Slashdot Mirror
TrollTech Responds To QT Accusations
David B. Harris writes: "On freshmeat.net Erik Eng of Trolltech responds to various accusations thrown at TrollTech about their QT libraries' licensing problems. Set the flame dial to low when you read this one, he seems sincere."
im going to lose karma over this but wtf iv got to say this
I am a Linux user not a programmer and I dont give a crap about licences
All i want is software that dont suck!
Face it as a desktop os Linux sux as bad as winblows
So what do the morons do talk world domination then attack
someone nice enuff to allow them to use a decent
toolkit that someone odviously thinks is
worth spending money on for free
To all the ungratefull idiots out there
GO JUMP IN A LAKE
To all the peaple that put there time energy
and skill into linux/*bsd/gnu/qt/kde/ect
to make them suck less thank you!
This includes the qt programmers thank you
for helping make free operating systems suck less!!!!!!
There is nothing illegal with linking GPLed code to non-GPL libraries!
As the article states, people were writing GPLed software against the uber-proprietary Motif libraries for years. Emacs made system calls on non-free OSes. There is nothing illegal about writing GPLed code which uses the Qt API and libraries.
One keeps seeing this, and it is wrong. It is pure FUD spread by some zealots--GNOME users? GPL fanatics? who knows.
Disclosure: I dislike KDE, but agree that Qt is much more attractive than GTK. I like the GPL and support it. Doesn't mean that I support fanatics in their zealotry or their lies.
I'm probably getting trolled... oh well....
1) It's not his job to deal with KDE or the GPLd code in KDE. He codes Qt. Not KDE. The issue wasn't KDE/GPL, he was addressed why Troll chose the QPL, not the GPL.
2) False. Some are employed my Mandrake, part or full time. Corel has some. Caldera has some. The vast vast majority are volunteers are non-affiliated. KOffice is funded by some company in Asia. KDE, unlike most open source projects, does not have a "benevolent dictator" to steer, anyway.
3) I have no inmmediate problems with the QPL, except for the GPL incompatibility. Compared to crap coming out like the APSL and SCSL, the QPL is a really nice license. I still, however, maintain that it's GPL incompatible (due to my definition of linking). It was nice to learn it's GPL incompatible because of their definition of linking, not from some secondary motive. The issue of the license should be taken up by the KDE team, not Qt. And the KDE team has shown great apathy towards licensing issues. I'm generally a "shut up and code" kind of guy, but this issue has been there for years, and hasn't been solved yet.
KDE and Qt are totally seperate. This guy is from Qt. We heard from Debian before. Let's hear something from KDE, they're the ones at fault in the minds of the Deian developers (and others), not Qt.
Windows 2000: Designed for the Internet. The Internet: Designed for UNIX.
You're both right and wrong about this one. From http://www.apache.org/ABOUT_APACHE.html:
The name Apache came from "A PATCHY server", because it was originally a bunch of patches to NCSA's reference httpd. So one could argue that Apache copied NCSA httpd, or one could say that it is an extension of it -- in which case Apache is the original product, and all the commercial webservers are copies of Apache.
But apart from that, you're quite right about OSS vs. commercial software: OSS is best at projects that meet a well-defined, pre-existing need. Innovation (real innovation, that is, not M$ "innovation") into new areas of software is typically done by companies with plenty of money to spend on R&D.
-----
The real meaning of the GNU GPL:
The real meaning of the GNU GPL:
"The Source will be with you... Always."
End o'story.
Actually it's a (mis)interpretation on the Debian developers' part; it's why, if they were taken to court over this, they would be run through the wringers for libel.
The GPL only deals with the code in question (the software that is licensed under the GPL) being used in other works, hence the derivative works argument. It *does not* deal with linking to the libraries.
Heck, one could argue that GNOME apps are illegal. X11 libraries aren't just not released under the GPL, and IMHO (and against RMS's NSHO) not really GPL compatible (at least in this particular misinterpretation) the X11 libraries are also not essential system libraries. Nope, don't try to say they are: you can run your system just fine without X at all.
So what does it all boil down to? FUD. I read through the licenses, and the only *potential* problem I found was perhaps the fact the QPL *demands* that any works linking to a QPLed piece of code be released free as in speech *and* beer. The GPL only requires that software be free as in speech. However, if you release your code for a fee, you'll just have to pay the (hefty) licensing fee for QT.
So, what does it all boil down to? FUD, aimed at those gullible enough to believe this without reading the licenses for themselves. Don't believe me? Go to http://www.gnu.org and http://www.troll.no and read the licenses. Print them out if you have to. But read them.
Coming from a non-X programming background, how hard would it be to implement the QT api as a C++ wrapper to GTK???
That would solve all the problems... KDE with GTK..mmmmm
Simon
Simon
The real linux_penguin has Slashdot ID 101961. Anyone else is an impostor. Including Bruce Perens.
Careful.
I could take one amendment of the United States Constitution, quote it, and tell you that alcoholic consumtion and/or production is illegal. Is it? No.
What's my point? Granted, the analogy isn't all that close. The point is this: you're quoting one specific part of the GPL and excluding other parts. You're also *completely* excluding the terms of the QPL. The QPL has terms that allows for QPLed software to be linked to by "open source" programs. The original licensing agreement called for software to either be GPLed or some other (freer) license.
IMHO (although IANAL) this is just FUD in an attempt to kill KDE. You'll notice the arguments are mostly against KDE. I really haven't heard too many arguments against, say, WMFinder (IMHO the best X11 file manager.) But, hell, it isn't "competing" against GNOME directly. KDE is.
AAAAAAAAAARGH!
LiteStep is illegal! You have to have VC to compile it!!!!!
Okay...where's the zealots with torches?
