Dish Network Dishes Source Code for DVR

An anonymous reader writes "According to Dish Network they are accommodating 'requests for the portions of the DISH 921 DVR source code that are subject to the GNU Public License, or GPL. In compliance with the terms of the GPL, we are making this source code available to the public to download. Please note that the DISH 921 DVR software also includes some proprietary elements that are not subject to the GPL. You cannot create a working DISH 921 DVR software build without the additional proprietary code. Do not replace or add any software to the DISH 921 DVR with items compiled from these source trees. Doing so will void all warranties and cause the unit to fail.'"

Comments

52 Megs of source?

[damonsmith]

Holy crud.

Re:52 Megs of source?

[ninji]

A company giving up GNU and GPL source without a fight and arugign it dosent exisist or creating excuses like everyone else? Speechless.

Re:52 Megs of source?

[hostyle]

Damn cat stepped on the Submit button before you could preview/re-read, eh?

Re:52 Megs of source?

[ninji]

Lol, while my cat does walk across my keybaord while im typing alot, I was just in a hurry, as I usually am and am now, and dont think it necissary (perhaps not considering the analretentiveness of slashdotters when it comes to spelling grammer and punctuation...) and yeah...

GPL != GNU Public License

[gimpboy]

Actually, GPL is the General Public License.

Re:GPL != GNU Public License

[Anonymous Coward]

It should be called the GPL Public License... :)

Re:GPL != GNU Public License

[Anonymous Coward]

Actually, that's the GNG license

Re:GPL != GNU Public License

[uss_valiant]

Actually, GPL is the General Public License.

Well, actually GPL is a recursive acronym for "GPL's a Public License".

Re:GPL != GNU Public License

[startleman]

If you follow your own link, you'll see that GPL is not simply the "General Public License."
It's actually the "GNU General Public License"

Re:GPL != GNU Public License

[gimpboy]

What does "GPL" stand for?
"GPL" stands for "General Public License". The most widespread such license is the GNU General Public License, or GNU GPL for short. This can be further shortened to "GPL", when it is understood that the GNU GPL is the one intended.

So GPL can stand for "General Public License" _or_ GNU General Public License. Neither of these include the "GNU Public License".

cool

[sabernet]

A larger media conglomerate fulfilling their end of the bargain? I think I like Charlie:)

On a more serious note, it's good to see some companies actually doing the right thing(aka the anti-SCO).

Re:cool

[Curtman]

A larger media conglomerate fulfilling their end of the bargain?

Are they though? I don't care to RTFA to find out, but doesn't this:

You cannot create a working DISH 921 DVR software build without the additional proprietary code.

... go directly against this:

You have a GPL'ed program that I'd like to link with my code to build a proprietary program. Does the fact that I link with your program mean I have to GPL my program?

Yes.

I thought this was what the LGPL was created to allow people to do.

Re:cool

[sabernet]

hmmm, touché

I have to admin my ignorance of teh intricacies of the GPL.

Re:cool

[Bradee-oh!]

This point is exactly what I was thinking the moment I read that story and they seem to be in violation.

However, I've always wondered (and maybe there's a simple answer for this but googling it would help only me and keep it out of the discussion) - What if you ran the GPL code and your proprietary code seperately (different processes?) and just had them communicate via IPC? Sockets, pipes, shared memory... whatever. It seems like a cheap way around that rule in the GPL but I fear that it is valid...?

Re:cool

[n0-0p]

No it does not violate the GPL if they are not binary linking. Meaning they are most likely running proprietary applications on Linux to provide the PVR functionality. In that case they only need to provide the source for modifications to GPL software. Perhaps you should RTFA before commenting instead of making a disclaimer about it; it might lower the noise floor a bit.

Re:cool

[Curtman]

hmmm, touché

Sorry, didn't mean to make a statement there, it was a question. I know the GPL has a clause that makes exemptions for operating systems. Maybe they plan to exercise that.

Re:cool

[NoMoreNicksLeft]

It's actually a cheap way to do it, that would alleviate all this stupid "GPL is cancer" fud.

The GPL isn't meant to be invasive, if you write your own seperate program, it is allowed to coexist with GPL programs on the same machine. Otherwise, IE users wouldn't be allowed to access an Apache website, for instance...

Re:cool

[Curtman]

Otherwise, IE users wouldn't be allowed to access an Apache website, for instance

That sounds like a great idea. Lets see how many IE fanboys are left when they can only talk to IIS servers. ;)

NOTE: This post is not to be taken seriously.

Re:cool

[Rasta Prefect]

Otherwise, IE users wouldn't be allowed to access an Apache website, for instance...

Well, actually Apache isn't GPL. http://www.apache.org/licenses/LICENSE-2.0.html.

Re:cool

[Anonymous Coward]

bundling software with linux is not "linking," sorry.

Re:cool

[Anonymous Coward]

No it doesn't. Maybe you ought to actually read it some time...

Re:cool

[Zeinfeld]

This point is exactly what I was thinking the moment I read that story and they seem to be in violation.

Could people stop the open source religious wanking for a few minutes and work out if the code that has been released allows something usefull to be done? Its not like anything in this thread has not been said a few million times already.

OK they have checksums, big deal, look for the file with the checksums in it or find where the public key is stored. I would guess they do this because they update the player over the air.

There are three important mods to a DishPlayer that would dramatically improve the utility. The first is to enable that external expansion port so that an external disk can be plugged in.

The second is to enable some form of serial port for command input so that the device can be set to record remotely.

The third is to have a program that allows files stored on the hard drive to be read and copied. An NFS daemon would be nice but something less would be acceptable.

I would guess that the reason that Disphlayer do not want people changing the code in the dishplayer is that there might be a way to disable the 'call home' feature on the modem. If that is their concern then it is easy enough to fix, in fact I will even tell them how to put the fix in place for free if they make the other modifications.

Re:cool

[Curtman]

if they are not binary linking

Thats what isn't clear. How they claim you cannot build the software without the proprietary bits, without linking.

Perhaps you should RTFA before commenting

Which FA?

Re:cool

[Curtman]

Maybe you ought to actually read...

Same to you buddy

Re:cool

[Arker]

They're ok legally if they aren't linking the code - just communicating through socket calls, etc. No different than running, say WordPerfect or Oracle on Linux - as long as the proprietary stuff is stand-alone and not linked it's not a legal problem.

However, if they are correct that installing binaries compiled from this code release will cause their PVR to stop functioning, that sounds to me like they aren't actually in compliance. If the proprietary stuff really isn't linked, and the code they released really is the code for the free software in the system, then binaries compiled from it (assuming the same compiler and settings) would be byte for byte identical with what's already on the PVR.

So, it seems to me they're lying about one thing or the other.

Re:cool

[n0-0p]

Sorry, I assumed a degree of familiarity with the backstory. Perhaps the editor should have provided some of that, instead of just the Dish Network link. Anyway, Dish Network releases source to the modded GPL software, but keeps the proprietary apps and drivers closed. They have done the exact same thing with their previous releases (721 for example); it's all legal according to the GPL. One can argue on the spirit of it, but they aren't breaking the law.

Re:cool

[gstovall]

That's what I think, too. However, note that the MySQL documentation explicitly states that it is their position that if your program is accessing the MySQL port, even remotely, then it is subject to GPL restrictions, unless you buy MySQL's commercial license. Now, perhaps this is just FUD to drum up buyers for their commercial license, but that's what they say...

Re:cool

[Curtman]

Sorry, I assumed a degree of familiarity with the backstory

No problem, I'm surprised there wasn't a link myself.

I'm just curious how they claim that you cannot build the software without the proprietary bits, and still be compliant with the GPL. That's their words, not mine.

You cannot create a working DISH 921 DVR software build without the additional proprietary code

Re:cool

[n0-0p]

I'm just curious how they claim that you cannot build the software without the proprietary bits, and still be compliant with the GPL. That's their words, not mine.

Because they are using "software build" to mean the entire software package including the OS, and all drivers and applications. This is a very common use of the term when referring to embedded systems. They are pretty much saying that most of the juicy bits are contained in seperate binaries in the form of proprietary applications and device drivers. This is completely within the bounds of the current GPL. As a side note, I've actually seen companies use CORBA wrappers to avoid GPL requirements; I consider that a much worse approach myself.

Re:cool

[Curtman]

That sounds reasonable enough.. Much like including an nVidia or ATI binary driver with a Linux distro. Much as I dislike it, it's perfectly legal. I guess it's best we don't give them too much flack for a bad choice of words. I think this device is on the whole a great thing. Hopefully it won't be too long before some hackers create free replacements for the proprietary parts.

Re:cool

[Anonymous Coward]

i think maybe it would work, it's just that they don't want to claim liability if you break your hardware.

Re:cool

[jimicus]

Or maybe they just don't like the idea of facing support calls which begin with "Hi, I just downloaded this source code of your website, I thought 'this looks pretty cool' and I wrote my own program to go on the box... and it doesn't work any more."

Re:cool

[shutdown -p now]

If what they say is true, then accessing a GPL'd web server with e.g. IE would break the GPL as well. In other words, it's nonsense and FUD.

Re:cool

[Lehk228]

they probably sign their binaries before sending them to the DVR, if you send just the binary unsigned the device won't run it (just my guess)

Re:cool

[Arker]

From what I've read, it sounds like the issue is very close to that. I'm told they modified the kernel to do some black magic when loading programs, and aren't releasing those modifications - so if you compile and install a kernel from the sources provided it won't work anymore.

If that's accurate, they still aren't in compliance with the GPL, and they made a very poor architectural decision that's left them with the choice of totally blowing their own security system or else continuing copyright infringements. Ugh. Hope they didn't do that. It's just hearsay at this point, but it sounds like something stupid enough to be credible.

Re:cool

[Anonymous Coward]

That isn't an "exemption for operating systems", it is an exemption for components that are "normally distributed with the operating system". It applies when you distribute, under the GPL, a binary for an operating system that the end user already has, which has those libraries and tools already on it. It doesn't apply in this case since they are shipping the operating system and the libraries to the end user along with the GPL binaries.

Re:cool

[Shadowlore]

That's what I think, too. However, note that the MySQL documentation explicitly states that it is their position that if your program is accessing the MySQL port, even remotely, then it is subject to GPL restrictions, unless you buy MySQL's commercial license.

IF it says it, it is wrong. However, I can't find where it states that so it's all speculation anyway.

Re:cool

[gstovall]

Gonna actually make me cite chapter and verse, eh? Well, good; that's the way to keep things honest.

From the MySQL manual, section 1.4.3.1 Using the MySQL software under a Commercial License:

You need a commercial license:
(other reasons removed)
When you distribute a non-GPL application that only works with the MySQL software and ship it with the MySQL software. This type of solution is considered to be linking even if it's done over a network.

So, IE would be off the hook, because it is not designed to only talk to MySQL. However, if I were to write an application and designed it to store configuration information in a remote MySQL database (my installer installed the MySQL database on a remote node to ensure that no MySQL APIs were involved), I would require a commercial license for MySQL if I did not want to GPL my program, even though I'm just using telnet to access the MySQL database, according to MySQL.

Re:cool

[h4x0r-3l337]

Even if they had supplied object files for the proprietary bits (which would be the accepted thing to do had this been LGPL and not GPL), or provided some other way for you to build the complete software, there still remains the "small" issue of how to get the software onto the device. There is nothing in the GPL that specifies that the manufacturers of devices that use GPL software need to make it possible for you to actually get your replacement software onto the device.

Re:cool

[Lehk228]

if they use the kernel to check the signatures it's not very secure, as a hacked kernel won't see itself as hacked, if the BIOS checks the signatures that would be a much better system

Odd caveat

[Shaper_pmp]

"Do not replace or add any software to the DISH 921 DVR with items compiled from these source trees. Doing so will void all warranties and cause the unit to fail.'""

Is this the normal "no user-serviceable parts inside" caveat, or does it suggest that they, in fact, haven't released all the modifications to GPLed code in their product?

Re:Odd caveat

[lachlan76]

Probably that they have included proprietary drivers.

Re:Odd caveat

[Anonymous Coward]

It's the latter. The 921/721 kernel does checksumming on various binaries upon mounting the root partition to make sure they haven't been modified, and that part of the code is MIA.

Re:Odd caveat

[Anonymous Coward]

I doubt they're giving away the details of how to decode the satellite portion, so you're going to be left with something that doesn't replace the original.

Unless FOSS is putting up their own satellites, it seems like a reasonable restriction given the cost of running their network.

Re:Odd caveat

[Anonymous Coward]

They are just trying to warn people that they will not get help if they screw things up for themselves. Manufaturers put the same type of warning stickers on diskmans and vcr's, but they are "servicable" with a little know-how.

Re:Odd caveat

[DaHat]

That and the fact that they look to be releasing only their Linux kernel, a few basic utilities and games... not the actual high level DVR software.

From the looks of it (and not surprisingly) there is no code related to the actual DVR functionality, nor interfacing with 8VSB or QPSK demodulators.

Re:Odd caveat

[rdc_uk]

This means what they said; their changes to GPL code rely on proprietary code. They don't have to, and hence have not, released the proprietary code. Therefore, you can't make a working system from whats relesed.

Example:

if this is GPL code:

int func_a()
{
return 3;
}

and this is your own PRIVATE code:

int func_mine_all_mine()
{
return 3003;
}

and you change the GPL code thus:

int func_a()
{
return (fun_mine_all_min() % 3);
}

You have to release your changed func_a to the world. But you DON'T have to release ANY VERSION of func_min_all_mine to anyone. And without that function, the new version of func_a does nothing but cause a compiler error.

Re:Odd caveat

[MrLint]

Proprietary check-summing? Perhaps more research is in order?

Re:Odd caveat

[Anonymous Coward]

A custom app written in C that uses no GPL'd code does not fall under the GPL.

Re:Odd caveat

[file-exists-p]

I am not a GPL expert, but what you are saying is likely to be incorrect. This would be a major loophole in the GPL (basically, put all your changes in a file my_changes.c, link with it and never give it).

I think you can not distribute a binary derivated from a GPL source code and which is linked (statically or dynamically) with a closed library. There may be exceptions though (or lot of GPL things would not dynamically link on proprietary UNIX).

--
Go Debian!

Re:Odd caveat

[resiak]

No, that's not true. If you're calling a function like that, you have linked the GPLed code to your homebrew code, and so the GPL's "viral" nature kicks in --- if you release the binaries to the world, you must release your own code under the GPL to anyone to whom you supply binaries.

The example you were looking for is something like the following:

#!/usr/bin/perl
# This file is covered by the GPL, blah blah.

sub frobnitzigate {
#return "No frobbing possible!";
return `a-binary-we-dont-want-to-gpl`; # Added by Evil Company(TM)
}

# blah

Now you don't have to release the source to a-binary-we-dont-want-to-gpl --- this is what the GPL calls "mere aggregation", and is (probably) what has happened with the DVR stuffs.

Re:Odd caveat

[po8]

Nope. Read your GPL carefully. If func_mine_all_mine() is part of a GPL-ed program, you'll have to release its source anytime you release the binary to that program.

Now there's some controversy about what counts as "part of a GPL-ed program." Shared libraries? Kernel modules? Driver scripts? RMS and the FSF legal team, Linus Torvalds, and others have all released opinions on this, but so far I'm not aware of any court cases that have defined these limits legally. It may be that DN's lawyers think that they have a legal case for keeping some code out.

Re:Odd caveat

[iCEBaLM]

If they're both included in the same program you do, ie: if the binary compiled by your source includes both functions then the source to both functions must be released.

Re:Odd caveat

[Lord Kano]

As someone who used to work for that company, I'd be willing to bet that the DVR will simply refuse to work if unblessed software is loaded.

The original DishPlayers *COULD* be upgraded with much larger hard drives than they originally came with. The company found out that people were doing this and updated the software so that the units would refuse to run if a hard drive larger than the original was in place.

They said that it was about compatibility issues, I suspect that it was because they wanted to force people to buy the newer, more expensive DVRs.

LK

Re:Odd caveat

[Anonymous Coward]

My opinion on this topic is that if I can't get the "non-GPL" parts in a functional product from you which doesn't contain the GPL parts, then every part of your product is subject to the GPL. Note that this doesn't apply to normal Linux-only programs unless they are exclusively distributed with Linux.

Re:Odd caveat

[Anonymous Coward]

That can be your opinion, but it's not a valid reading of the GNU license. The license doesn't talk about "functional products". It explicitly allows you to put GPL and non-GPL stuff on the same media/etc (see the text for the exact definition). LGPL stuff lets you link with the library as well. MPL lets you add files to a library/program that you don't have to release. BSD requires you to list their copyright/disclaimer. (30,000 foot overview of the licenses)

Re:Odd caveat

[Anonymous Coward]

A custom app written in C that uses no GPL'd code does not fall under the GPL.

Correct, if it's truly standalone. But if it links to GPL code then it does fall under it.

Re:Odd caveat

[Anonymous Coward]

It is my interpretation of "a work containing the Program or a portion of it". Firmwares are not mere aggregations. The stated point of the GPL is to allow users to modify the software. If there is no platform to run a modified version of the software on, then the user cannot use his rights. If firmwares were mere aggregations of GPL and proprietary code, then the user could modify the GPL parts and have a working system afterwards (provided he knows what he's doing). That is not possible with some of these source releases, so they are not in compliance with the GPL.

Re:Odd caveat

[Anonymous Coward]

Not the same AC, still you are missing the fact that the concept of a "working system" is not stated in the licence, nor is it part of the copyright law which the GPL's combined works provision depends on.

The FSF could have explicity stated what is allowed and not allowed, but then you would have a "EULA" rather than a "copyleft" licence.

Re:Odd caveat

[Anonymous Coward]

Hey, ever thought of the cost of writing the software which they're using without complying with its license? If they want to keep their system software secret, they should have written their own, damnit.

