are there plans to update / review the GLP? We all know that code needs maintaining and updating. Is anyone surprised that a legal document also needs updating? A lot is resting on this foundation.

Re:scenarios

[Karn]

How do you know what the author intended?

Noone forced the author to use the GPL.. He could have used BSD or LGPL but that's besides the point.

The original author's choice was the GPL, so it is only right that a company comply with the GPL.
If the company's product cannot legally use GPLed code then they are free to write their own code.

This may not be a bad thing if we lose

[commodoresloat]

One of the biggest corporate objections to using or building GPL'd code is the fear that they will have to release everything they own under the GPL. If Vidomi can successfully claim that their proprietary program is not derivative of GPL'd code (and it doesn't seem to be based on the discussion here), perhaps other companies will breathe easier about allowing GPL code into their businesses.

Re:Not to MLP, but...

[Bluesee]

It has already been argued elsewhere here that they in fact did not comply with the GPL. They complied with the LGPL, but not the GPL. That is an important point, of course, the crux of the argument. Vidomi seems to be playing dumb, and Lee seems to me to be morally and legally justified; he is outraged that Vidomi used his code in a way that violated his understanding of the license, which, IMHO, is the correct interpretation.

So I dispute your premise that Vidomi is legally in the clear here.

Look at it this way: If Lee instead decided to sell his software, and Vidomi took it and used it as a reasonably large element in their tool - strike the 'large' part, it's irrelevant - would you not expect that under traditional business law, Lee would be entitled to compensation? Vidomi would be sued and lose. Now, Lee entered into the GPL, which, if you read the text of the agreement (or any of a number of posts here), clearly points to exactly where Vidomi was in the wrong. Lee believes this, FSF should (I ain't got all day to follow this story), and a number of people here and elsewhere (hopefully on the bench of your local court) agree with them.

Not to MLP, but...

[Bluesee]

...I really like this retort, from the discussion page at Vidomi, entitled "You Just Don't Get It":

Let's say Virtual Dub wasn't GPL. Let's say Avery Lee was selling his software, and providing source code with each copy sold. Would it be right to copy his code into your own product without compensating him? If he demanded too high a price for his work, would it be okay for you to take it anyway? No. You would have to choose: Pay his price, or forego that code.

In this case, in exchange for using the code Mr. Lee is demanding you publish the rest of your own code under the GPL. Is the price he demands too high? If that is the case, then either you "pay", or you find some other code to do the same work.

Nice and Sweet. There it is in a nutshell. In fact, this leads me to ask the question: Can you write software that would use existing GPL programs without releasing it as GPL? There, I would think the answer would be "yes" in a court of law. But it Should be "no", by the above argument.

[OT] By the way, it has always been my opinion that the Law should be based on common sense. I dunno, it makes all this IANAL stuff somehow a little more irritating. YSNHTBAL!

Re:Vidomi's position

[isomeme]

So Red Hat bundling non-GPL and GPL apps on a single CD do not violate the GPL, and neither do you after installation, since both cases are 'mere aggregation'.

Ah, but you're forgetting that each of those applications is 'based on' Linux, in the sense of linking to its libraries (dynamic or static) and using its system services through them at runtime. None of them could run without those GPL'd components that make up Linux. Isn't that much stronger coupling than "mere aggregation"?

Again, it all depends on how you look at things. We tend to see OS libraries and services as being in a different category from app-specific libraries and services, but in terms of implementation and useage, this is a false dichotomy. The pragmatic implementation details of the two are identical.

So, why is it okay to build Apache httpd linked to Linux system libraries and release it under the Apache license, but not okay to build a DVD app linked to GPL'd utility code and release it under the app's license? It's a mystical convention we've all agreed to follow because it makes our lives easier, but lawyers and juries are unlikely to accept this arbitrary distinction.

--

Re:Vidomi's position

[isomeme]

There are no "Linux system libraries." Linux is a kernel. It exports functionality via system calls, not via any libraries. You are probably thinking of glibc, which has nothing whatever to do with Linux, and happens to be under the LGPL license which specifically permits this sort of activity.

Believe it or not, I know about glibc and its license. But...how is glibc implemented? In terms of kernel system calls. glibc is (in part) a wrapper around the kernel. So we have:

• The Linux kernel, which is GPL
• glibc, which is certainly a 'derived product' of the kernel (given that it is implemented through kernel system calls), yet is somehow LGPL
• Linux applications, which link with glibc to gain access to Linux kernel services, but may have any license themselves thanks to the glibc LGPL trick.

So, to clarify my original question: How goes glibc pull off the magic trick of being an LGPL library implemented using GPL underlying code?

--

Re:Vidomi's position

[isomeme]

Clearly if you write non-derived modules and distribute them individually, such pieces are not under the GPL. Once you package it all up for distribution everything falls under the GPL.

Not quite so clear as it might first seem, alas. The key question is what level of "packaging" creates a real connection between components. For example, Red Hat routinely distributes applications with non-GPL licenses (Apache, BSD, etc.) bundled with the pure-GPL Linux core product. Other than a few extremists, nobody considers this a GPL violation, despite the fact that the non-GPL apps cannot operate without access to Linux system services.

The line between 'bundling' and 'linking' gets fuzzy when discussing dynamic link libraries (and similar loose-coupling technologies). I personally feel that Vidomi has violated the terms of the GPL, but I'd hate to try to prove it in court.

--

Re:What's with the GNU icon?