Anyone?
Sigh...guess it's just KDE we want to nail to the cross, then.
PS GNOME is illegal too.
This is true of KDE as much as QT. In hindsight, the KDE developers would have probably chosen a different license, but now that it's popular, the QT and KDE license don't work so well together, so goes the argument.
The real problem is that one apparently needs a savvy lawyer right from the start, but none of these projects have the money to make this happen. Why should I need a lawyer to do free software? This is the real problem with the GPL and its intricacies. It needs to be either clearer or more flexible.
Yes, that would indeed make it GPL incompatible. :-)
The solution is to require that everyone who wants their patch to appear in the official Qt release assign the copyright to TT, much like RMS begs people to assign copyright the FSF.
If you ask me, it's a no brainer. TT should do this for its own good, because it benefits when KDE succeeeds, and KDE won't succeed unless this license issue is fixed!
Just one note about a GPL Qt. TT would have trouble incorporating patches from users since they will fall under GPL. They must have the patch authors permission to include it in their proprietary version of Qt. The code that TT has written is of course no problem. TT could add a clause to their GPL which says that they can use it in their proprietary version. But I am not sure if that makes the license non-GPL compatible.
But Motif is logically no more part of the OS than Qt is. They're both UI toolkits that are not required for basic system operation.
This is the one thing that bothers me about the GPL in the current situation. The prevailing wisdom seems to be that it's OK for non-GPL software to be built on top of GPL software as long as the communication mechanism doesn't involve mapping the GPL software into the same address space as the non-GPL software. If it's done by dynamic loading a GPL library, though, the result is apparently considered a derivative work. In real world terms, this seems like a stretch. The line between use (which is not covered by the GPL) and incorporation/modification (which is) is now blurred, and is becoming more and more so with time, as runtime loading and such become more and more universal.
Hold on, isn't that redundant? :D
Note: this may sound more funny if you know that I am French myself.
"The obvious mathematical breakthrough would be development of an easy way to factor large prime numbers." Bill Gates,
The GPL restricts distribution of a GPL program unless you can provide the entire source of the program down to virtually the OS level, including libraries and everything needed to make the program work, under terms no less restricive than the GPL, with the exception of components that are part of the operating system.
Wrong. The GPL prevents redistribution of a program unless you provide the entire source of the program. Period. As has been pointed out before, GPLed software can be written for any platform (where platform = hardware + OS + libraries). Even granted your mistaken belief, I challenge you to define `operating system.' Acc. to the strict and proper definition, the OS is only the bit which deals with resource allocation--it's not the shells, not the libraries, not the graphics systems, not the window managers, not the desktop environment; it's none of those. And yet GPLed software has been written under those circumstances. The license hasn't changed. Thus, if that GPLed software were legal (one hopes so, since that family includes emacs and gcc), then your phrase `operating system' is equivalent to my `platform,' and Qt, being a library, fits into that category. Ergo, GPLed programs may be linked against Qt.
This reminds me of the FUD spread by the anti-GPL folks, that software compiled with gcc could only legally be itself GPLed.
So, for your wrapup, that's essentially what KDE does when it takes other peoples GPL code, ports it to the old-style-licensed Qt, and then distribute it. Someone's property, made semi-proprietary through dependencies. That is the reason people have been rather upset.
It takes GPLed code, modifies it and releases those modifications under the GPL, as it is required to do. No law has been broken; the GPL has been followed. So where is the cause for complaint? That someone has modified one's code? Is that not one of the reasons for the GPL to begin with? It is no more semi-proprietary than code written against Motif or on a proprietary OS is.
With the coming of the New Internet Economy (feel free to hurl), Intellectual Property is becoming an important topic. As companies go to even greater lengths to ensure that nobody else makes a single buck from their "work", it has become equally important that the open-source/free software community makes sure that their own efforts are not abused. This, I presume, is why a new improved version of the GPL is on its way.
I agree, there is no need for endless rants and flames about the subject. However, the community must be able to voice their opinion, however annoying and bad PR this may produce.
Motif in Linux have the same problem as Qt (at least before lesstif). It was one of the reasons why lesstif was developed.
I can't tell exactly if Motif is considered a part of the OS in e.g. Solaris. I would consider it a part of Solaris, but that's just me.
Seems to me that TrollTech have the same right as everyone else to do what the hell they like with their software. Just like Sun or ... oh, come on, let's do some overkill ... like Microsoft. The only question for me is: is their software any good? If it's good, it's probably worth some money - though I can't see from here who's paying them any. Who are TrollTech's customers?
As the absolute total of newsgroup messages increases, so entropy rises, resulting in a rich cache of pat replies which can be invoked as incatations in place of arguments. In many forums, it is customary to accompany such incantations with insults, as these too are a valuable way of attacking whilst having nothing new to say. As this entropy reaches a maximum, the need for human involvement will decline below the level needed to sustain interest and people will revert to more traditional exchanges such as fist-fights and poison-pen letters.
No, your children are not the special ones. Nor are your pets.
So, for instance, I take work "myprog.o" and add "libqt.a" and create work "myprog". "myprog" is a derivative work of a GPL'd work "myprog.o" and a non-GPL'd work "libqt.a".
That is one opinion. IMHO it is incorrect. I've always thought that the whole linking issue has been viewed incorrectly. It comes from looking not at source code, but at executable files. The important thing is that GPLed code remain GPLed. The executables are generated by the compiler. What if an OS-provided compiler is used which links in a propietary library without which the program cannot run? It is insane to say that this is wrong. If the GPL does disallow this, then it should be modified.
I doubt that it really does disallow it. As has been pointed out before, Stallman himself has written code which had to link in proprietary libraries.