Re:Odd caveat

[Anonymous Coward]

That's why I stated that this is my opinion, my interpretation. The GPL just calls it a "a work containing the Program or a portion of it" and adds some rather vague limitations to that concept. What does and does not constitute a derivative work under the GPL is subject to interpretation, as can clearly be understood by reading the comments to this and every other GPL story. The preamble however states that the intent of the license is to allow users to create working derivatives of the work. This stated intent limits the interpretation of the aforementioned term. Interpretations which do not satisfy the intent are not acceptable. Therefore, a product which does not have any useful function without the GPL parts is, in my opinion, a derivative work and thus entirely subject to the GPL.

Re:Odd caveat

[jnik]

The exception is that libraries which are normally distributed with the operating system are OK to link against, even if they're closed source.

Re:Odd caveat

[TheKidWho]

no...

You can link to gpl code, but as long as the gpl code isn't inside of your code you don't have to release it at all.

Re:Odd caveat

[temcat]

When you're linking to GPL software, you effectively "have" the GPL code inside yours. So "no" to your "no".

Re:Odd caveat

[swillden]

so the GPL's "viral" nature kicks in

To be precise, it's copyright law's "viral nature" that kicks in. The GPL just doesn't give you a pass in this circumstance. Because the GPL grants other rights that are reseved to the copyright holder under the law, people sometimes wish it would in this case also, but it doesn't. By design.

if you release the binaries to the world, you must release your own code under the GPL to anyone to whom you supply binaries.

Actually, that's not your only option. You have at least two other options. One, you can contact the author(s) of the GPL code and ask them for another license that allows you to do what you'd like. Two, you can ignore the issue and simply infringe on the authors' copyrights. This option doesn't sound like a very good one, because it's likely to get you sued, but it's important to mention it because even if you do get sued you will still not be required to release your code. Usually, the owner of the GPL software will happily drop the suit if you simply stop infringing.

Even if the copyright holder wants something more, he/she/they is/are unlikely to get it, because proving that you caused financial damage is going to be very hard given that the software typically costs nothing.

Now, if your business is dependent on continuing the distribution, you may decide that it's better to release your source code rather than stop, but the choice is yours. Assuming you don't keep fighting until a judge is forced to make the decision for you. Who knows what the judge might decide to do? Probably depends on how much you irritated him during the trial.

Re:Odd caveat

[Anonymous Coward]

You miss the point. "Stated intent" has no bearing on copyright law. (And the intent here is arguable - is the intent to be able to modify the thing the software is part of, or is it to guarantee that modifications to the software are made public, or...)

Your interpretation is irrelevant. The legal interpretation is what matters, and that's also what the FSF lawyers assumed when they wrote it. They absolutely knew that this was a possible outcome (though their thinking was much more towards large computer systems where users run commands and executables, not embedded systems, so some of the wordings can lead to confusion among lay people).

Re:Odd caveat

[Random832]

Even if the copyright holder wants something more, he/she/they is/are unlikely to get it, because proving that you caused financial damage is going to be very hard given that the software typically costs nothing.

No punitive damages for copyright violation?

How about claiming you intended to charge $1000 per-seat for non-gpl licensing deals

Re:Odd caveat

[Alsee]

The 921/721 kernel does checksumming on various binaries upon mounting the root partition to make sure they haven't been modified, and that part of the code is MIA.

So what? The trivial thing to do is to stub out that test with a fixed "checksum-sucessful" result and simply load any existing files or your own replacements.

-

Re:Odd caveat

[Anonymous Coward]

That doesn't mean they haven't "released all the modifications to the GPLed code". The checksumming could include proprietary binaries that were written in-house, or is sensitive to being recompiled with a different compiler, different optimisations, etc.

Re:Odd caveat

[Anonymous Coward]

actually I bet that the encryption is probably some standard form of a well known scheme - probably 3DES. What is important is how the key exchange is done - their decryption is probably done by some proprietary hardware and it's the software for talking to that that they care the most about

Re:Odd caveat

[shutdown -p now]

I wonder if such functionality means that the kernel constitutes a single derived work together with the binaries (since it won't work without those very specific binaries). Which, in turn, would mean that they have to release the kernel source under GPL as well.

Re:Odd caveat

[Anonymous Coward]

Your interpretation directly conflicts with that of the FSF (never mind others). Among other examples, they note that a application program can run of top of a GPL'd OS without being forced underneath the GPL. You have to admit that a binary executable built to run on a particular OS is pretty darn useless without that GPL'd component. Yet the FSF note this as an example of a work that is not considered (by them) to be derivative.

Since you're not the author of either the GPL or the code in question, you're welcome to your own opinion, but it hardly matters. The "still works" test has no bearing on the GPL. Perhaps you should write your own license.

The GPL needs a lot of work to clarify the definition of "derivative work". Static linkage is the only clear case, and it is clear simply because the binary contains sequences of bytes directly derived from the GPL'd binary. The source actually has nothing to do with the derivative-ness; it's the fact that the linker has copied in chunks of copyrighted code, and so made a copy of parts of it.

All the other cases are an increasingly fuzzy spectrum. The FSF says apps on a OS are not derivative. However, they also claim that dynamically linking to a library is. But there's no real technical difference in the linkage methods here. Nor is there a lot of difference with using a form of IPC to invoke behavior from a server, another case excepted for being "derivative", yet every bit as dysfunctional in the absence of one of the components.

Re:Odd caveat

[Anonymous Coward]

I tried to clarify this by mentioning that programs are distributed without the operating system and can work (are in fact designed to work) on systems which have not been supplied with the product. Programs in firmware files are never distributed without their GPL environment and, as can be seen in this case, the user has no chance of modifying the code and ending up with a working product, unless the whole source is provided. A derivative work, in my interpretation, is something where you cannot (with reasonable effort) modify the GPL parts without rendering the product non-functional. This is a good measure of the independence of the two parts and since there is no legal precedence, it should be as valid an interpretation as any other.

Re:Odd caveat

[Faw]

I have checked DishLinux for a while (I wanted to add software of my own to the 721) but unfortunately it seems every executable is encrypted with PGP. The loader decrypts it and runs it.

Also there is a file in the root partition that seems to be a list of files on the disk, also PGP signed. When the system boots it checks and if there is any differences between the list and the disk and if there are any it does a full reinstall.

If anyone wants to know more go here and here . These are 2 threads on SatelliteGuys detailing what I found out while replacing the hard drive.

Re:Odd caveat

[MrLint]

well unlike most SW in embedded devices, it looks like these guys did their homework. So in essence, the only thing to do is a ground up rewrite, including handling of any evil(tm) sent down form the satellite.

Well...

[Paris The Pirate]

Some geeks out there are gonna be hella busy tonight.

Sounds like reverse psychology

[DrugCheese]

Don't try and make some code that'll fit in place of our proprietary code - YOU CAN'T DO IT!

Re:Sounds like reverse psychology

[hhawk]

It would be of course interesting to know if they are holding something back or just keeping their own code private... but doesn't things like the DMCA prevent "the community" from taking a look...

If it is something private, non-GPL code is it likely to be something small but critical like some driver? or is it likely to be something resulting from 1000's of staff hours..?

Perhaps we will never know.

Linux

[ErichTheWebGuy]

Apparently the DVR runs Linux under the hood?

GPL is viral no?

[Anonymous Coward]

I didn't think that you could mix proprietry and gpl code? How are they releasing parts of a source code tree - I thought it was all or nothing if you include GPL in your source?

I think there needs to be a monetary penalty for this sort of messing with GPL'd code.

Re:GPL is viral no?

[Ed Bugg]

I didn't think that you could mix proprietry and gpl code? How are they releasing parts of a source code tree - I thought it was all or nothing if you include GPL in your source?

Well kinda... What you do is write an abstraction layer, kinda of a hook into your proprietry software. You'll need to release the abstraction layer (The functions and routines that call GPL routines and use the variables from the GPL modules) but not the proprietry stuff.
That is perfectly legal (not neccaryly ethical) and many companies do it.
It would look something like this (assume code sprinkled in different source files

gpl_code() {
do stuff here;
abstract_api1();
}

abstract_api1() {
proprietry_func();
}

abstract_api2() {
do other stuff here;
}

proprietry_func() {
abstract_api2();
}

Re:GPL is viral no?

[ratboy666]

More to the point...

Vendor has custom hardware, that (say) has flash on it.

Decides to use Linux as the OS... The flash is preloaded with an updater and linux.

Of *course* the vendor puts the source for Linux on her site -- why not?

What is kept proprietary are the tools needed to generate and sign the flash load.

Now, you COULD try to reprogram the flash, by "cracking" the Linux installation, and reprogramming the hardware. But, if it breaks -- you own both pieces.

Ratboy

Re:GPL is viral no?

[icepick72]

It's ethical if you want to protect your IP. Also, the open source code project gets back all modifications to the code base. I hope the next version of the GPL would formerly recognize this relationship. The interests of both parties are protected.

Re:GPL is viral no?

[NovaX]

I don't believe that is exactly correct. You would still include GPL code in the proprietary code, just seperated slightly, but not as a seperate process. That would work for the LGPL, but not for the GPL. See this

At one end, the "mere aggregation" certainly makes it safe to ship GPLed software on the same media with your proprietary code, provided they do not link to or call each other. They may even be tools operating on the same file formats or on-disk structures; that situation, under copyright law, would not make one a derivative of the other.

At the other end, splicing GPLed code into your proprietary code, or linking GPLed object code to yours, certainly does make your code a derivative work and requires it to be GPLed.

It is generally believed that one program may execute a second program as a subprocess without either program becoming thereby a derivative work of the other.

The case that causes dispute is dynamic linking of shared libraries. The Free Software Foundation's position is that if a program calls another program as a shared library, then that program is a derivative work of the library. Some programmers think this claim is overreaching. There are technical, legal, and political arguments on both sides that we won't rehash here. Since the Free Software Foundation wrote and owns the license, it would be prudent to behave as if the FSF's position is correct until a court rules otherwise.

Re:GPL is viral no?

[Anonymous Coward]

Not that I don't agree with you, but you are missing the yackfucker's whole point.

He doesn't believe you have the right to keep your IP private. He thinks that all your IP should be open to the public. To him, it is unethical to keep it a secret.

He knows the rest of the world disagrees with him, so he drafted the GPL to allow like minded folk to start a mini-revolution. If I write some cool shit, and GPL it, and you want to use some of my cool shit, in your even cooler shit, great, but then you need to publish your cool shit to the world, and it grows. The point is, that sooner or later, if enough "starter cool shit" is out there, it would be virtually impossibly to write code that wasn't infected with the GPL, and so sooner or later, all source would be open.

This is the main difference between the yackfucker and other licenses like BSD style licenses. He doesn't want you to share, he wants it all. The BSD wants you to share what you want to share, and not share what you don't want to share.

Stated differently, Yackfucker doesn't believe in sharing, he is selling his IP for your "code futures", or probably better stated as "code options". BSD believes in sharing; once you have it, you are free to do whatever.

Re:GPL is viral no?

[Rakshasa Taisab]

I always wondered why so many companies managed to violate the GNU GPL... Now i understand, they read /.!

Re:GPL is viral no?

[KarmaMB84]

They could use proprietary modular drivers to get around the GPL.

Re:GPL is viral no?

[icepick72]

and so sooner or later, all source would be open

Hmm. That's a very interesting way of describing the situation. I like it. It provides a very high-level point of view that I had not been forced to think about before. Thanks. You know, I don't know if I'm that fond of the GPL after all. I think I would rather both worlds have their way, instead of one license being viral and forcing its way. Of course then I would have to argue equally against completely closed source software for the same reason. Very interesting indeed. I had better go and do some more learning about the open source licenses instead of relying on opinions (the GPL seems to be favoured throughout Slashdot).

Re:GPL is viral no?

[swillden]

That's a very interesting way of describing the situation. I like it. It provides a very high-level point of view that I had not been forced to think about before.

Keep in mind that there's a serious flaw in the AC's argument. Specifically, he said "The point is, that sooner or later, if enough 'starter cool shit' is out there, it would be virtually impossibly to write code that wasn't infected with the GPL". The last bit, that it would become impossible to write non-GPL code, is stated as though self-evident, but it's not, for at least three reasons.

First, it is possible for proprietary code to use GPL code to do "cool shit" without being a derived work. Any argument otherwise would likewise mean it's impossible to write a Windows program that isn't a derived work of Microsoft Windows. Yes, this may violate the spirit of the GPL but, IMO, it's important.

Second, it postulates a world in which there exist technologies that are only implemented with GPL'd code, and which are also too difficult for anyone to reimplement in a commercial setting. Given that complex, difficult code is easier to create in a commercial setting (because it's easier to focus a group of people very intently on one task when they're being paid for that task, and don't have to worry about how to support themselves), that's hard to accept.

Third, it ignores the fact that it is often possible to acquire a commercial license to GPL code, depending on the history of the code in question and the idealism of its owners.

I think I would rather both worlds have their way, instead of one license being viral and forcing its way.

This is a misapprehension on your part. The GPL is no more, or less, viral than any other copyright license, because it is in fact copyright law which is viral. Copyright law says that as the author of a particular expressive work, I have certain rights with respect to how that work is used. One of my exclusive rights is the creation of "derivative" works, where "derivative work" may be loosely defined as "any work that includes a copy of my work".

So, if I give you a copy of my work, you have no right under copyright law to create a derivative work unless I give you that right. Further, even if I give you permission to create a derivative work, because it includes my work, you cannot create additional copies without my permission.

None of that has anything to do with the GPL, it's all in copyright law.

What the GPL does (if I choose to license my code under it) is provide you with some of those rights, but it carefully avoids giving you the right to distribute my work under terms other than the GPL, allowing copyright law's basic restrictions to shine through in that case.

Compare this with a commercial source license. Suppose you got a copy of the source code for Windows XP, as part of a shared source license or whatever, and further suppose you decided to make and publish your own operating system that included some of Microsoft's code. What would Microsoft do? Microsoft would also rely on copyright law's restrictions to prevent you from doing that.

I had better go and do some more learning about the open source licenses instead of relying on opinions

Learning is always a good idea. In addition to the licenses, you should also read the law, since it provides the subtext to all of the licenses. It's Title 17 in the US Code, available at copyright.gov. Although it's long, you actually don't have to read all of it; it has a table of contents and it's pretty easy to see which parts matter.

Dragged kicking and screaming into the light...

[pieterh]

Some of the ways companies try to avoid complying with the spirit of the GPL, even if they comply with its letter:

1. The "Vogon" strategy... the source code is available on the web site but you have to go down the stairs, look in the bottom shelf of a filing cabinet behind a locked door with a sign "beware of the leopard" on it.

2. The "Proprietary pieces" strategy... the source code is released, oh yes, but with significant pieces missing.

3. The "Under development" strategy... coming soon folks, as soon as we get it ready.

All these are quite hard to sustain.

But what really amazes me is how slow companies like Dish are to understand the benefits that the GPL brings them. They are building on top of commodity software. They have access to hundreds of skilled engineers at little or no cost. These people ask nothing better than to act as a volunteer R&D department, in exchange for appropriate credit and possibly some long term kudos.

But no... instead we get these "compliance" releases, basically useless.

The key is this: if you are selling a device and your software is GPLd, you have created a platform and you can potentially sell 10, 100 times more if you provide a decent product at a reasonable cost.

Not only does it make excellent business sense to re-release improvements to GPL'd software as cleanly and transparently as possibly, but it makes sense to release proprietary software exactly the same way.

Re:Dragged kicking and screaming into the light...

[js3]

I find it a bit strange that a company that chooses to use open source code wants to hide the source. If they wanted to hide the source why didn't they just make their own code instead of using someone elses?

Re:Dragged kicking and screaming into the light...

[BiggyP]

the vogon strategy you suggest doesn't sound terribly vogon to me, more like terrestrial bypass planning.

Re:Dragged kicking and screaming into the light...

[Anonymous Coward]

But what really amazes me is how slow companies like Dish are to understand the benefits that the GPL brings them. They are building on top of commodity software. They have access to hundreds of skilled engineers at little or no cost. These people ask nothing better than to act as a volunteer R&D department, in exchange for appropriate credit and possibly some long term kudos.

What make you think Echostar wants outside help? What would they have to gain? We're not taking about a commodity hardware here; we're talking about a box that is useless without a satallite subscription.

Consider the heavy amount of security in this box. Signals from the satellite are encrypted. THey are decrypted by the box based on the user's subscription. The unencrypted stream is then re-encrypted with a key unique to the box before being stored on the hard drive. They've gone through a lot of trouble to make damn sure nobody can get access to the unencrypted data stream.

Given how locked down this box must be, why would Echostar ever want to encourage 3rd party hacking of it?

Re:Dragged kicking and screaming into the light...

[Bloater]

The problem is they are building on top of commodity hardware, with the only advantage being the software. A competitive market has no room for openness. In such a market, every dollar you grow is a dollar that your competition shrinks. You don't want that to happen the other way around. For everything you do that helps your competitor, your competitor grows a little faster. That means you grow a little slower or shrink.

Embedded consumer entertainment appliances are a very, very competitive market. That's why prices are so low, and there are so many small players. All those niche's have converged to one market leaving it over-crowded and over-supplied. Nobody can be expected to give up their share of that market without at least a fight.

Re:Dragged kicking and screaming into the light...

[pieterh]

You are suggesting that because (e.g.) the stream must be secured, that the user interface cannot be improved upon, translated, whatever?

The two issues can be separated, more or less easily. A secure application can run on top of an open platform, and vice-versa. It requires a clean API, documentation and run time binding (rather than build-time binding). All perfectly feasible. This is why, for instance, I can run Linux on proprietary hardware, something we all take for granted, but which is fundamentally exactly the same issue.

Further, a clean separation between the two makes GPL compliance easy while still encouraging people to play with and improve the open layers.

Re:Dragged kicking and screaming into the light...

[WhiskerTheMad]

Damn I wish I had mod points right now...

hehehe... Vogons... hehehe

Re:Dragged kicking and screaming into the light...

[Anonymous Coward]

You are suggesting that because (e.g.) the stream must be secured, that the user interface cannot be improved upon, translated, whatever?

The two issues can be separated, more or less easily. ... if and only if your security is 100% perfect.