[nagora]

Am I the only one who thinks that it looks like a turd with horns, shoes, and a blanket?

That's exactly what I think it looks like. I've never worked out what the hell it's supposed to be.

TWW

Wrong nut

[nagora]

Yes, yes, yes. I got the wrong TLA. Sorry.

TWW

A summary of the issue

[nagora]

1. The author of a work can place conditions on the uses of that work which involve copying it (that's why it's called copyright).
2. If you want to copy the work but don't want to follow the author's rules then you're out of luck and you'll have to do the work yourself.
3. Programmers can use the LGPL which says "dynamic linking is OK," or the GPL which says "Dynamic linking is NOT okay unless you open your sorce code".
4. SloMedia have dynamically linked to a piece of code which the author has said is not available for dynamic linking unless the linking program's source is released (GPL)
5. SloMedia should have written their own code if they don't want to abide by the programmer's wishes.
6. SloMedia should loose in court once the court is aware that the decision to use the GPL rather than the LGPL is a conscious declaration by the author that SloMedia's use of the code is not allowed.
7. SloMedia might have got away with it if the GPL'd code was not in DLL's which means that they personally must have copied the code at least into a DLL-wrapper.
8. The weak point of the GPL on this issue would be if they had simply said "If you happen to have made this code into a DLL and have it lying around, then our program will detect and use it". This is a difficult issue for the GPL since SloMedia would not then have done any copying of the work and it would be hard to apply the author's rules on them.
9. But they didn't.

Well, that's what it looks like to me.

RMS (the gun-nut) has an article on this topic which may be useful.

TWW

GPL v. LGPL

[IvyMike]

It seems that the code in question was released under the GPL, and to me, it seems clear that the GPL has been violated in both the letter and the spirit of the license.

I am tempted to argue that the more appropriate license would have been the LGPL. Yeah, yeah, RMS no longer likes the LGPL, but if I were creating a library, I still would LGPL it. However, I wasn't in charge this time :) and the choice of using GPL for a library has merits too.

Of course, it seems that Vidomi is confused: even though they might want the software to be LGPL'd, it isn't, and they are probably in violation because of the stricter GPL.

Re:GPL v. LGPL

[bay43270]

I understand the spirit of the license and the feeling everyone gets when they set it next to the LGPL, but if the GPL has a hole, it has a hole. If the GPL isn't legally binding in this situation, it doesn't matter what the writers meant or what their emotional state was when they wrote it. I know everyone here loves this stuff. I think open source software is a great concept, and I use it every day; but I can't help but think there is a real problem with the GPL. Creating a dll out of a GPL'd program is not a violation... everyone seems to agree to that. Linking to it requires the original headers, which theoretically constitutes a breach in the GPL.. but the reliance on headers from the calling code is simply a limitation of the programming language. Had the GPL'd code been written in Java, another programmer could wrap it into a Jar and treat it as a linked library. That code could be expected to be in the classpath, and accessed as if it were part of the same program. How can a syntactical limitation to the C programming language keep a language independent contract binding?

IRONY: ripping tools

[rerunn]

These guys make tools to rip movies... 'nuff said.

Re:It's funny...

[Kazymyr]

Actually, Avery mentions on his web site he can readily recognize large bunchs of assembly code he put into his app, which don't get changed by recompiling.

Re:It's funny...

[ichimunki]

Consider a recent case in the book world. "The Wind Done Gone" was enjoined from being published because it was purported to be a violation of the copyright restrictions on "Gone with the Wind" (the former is a retelling of the latter from a different perspective), even though I gather that there are no verbatim copies of any part of the original in the later work. The later work wasn't even considered a parody or satire by the court that issued the injunction.

If this is allowed to stand as a precedent-- whether or not we might agree with this ruling-- it makes a strong case for not being able to read any source code and essentially reconstructing it, unless the original source code is public domain or otherwise disclaims copy restrictions.

My personal opinion is that only the form of a work should be protected, so that story lines, algorithms, and basic functionality are all open season as long as it can be shown that there was no direct copying or technical derivation involved (like taking a picture of a painting or using OCR to get code from a book).

Why make an example of them?

[glassware]

Why is everyone so excited about "making an example" of a small company that is obviously trying to do the right thing?

Let me clarify: I observe that this company is small because their product is not widely known. I also observe that this company has gone to the extraordinary length to publish a "Credit Where Credit Is Due" page on their website, to specifically single out the software developers whose work made the program possible.

GPL aside, isn't this type of public credit exactly what we want authors of open-source software to get? It seems to be to be a pretty good world where, if I do hard work and create a valuable algorithm and open it to the public, I am recognized when people use it.

Yes, it stings when someone else makes money off your hard work: I've had it happen to my code, and I felt dumb for not making that money myself. But if I open my code to the world and yet limit its use to other people willing to open their code to the world, I'm not really opening my code to the world ... only to my own little close-knit community.

Thoughts?

Here's the problem

[jrumney]

Even if vidomi now come up with some scheme where they seperate the GPLed code from their own code sufficiently to comply with the GPL, they have broken the GPL already, which means they no longer have the right to use that code.

From the GPL:

4. You may not copy, modify, sublicense, or distribute the Program except as expressly provided under this License. Any attempt otherwise to copy, modify, sublicense or distribute the Program is void, and will automatically terminate your rights under this License.