The LGPL is currently needed, but the reason is that the solution of simply including source to statically-linked GPL libraries is not satisfactory; it is not possible to replace them. Dynamically linking, OTOH, solves this problem. The GPL should be modified to allow dynamic linking with distribution of source and static linking with distribution of source on platforms which still do not have dynamic linking. Then we can get rid of the GPL.
IMHO, the idea of `derived work' has been taken too far. A program is not a derived work of a library; otherwise I would have to pay Apple for every program I write on a Mac, or M$ for every program I write on Windows. A library provides the other end of an API.
The GPL does not prohibit deriving GPLed works from proprietary object code. Otherwise every single older FSF project would be GPL-incompatible.
That is one opinion. IMHO it is incorrect.
In that case, you are saying that RMS misunderstands the GPL, as RMS has repeatedly stated that position, and defended it as necessary to prevent the subversion of free software.
The important thing is that GPLed code remain GPLed. The executables are generated by the compiler. What if an OS-provided compiler is used which links in a propietary library without which the program cannot run? It is insane to say that this is wrong.
See this FSF text written by RMS. And point two of this one. And this other one.
It's supposed to be impossible to link a GPLed program to a library that can't be distributed under the terms of the GPL, and it's supposed to be impossible to link a GPLed library to a program that can't be distributed under the terms of the GPL. It was done deliberately. The only exception given was for OS components, and only because it was necessary to allow that in order to make free software available for those OSes.
(Yes, the definition of OS is fuzzy, and one can make the case that Qt is an equivalent of Motif, a system library for the X subsystem of the OS. That addresses this specific case, not the general.)
Steven E. Ehrbar
Obviously there are -- but still, developer time and/or developer productivity is important, and good development tools are worth quite a bit.
In any case, I don't think Qt is particularly competitive among commercial cross platform toolkits. Many of them are more mature and have better tools support (GUI builders, scripting, etc.).
So what would you suggest for a Linux/Windows app ? Qt is by a long shot the nicest toolkit I've seen on Linux.
There are several high quality toolkits out there without such restrictions...
Umm, you forget to list any restrictions. I can only find two restrictions that anyone can point to even vaguely onerous. Both of which, if you had read the article, are about to go away.
1) Patch clause. Gee, every OSS project I know of accepts only patches anyway. In any case, the only thing this restricts is forking.
2) No private distributions. If you distribute the software, you distribute the software. No if ands or buts.
I think it's valid to ask the question whether this is a good deal.
It's an excellent deal. Qt is Free Software if you're using it to write Free Software, but it's proprietary if you're writing proprietary. In a world where everyone keeps telling me "life isn't fair", this is a breath of fresh air.
And Qt is so expensive that I would have serious concerns about using it even as a purely commercial toolkit.
Have you priced out commercial toolkits? I don't mean those half-assed Microsoft offerings, I mean real-world high-quality tools.
A Government Is a Body of People, Usually Notably Ungoverned
I'm pretty sure a lot of the KDE developpers don't grasp the difference between Free software and Open Source.
So what is the difference?
I've read Richard Stallman's definition of Free Software, and I've read the OSS's definition of Open Source. As near as I can tell, anything that is Open Source must also be Free Software. Which one of RMS's four points does the OSS definition not cover?
A Government Is a Body of People, Usually Notably Ungoverned
OK well i bet penis bird person guy thing to it.
but what i dont get is why does everyone race to get fp?
what is so special about it?
can some ??"normal"?? geek out there explain this to me???
the only substantive reason a distribution would include Qt is so it can include KDE
Oh wow! This must be why Debian includes the Qt-2.x library! Thanks for clearing that up for me.
A Government Is a Body of People, Usually Notably Ungoverned
I agree with this fellow about KDE and such QT-based programs being fully GPL'd. I've read the GPL closely, and it does not make any restrictions of the compilation/linking of a GPL'd program. It DOES restrict non-GPL'd programs from linking with GPL'd libraries/programs, but a -> b does not equal b -> a. Since KDE(a) is linked against QT(b), eveything is good. No proprietary code is linking against a GPL'd program/library. QT is not linking against KDE, so where's the problem? Sure, QT links against LGPL'd libraries, but that's why they're LGPL'd. This license expressely allows proprietary programs to link against an LGPL-licensed program.
However, I this this person should really wake up and just say what everyone knows: They want to make money. They should just admit it. It's their software, and they can do what they want with it. Instead of bitching about it, make another library. Oh, wait, people did - GTK+.
So, just to wrap this up, I'd like to pose you a question. What if someone said that your nice, nifty, Free program, distributed under the GPL, should be proprietary? What if you had hundreds of people downright DEMANDING that you move to a draconian license? You'd be pretty upset, wouldn't you? Because it's your property. It's your blood, sweat, tears, and time. Everyone who is about to flame TrollTech should think about that.
Philippe
P.S.: I have a TrollTech-Free computer.
I'm thinking (with no particular evidence) that Trolltech is the company that may be involved in the rumored FSF GPL testcase.
Extremely unlikely. If it were true, RMS and his lawyers would walk out of the courtroom five minutes after they entered, then proceed to the men's room to wash the egg off of their faces.
The FSF can only sue on behalf of its own software. It does not hold copyright to any part of KDE or Qt. Furthermore, Trolltech doesn't even release anything under the GPL. What possible legal grounds could the FSF have against either of these entities?
A Government Is a Body of People, Usually Notably Ungoverned
The GPL says "normally distributed (in either source or binary form) with the major components".
You say "If the library in question is a standard part of the"
Now, the last time I checked my dictionary, 'with' and 'of' had very different meanings. One version or another of Qt is distributed with the major components of all free unix-like operating systems, including Debian. But even if it weren't, all you would need to do is distribute the Qt source code.