I am a long time Echostar customer, and over the years I've seen them take ever increasing measures to lock down their boxes. They (or their content providers) are completely paranoid on this issue. The DVR is question, the 921, initially came with firewire ports. Some content provder(s) complained, and Echostar deactivated the ports. Subsequent production runs omitted the firewire port completely.

Other posters have already covered the business reasons to not open up their DVRs. I just wanted to highlight the security reasons for them not to.

Re:Dragged kicking and screaming into the light...

[|<amikaze]

Survey says... Because it's cheaper!

Re:Dragged kicking and screaming into the light...

[technothrasher]

he problem is they are building on top of commodity hardware, with the only advantage being the software. A competitive market has no room for openness.

Exactly. The thing I've always wondered about is why all these companies aren't using *BSD for their OS. There's lots of support for embedded BSD and it makes this whole GPL problem go away.

Re:Dragged kicking and screaming into the light...

[geminidomino]

Can't be. There were no stairs into the cellar...

Re:Dragged kicking and screaming into the light...

[mqx]

"But what really amazes me is how slow companies like Dish are to understand the benefits that the GPL brings them. "

The GPL does not bring benefits to all companies.

"Not only does it make excellent business sense to re-release improvements to GPL'd software as cleanly and transparently as possibly, but it makes sense to release proprietary software exactly the same way."

It does not make good business sense when a cheaper competitor takes your work and sells a product at lower cost, in what is a market where cost prevails.

Your comments sound very "pro-GPL": in reality, software gives many businesses a competitive edge, and that's gained through the cost spent on making that software happen: businesses are relucant to give that away "for free" at the risk that a competitor will take the changes (at no cost) and produce a cheaper product with better margins.

For many businesses, GPL makes great sense as a consumer (i.e. to take the technology and use it), but usually less sense as a producer (to then give away the changes for free). Some businesses are smarter and realise how GPL can act as a strategic advantage: for example, Asterisk PBX: GPL has enabled a world-wide community of businesses around building and installing small to medium size PBX's.

From the few (two) products I have that were GPL based (and I obtained the Linux distributions for them), both were missing the source to key loadable modules and userland programs/tools: so it was possible to customise the platform, but not the proprietary functionality. This really doesn't do anything for the wider community, and certainly doesn't offer the total promise of "free software". About the only thing useful from these were a few low level drivers (clocks, eeproms, etc).

Honestly, from my point of view, the promise of free software has easily been defeated by manufacturers, and in theory, the GPL sounds like it offers a better outcome that BSD, but in practice, they pretty much well achieve the same results.

Re:Dragged kicking and screaming into the light...

[Anonymous Coward]

What you write is 100% correct, but it isn't a mitigating circumstance for GPL violation.

If the only way that they could have gotten commodity hardware to market with a competitive advantage was by linking directly to GPL code, then they have to release the source of their code too by the terms of the GPL license, end of story.

If releasing it means that they no longer have a product, fine, it means that they never had a standalone product of their own to sell in the first place. The GPL has allowed them to make some money quickly, and that's great, but now it's time to pay the piper.

TANSTAAFL

Re:Dragged kicking and screaming into the light...

[hedora]

How is this unlike the computer market? In fact, aren't products like Windows MCE or MythTV trying to take out a big chunk of the DVR market by providing a *platform* that runs software the user chooses?

If I want to install old atari or nintendo console games (or even ascii-art doom) on my cable box, I can do that today with MythTV. How long will it be until Dish's closed, non-consumer-servicable platform supports that sort of thing? What about the other 20-30 apps I'll want to use it for in the next 2-3 years?

I feel like the vast majority of these 'embedded consumer' hardware manufacturers have their heads in the sand. Unless they open their platforms, PC manufacturers will eventually put them out of business.

Re:Dragged kicking and screaming into the light...

[Anonymous Coward]

Right. Where is the poetry?

Re:Dragged kicking and screaming into the light...

[Anonymous Coward]

The GPL does not bring benefits to all companies.

Then they shouldn't have used GPL code. If they did use it, it's because it *DID* bring them some benefits that other code (or writing it themselves) would not have.

So, they did benefit from it, and now it's payback time.

Re:Dragged kicking and screaming into the light...

[Anonymous Coward]

What make you think Echostar wants outside help? What would they have to gain?

You misunderstood the parent. Echostar may indeed requuire no outside help, but they took it anyway, by using GPL code. And now it's payback time.

The fact that they didn't really need to use GPL code is irrelevant. They did.

Re:Dragged kicking and screaming into the light...

[torpor]

The key is this: if you are selling a device and your software is GPLd, you have created a platform and you can potentially sell 10, 100 times more if you provide a decent product at a reasonable cost.

Please provide examples of companies who have done this?

Re:Dragged kicking and screaming into the light...

[Bloater]

I agree, I was responding to the question of why they don't *want* to join in the community.

Re:Dragged kicking and screaming into the light...

[Anonymous Coward]

A competitive market will require openness if consumers get used to being able to modify their equipment, and I think this becomes more and more likely as time goes by. You may scoff, but look how big Napster got, and how much influence it had, and that was possible only because people could modify their PC to add that functionality.

In such an "open competitive" market, the GPL would be vastly preferred to BSD and similar by companies (for their own source code, that is - for importing code, it doesn't matter, but at least then they have the choice of using GPL code as a basis for their own if they're already planning on releasing their code as GPL). If they decide to release source code because there's a market demand for that, they don't want a competitor to come out with a closed-but-cheap product based on their work. The GPL forces a level playing field for all competitors, allowing the companies to be open and the customers to gain the advantage. Now companies compete on the basis of service and quality. Only shoddy companies should oppose that.

Re:Dragged kicking and screaming into the light...

[Anonymous Coward]

In such a market, every dollar you grow is a dollar that your competition shrinks.

This is assuming that it's a zero-sum game, and that the market is saturated. There is always the possibility of attracting new customers to the market by offering something that wasn't available before. Even if you competition follows suit, nobody need shrink. How many geeks bought an XBox or a PS2 to run Linux on it? Those were bonus customers.

Other than that though, your basic premise is true. How do you compete with someone who can offer everything you do, but without paying to develop it? Maybe the answer to that is that there is not a lot of money to be made in consumer electronics anymore, and a lot of the companies should leave the market. Leave us with a few top-end companies who pay for development by selling top-quality products, and a few cheap knock-offs who repackage their code on commodity hardware a few months later.

Nobody can be expected to give up their share of that market without at least a fight.

This is the dilemma. If someone has to leave the market, who is it going to be? Are they going to fight SCO-style or Apple-style? Apple has opened up new markets, and they maintain their foothold by having customers who LIKE them.

Re:Dragged kicking and screaming into the light...

[Anonymous Coward]

In particular, I'd like to see examples where someone can sell 100 times as many instances of dedicated hardware.

Xboxes are lousy examples because those are computers, and moreover the price on those is subsidized by the game sales. People want to use the Xbox for something else because they get a general-purpose computer at below fair market price, not because the software is open.

Computers are intended to be general-purpose machines. Satellite TV decoders are not. You're not suddenly going to create 100 times as many TV viewers, nor are people suddenly going to decide that they need a multi-hundred dollar processor to run their washing machine just because some guy posted a hack on Slashdot to do so.

So, forth with the examples showing how open source managed to transform a pacemaker into a automobile engine controller, or a house thermostat into flight control avionics for the A321, resulting in two orders of magnitude more sales of giant jetliners.

Re:Dragged kicking and screaming into the light...

[anon mouse-cow-aard]

On any open platform, you put in a little library between the GUI and the security layer. It reproduces the nice clean api, passes credentials down, and content up, copying it off to another device, un-encrypted, on the way.

The only reason you can run linux on proprietary hardware is because it isn't illegal to reverse engineer it. The DMCA coupled with the coming trusted computing platform architecture "fixes" that.

Re:Dragged kicking and screaming into the light...

[Anonymous Coward]

Not true at all.

Closed source has one purpose only: early obsolesence, and the hope you will re-buy the next model up, given the product is designed to have a finite lifetime. That or expensive spare parts(blades), consumables(printers), or add-ons(Office) or price fixing(console games).

Companies that hide schematics, repair bulletins, or in this case software, want you to buy again, even if the cause is only a 2 cent part.

If commodity hardware open sourced ALL and everything, the the oss community would buy it ,enhance it, and the others with closed source would be 'behind'.

What commodity makers FEAR, is retalliation and punishment of NOT winning a juicy contract, because some media mougul wants a locked down box, and not something where consumers may have their way , and avoid being 'branded'.

A second aspect is where a commodity maker, has worked out how to save costs by using OSS, based on a rack design, and wants to prevent knockoffs, because it has already decided that exemplatory support is not likely to be profitable.

The non-reprogrammable/upgradable list of el-cheapo MP3 clone players on the market, Vs Apple's Ipod says plenty. Those that have OSS drivers would be stand outs. The way to take on big league players is to OSS software, quicker, faster and better. Redhat and SUSE come to mind, although UBUNTU makes an interesting case study.

Re:Dragged kicking and screaming into the light...

[Mycroft_VIII]

The pvr's have one HUGE advantage over what you suggest currently.
You plug in the power cord and a couple others, turn it on, and use it. That's all
Now to do a pc pvr using gpl'd software you have to learn enough to put together a computer sytem, install some VERY un-user-friendly software, then configure it in ways joe user doesn't have the several weeks to learn (rember joe isn't a computer geek just learning a new linux app, he's probably never even dealt with with something outside of a gui).
Perhaps in a few years the game consoles will usurp pvr's place in what they do, but not pc's anytime soon.

Mycroft

Re:Dragged kicking and screaming into the light...

[tigga]

Exactly. The thing I've always wondered about is why all these companies aren't using *BSD for their OS. There's lots of support for embedded BSD and it makes this whole GPL problem go away.

The answer is simple - if there is no MMU in processor - you can't use BSD.
And there are companies that use BSD - there is no reason to slashdot them..

Re:Dragged kicking and screaming into the light...

[tigga]

In such an "open competitive" market, the GPL would be vastly preferred to BSD and similar by companies (for their own source code, that is - for importing code, it doesn't matter, but at least then they have the choice of using GPL code as a basis for their own if they're already planning on releasing their code as GPL). If they decide to release source code because there's a market demand for that, they don't want a competitor to come out with a closed-but-cheap product based on their work. The GPL forces a level playing field for all competitors, allowing the companies to be open and the customers to gain the advantage. Now companies compete on the basis of service and quality. Only shoddy companies should oppose that.

Why do you say "the GPL would be vastly preferred to BSD"?

You are daydreaming. Companies release software under open software license only if it does not relate to it's business. Opening software to it's competitors - it's business suicide.

How about cheapier (they don't hire programmers) competitor product based on your open source product?

And it's a wishful thinking about competition based on service and quality. You've forgotten about price. There is a lot of cheap junk... That's why Wallmart and fast food franchises are succesfull.

Re:Dragged kicking and screaming into the light...

[hedora]

This is all very true, but there's no real technical barrier toward building a user friendly linux PVR. (TiVo's are linux systems, after all!!)

Once MythTV hits a stable release, you could imagine selling preconfigured systems in a quiet, shiny box, and pushing out updates with apt-get. Granted, writing a UI that allows software to be installed is a bit tricky, but the linspire people seem to have already figured that one out with their 'click and run' warehouse. The real tricky is preconfiguring software so that it can be launched and (partially) controlled by the remote control / joystick.

Re:Dragged kicking and screaming into the light...

[Slacker]

Please provide examples of companies who have done this?

Ummm, well, I'd say that the Linksys WRT54G is a good example of this. Many more of those little blue and black boxes have been sold since the firmware was opened up for 3rd party development. The Linksys hardware isn't spectacular, but the community that has built up around it is.

If you'd like a second example, I'd suggest that the Asterisk open source telephony platform has been pretty widely accepted, and has greatly benefited Digium, the company that sponsored it's development. Other hardware companies have benefited as well, but over all Digium believes it is doing better than if Asterisk didn't exist at all.

There are lots of other examples.

Those that support the GPL tend to believe "a rising tide lifts all boats" and that every business poeple are in doesn't have to be a zero-sum game.

They call this compliance?

[McGregorMortis]

"Do not replace or add any software to the DISH 921 DVR with items compiled from these source trees. Doing so will void all warranties and cause the unit to fail."

The whole point of the GPL is that users can make modify the code. If the deriviative code they have released cannot be loaded without rendering the unit unusable, then they have clearly violated the spirit of the GPL. Maybe they've found some kind of loophole, I don't know.

If the device will not work without linking in proprietary code, well, then they gots themselves a problem. But it's their problem, not the GPL's. Either the proprietary code goes, or the GPL code goes.

Re:They call this compliance?

[Anonymous Coward]

They may have violated the spirit of the GPL, but there could be a good chance that their programs are completely stadalone from the GPL'd code, i.e. something like a smart card verifier running on top of a linux base

Re:They call this compliance?

[theonlyholle]

Pretty much depends on your definition of "linking in proprietary code". If the devices boots into Linux, but there are no applications to start (because they are proprietary and so you didn't have the source to compile them), so it doesn't do anything, that's not against the GPL - neither letter nor spirit of the GPL. You still get the source for the parts based on the GPL - but you'd have to write your own application software to run on top of it.

Re:They call this compliance?

[Anonymous Coward]

I bet the device will work just fine; but then you won't be able to connect to their network, ala xbox live and the modded xbox problem.

With the size of the cable descrambler market, I can see why they are paranoid of DVRs based on derivative works.

I think i'm gonna go play some "backup" games on my xb0x... yeah...backup games...

Re:They call this compliance?

[windows]

I can't comment on the device failing as a result of loading the code.

That being said, the part about voiding the warranty is hardly unreasonable. The unit is designed to run software that DISH Network distributes with it and not other software. Since GPL software is distributed without any warranty, and loading the "third party" software voids the original warranty, you would not have any warranty on the unit. I use the term third party software in a loose sense because it could be the same software but since they didn't compile it, they don't support it.

It's a little shaky, I agree, but not necessary as anti-GPL as it initially sounds. Minus the part about causing the unit to fail, that is.

Re:They call this compliance?

[matthew.thompson]

It's only their problem if the GPL code is linked to the proprietary stuff. If they pipe stuff through or don't directly interface the two then it's not an issue.

OR at least thats how I understood the GPL to work.

You should be able to discover from the code they're released what the case is but please don't tar them until you or someone else has done the research and had it ratified.

Re:They call this compliance?

[KarmaMB84]

You can take the code and use it in your own project and even write replacements for the proprietary code and make your own DVR boxes. It would be unwise to try to use the GPL to usurp control of the hardware and their support policies. They probably don't want people running modified binaries on their boxes, damaging them, and then just writing the original disk contents back and taking the box in to get a brand new one.

Re:They call this compliance?

[pla]

The whole point of the GPL is that users can make modify the code.

And in this case, you can modify the code. It just doesn't do anything without their proprietary codebase.

Simple hypothetical scenario - They wrote a totally closed program that transfers data from a hardware device (in this case, some form of PVR card) to a file.

They then made almost trivial modifications to GPL'd software to let it use those files, effectively giving them a "real" PVR suite for just the R&D cost of a single, nnearly trivial poll-and-dump program without violating the GPL.

Going a bit further, they could even have encrypted those files, giving the GPL'd code the ability to decrypt such data, given the key. And then kept the key as part of the proprietary portion (if you use GPG, do you have an obligation to release your key to anyone that might want to read your email?).


Now, does this violate the spirit of the GPL? Most of us would agree that yes, it most certainly does. You could argue (and DN certainly would, unless they want to take the "screw you, we pulled it off, go cry in your Cheerios" approach) that they did obey the spirit of the GPL, in that you and I and everyone else has every right to use and modify the GPL part of their effort to work with any already-existing data. Of course, we have no way to get or produce those files or the key to decrypt them, but hey, our problem, right?

This particular issue will keep coming up, and no clear answer exists to solve it. Even if, somehow, we could force them to release every last bit of code to their software, that doesn't change the fact that you'd still need their firmware. And if we could even stretch the GPL to force release of that, we'd still need their hardware. And if they had to release specs detailed enough to build your own? Well, do you have the ability to produce 10-layer boards? And, getting a bit more fantastic but still in the spirit of the topic, even if you do have that ability to make your own 10-layer boards (hey, on Slashdot, it would surprise me greatly if someone here couldn't perform any given currently-possible technological feat), what if the use of technology-X required something that only its creators can possibly have? "We have found a sample of element-217, the only such sample known in existance, under our corporate HQ while adding a new corporate swimming pool. It works wonderfully for controlling fusion on any scale, from battery-sized to power-plant-sized. Here, you can have every last bit of information about our designs, but gee, too bad we have the only sample of element-217."

Re:They call this compliance?

[mqx]

"The whole point of the GPL is that users can make modify the code. If the deriviative code they have released cannot be loaded without rendering the unit unusable, then they have clearly violated the spirit of the GPL."

You are wrong. The whole point about the GPL is to "keep the software free (free as in open, available, to the community)": this "guarantee of the GPL" refers to the software only, _not_ the product that the software may form part of.

It's entirely fair for a manufacturer to state that the warranty does not apply if you start "customising" the product with new code and builds. How does the manufacturer know that your cross-compilation of the firmware didn't produce a bad image?

Re:They call this compliance?

[arkhan_jg]

It's entirely fair for a manufacturer to state that the warranty does not apply if you start "customising" the product with new code and builds. How does the manufacturer know that your cross-compilation of the firmware didn't produce a bad image?

They only get to disclaim the warranty as far as the third party product caused the fault. Say you compile a new firmware, image it, and then your DVR won't boot any more. Chances are, your firmware's responsible and no warranty for you if you can't reflash it. On the other hand, if your custom firmware is working peachily and the power supply fails then they still have to replace the power supply, because your firmware wasn't responsible.

Same with car parts and printer cartridges; orginial manufacturers like to threaten voided warranties, but they can only do so when the fault is in the third party part, or a direct result of using it.

Re:They call this compliance?