Re:Vidomi's position

[mark_lybarger]

"If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works....Thus, it is not the intent of this section to claim rights or contest your rights to work written entirely by you; rather, the intent is to exercise the right to control the distribution of derivative or collective works based on the Program."

this merely says that if my program uses/extends gpl code, and also has legitimate non-gpl functionality (menu bar/tool bar/quit button don't qualify as legitimate functionality), then i don't have to release the code of my non gpl using program. i do however HAVE to release the sections of my code that use the gpl code.

Vidomi's position

[mblase]

...is readily available from this link on their Web site. Text follows:

We at Vidomi respect licenses and copyrights. From the beginning, we evaluated the GPL license to determine how to be in compliance. We have been explicit about how Vidomi can have increased functionality using GPL code, any revisions Vidomi made to GPL code have been licensed under the GPL and the source code for this GPL code is available for free at Vidomi (opensourcerelease.zip at http://www.vidomi.com/download.php).

The following language from paragraph 2 of the GPL explicitly concludes that it is possible for a program to in some way interact with GPL code and yet the proprietary independent and separate work not have to be licensed under the GPL:

"If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works....Thus, it is not the intent of this section to claim rights or contest your rights to work written entirely by you; rather, the intent is to exercise the right to control the distribution of derivative or collective works based on the Program."

We believe we are in compliance with these terms of the GPL (as well as others).

At the same time, we are interested in constructive feedback as to interpretation of the GPL license and Vidomi's approach to compliance with the license. We have been in on-going discussions with Avery Lee, author of Virtual Dub, for about a month, and just two days ago received from Avery more constructive feedback which we are now evaluating and will respond to soon. We welcome the community's input on this issue and will evaluate feedback to determine if we should make any adjustments to how Vidomi interacts with GPL code.

dean@vidomi.com

Re:Vidomi's position

[mvdwege]

Simple answer, straight from the license:

In addition, mere aggregation of another work not based on the Program with the Program (or with a work based on the Program) on a volume of a storage or distribution medium does not bring the other work under the scope of this License.

So Red Hat bundling non-GPL and GPL apps on a single CD do not violate the GPL, and neither do you after installation, since both cases are 'mere aggregation'.

Mart

Re:Vidomi's position

[mvdwege]

Glibc does call the kernel yes. It is able to do that because Linus has added a line to the kernel License stating that code using the normal kernel system calls is exempt from the GPL.

Mart

Re:Vidomi's position

[warmiak]

Which proves again,GPL is completely unusable in commerical settings.
If you want real and usefull code exchange between profesionals with no social or political strings attached go for BSD style license.

Re:Vidomi's position

[warmiak]

"It really pisses me off that these corporations think they can jam all sorts of crap into their EULAs to take away our rights to do anything with their software except exactly what they want us to do with it, but then they turn around and balk at the GPL. Seems like a bunch of hypocrites."

Nope, you got it wrong. It is GPL proponents who are outraged at corporation for their EULAs and yet they are using the very same methods.At least commercial entities do not claim to do have a higher goal like FSF does.
Who is a bigger hypocrite here ?

Re:Vidomi's position

[warmiak]

"The point is, the original copywrite holder *always* has the right to set the terms on how you can use thier copywrited material. If you don't like thier terms, don't use thier code!!!! "

Agreed. The problem here is that FSF claims to be on higher moral ground. They do pride themselves in vilifying other licenses and sometimes outright accusing others of limiting freedom ( a very serious charge.)

Re:Vidomi's position

[warmiak]

The problem here is that to obtain their ultimate goal of software freedom they in acted license that is not much different than "evil" they are trying to fight.
It might seem tired and overused argument but GPL really does force people into GPL version of what software freedom is about.

Re:Vidomi's position

[warmiak]

"As I've said before, anytime you use ANY copywrited material "

But the whole idea of FSF and GPL is to do away with copywrited software and any limitations that come with it. Yet you are stubbornly defending GPL using laws this license was designed to destroy.

Re:Vidomi's position

[warmiak]

"It's called pragmatism. "
Well, it doesn't work.
Most of commercial world simply disregarded GPLed software and went on using their own proprietary solutions. GPL software is often banned and, in any case, not that popular among users, let alone commercial programmers.

There is nothing unusual about that reaction, GPL is a direct threat to current model of financing software development and since, as of yet nobody managed to come up with better alternative, people are reluctant to embrace new, clearly unproven model (btw... FSF itself begs for donations, not an encouraging model to follow.)

Re:Vidomi's position

[dinivin]

And you have some way to back up this assertation?

I guess the real question is whether or not the code in question can "be reasonably considered independent and separate works."

Have you looked at Vidomi's non-GPLed code in question to confirm that it's using GPLed code?

Dinivin

Re:Vidomi's position

[dinivin]

Some of us don't believe the same legal interpretation of the GPL as Stallman. Frankly, linking to GPLed code, IMO, is not the same as a creating "derivative work," and with any luck that's exactly what a court would rule.

Dinivin

Automatic modification of GPL'd code possible?

[rxmd]

Would it be technically possible to write a program that takes source code as input and heavily modifies it (i.e. renaming of identifiers, restructuring of source, restructuring of call order (should be possible for given languages using a virtual-machine concept), insertion of loops, heavy optimizations and so on) so that the result has no similarity with the original?

It would be interesting to put it under the GPL (for the self-referentiality...) and use it to modify GPL's source code so that effectively, it's GPL'd-ness could no longer be proven or even ascertained :-)

Then one could create legally interesting situations.

Re:Get a life

[anichan]

If you make a library and release it under GPL and I use it in such a way that my code can stand alone, then it does not have to be released under the GPL. If, however, the core functionality of my program depends on your GPL code, then mine must also be release GPL.