A Government Is a Body of People, Usually Notably Ungoverned
Qt can't be used for internal/research projects w/o paying $1500
Bullshit. All they have to do is make it open source. Hell, they don't even have to go that far. All they need to do is give Trolltech a copy.
It wasn't too long ago when half the people here at Slashdot were excoriating Corel for having a closed beta distribution. Now suddenly when the license isn't the sacred GPL people are demanding the right to private distributions.
If you distribute the program you distribute the program. No if, ands, or buts. It doesn't matter if you distributed it to your coworker, you still distributed it.
A Government Is a Body of People, Usually Notably Ungoverned
Among other things, Carter wrote:
The draft license seen by me last before release of the final QPL was GPL compatible. I was proud of it. So, it seemed, was Troll Tech. And then the final license was released, undoing the parts of my work which made the license GPL compatible, but retaining enough to satisfy the definitions of "Free" many distributions (including Debian) use.
But the license issue remains. Qt is not non-free software. But it's not GPL compatible either. Some KDE core developers admit this privately, but won't do so in public because of the implications: that much of KDE is not legally distributable until they contact some people that are damned scarce these days and make the necessary arrangements.
There's a lively debate on that page already.
No, you're the jerk. You so caught up in Stallman's rhetoric that you fail to see true freedom when it stares you in the face.
'banbeans' chose Linux of his own free will. He was not under any duress or coercion. The fact that he was able to use Linux without legal or physical hinderances demonstrates that he was in fact free all along. He was free before he used Linux, he is free now, and he will still be free even if he chooses a 'non-free' OS. Maybe when you and your cohorts stop calling free men slaves you'll get more respect.
A Government Is a Body of People, Usually Notably Ungoverned
I've got no idea, maybe they were working on this change for a long time, maybe not...
In any case, it's clear why the article was published now, as a response to Carter's. What I fail to see is, how this necessitates the events described by you.
-- KDE programmer and computer science student in Klagenfurt, Austria.
The main changes we plan to make are to remove the patch clause and clause 6c.
The patch clause was meant to avoid leaving Qt users with slightly incompatible versions of the library without the possibility to tell where the code stems from. However, the Qt user community is now so large that we believe that this is less likely in the current situation. We also see that people tend to send their patches to us so we can include them in the official versions of Qt.
Clause 6c has been claimed to be the major reason for GPL incompatibility in the QPL. This clause gave us the possibility to ensure that companies writing internal Open Source software indeed release their source code to the general public.
The QPL version 2.0 will hopefully end the license discussions once and for all and get us all back to coding again.
Hopefully this action will be enough for the Debian folks to be able to include Qt-linked software. So... a Linux distro got that far to let a company re-license it's software? Really, talk about devotion...
But, as always, the real zealots will find something to complain about as it isn't adopted to their holy GPL. Probably some complaints about certain programming languages et cetera.
Really, if some of us would start being productive, a hell of a lot more could have been accomplished by now.
What for?
Besides, that would not alleviate the problems of Debian (I don't agrre with their interpretation of the GPL, but let's pretend it's correct for the moment), which is their claim that distributing binaries of GPLed programs linked to Qt is illegal.
-- KDE programmer and computer science student in Klagenfurt, Austria.
I want to write the app using Qt because I like the API. Is it ok to release only the parts which contain Qt API calls, and keep closed the rest (because I have some killer-algorithm in the app's engine which I don't want to publicy disclose). ??
Since the GUI sourcefiles of my app call other functions (and include the engine's headers) which are located in other files, the code I would release would of course not work. (does not compile because of missing parts)
My question: am I breaking the QPL by doing so ?
Or do I have to revert to a client-server model (GUI client opensource) and engine (=server) in order to avoid breaking the QPL ?
thanks for your infos.
GPL allows linking (dynamic and static) to libraries that are not GPL compliant if they are considered a part of the OS.
Define OS. This is a weakness in the GPL. Linking to anything from GPLed code should be allowed. IMHO linking anything dynamically to GPLed code should be allowed as well; a library simply provides the implementation of an API. Should a proprietary app be prevented from calling a GPLed ls?
It's really not TrollTechs job to clean up the licensing mess KDE made. The QPL meets all the requirements of the DFSG just fine. It's KDE which should have thought about the implications of using the GPL. I'm a Debian developer and enthusiastic KDE user and it breaks my heart to see two great projects not coming together because of silly trivialities. And it makes me angry that KDE is still trying to avoid and deny the problem instead expending a tiny amount of energy to fix it once and for all.
What you're missing is that there are many people who would prefer the GPL to be viral, not in the usual sense that it's accused, but in the sense that it infects anything which touches it. Is it illegal for a proprietary app to call a GPLed ls? Of course not; that's utterly absurd. So why do some think that dynamically linking to GPLed software should be illegal? Because they are fanatics.
I appreciate the GPL; I dislike proprietary software. But I am willing to co-exist. As long as my code remains free and open, I am happy. If my code is dynamically linked against but the source is included, I am happy. Free software will win in the end, for the most part. I'm sure that there will always be some proprietary software. Big deal. I needn't exterminate my enemy, but simply defeat him.
Actually Corel and Debian had a dispute with dpkg (the package tool in Debian). One of the authors had the opinion that the proprietary command-line interface of dpkg couldn't be used from a non-GPL GUI application. But RMS and others didn't have any complaints against it.
The command 'ls' is part of the POSIX standard, so there's nothing unique about that command.
Actually, in fairness, the GPL does make a special exception for linking againts system libraries. That term is not well defined, though.
Well my impression was that this is not the
whole problem. But as to the problem described
on the web site: I don't think it's real.