[mqx]

"They only get to disclaim the warranty as far as the third party product caused the fault."

It's not that straight forward, so I'm not sure that you're correct.

"On the other hand, if your custom firmware is working peachily and the power supply fails then they still have to replace the power supply, because your firmware wasn't responsible."

What if the new firmware used the platform in a different way so as to cause excessive power drain beyond the original design?

I'm not sure how these problems are handled, but I can easily see that there's a lot of scope for the manufacturer having to unfairly wear the cost here.

Of course, if the power supply fails and you send it back without mentioning that you reflashed the device, then fine; but if you take it into the service centre and the boot to find that it doesn't look like the original device as sold, then they may not be happy.

I don't think it's as clear as you suggest, there's a lot of grey in here.

Re:They call this compliance?

[peatbakke]

The whole point of the GPL is that users can make modify the code.

This is true.

If the deriviative code they have released cannot be loaded without rendering the unit unusable, then they have clearly violated the spirit of the GPL. Maybe they've found some kind of loophole, I don't know.

This is not true. The GPL's No Warranty section explicitly states:

THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.

In short, the GPL makes no guarantee that the code you receive is functional, only that you must release it upon request. If it doesn't run on the platform you have available, that's your problem .. but I'm sure the community would encourage you find a work around!

Regardless, it is perfectly legal for a proprietary system to checksum a GPL'd binary to ensure its legitimacy. Commercial security tools, such as Tripwire, do this already. Antivirus BIOSes will checksum your boot sectors. Heck, I'm sure there are LSMs (Linux Security Modules) out there that md5 binaries to ensure legitimacy.

Granted, you said "spirit of" ... but I think the GPL is pretty explicit in this particular case.

Re:They call this compliance?

[hesquibo]

Well, *suppose* they had some hardware-chip that would trigger a failure if it didn't get signal after xx minutes of operation. Now they themselves develop a closed source driver for that chip and include it.
I think they wouldn't have to let you see their driver but it could still be dangerous to run that chip without the driver......
IMO it's very utopic, but it would be a possibility where they would render the GPL part of the code useless and still could comply with the terms and distribute the modified GPLed code without having to "loose their property" or something (or whatever tf they think why they can't release it functional).

something like that... cheers :-S
P.S.:
> then they have clearly violated the spirit of the GPL
Couldn't agree more... but unfortunately spirits are hard to come by with material things as there are lawyers, money, ...

Re:They call this compliance?

[evilviper]

they have clearly violated the spirit of the GPL

No they haven't, the GPL just doesn't happen to work the way you want it to.

Go make your own license that forbids everything you dislike, and try to convince the world to switch over to it. Otherwise, give it a rest.

Re:They call this compliance?

[Anonymous Coward]

I don't think the "void your warranty" part is the problem, that part is clearly in line with the GPL. It isn't eve a little shaky.

Re:They call this compliance?

[Anonymous Coward]

The warranty isn't the problem. The fact that they have made "the software" into a monolithic block that can't be modified, AND that block contains GPL code, is the problem. You're absolutely right that the GPL doesn't require them to warrant the modified software, and absolutely irrelevant as well.

The GPL does require that you receive all of the code, makefiles, scripts, tools to build the binary (or the tools are already "distributed with the operating system", but that's not the case here). It absolutely does require that the source code you receive is as functional as the binary they distribute.

Re:They call this compliance?

[Anonymous Coward]

It isn't about the damned warranty! It is about the freedom of the person who received the GPL software to modify it. That means modify the binary on the device that runs the binary, and have it work, NOT just be able to have the source code and be able to port it to something else. Whether the rest of the code can be kept proprietary or not is up for argument, that's where the "mere aggregation" and "derivative work" and "runs in the same memory space" and "communicates over sockets" comes in. But absolutely clear is that the GPL portion MUST be made available to be modified and have the rest of the programming continue to work (to the extent that the modified code "works").

If the proprietary code is so closely linked to the GPL code that it fails because of a trivial change (or even fails because you recompiled it with NO change), then it is obviously so closely linked to the GPL code that it is a "derivative work", and thus the whole thing is a copyright violation unless the source code for the rest of it is released.

Re:They call this compliance?

[Alsee]

Their website says:
Do not replace or add any software to the DISH 921 DVR with items compiled from these source trees. Doing so will void all warranties and cause the unit to fail.

One way or another they are lying.

Option (1): They are in violation of the GPL.

Option (2): You can in fact compile the supplied source code and use it to replace EVERY file in the DVR which is covered by the GPL and it WILL work. You simply leave any non-GPL files in place.

-

Re:They call this compliance?

[Anonymous Coward]

that's not against the GPL - neither letter nor spirit of the GPL.

Well, I'd say it's against the spirit but not the letter. I think the distinction you are pointing out is that it isn't squeezing through some loophole or bending the words, it is fully compliant with everything that the GPL asks of them. They've met the minimum requirements, but no more.

The spirit of the GPL is about sharing to grow the community. Share and share alike. (Or the RMS view: proprietary is evil). If they are offering a few kernel modifications that don't help anyone out without their custom applications, then they aren't really helping the community at all. To fully meet the "spirit of the GPL" they would release all their code freely, including documentation and build instructions. I can understand why they wouldn't want to do that.

My guess is that they make that statement to avoid having to support hacked boxes, which is perfectly reasonable. Someone would have to do a lot of verification work to make sure of that, though.

Re:They call this compliance?

[windows]

The shaky part is that they're voiding the warranty on the hardware and the software if you install software you compile. It makes plenty of sense to not have any warranty on the software but the shaky part is that the warranty on the hardware would be voided as well. It's not totally unreasonable, but I think it's shaky.

Re:They call this compliance?

[Anonymous Coward]

"Growing the community" is not the spirt of the GPL, it is just a good side effect. The spirit of the GPL is "freedom to tinker". I should be able to modify the software and hardware I buy to fix errors, change behavior and add features. "Freedom to share" is also important, because without it, freedom to tinker becomes less useful.

"Share and share alike" isn't really part of it, though it is often the way the GPL is used. If I adapt a GPL program and sell it to someone, I have to provide that person with source code and I can't prohibit further redistribution. However, neither that person nor I is required to give my changes to anyone else. There's no "share and share alike" requirement.

Re:They call this compliance?

[julesh]

If the devices boots into Linux, but there are no applications to start (because they are proprietary and so you didn't have the source to compile them), so it doesn't do anything, that's not against the GPL - neither letter nor spirit of the GPL. You still get the source for the parts based on the GPL - but you'd have to write your own application software to run on top of it.

Or keep a copy of their applications from the originals provided and install them on your rebuilt system. This is a process which only involves making a transient unlicensed copy -- when you return it to the original device the copy is still licensed (unless you have agreed to specific terms which would terminate your license to it in this case) -- and is therefore probably legal in most jurisdictions. Check with an IP lawyer first, if you're worried, but my belief is that this is fine.

Nice to see

[Anonymous Coward]

This is a little off topic, but it's nice to see Dish with a little honesty for a change. When I started EFT service after being with them for 2 years. They turned my service off for non-payment twice after the EFT had started withdrawing funds. Like clockwork, every 2 months it was turned off. When I called them, they turned it back on, but then I got charged a reconnect and late fees.

Second time I told them not to turn it back on and I was cancelling service. I turned around and got DTV that day and canelled any EFT from dish.

Long story short, 6 months later I got a collection letter for $400. Turns out they turned the service back on and charged me for 2 months of service I didn't use plus the receiver I sent back to them. The collections still there with a note saying that it's invalid. After numerous conversations, they still swear that I owe them 2 months of service and a receiver even though I've sent them records of the receiver sent back to them via UPS and the start date of my DTV service.

I've spoken with 9 different Dish customers that have had similar cases but kept service. I also had a friend that took a job with them. They even lied to him about working around his school. He ended up having to quit because they wouldn't schedule him around his classes.

Anyone paying Dish for service deserves the screwing they're getting.

Re:Nice to see

[uofitorn]

Sounds like they're up to their old tricks to me
http://developers.slashdot.org/comments.pl?sid =149501&cid=12529298

Re:Nice to see

[Nezer]

This is an interesting account, although off-topic.

What I find really interesting is the case you describe above is *exactly* the same issues I had with DirecTV that led me go with Dish Network in the first place (it also helped motivate me when Dish started carrying the Sirius music streams but I digress). I've had no issues with Dish (except when I don't pay my bill and that's certainly not their fault).

Don't be surprised if DirecTV starts doing the same thing.

These companies are large, have highly automated and complex billing systems and often don't have employees that give a rat's ass about the customers. This is partially the fault of management and shareholders but, ultimately, is the fault of a market share that gladly hands over $30-90/month because they can't wait for the next season of "The Sopranos" to come out on DVD (hint: if it's not just as good 2 years later, it's probably not worth watching). Because of this and our need to keep up with those damned Jones' folks we (the consumer) tolerate this sort of activity.

If it really gets your goat there's always cable, or, even still, the free-to-air signals that don't cost you a dime (outside of whatever portion of taxes they get in subsidies and whatever products you buy that support the consumerism-based system). Remember this is a service you choose to pay for and you can stop supporting it at any time if you are that unsatisfied. Is television really *that* important to your well-being (especially when it is clearly causing you distress)?

Re:Nice to see

[Blakey Rat]

That always happens with big companies.

Person A has a terrible experience with ComCast so moves to Time-Warner. Person B has a terrible experience with Time-Warner and so moves to ComCast. What neither person realizes is that the odds of having a terrible experience is about the same on both services... they just had bad luck on one, and good luck on the second. But no company has 100% excellent customer service all the time.

What really bugs me is that these people, not realizing this, then go online and shout out, "don't sign up with DishNetwork! I had a bad experience!" (I guess it evens out with people who have a similar bad experience with DirectTV, but still.) Instead, they should be saving, "go with the company that has the features you want for the price you want."

Doing so will void all warranties

[a_greer2005]

warranties? warranties???? This is slashdot! WE DONT NEEDNO STINKING WARRENTIES!!!

Re:Doing so will void all warranties

[republican gourd]

You, my friend, are experiencing one of the common blunders! Besides fighting a land war in asia (assuming you are an american), you are also mixing WARRANTIES up with BADGERS!

Re:Doing so will void all warranties

[Anonymous Coward]

Ah, whatever, just wait the 30 days or so that the warranty lasts these days.

Violating the GPL

[Trevin]

Do not replace or add any software to the DISH 921 DVR with items compiled from these source trees. Doing so will void all warranties and cause the unit to fail.

This is an obvious violation of the spirit of the GPL. From the Preamble:

The licenses for most software are designed to take away your freedom to share and change it. By contrast, the GNU General Public License is intended to guarantee your freedom to share and change free software ...

When we speak of free software, we are referring to freedom, not price. Our General Public Licenses are designed to make sure that you have the freedom to distribute copies of free software (and charge for this service if you wish), that you receive source code or can get it if you want it, that you can change the software or use pieces of it in new free programs; and that you know you can do these things.

Of course, they are well within their rights to say that modifying the software will void their warranty, but they can't forbid you to modify the software, and they shouldn't rig the system so that it intentionally fails if the software is modified.

Re:Violating the GPL

[jericho4.0]

Modifying the software doesn't void the warranty, modifiying the hardware does.

Re:Violating the GPL

[Luthair]

A warranty is for manufacturers defects, you modifying the software obviously wouldn't qualify.

Re:Violating the GPL

[l2718]

There may be a GPL violation here, actually. They say:

You cannot create a working DISH 921 DVR software build without the additional proprietary code.

There are two ways to interpret this statement; I think they are honest and mean the first, but someone (not me) might want to verify that:

1. The OS for the device derives from GNU/Linux/etc and is covered by the GPL; they run their written-from-scratch DVR software on top. The code they released will compile and run, but will not give you a DVR by itself.
2. In order to get the OS to run on the hardware, or perhaps even to compile it, you need to add in the proprietary elements they are not releasing. This would violate the terms set in the last two paragraphs of Clause 3 of the GPL.

Re:Violating the GPL

[DemonUrge]

Odds are their proprietary code dynamically links against LGPL libraries, in which case they have no obligation to release their proprietary code. As long as they don't statically link against LGPL code or dynaically link against GPL code, all that they have to do is make source available for the GPL and LGPL code that they are distributing ( in the manner stipulated by the two licenses ). And if you look, most everything under /lib is LGPL.

The interesting question to me regards the status of leased appliances. If you lease them does that still constitute a distribution of GPL software? The licenses are sadly unspecific in regards to what constitutes a distribution.

Re:Violating the GPL

[jesup]

You're assuming it's a monolithic executable. I'm sure it's not. It's proprietary stuff running on top of (possibly modified and released) Linux, with some LGPL libs (released I'm sure with any mods) and some GPL executables that are part of Linux userspace.

Totally within the bounds of GPL.

Re:Violating the GPL

[GlassHeart]

Modifying the software doesn't void the warranty

Of course it does. The modified software can destroy recorded programs, stop working entirely, or in extreme circumstances even destroy hardware*. I don't think it's unreasonable at all for them to disclaim any responsibility when you do that.

The actual hackers probably know the chances they are taking, but somebody who just downloads the patched software will probably blame Dish if things break.

* One example was the IBM EGA, which had registers that can be reprogrammed to permanently damage your monitor.

Re:Violating the GPL

[Nezer]

Besides this argument being flat-out wrong I feel obligated to point out to you that software doesn't exist in a vacuum. Re-arranging the bits on whatever medium stores the executable code is, ultimatly, a hardware change.

Re:Violating the GPL

[Anonymous Coward]

To void a warrenty a company must establish:
A user caused a fault do to using the product in a way it is not designed or marketed.
You can not disclaim a warrenty over the whole product based on a user's action.

Re:Violating the GPL

[tgibbs]

The licenses for most software are designed to take away your freedom to share and change it. By contrast, the GNU General Public License is intended to guarantee your freedom to share and change free software

What the GPL specifically does not promise is that your altered software will be useful for any specific purpose. You are free to alter the software, but that does not mean that Dish is required to support hardware with user-altered software, or make it easy for users to alter the unit's software.

Re:Violating the GPL

[Anonymous Coward]

If that were the case, then modifying the (L)GPL portions of it wouldn't make the whole unit fail. That they've obviously built it as a monolithic application that can't be modified by replacing individual parts of it means the whole thing IS a derivative work of GPL code.

Even for those parts that are LGPL, it is required that you be able to modify the LGPL library and have the proprietary program use your modified version, either by dynamic or static linking. In the case of static linking, that means they must provide linkable object code for the portions that link to the LGPL static libraries that they don't provide source to.

They must provide all scripts, makefiles, and tools required to produce the binary they distribute - that would include any private key used to sign the binaries (or a way to modify the proprietary part to use a different public key, or something). They can't release source code that doesn't have the same functionality as the binary. How much of the binary can be kept proprietary may be up for argument, but those portions that are GPL clearly MUST be able to be modified and still have it work.

Re:Violating the GPL

[Anonymous Coward]

Violation of "the spirit" of the GPL isn't violation of the GPL.

they can't forbid you to modify the software

They haven't done so.

they shouldn't rig the system so that it intentionally fails if the software is modified.

Shouldn't in what sense? Legally they can do it if they want to, and the GPL doesn't forbid it.

Re:Violating the GPL

[Mycroft_VIII]

Also might depend on the specific definition of 'lease'.
I've seen that word mean everything from rent, to buy with payments the downpayment being the last one. And several even odder variations.

Mycroft

Re:Violating the GPL

[tigga]

Of course, they are well within their rights to say that modifying the software will void their warranty, but they can't forbid you to modify the software, and they shouldn't rig the system so that it intentionally fails if the software is modified.

How do you think - do they want their customer support to help cusomers who botched their software? Like somebody compiled software with wrong libraries, wrong flags, wrong compiler version etc.. And there is no installation procedure, is it?

BTW everyone is free to violate spirit of the GPL. GPL is about a freedom, right? ;)))

Re:Violating the GPL

[jesup]

I would assume their apps are dynamically linked against LGPL libs (if their apps directly use LGPL libs at all).

Realize their "warning" that it won't work may not be true in practice. (Though it violating your warrantee probably is true.) If you modify a lib and it accidentally deletes all your saved programs, it's not their problem.

More to the point, I suspect the anti-service-theft code makes it effectively impossible to replace a library - there's probably a disk and/or memory checksum that considers any change that isn't signed with their private key to be an attempt to hack the unit and steal service. This checksum process probably is outside of the app that uses LGPL (for example, burned in ROM/flash within the processor or chipset). So you can modify the library or recompile linux, etc and replace it - but the unit will probably simply fail to even attempt to start. Or the code is encrypted on the disk with a key you don't have access to.

LGPL has limits - which is part of why Stallman didn't like it to begin with. GPL has limits too; many of these things I mentioned above apply to GPL code as well.

IANAL, etc. Yes, some of these things may (depending on your point of view) violate the spirit of the GPL/LGPL. But be careful trying to decide whether they've violated the legalities of a GPL/LGPL copyright; it's not so simple, especially with devices with firmware (flash, disk or otherwise). It does sound like they're following the legal requirements of the GPL/LGPL as needed. The anti-hacking code in the processor/flash/disk probably will make it impossible to modify the code in the actual unit (and have it receive service!) You may be able to replace the code on the unit and use the hardware. Getting it to then receive encrypted broadcasts and record/show them will probably not be possible.

Just my guess. I haven't even looked at what they released.

Signed binaries?

[CdBee]

Could be the apps in the PVR are digitally signed to prevent them being replaced / infected. A home-compiled module would lack the necessary signature.

No

[Kjella]

GPL is viral no?

Look at nVidias drivers. There's nothing but a stub in the kernel. The rest is in userspace and closed. Everyone seems ok with that. And any user-space program running on top of Linux can be non-free/closed.

Just because they are required to share all code that is being linked to, doesn't mean it is complete or in reality useful. That being said, I don't know anything about what this specific code is missing.

Kjella

Re:No

[|<amikaze]

Everyone seems ok with that.

Not really. There's a quite vocal minority that is really choked that that's happening.