The example I could think of would be an audio program. Say that I wrote code to play voc files, and included a GPL library to play mp3s. Since my audio player don't require the GLP code to do it's thing (you don't *need* to play mp3s, it's just a nice bonus), the code doesn't have to be GPLed.

At least, that's how I understand it.

Re:Get a life

[anichan]

I know it's kind of late, but how does this work for a common lib, such as glibc? glibc, afaik, is GPL, but non-gpl programs use it. When does the functionality provided by a library bring it up to the level of forcing programs using it to be GPL'ed?

What's the worst

[CoreyG]

When source is released under the GPL I'm assuming that it's been released under an implied version of the GPL, correct? If I release code under GPL 1.2, can somebody later release code based on mine under GPL 1.0? I suppose then it would be possible for any company to steal code from another GPL'd source if that source was released under a "faulty" license.

But if there are no versions, then there is no harm done. The license needs to be rewritten.

But if all licenses are unenforcable, then I guess we can start copying Windows and publishing benchmarks about Oracle, right?

Re:What's the worst

[Tech187]

So, if the GPL is completely declawed in court, the FSF has to scamper to come out with a new version. However, anything that was ever released under the earlier, now declawed, license, can be used however the court has reinterpeted the licence to read. In other words, the FSF will be scampering to update the license, and every software developer with code released under the older GPL will be scampering to add funtionality that will justify anybody using the version with the new revised license.

Sounds like a free for all to me, and the potential for much, much forking.

Re:What's the worst

[Snootch]

Nah - the GPL stipulates that you can only release under that version "or any later version at your discretion" - allowing for loopholes to be easily closed, not that the legalese the GPL is written in seems to have that many holes...

43rd Law of Computing:

Dynamic Linking, Static Linking, Separate EXE

[strags]

Dynamic Linking, Static Linking, Separate EXE - it really shouldn't matter. No matter how Vidomi tries to weasel out of describing their product as a derivative work, the fact remains that they make use of the functionality provided by VirtuaDub.

It seems fairly clear that they are trying to get something for nothing. The excuse that they "have not heard back from the FSF" is a ridiculous stalling tactic. If they wish to act in good faith (and they claim that they do), the correct thing to do would be to remove all traces of GPL code from their product until an agreement is reached.

I've checked their forums - every one of their posts is of the form "How can we bring our product into compliance?". Every response is of the form: "Simple - remove all the GPL code, or release the source to your product!". They seem to ignore all these posts, and come back with "What if we tried weaselling out of our responsibility using sneaky tactic X? Would that work?".

This is my compiler

[graveyhead]

It is propriatery IP, and you may not use it for any purpose, unless you pay me $1,000,000. Also, I absolutely refuse to GPL the code.

#define NOT_GCC = "gcc "
#define NOT_GCC_LEN = strlen(NOT_GCC)

int main (int argc, char **argv)
{
int arglen = strlen(argv[1]); char *cmd = malloc (arglen+NOT_GCC_LEN+1);
strcpy (cmd, NOTGCC)
strcpy (cmd+NOT_GCC_LEN, argv[1])
cmd[NOT_GCC_LEN+arglen] = '\0';
system (cmd);
}

Well, your fingers weave quick minarets; Speak in secret alphabets;

Don't be so sure

[phr1]

I'd be very careful if I was the plaintiff. Somehow I expect to see Craig Mundie on the Sunday gasbag news shows commenting on the case, and expensive anti-GPL lawyers appearing for the defense, funded by unknown sources but who all seem to come from the area of Redmond, Washington. :-(((

Re:Lately?

[update()]

...and come to think of it:

the developer is planning to file suit with the aid of the Free Software Foundation

Where does it say that Avery Lee is planning to sue, with or without the help of the FSF? The closest thing I see on the linked page is "As many have suggested, I've contacted the Free Software Foundation and asked them for help in enforcing the GPL." Is there more information anywhere else?

Unsettling MOTD at my ISP.

Re:Lately?

[update()]

Sing? You mean like this? Why don't we start with just a lawsuit and escalate from there?

But anyway, the site doesn't even say the FSF is officially getting involved, let alone that the case is going before a judge anytime soon. I'm thinking "enforce" means "Send a cease-and-desist letter and explain their interpretation in more detail."

Unsettling MOTD at my ISP.

Lately?

[update()]

From the submission:

And the "recently" links to a 9 month old story. Come on guys, we've had plenty of false alarms more recently than that --like yesterday!

Unsettling MOTD at my ISP.

Re:Lately?

[Tech187]

Judging from what the Free Software Song sounds like, some might choose the lawsuit if offered a choice.

The problem is...

[infiniti99]

That it is still a violation regardless of whether or not it is a dynamic or static link. That's why there is an LGPL. Think of Trolltech and Qt right now. Their library is GPL, and almost every application is dynamically linked to it. I'm sure the outcome of this case will be *very* important to them. If the GPL doesn't hold water for dynamic linking, then we may never see another GPL'd Qt.

People/companies depend on the GPL. And Vidomi needs to play by the rules.

-Justin

Re:"Let's hope Vidomi is correct" ??

[infiniti99]

Dynamic linking could be an easy-out for companies, and this would not be good for the GPL. If a company wants to extend a GPL program or include some its code, they could just put the guts into a shared library and now they have circumvented the GPL. This is BAD. The viral effect is now completely gone. We might as well license everything by the BSD license then.