Sure the GPL may contain restrictions the author
doesn't actually want. This is only relevant if
that license is the entire basis of your contract
with the author. Contracts in general do not
have to be in written form. Neither do amendments
to contracts.
E.g. waving to a waitress with two fingers
can create a contract: "I want two beers" (well
assuming the right environment).
Also drinking a beer which has been put in front
of you creates a contract - now you have agreed
to pay for it. (This is true even if you didn't
actually order the beer.)
The same thing applies with KDE. The programs
calls the QT library. That means the author calls
for that library to be included. So that
constitutes permission to link against it.
That's all that's needed.
It gets more interesting if the KDE code contains
other GPL code. The example I've been told
about is kghostscript, I assume there is more.
(Is there a list anywhere?)
Then if you feel that QPL+GPL doesn't work
(I'm not really clear on this) you'd have to
get the author of ghostscript to agree to the
linking. Or politely ask the Troll people to
make a change to their license.
However I don't see that changing the KDE license would:
- be necessary in any way or
- provide any benefit at all
I stand corrected.
-Denor
OK, this will be a real big example. Exageration are intended to show the point. Take a big company, said MS. Let this company removing there dear Explorer and replace with, said, Konqueror. Let this company use there dear "Embrassed and Enhanced" practice and modify it so Konqueror can now display multiple new MS-XML proprietary extensions through a DLL. Patches for Konqueror are distributed normally under the GPL, but no code for the libraries which even required a special registrations and agreement to not used it under anything else than MS products. Aren't this fare? (1)
I think not. Konqueror is licenced to be free. Free to use by anyone on any OS. The example addon made by MS in my examples just made Konqueror-with-MS-XML a part of MS with no regards for the intentions of the original authors, which should be clearly disclaimed in their license. They treat their code mostly like public domain code, breaking the usability of the code and crippling the software for use on other OS. I really feel like KDE (not TT) take all this issue the same way. They take GPL software, add dependencies over a non-free library (remember that most KDE 1.2 are release over Qt 1.x which aren't under the QPL), and doesn't even ask for permissions of the original author. They choose to ignore the warning of several people, mostly because they don't care about all this and don't believe it there are any issue. If there are no issue, why they simply don't add the "exemption clause" stating that you can link the program (including ghostview) against Qt?
Instead of simply make their lessons KDE choose to don't care about and then complain about conspiracy from Debian because is the only distribution who refused to engage themself into distributing codes where the license said you doesn't have the right to do so. KDE simply refused to clarify this issue, preferring hide their heads in the sand of their own interpretation (are they lawyers? have they ask the original authors?).
I'm really happy to see TT trying to make the QPL 2.0 compatible with the GPL but this doesn't change the arrogant attitude that KDE has with regards to the intentions of the authors, stated by the license under which the code is distributed. They modify it and distribute it, adding dependencies with a more restrictively licensed library, without even asking the original authors if they can, especially when the issue was often points by many as not that clear as they claim. The simple fact to trying to reach those authors would had prove their respect to free software authors but their mailing list show that even initiative to do that by some members was clearly seens as a waste of time.
For this, I avoid KDE the best I can, whatever the technical merits of their desktop, just like I avoid MS: they both behave badly regarding usage of other people code, making cheating a common practice that we should ackowledge.
(1): I know that the GPL has loophole permitting this. MS just have to distribute the DLL with their OS and sell Konqueror-MS-edition a separate bundle. Is not my intention, neither my hability, to analyse such text. I will try to make my point by defining the licenses as a (fuzzy) claim of intention from the author of the software. When the intentions became ambiguous, it should be the responsability of the authors of derived works or the distributors to ask the original authors for a clarification. This task is necessary to improve the license and not build our relationships around misinterpretation.
Fabien Niñoles - Debian Maintainer
From what I can see, it would be much harder to release proprietary QTized code that used GPL stuff too (almost impossible, I'd say).
I don't get mad at MS for releasing non-free software. I get mad at them for releasing shitty non-free software and trying to pretend that it's the best there is.
`ø,,ø`ø,,ø`ø,,ø`ø`ø
Free Software: Like love, it grows best when given away.
So what is the difference, exactly, between a command and a function in a shared library? Both implement an API; the one in a shell, the other in some other language.
And the bit about ls is exactly my point. It's an implementation of an API. Can non-GPLed work use the GNU extensions to ls?
Allow dynamic linking; disallow static linking. Static linking defeats the ability to rebuild th binary, but dynamic linking allows any library in the background. If we wanted we could write a GPLed Qt-alike.
Actually, that sounds quite fair. We would have the patches; we could write a free library to implement that API. It's not as though it is some undocumented API; we would have the code, and thus have all the docs we need.
Today, any company that gives away some source code clams to be an Open source company and, not counting the community's opinion on it, they are. So they release "open source software under a licence which *RESTRICTS YOUR FREEDOM* and some outside developers start to work on the code but the company still profits on his work in an exclusive way because the licence is not a Free(freedom) Software licence.
It's speculative prediction on my part.
In the context of competing products, like KDE and GNOME are, as long as both a viable, they will eventually become indistinguishable by the users (like CocaCola and PepsiCola, McDonald's and Burger King, IE4 and NC4, Intel and AMD, etc.).
Imagine a day when both GNOME and KDE are in that state. Which one would people choose to use? The one with a licensing shadow hanging over its head, or the one that's GPL/LGPLed free and clear?
For the 90% of the people who don't care, it'll be half and half, so 45% for KDE and 45% for GNOME. For the 10% who cares, it'll go GNOME! So GNOME gets 55% while KDE gets 45%.
Granted this day won't come within this year, or the next, but as long as GNOME is alive and well and attracting developers, it will eventually come.
The current technical merit (which KDE/Qt seems to have a lot over GNOME/gtk+) doesn't really matter in this argument.