Even Linus is pretty sketchy about the closed-source binary drivers. http://kerneltrap.org/node/1758

Re:No

[Anonymous Coward]

So you can quarantine it, doesn't make it any less viral.

Kernel .config file

[paulproteus]

They did release the .config file used to build their kernel. If you don't want to download their whole "fLinux" source tree just to look at it, I put it up on my web space. Give it a look-over; I imagine this would be the easiest component of the device to replace. Modify the kernel with e.g. Sebek and you should be able to get a good idea of what's going on under the hood.

Also, the GNU_Source_Code.zip includes fLinux.tar.gz and games.tar.gz, so if you get the zip file the other two are redundant.

Re:Kernel .config file

[random coward]

Looks like they have video for linux enabled, but none of the drivers...
Hmmm I wonder if they modified the kernel to use a propriatary driver but didn't release it wich would be a violation of the binary only driver policy...

They also enabled a few of the kernel ciphers.

They are afraid that if people modify the dish reciever they can get free Dish Network service. If they designed the reciever correctly taht would be impossible to do. Anyone want to bet they did not design it correctly?

Anyone want to bet that the checksumming is for upgrade/installation of the software and not for running it?

Proprietary BIOS

[tepples]

A PC will not even boot without running at least some proprietary code. Upwards of 99 percent of desktop PCs come with a proprietary BIOS.

Re:Proprietary BIOS

[wasabii]

www.linuxbios.org

Re:Proprietary BIOS

[tepples]

What is linuxbios's market share?

Re:Proprietary BIOS

[millette]

Hey, mind share counts too !

Re:Proprietary BIOS

[Anonymous Coward]

That a BIOS can load a GPL program doesn't make the BIOS a derivative work, thus it isn't a GPL violation to have it be proprietary. However, if someone created a PC that would only load signed binaries, it would probably be a GPL violation to release binaries based on GPL source code that was signed, without also releasing the ability to sign your own modified binaries. The PC itself wouldn't be infringing, but the release of the binary would be.

The Apple approach...

[Jerry]

This sounds familiar.

The KDE-Konqueror team was elated two years ago with the news that Apple was going to use KHTML in their browser. Like the CEO's of new software startups that were wined & dined by Microsoft before they were deflowered of their innovative software, the Konqueror coding crew now knows what it is like "working with" a company whose motivation is not community but profit if not just plain greed.

Apple used KHTML to build their browser but made calls to their own proprietary API, resulting in source that could not be effectively back ported to KDE's Konqueror. The effect of this software 'sharing' of GPL code was that a corporation selling proprietary hardware driven by proprietary software was able to exploit GPL software WITHOUT actually allowing the FOSS community to benefit from their exploitation. The "two street" of code sharing was downhill on one side and uphill on the other ... in fact a wall. Apple's OS now includes a browser built on Konqueror GPL code but, IN VIOLATION OF THE GPL, the enhancements to that code were added in such a way that the FOSS community cannot benefit from them. This effectively turned the GPL license into a BSD license! The GPL needs to be modified to address this subtle form of code theft.

The only difference between Microsoft's "thought thefts" and Apple's is that Apple used a little more finess and polish. The RESULTS are identical, however, which puts them both in the same class -- corporate thieves.

Re:The Apple approach...

[jvd]

Yes, the community can benefit from it. Not in they way they might like, but still, they can still benefit from it. The source code is available, they can use that source code to get ideas. What Apple is doing is perfectly fine and above all, legally and morally OK.

Re:The Apple approach...

[Anonymous Coward]

What Apple is doing is perfectly fine and above all, legally and morally OK.

Go fuck yourself.

Re:The Apple approach...

[Anonymous Coward]

What Apple is doing is perfectly fine and above all, legally and morally OK.

Legally? probably. Morally? no chance. What's that in your mouth? Oh look! it's Steve Jobs's cock!

Happy Fun Ball

[flood6]

You cannot create a working DISH 921 DVR software build without the additional proprietary code. Do not replace or add any software to the DISH 921 DVR with items compiled from these source trees.

Do not stare at DISH 921 DVR, do not taunt DISH 921 DVR, if DISH 921 DVR begins to smoke, seek shelter and cover head.

Re:Happy Fun Ball

[Anonymous Coward]

I'm just glad the dish network folks haven't noticed all the code we snuck past them. The antenna array is go. And speaking of the antenna array... if you enter simply 'antennaarray' in Firefox's address bar you'll get a document returned entitled 'THIN CERAMIC FERROELECTRIC PHASE SHIFTER FOR STEERABLE MICROSTRIP PATCH ANTENNAARRAY'. Isn't that cool?

Re:Happy Fun Ball

[Anonymous Coward]

a PDF document that is

Re:Happy Fun Ball

[rob_squared]

The first rule about DISH 921 DVR is that we do not talk about DISH 921 DVR!

Would you at least...

[Kjella]

...try to get a clue, before you post? That code must be released under the GPL, hell even the LGPL as it is an internal change. There are other ways like having a kernel stub which are legal, but I think their function would be completely lost on you.

Kjella

Heh...

[Anonymous Coward]

Maybe this receiver won't suck quite so badly with open-source user input! As it is, it's a $500+ boat anchor!

Don't believe me? Go visit the 921 user support forum over at dbstalk.

Not viral, but derivative work = GPL

[nietsch]

The GPL is not viral in that everything that comes into contact with it gets ill/becomes gpl. There is absolutely no problem for a non-free program to run an a free OS, so dish can use a linux kernel for their os, and run their own non-free software on top of it. That is at least, if you regard the box a a 'regular' computer that users can run their own programs on.

However if you regard it as an appliance, where the only program the user can run is non-free, maybe then you can regard the whole appliance as a derivative of linux => GPL. Also if their proprietary code uses specific linux calls that are not offered by any other OS, one can still argue that the complete product is a derivative work.
Let's hope they are not using any netfilter code in there, otherwise Harald Welte (of gplviolations and author of said code) might give them some 'free publicity'. ;-)

Re:Not viral, but derivative work = GPL

[Anonymous Coward]

It is called a metaphor dumb fuck.

Anyways I would say that the GPL or the GNU Public License fits right in with the definition of viral.

http://en.wikipedia.org/wiki/Viral
Viral phenomena are objects or patterns able to replicate themselves or convert other objects into copies of themselves when these objects are exposed to them.

Are there headers / dynamic libs that require src?

[Anonymous Coward]

Are there dynamic libraries or header files provided in Linux(RH9) that would require commercial source code to be released to comply with the GPL? If a product has #include or -llib, does doing so require a companies source code to be released?

If so which libraries and is there a roadmap of them?

Note this question does not include statically linked libraries.

Is there a way to find them with RPM? That would be nice.

More actually:

[nietsch]

fLinux (34,784 KB)

GNU Source Code (54,281 KB)

games (4,258 KB)

So it is more like ~95Mb of source. That crud has reached nirvana even!

Re:More actually:

[Anonymous Coward]

Not really, the 1st and 3rd packages are included in the 2nd archive...

Re:More actually:

[nietsch]

Even then, it is still more ?

Re:More actually:

[Directrix1]

Kudos to Dish, can we submit open source games and such to them?

DISH 721 DVR source code

[Xeo 024]

If anyone is interested, the DISH 721 DVR source code is also available.

But, it is not a violation.

[beldraen]

They used GPL code. Anything they modified, I presume, is re-released. The fact that they call external code that you cannot have access to is your problem.

If you give people a free hammer to use any way they wish, you cannot be justified when they make a building that you are not allowed to enter. GPL gives people the freedom to use it as the people see fit as long as they do not try keep the source code to themselves. Apple has not.

What you are really mad about is that you want anything the GPL code to which it is linked to be free as well. If this is what your idea of GPL means, then I can understand why Gates calls GPL viral. Your idea is viral and Apple's proprietary code is not yours to have. KDE's team was naive to expect they would get a free lunch in return for handing out free lunches. The GPL's purpose is to keep the code free, not its use.

Re:But, it is not a violation.

[Anonymous Coward]

-1 Flamebait?

(I'm not the same person)

This is all true... how are facts considered flamebait?

Re:But, it is not a violation.

[Anonymous Coward]

Moderators, this has been modded Flamebait, while other more vulgar posts have been left alone. This shows a clear act of fanboy moderation. Please mod it back up - it's a perfectly interesting post.

Re:But, it is not a violation.

[Pollardito]

hiring OSS developers is harmful in the short run, but helpful in the long run in that it makes OSS development more attractive to others who might use it to get a job or get a better job. it essentially adds another reward for all that time spent, and that's motivational

Re:But, it is not a violation.

[Phroggy]

KDE's team was naive to expect they would get a free lunch in return for handing out free lunches.

Correction: KDE's users were naïve to expect that they would get a free lunch in return for KDE developers handing out free lunches.

The KDE developers understood the license, and its ramifications, when they chose to release their code under it. They have acknowledged that Apple is playing fair, and is completely within their rights. However, Apple has not gone as far above and beyond the requirements of the LGPL as some Konqueror users expected, and THOSE people are upset. The developers are a bit disappointed, but they're not upset at Apple.

Re:But, it is not a violation.

[Phroggy]

They are hiring OSS developers,

Like Jordan Hubbard, who is continuing to work on FreeBSD in his spare time, but also has a good paying job doing work he presumably enjoys and putting food on his table. Yeah, giving him a job was a real blow to the community.

forking OSS projects

Like forking WebCore off from KHTML to produce the first browser to pass Acid2, and inspiring interesting (if bizarre) things like the Gtk+ WebCore Project, which is a port of Apple's fork to Gtk+. Meanwhile, Konqueror has gotten a lot of positive press, for example:

When we were evaluating technologies over a year ago, KHTML and KJS stood out. Not only were they the basis of an excellent modern and standards compliant web browser, they were also less than 140,000 lines of code. The size of your code and ease of development within that code made it a better choice for us than other open source projects. Your clean design was also a plus. And the small size of your code is a significant reason for our winning startup performance...

Meanwhile on the Konqueror News page you'll see things like "ships with most of the khtml improvements which Apple supplied" and "this release ships more WebCore merges." Yeah, sure looks like a lot of harm was done there!

selling their expensive hardware

Hardware like the cheapest PowerPC system you can buy.

what they claim to be the best OS/Desktop ever.

Show me an OS vendor who doesn't claim theirs is the best ever.

Meanwhile they contribute crap back to the FOSS projects

Crap which those projects then incorporate into their future releases, because the project maintainers think it's a good idea.

A lame way to sell hardware apple.

How would you suggest they sell their hardware?

Re:But, it is not a violation.

[Anonymous Coward]

The difference between what Apple has done and DN has done is that Apple's source code can actually be modified and installed and it will work. You can change the behavior of the GPL code and make Safari use that changed behavior. According to TFA, the DN software "will not work" if you try to do that. Neither will provide a warranty for modified stuff, and there's no problem with that.

Re:But, it is not a violation.

[tigga]

Correction: KDE's users were naïve to expect that they would get a free lunch in return for KDE developers handing out free lunches.

You know it's not kosher to paint KDE's users as idiots.

Re:But, it is not a violation.

[Phroggy]

You know it's not kosher to paint KDE's users as idiots.

Well, the word I used was naïve; the word Zack Rusin used was clueless.

However, perhaps I misspoke when I said KDE users - I should have said jackasses on Slashdot. :-)

Open Source FUD?

[screwthemoderators]

Perhaps they have violated the spirit, but I don't beleive they have actually violated GPL. The requirement is to share code, not share code that can be "effectively backported." They have given proper credit to Konqueror and the developers, and that's all the "FOSS community" really asks for. This is like the "I bought her a nice dinner and she won't put out!" argument. It may be irritating, but its not theivery.

Why?

[PrivateDonut]

I know that every step towards an OSS model is a great, but can this be considered as anything more than Dish Networks trying to cover its own ass? They have released the GPL'ed bits, but are they relevent to anyone except Dish Networks (who has the proprietary bits as well)? What I am really asking is; Do they care about themselves, the community or both*? *both being mainly themselves.

While you're fuming...

[Kjella]

A "preamble" or "spirit" is not a legal obligation.

a) Apple is under no requirement to release individual patches, nor any big patch at all. Only the complete source.
b) Apple is under no obligation to give anything at all to the KHTML team (only to those who got their browser as part of a Mac).
c) The source files as they stand are obviously the "perferred form" of editing the source code internally in Apple. The GPL does not cover any other form of material, such as structural documents, documentation, bug database or any other material required to understand the source. If anything, it is an oversight of the GPL, not Apple.

In short, you are asking for kindness above and beyond their obligations. They have forked the project, and I don't see why they should have to maintain someone else's code tree. Are the *BSDs required to make compatible patches, because they came from the same source? No. I assume Apple has released everything they are required to under the GPL, and so your allegations of closing the source like a BSD license is groundless, as are your allegations of a GPL violation. If I had mod points I would mod you as -1, Troll, UID not withstanding.

Kjella

Re:While you're fuming...

[Anonymous Coward]

If I had mod points I would mod you as -1, Troll

Because, according to your own summary he makes a point you don't agree with ?

Thanks for clarifying the position on how mod points are abused on Slashdot to silence criticism on Apple, Google or BBC related topics.

Re:While you're fuming...

[Anonymous Coward]

I love the way this is modded straight up, and anything that questions Apple is modded straight down. This is the problem with slashdot, there's no real 'groupthink' as such - no matter what you say you'll get someone who disagrees with it.

The problem is that the majority of people who get mod points seem to be stupid dicks about it and use it to promote their own pathetic agenda.

Go on, mod this down too you fuckers. At least you won't spend that mod point on censoring someone else's opinion.

Re:While you're fuming...

[Anonymous Coward]

It was modded stright up because he's right.

I personally *dislike* Apple, I don't like many of the things they do, but this post was spot on as far as the KHTML situation goes.

What Apple has done is not just legal and in compliance with the GPL, it's entirely *reasonable* too. Hell, there is now a kickass browser engine available to anyone under GPL, and some people are whineing because Apple won't fix KDE's broken one for them.

Re:While you're fuming...

[10101001 10101001]

b) Apple is under no obligation to give anything at all to the KHTML team (only to those who got their browser as part of a Mac).

I wasn't aware that Apple was including the full source code of Safari every time they distributed it to users. If they were, then the KHTML team could simply d/l a copy; if one on the KHTML team owns a Mac, all the more clear the obligation.

Now, if Apple isn't releasing source code every time, then they must fall under 3b, requiring a written notice to offer source code to anyone who asks. So, if KHTML team asks, Apple is obligated to reply with the source code.

Your other two points are quite valid, though. And I'd assume (c) was partially an oversight and partially a realization that most hackers wouldn't keep much of a record. The last thing I think that would be wanted is to dissuade authors by making it even more difficult to be compliant--I'm sure at least some of my programs aren't properly GPLed because they don't print out notification at runtime or aren't properly documented when changes were made.

Microsoft Marking is viral, no?

[Anonymous Coward]

GPL is not viral. Microsoft/SCO FUD is viral.

Re:Nice to see (Helpful, but way off topic)

[Anonymous Coward]

Under the Fair Credit Reporting Act it is up to them to prove that you didn't cancel service and not the other way around. In a court case with your shipping information you have the preponderence of evidence. Send them a certified letter telling Dish they have 30-days to send you their evidence or they must cease all collection efforts. Be sure to reference the Reporting act and be sure to detail the damages you have suffered due to the lower credit score and higher interest rates they have caused. The Act specifies that recovery of damages is permissible. If they do not comply, talk to an attorney as you have a case. They count on people not knowing the rules....

www.linuxbios.org ... but

[screwthemoderators]

http://www.linuxbios.org/ You're point is well taken, however... there's no reason why they shouldn't cover their butts on this. happy hacking!

Re:www.linuxbios.org ... but

[LiquidCoooled]

the link to linuxbios appears to be dead. Can't see a reason why, although it has been down for a while (at least long engough for google to use the current nonesense page in the cache) Archive.org has a history from 2000 to 2004, but the latest 2004 ones are not connecting. Is linuxbios doomed? Is there an alternative?

Different executables, you mean.

[roystgnr]

That "abstract_api1" better be calling system(" proprietary_prog args") if you want to be GPL-compatible. What's more, you'd better be sure that proprietary_prog doesn't even include more GPL header files than fair use allows.

What you're describing is LGPL-compatible: with the LGPL you only have to release code to a library and modifications to that library. With the GPL, you have to release modifications to the entire derived work.

Now, you might be able to convince a court that sufficiently abstractly linked code isn't a derived work. I'd at least make sure nothing is statically linked in that case, but I know RMS considers even dynamically linked binary-only code to be a GPL violation - who knows what a court would say. I know Nvidia's walking on thin ice with this one - Linus considers binary modules that don't require core kernel changes to be GPL-compatible, but IIRC there are other kernel developers who disagree and just aren't angry enough to sue about it.

Re:Different executables, you mean.

[swillden]

I know RMS considers even dynamically linked binary-only code to be a GPL violation - who knows what a court would say.

Indeed. I've talked to attorneys who have studied the issue and come down on both sides of that question. And none of them wanted to try to argue it in court because they said there's simply no way of knowing what the outcome would be. Good lawyers don't like that sort of uncertainty.

Longer lasting = added value

[Alwin Henseler]

But what really amazes me is how slow companies like Dish are to understand the benefits that the GPL brings them. (..) Not only does it make excellent business sense to re-release improvements to GPL'd software as cleanly and transparently as possibly, but it makes sense to release proprietary software exactly the same way.

I think most businesses still underestimate the added value it gives to customers, if products can be extended/upgraded in 'unintended' ways. I suppose the thinking may be something like "this year, we sell you something with a fixed set of functions, so if you want something extra that we come up with next year, then you'll have to buy again". If customers can upgrade and extend functionality themselves, the product may last a lot longer, and businesses may feel they lose out on sales of newer products. And maybe business feel that added value for customers != added value for the company. Note this is a lot of speculation on my part.