Also, certain programs make more sense to be dynamically linked. Like most libraries. Does this mean there can be no GPL libraries? Do you know how screwed over Trolltech would be if the GPL didn't apply to dynamic libraries?

Dynamic linking MUST be covered by the GPL (and it is) otherwise it could be used as a loophole.

-Justin

No harm intended/done?

[bziman]

According to this note posted last week on the Vidomi web site, they will "no longer provide an option of including the GPL code with the Vidomi code in the same download and installer." Furthermore, they say they have "contacted ... the Free Software Foundation (FSF) last week to request their input on this situation." And finally, they say, "in no way do we mean to disrespect Avery Lee".

It seems to me that their product relies on the features of readily available libraries, and for the convenience of the users, they were shipping them together.

Imagine you write an application in Java... the Sun JVM is freely downloadable, so you can expect the user to go get it himself. But for the convenience of the user, some companies bundle the JVM with the product. Not a really big licensing issue. And besides, for the GPL components, they are making the source code available.

This is like saying that any application that runs on Linux must be GPL'd because it relies on the open source platform. As much as we'd like that, the license does not actually require it.

I think this will be worked out without any litigation and that when everyone gets together and talks about it, they'll find everyone is within their rights.

--brian

now THIS I am excited to see!

[Arethan]

I feel kind of bad for the developer that is having his code used, but if this pulls through and he collects damages, then this will server as a warning flare to all corporations that are currently (or planning on) exploiting the open source community.

I'd like to see the actual figures he collects though if he wins. Generally, the idea is you sue for lost petential revenue. Well, open source software is generally given away for free these days, so the best I imagine the judge will make them do is release their source back the community. Which really won't hurt the company as visibly as paying royalties to the developer for the life cycle of the product. :)

DLLs and COM

[rabtech]

What if I take GPL'ed code and create a COM component out of it. I releast the full source.

3 months later, I decide to start writing a closed-source app and I find the component would be useful.

Should my application now be covered under the GPL?

In cases like this, I think the courts must consider if MOST of the app's functionality is relying on the component, or just a small part. If it is a large part, then the program can be considered a derivative of a GPL'ed work, rather than a regular program take advantage of a small GPL'ed library.

Of course if they ship a binary that doesn't include ANY GPL'ed code in it, but the user is required to download the GPL'ed library to make the program work, tough beans. Otherwise, any program making use of any GPL'ed library on any Linux system must be GPL'ed, or it is in violation of the license.
-------
-- russ

"You want people to think logically? ACK! Turn in your UID, you traitor!"

NO ELECTRONIC THEFT (NET) ACT OF 1977:

[Courageous]

The NET Act was signed into law by President Clinton in December 1997, making it illegal to reproduce or distribute copyrighted works, such as software programs and musical recordings, even if the defendant acts without a commercial purpose or for private financial gain. If the defendant reproduces or distributes 10 or more copyrighted works that have a total value of more than $2,500, he or she can be charged with a felony, and faces a sentence of up to 3 years imprisonment and a fine of up to $250,000. A defendant who reproduces or distributes one or more copies of copyrighted works with a value of more than $1,000 can be charged with a misdemeanor, and face up to one year in prison and a fine of up to $100,000.

Get a life

[unicaller]

If they kept the GPL'ed code seprate from their code ie use GPL'ed librays and wrote there own interface compiled as seprate files then the files that DO NOT contain GPL'ed code would not be subject to the GPL.

If I write a prograg that uses one file(compiled) that has GPL'ed code then, that file is subject to the GPL any other files that do not contain GPL'ed code are NOT SUBJECT to the GPL.

Re:Get a life

[mvdwege]

Read it again. If your code is not essential to the functioning of my program, then the fact that your code is GPL'ed is moot. So it is possible to write GPL code that extends proprietary code without invoking the GPL on the original code.

What is necessary is of course a huge grey area, but in this case it is fairly clear cut, as the extended functionality in the GPL'ed DLLs is advertised as an essential function of the program, thus the program must be GPL or it can't use the DLL.

Mart

Re:Get a life

[CathodeJack]

If you make a library and release it under GPL and I use it in such a way that my code can stand alone, then it does not have to be released under the GPL.

This is only true if the library is released under the LGPL, not the GPL. The LGPL was created specifically because the GPL does not allow this sort of behavior.

Avery's code is quite clearly GPL'd and not LGPL'd. Morover, it was not released as a library. Taking Avery's code and making a library of it does re-license it under the LGPL. Vidomi should release all their code or recall it completely. Anything else is intellectual property theft.

Re:Get a life

[CathodeJack]

I know it's kind of late, but how does this work for a common lib, such as glibc? glibc, afaik, is GPL

Simple. Glibc is not GPL. It is LGPL. Check out gnu.org to learn what the difference is between the two and why the distinction is so important.

Re:Get a life

[Tech187]

Not necessarily.

Non-GPL header files may have been contrived. I see no reason why that wouldn't be acceptable.

Re:scenarios

[Lonath]

So, that would mean that every copy of linux would have to be returned and destroyed, and every webserver running linux would have to shut down immediately? That wouldn't happen. What might happen is that the "viral clause" might be destroyed, but that's unlikely. Considering the crap that other companies put into their EULA's, the GPL requirement is actually quite minor. What would be interesting is if the viral clause is declared to be too restrictive, could that precedent be used to attack other EULAs?....

Re:scenarios

[Lonath]

But did other, more restictive, EULA, has been in court?