Like I said in the first post, the best Trolltech can do is to make Qt at least as free as gtk+, and hope to get even at the end. But I'm afraid that they can't do that until too late. In a sense, the QPL was already too late.
(Netscape gave up on the $29.99 revenue for every copy of their browser too late. Motif released it's (semi-open) source too late. etc.)
These are no accidents. Proprietary technologies won't have a reason to lower their price until a competitor come along. But if a competitor gives out the same functionality for free, then the proprietary technology is hosed! Because they CAN'T afford to make their product completely free. They have to attach a little string to their 'free' product, because the livelihood of every employee in the company depends on it. It is that final string, however short, that separates them from their truly free competitors.
-- Weiqi Gao weiqigao@speakeasy.net
Then nearly anything can be compatible with GPL. Either you concede the point that 2. doesnt apply for dynamic linking or everything has to be licensed under GPL. There is no middle road.
It is not illegal by any means to compile and link programs against Qt. The GPL has no restrictions on this.
What it is not legal to do is distribute such programs, because Qt violates one of the terms of the GPL through its discriminatory licensing. It can be made legal again by appending an exception to the GPL.
I develop a GPL'd product for the MacOS which uses PowerPlant, a proprietary toolkit. But my program is perfectly legal to distribute. How, you ask? Because I add the following to the license:
Now, Qt does have a pseudo-Open-Source license, and the code can be distributed (unlike PowerPlant's). This makes a Qt-style exception even easier. In fact, it would take no more than two sentences...
...and all of the licensing isues magically vanish. It really is that simple. Why the KDE team can't swallow their pride and admit that they made a very tiny (and understandable) oversight in their choice of license is beyond me. It's not like it would be that much work; I could do it with a Perl script in five minutes, including the time it takes to write the script. They never struck me as being too arrogant to admit their mistakes; I've seen them do it in the past. If they've kept decent records, they should still have contact information for the developers, not a single one of which would refuse to make the program legally distributable.
Honestly. This issue is so simple, and it can be resolved in a manner that leaves everyone happy, except maybe RMS. What's the big deal about?
Have you ever tried to maintain a Perl system? (Hint, try "perldoc Test".)
Have you ever compared how much nicer it is to have an automated test-suite than it is to not have one?
Trust me, any large distributed project that is used across multiple platforms can benefit from a test-suite. Making it easy to build one is a good thing. Being able to drive an interactive gui program that you may not have source for from a command line is in and of itself darned useful. (The lack of an integrated ability to be so driven during batch processing is a major drawback of most of them, and is one of the things that I dislike about having to work with stuff that was thrown together in VB.)
OK, now that I have addressed usefulness, let me address Debian. Debian's problems would be totally resolved by this. The program would be distributed with dynamically linked against a program with no license problems. The fact that most users would actually link it to Qt after you distribute it is just fine. The GPL doesn't prevent you from doing that, it only prevents you from distributing it that way.
Cheers,
Ben
My usual seat in the cluetrain is at A HREF="http://pub4.ezboard.com/biwethey.ht
Why are you guys harping on TrollTech? Aren't the KDE guys the ones who are doing the illegal action by incorporating non-GPL compatible code? It's like sueing Sun if Linus incorporated non-GPL Solaris code into Linux.
A deep unwavering belief is a sure sign you're missing something...
It's KDE's developpers that the free software community should be pointing fingers at. One of the main thing advocated by the FSF is to value freedom over convenience and the GPL serves as a legal groundhold to that philosophy.
The problem is that I'm pretty sure a lot of the KDE developpers don't grasp the difference between Free software and Open Source. Why would anyone who understand what they're dealing with choose to put oneself in such a (legal) mess?
Unless of course you're one of the KDE developpers that work for TrollTech. Then, you know it't good for you if every developper that wants to code a GUI app to the GUI's native toolkit has to pay you if they want to develop closed source. It's really that simple yo know. You develop a desktop that depends on you proprietary software and try to make it become a standard on Linux; then you sit tight as you watch the money pouring in.
Wait a sec. If KDE's main developper is working for TrollTech, that make them as guilty as anyone. This is just plain parasitic.
nasty nasty nasty this is exackly what im talking about
Atleast I had the guts to put my nick on my post
and im sorry if the only advantage linux
has going for it is the gpl then it deserves to fail
No i dont think the only advantages to linux are the gpl or I wouldnt use it
Linux is one heck of a server os and i use it for that
but as a desktop os it isnt ready yet
and being an ungratefull snot
nosed name calling pup like you
when someone offers us something isnt going to help that at all
Again I would like to thank all of the peaple that put thier time energy and skill into making
linux/*bsd/qt/kde/gnu suck less:}
Actually you forgot something else. Companies tend to produce two other good things. :), Visual Studio vs. KDevelop, MSVC or Intel C/C++ vs. GCC, etc. Actually, about the only best in class Open Source software that is widely used is Apache, and maybe Samba.
A) Best in class software. If I have the jingle for it, I am usually much better off buying commercial software because they tend to be superior products. Photoshop vs. GIMP, 3D Studio vs. Blender, Solaris vs. Linux (for my quad proc SPARC box
B) They tend to innovate more. I've yet to see Open Source software that really redifines a market segment or brings something new to the table. All the high profile stuff is essentially a free implementation of stuff that's been done before. KDE is a CDE-workalike. Even though its better, its still nothing new. Same thing for GIMP, Apache, KDevelop, Blender, Gnumeric, Bonobo, KParts, Konqueror. None of these things really have any new features that make you step back and go, "wow, I never expected software could do that!" Even MS induced those feelings the first time you embedded an OLE object into Word. Even something like Berlin, which is pretty nifty, doesn't have anything that's been done before.