What they fail to realize, is that a longer lasting product is one that is a) more appreciated, and b) is worth more (=could be sold at a higher price). It's no different from the pricing difference between a cheap model car, and say... a <insert your favourite top-of-the-line car maker here>. The latter is a lot more expensive, but still very much worth its money. Not only because of the name, but by including quality parts, more attention to detail, better service, and so forth. Both cars will get you from A to B, but the expensive model may even be cheaper... in the long run, when you drive enough miles in it. And by breaking down less often, and give you more time before you need a new car.

The 'quality' part gives a product a higher profit margin, and the 'customer satisfaction' keeps customers coming back. So that added customer value does equal added value for the company.

It just shows that few companies understand this, or prefer to go for the quick buck as opposed to keeping their customers happy for a long time.

Re:Longer lasting = added value

[The Cisco Kid]

Not only that, but the 'tinkerers' will develop the new features *for* you, that you can then take, incorporate into your own release, and sell it to the non-tinkerers.

Re:Longer lasting = added value

[InvalidError]

One major side-effect of opening sources is that it also opens your market to every wannabe out there. Since the PVR software is the only added value on top of commodity hardware, releasing the PVR software's sources would allow everybody to market their own clones.

Releasing sources for non-profit stuff is not a problem but releasing sources which represent your only value-added competitive asset in a highly volatile market is fundamentally suicidal. The commercial market place is a hostile, cannibalistic place, showing too much flesh to the local piranhas is dangerous.

Re:Longer lasting = added value

[evilviper]

Since the PVR software is the only added value on top of commodity hardware, releasing the PVR software's sources would allow everybody to market their own clones.

There are a couple big flaws in your theory.

First, the hardware isn't commodity. It is partly custom and propritary. I don't see 3rd parties selling Dishnet recievers at all, let alone PVRs.

Second, everybody is ALREADY marketing their own clones. PVRs are not a great mystery, and not terribly complex. Opening up their source code would not result in a larger number of PVR makers by any means. As Tivo has shown, it's very hard to survive in the PVR market, even if you keep your software propritary.

Re:Longer lasting = added value

[InvalidError]

Exactly.

Surviving in the marketplace is already hard enough without open-sourcing the only distinctive trait: the proprietary bits.

Once the software is open-sourced, adapting it to work with commodity hardware becomes trivial unless the software extensively used special/undocumented hardware features.

The Dishnet and other boxes might not be using 100% stock parts but their fundamental architectures are most likely awfully similar, there may be an infinite number of ways to push bits around but only a select few make practical sense.

i.e. non-compliant

[Dwonis]

You cannot create a working DISH 921 DVR software build without the additional proprietary code.

That doesn't sound like GPL-compliance to me. From the GPL:

The source code for a work means the preferred form of the work for making modifications to it. For an executable work, complete source code means all the source code for all modules it contains, plus any associated interface definition files, plus the scripts used to control compilation and installation of the executable.

Re:i.e. non-compliant

[Anonymous Coward]

Actually, it's completely compliant.

Re:i.e. non-compliant

[evilviper]

The source code for a work means the preferred form of the work for making modifications to it.

What does this have to do with anything? You are free to make modifications to the code all you want... That does NOT mean they have to give you free license to propritary software that is necessary in addition to it, and it sure as hell doesn't mean they have to make their hardware work with anything you might want to run on it.

They are meeting their GPL obligations 100%.

Re:i.e. non-compliant

[Alsee]

They are meeting their GPL obligations 100%.

They said that compiling the supplied source and replacing that portion of the code in the machine causes it to fail.

One of those two statements is false. Either they lied that it won't work, or you are wrong *and* they lied about complying with the GPL.

You are certainly correct that they do not need to give you any source to non-GPL files. You simply leave any non-GPL files in place.

-

Re:i.e. non-compliant

[evilviper]

Their words are quite vague, and I doubt anyone here has ANY knowledge about the hardware in question. There can be a great many other reasons for that to be the case, and/or they may have been overly vague about it.

It's entirely possible that the machine verifies checksums on the files, or that the propritary bits cannot be replaced seperately. We are talking about an embedded system here, not a normal general purpose computer, and everyone is going apeshit for no reason.

Re:i.e. non-compliant

[Lehk228]

if the digitally sign/encrypt their binaries, as I have said earlier in the discussion, they would be in full compliance with the GPL and accurate in saying the binaries compiled from this source would not work.

Test your linuxbios.org link

[ScrappyLaptop]

...clicking on which brings you to a page with only the message:

This Page Intentionally Left Blank

Nice link, sparky!

Looks like the main page is now on a wiki:

http://wiki.linuxbios.org/index.php/Main_Page

;)

This page intentionally left blank!

[screwthemoderators]

http://freebios.sourceforge.net/ I'm not sure why the site isn't working (faqs and archives aren't up either. What's going on there?

TO MODERATORS

[Anonymous Coward]

This post had so much controversal moderation, that is thus interesting and must be moded up.

violates intent of the GPL

[cahiha]

Do not replace or add any software to the DISH 921 DVR with items compiled from these source trees. Doing so will void all warranties and cause the unit to fail.'"

Almost the whole point of the GPL is that you can do exactly this: you should be able to change the behavior of GPL'ed software components and replace the existing versions of it.

I suspect future versions of the GPL are going to try to limit these kinds of abuses: if you distribute systems containing GPL-derived binaries, you must ensure that people can reasonably replace your GPL'ed software components with components they recompiled. You should not be permitted to use either cryptographic means, warranties, patents, or proprietary development tools to prevent that.

Re:violates intent of the GPL

[l2718]

Almost the whole point of the GPL is that you can do exactly this: you should be able to change the behavior of GPL'ed software components and replace the existing versions of it.

But they are giving you exactly this ability; they are simply advising you against doing it in practice, because they won't offer you any support if you muck things up. The GPL says you have a right to hack this code, not that you have a right for technical support while doing that.

I suspect future versions of the GPL are going to try to limit these kinds of abuses: if you distribute systems containing GPL-derived binaries, you must ensure that people can reasonably replace your GPL'ed software components with components they recompiled.

Most likely, there is no abuse here -- you can replace components of the Linux installation in any way you want. They are simply not guaranteeing that this will still work with their proprietary DVR code. Note that since GPL'd code comes with no warranty, I'm not sure what you mean by abusing warranties related to it. Certainly they cannot warrant that their proprietary code will work with whatever modified kernel you choose to put in there!

Re:violates intent of the GPL

[cahiha]

But they are giving you exactly this ability; they are simply advising you against doing it in practice, because they won't offer you any support if you muck things up.

Saying "replacing any shipping code with your own will void your warranty" would OK. But that's not what they are saying.

They are saying "Do not replace or add any software to the DISH 921 DVR with items compiled from these source trees. Doing so will [...] cause the unit to fail." No "might" or "may", but "will", implying that there is some checksumming going on there that explicitly makes the unit fail if you attempt to hack it in any way.

And it doesn't matter whether their threat is empty or not--risking several hundred dollars is too much to find out.

Re:violates intent of the GPL

[XenonOfArcticus]

I can clarify. Replacing the supplied GPL-covered parts of the 721 (and I assume 921) Linux OS with versions you have compiled from the Dish-network-released source, will fail.

The OS has been modified to prevent this. Those modifications have not been released. I had a page to cite about all of this, but it seems to have disappeared from the internet. There was a page about digging into the 721's guts, probably on DBSForums.com long ago.

It's not an idle threat. It is enforced by software. Replacing GPL code with GPL code will break the system, because the GPL'ed code you are replacing is not 100% the same as what is generated by their 'released' source. This, I believe is where the violation comes in.

Wouldn't have this problem if...

[telemonster]

they developed on BSD.

Re:Wouldn't have this problem if...

[Anonymous Coward]

Only a BSD developer would think releasing source code is a "problem".

/me goes back to his Linux system, which manages to be doing OK despite that moron Linus having decided to use the nasty nasty GPL

Re:Wouldn't have this problem if...

[Anonymous Coward]

Only a GPL developer would think not releasing source code is a problem.

Re:Wouldn't have this problem if...

[Brandybuck]

I agree. I think the Linux community is shooting themselves in the foot by pissing on every embedded systems manufacturer that decides to use Linux. Hell, not even the dominating and subjugating evil nasty nazi <gasp$gt; proprietary embedded OS vendors treat their customers like rotting pond scum. But Linux advocates do. Like a dog that's been kicked one too many times, eventually the commercial embedded users are going to look elsewhere.

Re:Wouldn't have this problem if...

[Anonymous Coward]

What? Do you realize you just said that Linux advocates treat their customers as scum?

Who is a customer of a Linux advocate, and how much does a Linux advocate product sell for these days?

What part of 'free' in 'free software' don't you get?

Re:Wouldn't have this problem if...

[swillden]

Wouldn't have this problem if... they developed on BSD.

Although funny as yet another tongue-in-cheek shot in the long-running GPL vs BSD argument here on slashdot, this is actually a very insightful point.

Why *didn't* they use BSD? They certainly could have. They might have had to work a bit harder to get drivers for all of the hardware in the PVR if they'd chosen BSD, some other tweaks may have been necessary, and performance might not have been quite as good (or maybe it would have -- hard to know). Overall, though, there really is no fundamental reason why they could not have chosen BSD over Linux and avoided such questions.

So they had a choice between a no-strings-attached, do-whatever-you-like BSD and Freedom-is-not-completely-free Linux, and they chose the latter. Why? Because Linux offered them some value that BSD did not. I don't know what it was... probably ease of use, mainly, in that lots of people have already done the work of making Linux run in lots of embedded environments (and that takes a *lot* more than just compiling it for the right processor), including the one they wanted to use.

So, they took the option with greater benefits, and a higher price, and then tried to (may still be trying to) skip out without paying the bill.

I might have *some* sympathy if there weren't a plethora of commercial alternatives they could have used, and especially if there weren't the "evbarm" port of NetBSD which they could have used without any cost. As it is, all I can do is hope that it becomes possible to make them provide the rest of the code.

Re:Wouldn't have this problem if...

[Anonymous Coward]

> I think the Linux community is shooting themselves in the foot by pissing on every embedded systems manufacturer that decides to use Linux.

No, dumbfuck. We haven't asked any manufacturer to steal code. And after all, all we ask for is a release of source code of the GPL parts. Thats not much obligation.

If they dont like this, they can always reprogram the parts like the kernel themselves or buy commercial parts.

Re:Wouldn't have this problem if...

[dvdeug]

eventually the commercial embedded users are going to look elsewhere.

Go for it. If you don't want the deal we're offering, go someplace else. I hardly see why it's a big deal if someone can't get a free ride on the exact terms they want. If you want a BSD license, use a BSD system, or any other system with similar licensing.

not even the [...] proprietary embedded OS vendors treat their customers like rotting pond scum.

They get paid. Start failing to live up to your agreement with these vendors--stop paying the price that was charged--and they'll stop treating you like the customer you aren't and more like the thief you are.

Re:Wouldn't have this problem if...

[Brandybuck]

If you don't want the deal we're offering, go someplace else.

That's exactly what I said! If you keep pissing on commercial hardware manufacturers, they WILL go someplace else!

By "pissing" I don't mean "suing for license violation". Dish isn't getting sued, and in all probability is abiding by the terms of the GPL. But it's quite clear to them that the Linux community wants them to go fsck off. Wonderful advocacy that is.

Re:Wouldn't have this problem if...

[Shadowlore]

I agree. I think the Linux community is shooting themselves in the foot by pissing on every embedded systems manufacturer that decides to use Linux.

Which, of course, explains why more and more embedded devices are moving to Linux.

IF the company is in violation of the license agreement, then it is NOT pissing on the company to point it out, questions it, and if needs be to take them to court over it.

Do you seriously believe that if you got access to WIndowsCE source via a license, then violated the license that Microsft would just go "aww shucks fellas" and move on? No. They would rightfully have an actionable case and could rightfully pursue it. If they have reason to believe, particularly by statements made by the company, that the license agreement may have been violated, then they have reason to question.

The same is true with regards to the GPL, or even the GPL, or anyother license. And if you think that the "evil nasty nazi proprietary embedded OS vendors" do not take steps to determine if a license has been complied with let us just say you are entitled to your beliefs and opinion, but that doesn't make them correct.

Re:Wouldn't have this problem if...

[MikeBabcock]

Exactly my thought -- I'm glad people want to use Linux in their devices, and if its unmodified, they have nothing to lose by releasing source code.

If however they gain from my code, or Linus', or someone else's, they need to release their modifications to their customers just as the rest of us released our code.

Re:Wouldn't have this problem if...

[evilviper]

If however they gain from my code, or Linus', or someone else's, they need to release their modifications to their customers just as the rest of us released our code.

Fortunately, the GPL doesn't say that. They can gain from your GPL'd code without releasing anything valuable back into the community, so why should modifications make a huge difference. It really adds to the burden, as they have an invisible line they can't cross, or else they have to essentially pay a fee.

I am also surprised so many companies are using GPL'd code, when the BSD equivalent is under a much better license for them, and certainly has all the same features they get.

Re:Wouldn't have this problem if...

[tigga]

No, dumbfuck.

That's the spirit! Do you know those words use would denigrate linux advocates image?

We haven't asked any manufacturer to steal code.

Nobody asked to do it. BTW open software impossible to steal.

Re:Wouldn't have this problem if...

[Anonymous Coward]

Treat like rotting pond scum?

The reason why Linux is so great for embedded use _is_ GPL. Every vendor who made Linux a bit better, had to share it with others, thus adding to the greatness of Linux. *BSD vendors kept the modifications for themself, and thus BSD's evolved a lot slower, even netBSD is no longer the most portable OS. Try lwn.net for more mature discussion with less jumping to conclusions.

For example, dishPVR uses XFS. If Linux had a "friendlier" license than GPL, SGI would have simply distributed a xfs.o binary for it's only customers, and dish networks would not have had such a great filesystem at their usage.

Now, ignoring various blathering slashdotters, the thing most reasonable customers who have bought a embedded Linux based system can require from the vendor is simply to comply with GPL, which DISH appears to be doing.

And please dont equalize /. = linux community. /. is the junkyard of spoiled kids, trolls and lack of clue. I'm pretty sure many slashdotters would not be happy even if the embedded vendor gave all sources _and_ hardware for free.

Re:Wouldn't have this problem if...

[bill_mcgonigle]

Dish isn't getting sued, and in all probability is abiding by the terms of the GPL. But it's quite clear to them that the Linux community wants them to go fsck off.

We want them to comply with the GPL so we can make enhancements to the products that we own. So, yeah, there's an agenda, but it also serves to keep Dish Network's paying customers onboard.

The only fsck I'm interested in is the weird-o filesystem they have put on my PVR.

Re:Wouldn't have this problem if...

[dvdeug]

That's exactly what I said! If you keep pissing on commercial hardware manufacturers, they WILL go someplace else!

I fail to see why I care. Why is it an advantage to me that they use Linux?

But it's quite clear to them that the Linux community wants them to go fsck off.

Even if it was clear to them--and I hardly think Slashdot is a clear indicator of anything--they don't use Linux because it makes the Linux community happy, they use Linux because it provides certain advantages to them. If they wanted to use BSD, they would have used BSD.

Re:Wouldn't have this problem if...

[MikeBabcock]

Actually, the GPL *does* say that.

If you modify my code and distribute the results, you must allow the recipient access to the source code, including the changes you made.

Re:Wouldn't have this problem if...

[evilviper]

You've completely misread my post.

I don't think you read the Apple case

[arete]

As I understand it, Apple released all of their changes - enough that you can completely build Safari from that source. What they didn't do was spend extra time making it easy to merge back into KHTML.

From what I know of Safari and Konq from a user point of view they ought to be porting Safari to other platforms more than they ought to be trying to pick and choose each patch.

This Dish thing - it depends on details of how linked they are. Details I certainly don't know at this time.

Re:Dish is for Faggots

[APE992]

Comcast ruined Techtv. I don't think so senor.

Sounds like a witch hunt!

[fluffy99]

Use a modified Linux as the OS doesn't mean you have to release the source code for all the programs that are bundled with it. That's really stretching the already vague GPL license. Keep taking that approach and soon businesses will believe Bill Gates and start shying away from GPL software for fear of rabid slashdoters threatening lawsuits.

Only an idiot would think giving away the source code to their program would benefit Dish Networks. It would be suicide because you would suddenly see third parties out there selling cheap clones of their systems.

Re:Sounds like a witch hunt!

[Reverend528]

Only an idiot would think giving away the source code to their program would benefit Dish Networks

yes, because clearly all of their profit comes from the sale of these DVR systems.

Re:Sounds like a witch hunt!

[Anonymous Coward]

Chances are they have have used some GPL'd code to verify their subscribers rights to content and hard coded a checksum test because they don't want that code changed. Maybe IO routines for the smartcard or encryption modules.

nVidia kernel interface

[Alwin Henseler]

Look at nVidias drivers. There's nothing but a stub in the kernel. The rest is in userspace and closed.

Yes, and the reason there is a seperate stub to interface nVidia's drivers with the kernel, is because nVidia doesn't want to risk having to open-source their drivers because they're interfaced with GPL-ed code.

Whether the GPL should include the interfaces as well, is up to discussion. Some GPL advocates or lawyers may think yes, some may think no.

Personally, I think licenses/copyrights should exclude any kind of interface, and be limited to to hardware/software that implements them.

Ofcourse, there's no law against not documenting 'internal' interfaces, describing interfaces in proprietary documents, or under non-disclosure agreements. Regardless of how one would feel about that.

Echostar could use some help

[baomike]

I don't know about the code in their boxes but they could use some help in programming their web pages.
They must have a bunch of people using Front Page and no one talks to anybody else, dead links, buttons that down load instead of activate, screwed up display , etc.

no it isn't sinister, it's an entirely valid point

[Anonymous Coward]

What this means is that they are prepared to offer warranty for the product 'as they supplied it' or when alterations have been made by authorised person only.

This kind of proviso is entirely standard, not sinister or against the gpl in any way.