I don't know of other cases off hand, but they are much more restrictive.

And remember, the GPL is *not* an EULA (*end user* license agreement) , it's a license for the developer(s).

I have to disagree with you on this one. If you get a binary of a piece of GPLed software and use it as you would be required to use a piece of "regular" proprietary software...i.e. don't try to figure out how it works, don't give it away to anyone else, just use it as written, then you can't run afoul of the GPL. However, the GPL allows you to do many things with the software that a standard EULA prohibits. So, I do see the GPL as an EULA in that sense.

Wrapping GPL code with CORBA

[Expresso]

Consider if I take any GLP'ed program and wrap it with IDL interface. Then I give away the source for that CORBA service and release it under GPL. Now the question: If someone wants to use that CORBA service from a proprietary CORBA client, does the CORBA client need to be published under GPL also? I don't think so... CORBA service and CORBA client are clearly two different programs. I think GPL license should be more accurate on issues like this...

The abstraction factor

[rfredell]

It seems to me that the plugin\COM\CORBA paradigms might have enough of a decoupling between the main program and the GPL'd code to not violate the GPL. Each of them provide a generic abstraction to allow third-parties to extend, or even provide, core functionality. Here are a few examples of this.

Example 1. An application developer writes an application that allows given pieces of functionality to be provided by whatever module the user specifies via a framework and a registration mechanism, for instance a spell-checker in a document editor. Now the developer writes code for two spell-checking modules. The first is completely proprietary written from scratch, the second is based on a GPL'd spell-checker. The developer releases full source for all changes made to the GPL'd version as well as makes the framework API spec public, but the rest of the application remains closed-source. Does this comply? I'd say yes, assuming the GPL'd spell-checker isn't distributed with the application.

Example 2. A user of, say, MS Media Player comes across a audio|video type that doesn't have a MS codec for it, however there is a GPL library that supports it. They code the necessary changes to the GPL library to make it usable as a codec. They release all my changes to the source as GPL as well. Is MS now violating the GPL? I'd say no. Is the author violating the GPL? (Note: The author has no affiliation with MS other than using it's product.)

Example 3. A developer writes a client/server app. The server component is merely a GPL'd database with all supporting business logic coded as stored procs. The client talks to the DB through ODBC or some other framework. Is the client required to be open-sourced? I'd think not. How about the stored procs? I'm not sure...maybe. It's been argued that level of functionality plays a role. What if, in Example 1, my document editor only worked with plain ASCII. Spell-checking now becomes a rather significant portion of the functionality. Does this change the rules? Is the document editor now in violation of the GPL?

It's also been argued that advertised and/or core fuctionality plays a role. MS Media Player's core functionality is the playback of audio/video media, and the GPL'd codec in Example 2 provides core functionality if you are playing that particular media type. Is MS in violation at the point the codec is loaded prior to playback? If so, this could be a rather sinister way to force closed-source apps to open their source. If nothing else it might cause closed-source developers the headache of examining every dynamically loaded module used and rejecting modules known to put the closed-source app at risk.

Given my (somewhat contrived but entirely possible) examples it seems to me that at some point if enough of an abstraction is introduced, either through an open framework whether API(e.g. plugins), library(e.g. COM, CORBA), or protocol based(e.g. HTTP, SOAP), between the closed-source and GPL'd code it could provide closed-source developers the means to legally use GPL'd code and keep their proprietary code closed.

Re:The abstraction factor

[rfredell]

I never said it was for me to define, but one needs to have some idea of what the GPL really says before releasing code under it. My opinions and assumptions may be flat out wrong, in which case any code released under them may need to be opened or re-written or just tossed. Since I have yet to ever develop using any GPL'd code in any way, shape, or form it's all academic, but I may face these issues in the future. If I do, I'd like to be prepared.

How can _anyone_ release code under any license without examining all sides of the issues involved? What RMS intends the GPL to say and what it will be interpreted to say by a judge are not necessarily going to be the same thing. Here I have provided some examples that could indicate that defining what is and what is not derived work is not black and white in the presence of abstracted frameworks.

What gets decided in a court of law will most undoubtably be influenced by the arguments of not only lawyers, but also by the tesimony of "expert" witnesses. And you can bet your last dollar that the definition of what derived work is, and what constitutes linking, is going to be a significant factor.

It's funny...

[Dancin_Santa]

but it's true. Binaries can't really be diff'd for similarity. You can say I stole your code, but I can say I just reverse-engineered your code without ever looking at the source. A few perl scripts later, and I've got a set of source files that don't look anything like the originals. Any similarities can be chalked up to similarities in the problem domain.

Dancin Santa

Re:It's funny...

[Ayende Rahien]

However, reading the algoritm/way of doing it is not covered by the GPL.

Re:It's funny...

[Snootch]

True - but I doubt their official copies are the only ones that exist!
43rd Law of Computing:

Test of GPL or Copyright Violation

[mheine]

I've always assumed the derived work in the GPL followed Copyrights definition of derived work. Considering that it appears that Vidomi is in the clear. The derived works, the .dll(s) under contention have source code available on their site. The fact that their proprietary product cannot perform without GPL'd .dll does not make it a derived work any more than a closed work using Linux System Calls makes the work derived form Linux.

It's NOT 'virtuadub' its 'VIRTUALDUB'!

[Cyberia125]

Sheesh, did't you even TEST your own link, Slashdot?