A deep unwavering belief is a sure sign you're missing something...
OpenDOC was released to developers a few months after OLE applications were already on shelves. Unless there was another OpenDOC before the Apple on, I'm pretty sure that OLE came first. (Though IBM's SOM predated COM.)
A deep unwavering belief is a sure sign you're missing something...
Speculative indeed!
Though your reasoning seems to be more sane in this post than in your initial one, I still fail too see how it could be too late NOW, after you've just stated "Granted this day won't come within this year, or the next,...".
Especially since "this day" only means 55:45, according to what you said. *wonders*
-- KDE programmer and computer science student in Klagenfurt, Austria.
GPL allows linking (dynamic and static) to libraries that are not GPL compliant if they are considered a part of the OS. Hence, an application in windows that uses WIN32 API has no problems. Qt is not part of the OS on any system, not even Corel Linux.
So it is not okay for one to link against the QT libraries, because they're not part of your system - but it's okay for them to link against, say, the Metro-X libraries? I fail to see how the X Server is in any way shape or form part of the OS, yet it is okay to link against a decidely PROPRIETARY program.
The copper bosses killed you, Joe. 'I never died', said he.
If I have the jingle for it, I am usually much better off buying commercial software because they tend to be superior products.
The super-configurable Gnome panel, with its myriad applets, has been making my life easier and desktop prettier ever since October Gnome, and is further improved in Helix Gnome. Nothing it does is 'innovative' per se, but it is the best panel I've ever seen, anywhere.
None of these things really have any new features that make you step back and go, "wow, I never expected software could do that!"
Debian's APT is the slickest installer/updater/package-manager in the universe. I dread the thought of going back to RPM-based distributions (let alone Windows).
As long as they try to beat GTK by enhancing QT I am GLADE ;). This is the spirit of capitalism, competition, and this is why having both KDE and Gnome is good, as long as everybody plays fair.
"The obvious mathematical breakthrough would be development of an easy way to factor large prime numbers." Bill Gates,
GPLv2, end of section 3, starting halfway through second-last paragraph.
"However, as a special exception, the source code distributed need not include anything that is normally distributed (in either source or binary
form) with the major components (compiler, kernel, and so on) of the operating system on which the executable runs, unless that component
itself accompanies the executable."
Pretty clear. If the library in question is a standard part of the operating system you are designing for, you don't have to include source.
If you are shipping the library with the GPL'd program, then you DO have to include source.
So.. if libqt is part of Mandrake 7.1, I don't see a problem with saying that you are coding for mandrake 7.1, and that your software is designed for mandrake 7.1.
It is a bit misleading.. it should say 'intended platform' instead of 'platform on which the software runs', as that is non-specific.
Except, if QT is considered a standard part of the OS(and if the distribution includes it, then it is), then they are free and clear.
From the GPL (Section 0):
'The "Program", below, refers to any such program or work, and a "work based on the Program" means either the Program or any derivative work under copyright law: that is to say, a work containing the Program or a portion of it, either verbatim or with modifications and/or translated into another language.'
(emphasis added)
This phrase makes it seem clear to me that a non-GPL'd library cannot be linked to a GPL'd program (which is then distributed).
"...Stallman himself has written code which had to link in proprietary libraries."
I'd like to see some references on this. Make sure that we are talking about STATIC linking, too. Dynamic linking against non-GPL'd libraries is probably OK.
The key here is distribution. When you ship a GPL'd binary, the entire source code to make that binary must be included. If you've linked a library staticly, then the source for that library must be available. Another quote (Section 3-3a):
You may copy and distribute the Program...in executable form...provided that you also...[a]ccompany it with the complete corresponding machine-readable source code
Again, if it is in executable form AND the libraries are static then the "complete source code" includes the library's source code.
--
Linux MAPI Server!
http://www.openone.com/software/MailOne/
(Exchange Migration HOWTO coming soon)
*The QPL did not kill Harmony. Harmony programmers were hired by RedHat for whatever reason. I won't speculate on why, but you can...
*Gnome and Gtk are in no way comparable to Kde and Qt. Yes, with "October" Gnome and now with "Helix" 1.2 Gnome is stable in the sense that it doesn't crash so much. But Gtk is nowhere near Qt in quality or design or ease of use for application developers. Gtk and Gnome apps show this. Gtk is simply not professional quality and it would have to be redesigned from scratch to become a professional quality toolkit. Being built on Gtk, Gnome can never really be competetive with Kde and will remain attractive mostly to GNU zealots and commercial entities wanting to profit from free software as explained below. Please read on....
*You may be correct that Gnome may become the default desktop for Linux but not soon. If that does happen it will not happen for the reasons you cite above. It will happen because:
1. Kde will have attracted more new users to Linux by offering a desktop with usability and power and the apps most users want. Kde will do the hard work for Gnome, which gets a free ride on Kde's coat-tails, just as it always has.
2. It is easier for commercial (proprietary) apps to integrate with or interface with Gnome because its license allows that. Apps that use Kde must be non-commercial and open because of the QPL. When linux desktop market share approaches 20% commercial companies will want to use LGPL'd libraries for that purpose. Right now the linux desktop share is around 5%. Within the coming year, mostly because of Kde2, it may well reach 20%.
As Gnome is already heavily backed by several large commercial entities (but not by most distributions of Linux, which will continue to use Kde as the default desktop) Gnome will probably be the desktop of choice for closed, proprietary apps. It was designed for just that purpose - for corporations to make money off free software with proprietary apps.