If DISH said people could go ahead and mix home compiled stuff to their product, they'd end up in a warranty nightmare. The simplest and most sensible option is to say 'don't do it'

Something fishy

[XenonOfArcticus]

I own a DISH 721, which is closely related, and also runs a Linux OS.

I've always wondered how they are able to make this work, legally. Seems to me that you can't compile a working Linux installation out of what they have provided. They have made modifications to the Linux kernel itself. They have presumably made changes to call unreleased code from within Linux. Ity seems to me that that unreleased code is bound to the GPL. The LGPL allows non-LGPL code to be called form LGPL code, but the full GPL doesn't. So, if the DISH Linux is doing anything other than calling user-space programs, I think they're in violation.

And, my understanding is that they are doing far more than calling user-space code. They have modified the program loader to refuse to load any programs that aren't code-signed. To accomplish this would require modifying the actual kernel, AFAIK, and those changes appear to have never been released (completely, at least). They can't withold the key. Maybe the key is loaded at runtime from the hardware flash, but I kind of doubt it.

Why do I have a problem with this? Because my 721 ought to be able to run Firefox and Thunderbird, damnit, and it can't because of some paranoia. And play MP3s. And have a decent screensaver. Yes, I understand the need for video datastream security, but that's all already taken care of in hardware. Why do you need to lock up the whole OS?

Re:Something fishy

[DiscoOnTheSide]

Because my 721 ought to be able to run Firefox and Thunderbird, damnit, and it can't because of some paranoia. And play MP3s. And have a decent screensaver

Why the F___ do you need to run FireFox, Thunderbird, play MP3s and install screen savers on your god damned DVR!?!?

I swear, this is why companies are hesitant to use Linux in stuff, because there's people who demand to know any trade secret stuff for FREE because they feel they NEED it to do shit the device was never designed for. I love Linux as much as the next guy (most of my servers are Gentoo, Red Hat, and FreeBSD) but this is seriously going to drive people to the FreeBSD end of things because their licence isn't full of whiny people who feel they deserve everything for free. I can't STAND the people on here who think the world should come barreling to their doorstep for FREE because they have a notion of "Free Software." Ya know what? Free software is great. I use quite a bit and give back when I can. As a programmer and IT guy, I'm also very fond of a notion called "Compensation for my effort and time" and if all the world is free software, well, there goes half the industry because no one's getting paid for writing software. But thats a completely different arguement entirely.

Companies like Dish Network LOOSE MONEY because people crack the cards and get free services. They cut down on this by making their units "Black Boxes" or part of them, in this case... If you can run ANY old code on one of these things, the card cracking will SURGE as you can put all your nifty hacking tools STRAIGHT on the box... companies like to make MONEY and can't do that if shit happens. Don't like it? Don't buy their stuff. Go out and make your own Dish Network compatible DVR out of a bunch of PCBs and solder, code your own Linux for it, release it for free with no compensation, and Dish will go use other, more propritery shit and you'll complain some more.

Sorry for this flame/tangent, but the attitudes on here for crap like this and Apple's work with KHTML just makes me sick.

Re:Something fishy

[evilviper]

because there's people who demand to know any trade secret stuff for FREE because they feel they NEED it to do shit the device was never designed for.

Making a device do things it was never designed for is a WONDERFUL thing, even if you happen to see it as pointless. I have a lot of older devices that I've put back into service doing completely different jobs than they were designed for, and it saves a lot of money, time, waste, etc.

As for demands, you clearly have no perspective on this. I saw George Foreman on TV the other day, saying that people come up to him and ask him to fix their grills (with his name on it) as if he makes them in his garage or something. My point being, if they had used Windows CE, they would have a bunch of Windows idiots contacting them, asking how they can run Photoshop on their DVRs. Having Linux idiots instead suggesting that they release all their propritary code is surely such a small group that they don't even notice against all the other idiots.

When you run a big company, you just have to deal with a lot of idiots, and the Linux idiots (I'm refering just to a small number of misinformed zealots, BTW) go unnoticed, I'm sure. Send them the pre-formated "Thank You for your concerns" letter, and ignore them.

Re:Something fishy

[DiscoOnTheSide]

I will give you the point of hacking devices and making them do odd things. Yeah, thats fun. But it's just the people that act like its the last amendment on the Bill of Rights that no one knows about that says "You're allowed to act like a whinny asshole if something doesn't meet your unreasonable expectations".

And I'm not too sure about that whole Windows people wanting to run Photoshop and etc etc... MAJORITY of the posts on this article that I glanced through were of the tone that "Well, that code runs on linux! WE MUST SEE IT OR GOD WILL COME DOWN AND SMITE YOU!!!!111" It reminds me of the movie "Little Nemo" and the seagulls. "MINE! MINE MINE! MINE!"

Re:Something fishy

[Anonymous Coward]

You can compile a working Linux installation. What you can't do is build a working DVR from only the sources on their web site.

There's nothing wrong with this. They've released any mods they made to the GPL stuff they use. Their contributions to the open part are available to the community. Their own app containing the DVR functionality is not GPL, and thus they have no compulsion to release it as GPL (assuming they've structured the code properly).

If you want a DVR, write your own DVR code. Release it under the GPL if you like.

Re:Something fishy

[XenonOfArcticus]

Get over yourself. Why do I want to run Firefox on my PVR? Because this is Slashdot. Because we all want a home media convergence device. Because that's the promise of a media platform. The 721 is the perfect platform for this, if it wasn't under lock and key.

I don't think I _deserve_ anything that I don't actually deserve. I'm not asking for ANY trade secret crap. I think the rules, as they are stated (the GPL) require you to make all of your Linux changes public. If you don't like the rules, don't use Linux, and I'll be less interested in your non-Linux-using device. But if it runs Linux, I believe there are certain ground rules you gotta obey. I think Dish has broken one of them. I don't care if they keep their DVR application and their secret decryption code closed. It was never under the GPL, they can do what they like. But you can't close up an OS that was open and free under the GPL.

Yes, Dish loses money (LOSE, one O, got it?) from piracy. But you know what? It's not any easier to crack a Linux DVR than it is to crack any other nagra1 or nagra2 device. They don't crack the OS, doofus. They have to crack the encryption on the video stream. The OS has nothing to do with it, and running hacking tools on the Linux OS isn't going to get you squat for an advantage. The 721 is powered by a National Semiconductor Geode GX1 CPU, which is a Pentium-class at about 180Mhz. If you think this is a great encryption cracking platform you're smoking crack.

Why do I want to run a web browser on it? Well, for one, Dish Network originally announced we'd be able to do JUST THAT. It never was delivered. Being able to run our own Linux apps would help make up for promises never made good.

Why do I think open Linux would be good for the 721 and not bad? I dunno, why don't you look over at Tivo land, where their Linux OS _is_ open and moddable, and they have piles of happy users buying Tivos and doing cool stuff with it. I don't see Tivo freaking out about it, even with DirecTivo, the DBS/satellite analog to the 721.

I'm not asking for Dish to bend over backwards to give me free stuff. Quite the contrary -- do you know how much effort they had to put into making a Linux that _wouldn't_ run normal Linux apps? Making it run like a normal Linux would have been LESS work, not more.

In short, don't vent your uninformed drivel without knowing the situation you're dealing with. I'm not some "everything should be free because I don't want to pay for it" zealot. I paid a large amount of cash for the 721. I'm a consumer who wants to get capabilities they thought they paid for -- a web browser -- and an OS they thought they already had -- an open Linux. If Dish doesn't like GPL terms, they shouldn't play GPL. But, you can't shake the devil's hand and say you're only kidding.

The KHTML thing is totally different. Apple is adhering to the letter of the GPL, even if the KHTML guys don't like how they do it. I don't believe Dish is, period.

Re:Something fishy

[DiscoOnTheSide]

I fail to see what's stopping you from hooking up a regular ol' computer up to your TV and using that? Why must everything run Firefox, Apache or X? Your computer was made for it, Linux was made for it. Go to it. MythTV or some such... plus who says this thing has the processor and RAM to RUN Firefox? You'd probably have to port the software to run on this special platform, and it'll perform poorly as it wasn't up to the task.

Why use a bloody DVR for a web browser on your TV when you can get a computer and hook it up... 95% of graphics cards out there have S-Video out, whats the hold up? Hell, a Mac Mini, a TV tuner, and SAN, and you could have a nicely set up distributed media setup for your house and it'll save you time and energy.

And for the record, telling the GP to "get over himself" is more ironic than you know, and if you need to nitpick his spelling, then you've already lost points in credibility.

Re:Something fishy

[bill_mcgonigle]

Go out and make your own Dish Network compatible DVR out of a bunch of PCBs and solder, code your own Linux for it, release it for free with no compensation, and Dish will go use other, more propritery shit and you'll complain some more.

If you could noone would be interested in this. We'd hook up a DVB decoder with a smartcard reader and use the smartcard Dish sent us to do so (legally). But you can't. People who have started down this path have allegedly received letters from lawyers and don't have the funds to fight a court battle, so they quit. At least that's what I've been told in some Myth/DVB-sat discussions. There's a community doing DVD OTA in Europe but it's not directly compatible.

You're WRONG, GPL covers linked code

[Anonymous Coward]

But you DON'T have to release ANY VERSION of func_min_all_mine to anyone.

You're 100% WRONG. The GPL is extremely explicit about it, direct linking of new code to GPL code entails bringing the new code under the GPL. This applies whether the new code is in the same file as in your example, or whether the new code is in a separate file. In both cases, the source of the new code has to be released.

Don't like that? So don't use GPL code ... and good luck coding everything from scratch. :-)

In that case . . .

[hawk]

I am a lawyer, but this is not legal advice. I am encouraging you to get some from someone licensed in your jurisdiction, however.

*If* you can show that you returned it, properly (by credit reporting agency standards) disputed it, and that it's still showing, you should be talking to an attorney who handles credit report problems.

They love these.

There probably won't even be a consultation fee . . .

hawk, esq.

Re:Mere aggregation

[RedLaggedTeut]

I'm not really sure whether your example, including proprietary code by calling the shell, is exactly the thing that can be called "mere aggregation".

Suppose, for example that you had a perl compiler which would somehow parse the stuff in the backtick operator and would create a single binary from it.

Then, you would just end up with one program and the GPL would apply. For your example, I admit my argument is somewhat odd, but for a lot of other situations, for example when a C program calls some scripting, e.g. in lua, then the argument is not as clear cut.

IMHO, if you took it to court, it would depend on how much of the functionality of your program was implemented by the called function, i.e. if you asked me, just calling proprietary code indirectly thru backticks would not be enough to separate the work, but handing lots of parameters to the function called might make it clear that this function is really called at the discretion of the caller. Another good question would be, is the called subprogram useful without the caller, or is it so specific that it is an extension.

But linux isn't really GPL

[hawk]

Whether it says so in the license or not, Linux isn't really GPL, but QGPL (quasi-GPL). There has been long-standing permission for binary, closed source drivers. Nearly all of the codee has been contributed during this period. The actions of the authors supercede the letter of the license.

hawk, esq.

Whining...

[Anonymous Coward]

Whining like this is why more companies should look at using BSD licensed code instead of GPL.

Many companies using BSD licensed code give back to the community. If they don't want to, that's cool too as development will continue.

Re:Whining...

[Anonymous Coward]

"Whining like this is why more companies should look at using BSD licensed code instead of GPL."

i never understand this line of thought. why should companies "use" anybody else's code at all? they can write their own code if they do not like the conditions under which the code they build their derivative work off of is provided.

basically, consumers of oss do not get to dictate the terms under which that software is created. that's the whole point of oss.

Don't suppose...

[NoMoreNicksLeft]

Anyone knows if the nagra2 (yellow) cards are hacked?

Re:Don't suppose...

[Anonymous Coward]

Nope!

Re:Don't suppose...

[bill_mcgonigle]

Anyone knows if the nagra2 (yellow) cards are hacked?

That kind of request really makes it hard for those of us trying to legally use the Dish signal to avoid dispersions. Please take it elsewhere.

Re:Don't suppose...

[NoMoreNicksLeft]

It was meant as a troll. Besides, what is a 4 digit uid doing making me a friend, anyway? ;)

Re:Why bother

[Anonymous Coward]

Okay. =)

Dilbert

[Anonymous Coward]

The thing I've always wondered about is why all these companies aren't using *BSD for their OS. There's lots of support for embedded BSD and it makes this whole GPL problem go away.

The reason is simple.

Techies write code, and their main interest is in functionality. With rare exceptions, they won't choose imported code on the basis of its license.

Non-technical people market products, and they have no clue about what goes into them.

It's a partnership made in heaven .... er, hell. :-)

Depends if they've linked to GPL code

[Anonymous Coward]

If those applications of theirs are truly standalone then you would be right. On the other hand, if their applications are not actually standalone but link to GPL'd libraries then the GPL applies just as much as if they'd modified GPL code directly.

The only way that they could get around it legally would be to decouple their code from the GPL code in a GPL-approved way .... eg. through a socket interface.

Re:Depends if they've linked to GPL code

[Anonymous Coward]

You're assuming that libraries are GPL'd, as opposed to LGPL'd.

If you link a proprietary program against the LGPL'd C/C++ library, I believe you can distribute the binary without invoking any obligation to put the proprietary program under GPL/LGPL-like terms.

Re:Depends if they've linked to GPL code

[Anonymous Coward]

Yes, only the LGPL portion of a proprietary binary has to be released as source, BUT the rest of the program has to be able to use a modified version of that library - either with dynamic linking or by providing the proprietary part as a linkable object file. Countermeasures to prevent this clearly violate the LGPL.

For the portions that are GPL, if you make it so the product doesn't work at all if you modify the GPL portion, I think you've made the case that the entire thing is a derivative work. If you can remove the GPL portion entirely and the rest still works, it is obviously "mere aggregation". If you can replace the GPL portion with modified versions made from the source code you received, you can make a case for it not being a derivative work (depending on how closely tied the GPL and proprietary portions are; this is where the arguments over communicating by sockets or running in the same memory space start to come in). If the only portion that is GPL is the Linux kernel, and you can replace that with a modified version of the kernel, then that is clearly allowable (as the Linux license makes explicit that running an executable does not make the executable a derivative work).

The PRIMARY purpose of the GPL, as explained in the GPL itself is to allow the end user the freedom to modify the software that they receive, to better suit their needs and to understand how it works, and to be able to redistribute that software and anny changes they make to it. It isn't a "payback" to the original authors, it is freedom for the end user. If they've released source that can't be used for that purpose, they have not met the requirements of the GPL.

What about the 522?

[Anonymous Coward]

The 522 DVR is also Linux-based, and much more widely deployed than the 921.... What about it's source?

Recompilation stops system from working?

[Anonymous Coward]

GPL v3 needs to make it clear that replacement of the released GPL'd parts of a proprietary system by recompiled (but not modified) versions must be permitted, ie. the system should continue to work.

Without that key clause, access to the source code under GPL is worthless.

Okay question.....

[wo1verin3]

Software vendor XXX is releasing a commercial piece of software but includes a GPL component like LAME.

Since it is a codec they are using do they need to release source for LAME (+modifications) in addition to source for their entire application?

Re:Okay question.....

[Anonymous Coward]

People don't realise that, the more that the community appears to 'force' someone to open up previously proprietary code (such as in the instance of using LAME, for example - it would be trivial to use LAME without linking to it, since there's a command line interface that could easily be scripted in a standard fashion). And yet you seem to slyly advocate them being required to release the source for their entire application.

People like RMS are too rabidly anti-proprietary software. And until that changes, F/OSS will always be a niche within a much larger commercial software world. Sure, F/OSS is great for the people (I run a small personal Linux server myself, and I love it). But companies don't like to buy in to people telling them they HAVE to open up their source. Why do you think Tivo hasn't GPL'd their actual UI code, even though they run on Linux? Someone like RMS would advocate that they should.

Re:Okay question.....

[julesh]

Software vendor XXX is releasing a commercial piece of software but includes a GPL component like LAME.

Since it is a codec they are using do they need to release source for LAME (+modifications) in addition to source for their entire application?

If they are linking directly to LAME, then they need to release their code under the GPL to keep this legal.

If they are linking to a codec management layer that has a LAME-based plugin (e.g. Windows ACM) then they don't.

(My opinion only, IANAL, get legal advice if this is important, etc.)

Re:Mod parent up

[Anonymous Coward]

Safari is open source under GPL.

How is that taking more than giving?
They've taken KHTML, made it better, and the better version is available under GPL, this is how GPL is supposed to work.

Okay, so is this actually useful?

[Anonymous Coward]

Serious question, not a smart-alec response. For those of you who know what this gobbledegook actually means, could somebody use this code to hack their Dish box for a useful purpose? I don't mean stealing service without paying here, but rather something like remote scheduling or video extraction. The latter in particular would make it worth my while to give Dish a shot...

I really like my dish network dvr,

[Odocoileus]

While were on the subject, does anyone know how to make use of the data port on the back of the 921?

GPL: Mere aggregation

[Thomas Frayne]

The example is meant to be a case of mere aggregation, as described in the GPL. The example is so short that it may not get the point across. Here is some more information.

http://www.gnu.org/licenses/gpl-faq.html#MereAggre gation says: --- "Mere aggregation of two programs means putting them side by side on the same CD-ROM or hard disk. We use this term in the case where they are separate programs, not parts of a single program. In this case, if one of the programs is covered by the GPL, it has no effect on the other program.

Combining two modules means connecting them together so that they form a single larger program. ... What constitutes combining two parts into one program? ... We believe that a proper criterion depends both on the mechanism of communication (exec, pipes, rpc, function calls within a shared address space, etc.) and the semantics of the communication (what kinds of information are interchanged).... pipes, sockets and command-line arguments are communication mechanisms normally used between two separate programs. So when they are used for communication, the modules normally are separate programs. But if the semantics of the communication are intimate enough, exchanging complex internal data structures, that too could be a basis to consider the two parts as combined into a larger program." ---

The given examples are probably meant to be separate programs invoking each other by exec, and with no complex internal data structures passed, so a distribution that contained both would be a mere aggregation.