License ~=? EULA

[Xesdeeni]

Doesn't the GPL stand on the same ground as all the ridiculous new EULAs floating around? I mean, I bought a pieces of software that said I have to a) register to use (or it expires in 7 days), and b) that I can't sell or even give away when I don't want it any more! Apart from the obvious (What happens if the company goes out of business and I have to re-install the damn thing?), if they can actually enforce that kind of garbage, then I don't see how anyone can legally voilate the GPL either. I'm sure Vidomi has a nice EULA of their own. It seems like Open Source wins if GPL triumphs, and if not, then consumers can tear up their EULAs! Xesdeeni

Re:If Vidomi Wins...

[Ayende Rahien]

No, if they win, then get ready to see GPL stuff in a lot of products.
Including, dare I say, Microsoft's.

Re:scenarios

[Ayende Rahien]

But did other, more restictive, EULA, has been in court?
And remember, the GPL is *not* an EULA (*end user* license agreement) , it's a license for the developer(s).

Pity TrollTech

[Ayende Rahien]

If they are correct, and dynamic linking doesn't create a derived work, then no one would have to pay TrollTech for closed source Qt applications.

Re:I had a (somewhat) similar issue - it gets murk

[Ayende Rahien]

Linus was wrong. (Bye Bye Karma... )

The GPL says that if you bundle the GPL code with non-GPL code, then you must GPL the whole thing.

Qoute from the GPL:
If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.

Patents apply to ideas

[Spamalamadingdong]

It's copyrights that apply to implementations; if implementations were the only things protected by patents, everyone who wrote their own LZW implementation would be immune to the Unisys patent on the GIF format, and there would never have been a "Burn all GIFs" movement.
--
Having 50 karma is an itchy feeling; I know I'll get

Re:scenarios

[azrix]

What? If the FSF loses, then everyone knows the GPL isn't a valid, legal license the way everyone wants it to be. So they write another one. Most GPL licensing that I've read reads like "this software is released under version 2 of the GPL or under any later versions of the GPL." So, then, all GPL code is immediately covered by the new license, which would then have to be tested again in court. And if the FSF wins, they don't gain much? Have you ever heard of legal precedence? If they win, it makes it that much more easier to defend against other violators of the GPL. They set a legal precedence of the GPL being a perfectly valid and defendable licese. That seems like an awful lot to me. A lot more than "GPL source will be liberated from the commercial DVD ripper", don't you think? It doesn't matter who wins. The FSF can't really lose. They just get setback a little bit if they lose. But there's an awful lot to gain if they win.

Not VirtuaDub...

[torian]

It's VirtualDub, not Virtuadub. And the homepage is here not virtuadub.org. Do the slashdot editors even click on the links submitted to them anymore?

Maybe this is what that nice Ms Mundie meant?

[LaminatorX]

This scenario sounds oddly familliar:

" Some of the most successful OSS technology is licensed under the GNU General Public License or GPL. The GPL mandates that any software that incorporates source code already licensed under the GPL will itself become subject to the GPL. When the resulting software product is distributed, its creator must make the entire source code base freely available to everyone, at no additional charge. This viral aspect of the GPL poses a threat to the intellectual property of any organization making use of it. It also fundamentally undermines the independent commercial software sector because it effectively makes it impossible to distribute software on a basis where recipients pay for the product rather than just the cost of distribution."

..and..

" The GPL asserts that any product derived from source code licensed under it becomes subject to the GPL itself. When the resulting software product is distributed, the creator must make all of the source code available, at no additional charge. This effectively makes it impossible for commercial software companies to include source code that is licensed under the GPL into their products, since by doing so, they are constrained to give away the fruits of their labor."

If I were a conspiracy theorist, I'd think the timing was by design somehow, rather I think the timing of Craig's speach was in response to this sort of thing boiling in the industry for some time now. Here's the Mundie Text and the /. article.

Re:Get a clue

[CathodeJack]

If they kept the GPL'ed code seprate from their code ie use GPL'ed librays and wrote there own interface compiled as seprate files then the files that DO NOT contain GPL'ed code would not be subject to the GPL.

If I write a program that uses one file(compiled) that has GPL'ed code then, that file is subject to the GPL any other files that do not contain GPL'ed code are NOT SUBJECT to the GPL.

WRONG!! Wrong wrong wrong!

The "How to Apply these Terms to Your New Programs" section of the GPL specifically mentions the subject of linking:

This General Public License does not permit incorporating your program into proprietary programs. If your program is a subroutine library, you may consider it more useful to permit linking proprietary applications with the library. If this is what you want to do, use the GNU Library General Public License instead of this License.

Avery Lee did not release his code under the LGPL, and did not release it as a library. It was released under the GPL. Therefore, any code linking into his code must be released under the GPL. That's how it is, whether Vidomi likes it or not. Taking Avery's code and turning it into a library does NOT change the code license from the GPL to the LGPL. Morover, in their initial release, Vidomi clearly bundled Avery's GPL code with their own as one package. That means, whether Vidomi likes it or not, their code is already licensed under the GPL, and they MUST released their source code. Once GPL, always GPL.

Oops. Need to watch my typing.

[CathodeJack]

Taking Avery's code and making a library of it does re-license it under the LGPL.

That "does" should be read as "does not".

Crap.

[Computer!]

Finally, Open Source has a business model: the lawsuit!

Re:Idea

[Tech187]

Hmmm.

So it's got to be fundamental, hard to reproduce, and functional?

When will you have time to write the part of the code that people will care about enough to download and run?