There is a continuum of possibilities for mere aggregation versus "parts of a single program". For cases close to the border, which side applies would have to be decided by negotiation or by a court.

MythTV

[Anonymous Coward]

Hi I'm the Anonymous Coward that posted this. Will this source code aid in the development of a DVR that can use DISH NETWORKS signal through a MythTV box?

ExpressVu PVR

[PunkPig]

Do they have the source code for the dishnetwork version of the expressvu 5900 PVR? The code it runs needs some work. Missed recordings, broken search, broken guide, etc.

Re:Mere aggregation

[NickFortune]

I'm not really sure whether your example, including proprietary code by calling the shell, is exactly the thing that can be called "mere aggregation".

Oh, really? If the backtick operator doesn't qualify as a mechanism of mere aggregation, then it's hard to imagine what could qualify.

The backtick operator can execute any executable on the system, to say nothing of arbitary shell scripts and sequences of shell commands. If backticks can "infect" code, so can bash. I don't think even the most fanatical free software zealot has ever pushed for that interpretation.

It seems to me that it's all a question of interfaces. If someone designs GPL software with a well defined interface, then you're free to write non GPL software that uses that interface. If there is no interface and you're changing the files in the original software, then the GPL applies.

Kernel modules are far more intimately connected to GPL software than an arbitatry executable is to Perl. And the Linux kernel development community seems count a healthy contingent of Free Software devotees amongst its number. Yet few if any people seem to consider that proprietory kernel modules are "tainted" by their association with the kernel. Quite the reverse in fact.

Nor does the compilation issue change matters. If I can write a compiler and use it to compile code that you have copyrighted. However there is no licence in existence under which that process of compilation grants me the the copyright to your code.

cause the unit to fail

[sl4shd0rk]

yeah right... I'll just compile this all as a kernel modules and modpro.. *poof*

Their advantage is programming, not software

[bill_mcgonigle]

The problem is they are building on top of commodity hardware, with the only advantage being the software.

They're not making much money on hardware if any. I got my DishPVR 508 for $199 a couple years ago with an 80GB drive. I think that was about $120 worth of drive from NewEgg at the time.

Dish makes their money from programming - each new customer costs almost nothing to add. They're probably losing some money on hardware to get customers.

To be perfectly clear: Dish is playing this stupidly.

If they were to GPL the whole damn thing somebody would go make a MythTV version that could take their smart card and allow paying customers to decrypt the DVB stream and they'd have a better PVR then they're producing today which is buggy and little more than a VCR with a hard drive. Oh, and they wouldn't have any costs in the software besides QA. Even better, they'd assign their team to add features to MythTV and have a win-win kick-ass development strategy. Sure, DirectTV may decide to crib it but they'd have a large early-mover advantage and at this point DirectTV is looking like it's got the better DVR strategy anyway, so what's there to lose? Charlie Ergen is a guy who knows how to take calculated risks.

Customers like features. When they get this they'll get more customers. You'd think this was rocket science - I guess they have all their rocket scientists busy with satellite launches.

Charlie, give me a call - I'll help you get this rolling.

Re:Their advantage is programming, not software

[Mycroft_VIII]

Problem is the people putting out the content would freak if there wasn't absolute controll over said content. The minute things look like the content could be on more than one device at a time they pull every stop out to pound the partie responsible into a smoldering crater.
Also they stand to make money not just off subscribtions, but upgraded units. "BUY the NEW 44k-zillion unit and get an additional 5 hours record time and new dayglow digital readouts on the front pannel and alarm clock feature!".

Mycroft

Lies, damn lies, and posts about the GPL

[Rogerborg]

> if you release the binaries to the world, you must release your own code under the GPL to anyone to whom you supply binaries

And you can quote the part of the GPL that mandates that requirement, exactly that requirement, and only that requirement, right? I can see at least two things wrong with what you just said. Do you know what they are?

Re:Lies, damn lies, and posts about the GPL

[resiak]

In the context of a summary, I can't see anything wrong. I'd (honestly) like to know what it is that you disagree with.

Linking creates a derivative work

[Rogerborg]

Oh, just read the damn license. I'm tired of explaining this crap to you cretins.

Re:Linking creates a derivative work

[Mr. Slippery]

Oh, just read the damn license.

The GPL talks about "forming a work based on the Program". It does not discuss linking.

The FSF has an opinion that linking, static or dynamic, creates a derived work. But their opinion is worth no more that yours or mine - I find it unconvincing that dynamic linking necessarily creates a derived work.

Drawing the line on what constitutes a derived work in software will probably left up to the courts, poor bastards.

I approve of this message

[Rogerborg]

It's nice to see an adult posting here.

Re:I approve of this message

[evilviper]

Thanks. Though I'm sure my message will be summarily modded down to -500 any time now.

Violates the spirit of Free Software

[Krehbiel]

Do not replace or add any software to the DISH 921 DVR with items compiled from these source trees. Doing so will void all warranties and cause the unit to fail.

Well I seem to recall that the whole POINT of the GPL is that you CAN CHANGE the software you receive, to fix it or enhance it - not merely to study it and copy it. Now they add this clause to make this particular thing impossible.

This is a big loophole in the GPL that I hope they address in GPL V3. You can get some device which contains some GPLed firmware, and though you might even get the full source for the software, it's impossible to replace the firmware because (a) it's technically impossible to replace (masked ROM), (b) because some non-GPLed watchdog is checksumming the GPLed part to make sure it doesn't actually change, (c) the protocol for flashing new firmware is wrapped in a cryptographic signature thingy (d) whatever else might make it tecnically impossible to actually MAKE CHANGES to the code...

Re:Violates the spirit of Free Software

[man_ls]

They warranty the device as shipped. If you screw with it, even if you are legally permitted to do the screwing, they don't have to help you out with it.

They didn't say don't do it. They said if you do it, it probably won't work, and we're not going to help you fix it.

Re:Violates the spirit of Free Software

[Krehbiel]

Read it again. It's pretty clear.

Do not replace or add any software to the DISH 921 DVR with items compiled from these source trees. Doing so will void all warranties and cause the unit to fail.

I didn't see "probably" in there. I saw "will."

LGPL, not GPL

[Luke-Jr]

KHTML is LGPL, not GPL. These reasons are exactly why RMS has been arguing that the GPL should be used instead of the LGPL on *all* new software.

That is not a metaphor

[nietsch]

That is just downright name-calling.

If you don't like it, don't use it. Nobody will force you to sign a NDA or sue you for looking at the code. The only reason your code should become GPL-ed is when you actively use GPLed code in a product you distibute. Calling that viral would be on equal terms with calling water poisonous: only if you inject very large amounts directly into the bloodstream.

The 'viral' term was a stroke of genius of those opposing Free Software. That meme relies on hotheads like you for its propagation. Wouldn't that be more viral?

Not necessarily a GPL violation

[Pedrito]

Many people are claiming that the comments Dish has with the code suggests the code violates the spirit of the GPL. While it's possible (I haven't looked at the code), this is not necessarily the case, as others have already pointed out.

I suspect the dish runs on Linux and has a number of modules that work together much like other complex Linux applications. An good related examples is MythTV. MythTV isn't one module. It uses some existing applications and tools and then provides its own.

Whatever GPLed modules they modify, they're required to release their source changes to. But any completely custom modules they've written can remain proprietary without being released even if they interact with the GPL'ed modules.

What their comments suggest is that without their proprietary code, you cannot create a working Dish application.

On the other hand, if you want to take their code modifications and using other GPL'ed or even your own proprietary modules to create a DVR (probably using different hardware), you're free to do so.

This neither violates the spirit of the GPL nor the letter. As long as they're releasing the changes they've made to GPL'ed code back to the community and that code isn't obfuscated or otherwise rendered "unusable", they're completely within the letter and spirit. Now, keep in mind, usable doesn't necessarily mean usable on THEIR hardware. It just means that it is operating code.

Off topic - PVR 350, XV, and white lines

[Anonymous Coward]

Okay, this is off topic but it's something I feel I should share, because Slashdot often comes up in Google searches when researching problems.

I built a PVR with dual Hauppauge Win PVR 350's, the idea being able to record two while watching a third. Anyway, when running MythTV with XV and playing back recorded shows there was jitteriness and white lines that weren't actually a part of the recording itself. Finally, we managed to rundown the NOXV='1' flag when starting MythTV from a command line. That, and a setting in the menus (which I don't recall offhand) cleared it right up.

Here's hoping that someone will find this post when researching the same problem.

P.S. I can't have Dish (nothing facing south) so I use Comcast, but I have a friend who does. They seem to treat him well enough, though I'm certain that the horror story above (guy getting $400 collection notice) is true too. It's a mixed bag with most companies, in fairness.

Ugly mother of crap!

[vvaduva]

I am so glad i ditched them and went the HD TiVo way with DirecTV! Not only is this unit a piece of crap (technically speaking), but it's ugly too!!

Can one piecemeal GPL apps?

[Anonymous Coward]

I thought once you put out binary that uses GPL that it all is GPL. If you only have to release the source to what is already released (source), what's the point?

What we need is a list of all known GPL violators

[jonwil]

There should be a list out there of all companies known to be violating the licence on so we know which products are in violation (and can avoid buying those) and which companies are in violation (and can avoid buying from them). This would be especially focused on hardware + software combos (e.g. routers, PVRs and other things that use OSS)

Mabie, to ballence it out, a list of the companies (hardware companies for example) that ARE in 100% complience with the GPL could be included.

I think there's one loophole

[roystgnr]

I've seen case law (wish I remember where) saying APIs aren't copyrightable - you can't necessarily just #include "whatever.h" into your program, but you can rewrite a "my_whatever.h" to declare the same symbols and binary compatible structs. If that's true, if a GPL program has an existing binary plugin loading interface, and if you're willing to do a bit of redundant work, you can compile a module that meets that interface and isn't legally a derivative work.

Can you redistribute that plugin without source code along with the GPL program? I have no idea. But I'll bet you can redistribute the plugin alone, never distribute the GPL program, and be in the clear. Remember, you can't sue someone for "GPL violations" - nobody has to agree to the GPL. GPL violations are only important because they remove a defense against copyright infringement lawsuits. So, if you're never going to redistribute copyrighted material in the first place, why worry about whether you would be licensed to do so or not?

Slashdot community != Linux community

[goldfndr]

There are plenty of non-Linux people here just to contribute flames. If you found an article in which a "Linux community" was "pissing" (whatever that means) on commercial hardware manufacturers, you really ought to cite it.

But most likely you're just trolling again.

Cat probably types better...

[leonbrooks]

...as if that's a prerequisite for a /. post or anything. (-:

OK, let's ask.

[leonbrooks]

If I get a reply from them, I'll post it here.

Sorry to hear your bad experiences...

[RoverDaddy]

but we have only good things to say about Dish here. I have been more than happy with Dish since we signed up about 8 years ago. No problems with equipment, installation, billing, etc. I saved a bundle switching from cable, now I have a PVR that we love, and even at current prices we're doing much better than we ever could have with cable. YMMV
OTOH, all those asinine commercials from the cable industry that try to tell you satellite TV is nothing but trouble, just piss me off.

So you've never seen lawyer speak before?

[RoverDaddy]

If they say 'probably', and you go off and break the damn thing, you might come back and say 'you didn't try hard enough to warn me'. If they say 'it will break', you don't have a case. Same reason every damn cup of coffee we buy now has to warn us that it's 'hot'.

Yes!

[RoverDaddy]

You spell it out very well. Also, I think the unspoken undercurrent here is largely one of "Hey, I want to rip those vids from my PVR, and damn you to hell for stopping me". Well, if Dish didn't do it's absolute best to lock down those streams, you wouldn't have a PVR that basically you didn't even pay for, when you compare the deals they offer with the real cost of hardware. And to the GP saying the security is done in hardware, I wouldn't assume that without digging in there myself. I suspect at least some of the code they aren't releasing is essential to the DRM.

http://freebios.sourceforge.net/

[screwthemoderators]

http://freebios.sourceforge.net/ Oops! Sorry about that. I'm not sure why its moved, maybe financial reasons? http://www.openbios.info/ is an attempt to implement open firmware for x86 and other architectures.

Well...

[oldosadmin]

Dish Network has proven time and time again to be less evil than most cable companies and definately less evil than Direct TV.

Checksumming, silly

[oldosadmin]

You people should read the previous comments.

Dish uses a checksumming process. If you upload a binary that's not matching the checksum, it bails. I doubt this is about linking -- it's probably more about an executeable modified *at all* not running on the box due to checksum failure.

Encryption

[lazysquid]

Does anyone know if the filesystem is encrypted? Are the files formatted in some unique format?

Does that explain why the Linux kernel isnt Micro?

[RedLaggedTeut]

Maybe that explains why the Linux kernel isn't a microkernel .. R.M. Stallman probably would get a heart attack if proprietary drivers could be distributed along with the kernel when they were included as separate works, attached by a clean interface.

But really, what constitutes an interface? You could detach many kernel modules, drivers and code by calling a separate process, like in backticks operators, it would just be horribly slow. But maybe not so slow if you had optimized the process, like using some form of interprocess communication.

So really, I think that just considering the backtick operator as a form of separation would break as soon as someone puts his heart into breaking it.

Let me try to follow your argument - consider the LGPL for libraries - if all you'd need was a clean interface to establish a separate work, then there would be no need for the LGPL at all, since the library would already be a separate work under the terms of the GPL.

And yes I understand the difference between compiling and interpreting/linking, but I also understand that most of the stuff you can interpret can also be compiled (gcj,php).

Re:Mere aggregation

[julesh]

The "mere aggregation" clause in the GPL is intended to allow unrelated software to be distributed together, which is quite clear if you read the entire sentence it is found in:

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.

The problem is defining what "another work not based on the Program" means. It is unclear in your example whether or not your program is based on the program it executes or not... this is a difficult concept that has yet to be decided by case law.

The FSF's interpretation of this is available here.

The GPL kicks in on distribution

[NickFortune]

But really, what constitutes an interface?

I would have to say intent. The intention of the backticks operator is to allow a perl script to run shell commands that may be completely unrelated to perl in any way whatsoever and to do so without modifing Perl itself in any manner at all.

if all you'd need was a clean interface to establish a separate work, then there would be no need for the LGPL at all, since the library would already be a separate work under the terms of the GPL.

We may be talking at cross purposes here. Running a program via perl's backtick operator isn't going to change the licence of the command invoked thereby. The GPL only kicks in if you modify and then distribute someone else's code.

On the other hand, suppose you were to take a GPLed application that used a mixture of perl and binary executables with the binaries called via backticks. Suppose further that you relace one of the original binaries with one of your own devising. Then, in that precise narrow circumstance, the application author might be able to call for your source to be released under the GPL.

If that's the scenario you meant, then yes, it could be considered a grey area.

As for libraries: they need a different licence because you implicitly distribute them when you distribute an application that has been compiled using them, and in order to distribute them, you are required to abide by the terms of the licence. It's a descision you take at the time of distribution, and nobody is forced into anything. Nevertheless, the LGPL makes this a non-issue in the majority of cases.

With backticks there is no distribution of anyone's code. All the backtick does is issue a command to the OS which may or may not be able to respond, and may respond with a variety of different executables depending on what packages are installed and the users PATH setting.

It may be that you had in mind some specific combination of authorship and distribution that I've failed to cover here. If so, I invite you to explain it in a bit more detailm since it's clearly gone over my head. If not, I think proprietory code is safe from the backticks operator, at least for this version of the GPL.

Re:Mere aggregation

[NickFortune]

Good point. I must confess, I hadn't realised that mere aggregation was being used in such a precise manner.

All the same, the backticks are still proprietory safe. Well, unless you are distributing someone else's GPL'd perl app and replacing one of their binary driver modules with a proprietory one of your own. Even then the FAQ suggests that intimate semantics and/or complex data structures would need to be involved.

My point is there is not much difference

[RedLaggedTeut]

My point is there is not as much difference between a backtick and a function call as people amke it out to be.

For example, you could imagine that one could do, in pseudo-sourcecode:

?php

doA();
include('B.php');

or you could do:

?php

doA();
echo shell_exec('env '.passArgs($_REQUEST).' php.exe B.php');

With the shell_exec(=backticks), you would invoke the system, but the program would still do the exact same thing. It is illogical if the first version spreads the GPL, while the 2nd version may call proprietary code.

It's a superficial similarity

[NickFortune]

My PHP is a little rusty, but The analagous perl mechanism IIRC is "use" or "require".

As in your PHP example, you can effectively write "require" in terms of backticks, turning

require foo.pm

into

"$foo = `cat foo.pm`; eval "$foo";"

Now I'll concede that in such a case a case might be made (subject to my earlier qualifications) for the GPLification of foo.pm.

However -- and this is a big however -- the backtick mechanism is not the vector of contagion. The factor that may result in the invocation of the GPL is the inclusion of the contents of foo.pm into the script being executed. The backticks are being used to read the file being included, but the insertion of the code into the "address space" of the script is performed by the eval mechanism.

Backticks and function calls can be used to achieve similar results, and so I suppose it's tempting to consider that they do the same thing. However they work in profoundly different ways that make them quite distinct for GPL purposes.

Re:Mere aggregation

[julesh]

Even then the FAQ suggests that intimate semantics and/or complex data structures would need to be involved.

Yes, but there's no reason this couldn't be achieved through a pipe.

I think the way I would see it is this: if it is practically infeasible to replace the GPL subprocess with another implementation of the same interface (regardless of whether any implementation exists at the moment... it's the practical possibility that one may be created I'm thinking of), then the tie is probably too close and the new program is a derivitive work of the GPL one. But, IANAL, etc.

Re:Mere aggregation

[NickFortune]

I think you need a better word than "infeasible". As it stands you seem to be saying that if the GPL only applies in cases that will never happen, and that if they somehow happen anyway, that can be used to prove that it wasn't really infeasible, and so the GPL didn't apply in any case.

I appreciate this is probably not what you meant, and I believe I agree with what I think you meant... but I'm not sure how to better articulate the idea, so it's difficult to be sure.