Re:scenarios

[warmiak]

"The GPL source will be liberated from the commercial DVD ripper. "

Sure, it will be liberated and left to "rot" NOT BEING used by thousands of people just for the sake of some social goal.
I am sure the author's goal was to write code that won't be used by anybody ...

Re:scenarios

[warmiak]

The point here is not that GPL is legally wrong.
All I am saying that GPL which was designed to spread freedom is actually preventing people from reusing code.
Mind you, nobody was trying to steal GPL code here. It was and is available for download. It is that insane requirement that anything linked with GPL libs has to be distributed as GPL that is preventing code sharing.
It might be legally sound but sure does not sound like freedom, does it ?

Big Probelems...

[gnovos]

Ok, let's say I have a library from microsoft.com, and a library from GPL.com, and I use both of them in the same program, does that make the Microsoft library GPL'd?

If not, then how is this different than just writing my own closed-source library and then trying THAT to a GPL'd pice of code?

A dark day for the GPL

[foonf]

Based on the evidence available, it seems that this program is indeed not in compliance with the GPL. The FSF's position on GPL'ed code linked to proprietary code is pretty clear.

But thats the problem. There is a good possibility that the GPL could be interpreted as not legally binding. Certainly quite a few companies involved in previous cases have argued this (QT, for instance) Considering the pro-corporate bent of the US legal system right now (and particularly the supreme court) I can easily see this case ending with the GPL being essentially nullified, leaving vast amounts of GPL code out there open to being incorporated into proprietary software, with no protection for the original authors.

This isn't a problem now even though the legality of the GPL is questioned. Corporations (like trolltech) observe the license because they do not want to offend the community. But if they are given explicit legal permission to plunder GPL code, what makes you think IBM, etc. wont?

scenarios

[zoombah]

if FSF loses: Other companies will trample all over the GPL. They'll still maintain a certain discreetness, and any power that the FSF has in court will be severely limited. Possible repercussions on impending yro cases by association.

if FSF wins: not much. The GPL source will be liberated from the commercial DVD ripper. What else could the FSF et al ask for? Not money..

a GPL DLL issue I once pondered

[sevinkey]

Although this isn't the case with this particular issue, would creating a plug-in DLL that uses GPL'd code and is released under GPL for a non-GPL'd application be allowed under the license, in the case that the GPL'd plugin isn't required for the app to function.

Example: I was working on a networked-based emulator (until someone beat me to it) that used plugins to emulate each of the consoles, one being the Stella 2600 engine (GPL'd). I was considering not releasing the networking code, but the code to the plugins.

I've changed my mind on not releasing the code since then, but I was curious if anyone else thought this would be a valid use of GPL code since the program could run any plugin, GPL'd or not. BTW, if anyone's actually interested in the code, email me and I'll post it somewhere, since I'm not using it anymore.

Pluggable Module Issue?

[cnelzie]

This seems to be quite like the issue of having a proprietary pluggable module that can be plugged into a GPL piece of software. A good example of this is the fact that the Linux Kernel can have proprietary driver modules plugged and loaded on the fly for pieces of hardware.

This sounds similiar to what is happening with this proprietary code. The only major diference is that the "pluggable module" in this case is a piece of GPL'd software. Of course it is a DLL. The issue then, would be can this software perform its actions without having this DLL?

If it can be proven that the software can function without the DLL than the issue is a moot point. Of course if the software requires this module than it is possible that this will spend some time in the courts as they figure this out.

Re:About fscking time

[Snootch]

You guys need a system where if X percent of your readers submit the same story you automatically run it... (trails off into anti-/. rant)

Now that *would* be an idea! Perhaps somehow linked to karma...though i imagine it would be a nightmare to implement, and that it would be far easier to just keep a close eye on the story pool.

43rd Law of Computing:

Re:GPL code is proprietary

[Snootch]

Should've used code released under a truly free license, like the BSD license, instead of the GPL

Not at all! The GPL is deliberately designed not to be proprietary, and for it to be impossible to become so. Look what happened to BSD-licenced, well, bits of BSD - they ended up in bits of Windows. Companies don't like people nicking their stuff and making money out of it, nor do Free Software programmers...

43rd Law of Computing:

Update from Vidomi

[deanvidomi]

We are attempting to deal with this GPL issue openly and with integrity. We emailed the FSF almost two weeks ago, requesting their input, and have not heard from them. We are disappointed that so far we haven't been able to engage with them on this issue. We are actually hoping that the increased exposure of this issue on Slashdot will help move this forward. Just yesterday I emailed Avery Lee, author of virtualdub, with an update on our recent activities and an outline of another possible change in how Vidomi interacts with GPL code, with the goal of determining if he felt this approach was GPL compliant. We recognize that Avery's opinion is not binding on anyone, but it is certainly something we value. We've also been communicating privately with a number of people who have expressed concern about the issue and have been willing to offer their constructive input. We welcome further constructive input and hope that this situation is fully resolved very soon. dean@vidomi.com

OT: MODERATE PARENT UP

[absurd_spork]

It's absurd how much FUD slashdot is spreading these days. Isn't it about time you guys revisited your submission process, and the way stories get posted.

What Slashdot definitely lacks is an equivalent of Kuro5hin's Meta forum where Slashdot-relevant things could be discussed. Then one wouldn't have to file legitimate complaints like this one under other topics, they wouldn't get moderated to Offtopic or Troll, and maybe they'd even get read...

Otherwise, there can never be a sensible discussion among Slashdot users how Slashdot community problems might be solved.