I wish none of the above were the case. I wish that Gnome/Gtk were a professional quality toolkit and desktop environment competetive with Kde/Qt. But there is really little incentive for Gnome/Gtk to become what it should become. Gnome/Gtk will probably be able to get by with being just "good enough" as a foundation for inefficient, bloated and poorly designed applications and subsystems which commercial software vendors are used to. They have dealt with MS for years, and so are well prepared for Gnome/Gtk.
Fortunately, for the immediate future, Kde will retain dominance. Kde2 will be the biggest drawing card Linux has for home and corporate desktop users during the coming year. You can expect the Linux desktop user base to triple or quadruple within a year thanks to Kde. This gives Qt some breathing room to make its license more attractive to commercial developers who want to use Kde or better integrate with a Kde desktop. I think eventually Troll Tech may just fork Qt into separate commercial and free versions and give the free version to Kde to maintain and develop under GPL. They won't acknowledge that just yet. Kde does need the ability to better integrate commercial applications which presently can't use Kde as a foundation for their proprietary apps.
But none of that really matters. Gnome and Kde will both become irrelevant in a few years anyway. I had a look at the new Amiga SDK for Linux. The Amiga desktop subsystem will run on any OS, not just unix-like ones, and I was shocked by how good it is. It uses a Java byte code implementation and some other low-level tricks to achieve this, and it is VERY easy to port posix compliant apps to the new Amiga and/or to code new apps for Amiga. This system is truly revolutionary, not just evolutionary. Like the original Amiga was in its time. It works, right now, with Linux. Corporations will love it. It meshes beautifully with their plans for thin clients. Users will love it. GNU zealots and Troll Tech will not. Compare $100 for the Amiga SDK to many thousands of dollars per programmer to use Qt professional edition. Troll Tech can no longer get away with its exhorbitant pricing scheme because it finally has competition which is very reasonably priced and platform independent in a way it can never be. The best thing about the new Amiga system is that it offers a smooth transition from the unix desktop to something new and revolutionary. It also offer the same to those currently using MS Windows.
Linux and other free unixes in their pure form may remain the system of choice for free software zealots and for servers, but Linux + Amiga and then just Amiga will replace the MS desktop for most corporate and home users. Gnome, Gtk, Kde, Qt and Microsoft will become irrelevant, and it will happen within just a few years. Other systems which use linux mostly to bootstrap Java for thin clients will fill in the gaps for corporate networks that Amiga is not suited for. I am afraid to say that Java and/or what evolves from Java is the future, not unix. MS sensed that Java could do that, but the manner in which Java will realize the darkest dreams of MS will be totally unexpected.
You heard it here first!
Moderator - please mod this up to at least a 1 so people will read it. I have reasons for posting anonymously.
The GPL permits making proprietary derivative works, as long as they are not distributed. The big difference is that Troll Tech wants to derive revenue even from internal usage (research, internal applications). Effectively, that means that every developer at a commercial site needs to get a commercial Qt license, an expensive proposition, even if they never produce a money making product.
Please check your facts before spouting such utter drivel. The Qt foundation is anything but garbage. It is true that TT hired one of the KDE representatives, but you fail to mention that the representative then stood down and another took his place. The foundation still exists and still acvhieves it's purpose - ensuring that Open Source developers will always have access to the Qt source no matter what happens at Troll Tech.
It is not true that the GPL prohibits linking non-GPL code, the situation is nowhere near than simple. You would have to delete every GPL graphical program you have if your reading was true, because Xlib isn't GPL.
I develop a GPL'd product for the MacOS which uses PowerPlant, a proprietary toolkit. But my program is perfectly legal to distribute. How, you ask? Because I add the following to the license...
... but nobody can add any GPLd code to Millenium's-Almost-But-Not-Quite-GPLd code without falling afoul of GPL 2b) again ... because PowerPlant is part of the whole, and although you can certainly grant an exception for your own Not-Quite-GPLd code, you can't grant an exception for anyone else's GPLd code.
... this accomplishes nothing, really, other than setting up Yet Another Schism, between The Real GPL(tm) code and Millenium's GPL Except For PowerPlant(tm) code.
Actually, that doesn't cut it, because that exception is fairly obviously (OD: IANAL) a restriction of the GPL as defined in section 6. It works as long as nobody adds anything non-original to your distribution, true 'nuff
So
I'm also quite mystified why on earth you think it's sensible to make whatever this PowerPlant-based project is GPL anyway, since that very basis would seem to make it philosophically unacceptable to anyone who thinks the GPL is a useful license.
Let's face it, basing GPL code on non-GPL frameworks is a pretty bloody straightforward violation of the spirit of the GPL, isn't it now? Whether you can come up with sneaky interpretations to pretend to justify it, the fact remains is that it is not part of RMS's political manifesto, as anyone who thinks about it for a minute rather than scurry around trying to twist the GPL until a loophole appears will realize.
Really? I've certainly not seen that definition at www.opensource.org. It might be true for some developers, but not all, and I would argue, not even most.
Some counter-examples:
- RMS - "I don't like proprietry software. I don't want to use them so I won't subject you to them either. Get rid of proprietry systems because I don't like them.", It really is a selfish viewpoint (actually selfish for the masses). RMS didn't release GNU tools because he didn't want money for them, he did it because his ideal world is full of free software, and he wanted to live in that world. That's cool, but it's not charity, it's idealism. There's a difference.
- Tivo - I will release my linux kernel changes as open-source because the licence forces me to.
- BeUnited - We release open-source because then we get free support from Be engineers.
- Slashdot - We release open source software because the community has produced the tools that we run the site on, and we think that it is appropriate to give back to that community. You scratch our back, we scratch yours.
In all those cases, the developers have an agenda (generally quite a valid one) they are seeking to acheive, and going opensource works to that end.I really can't see how Trolltech is any worse than the above.
Your suggestion that TT only provides free versions of QT for their own benefit, it IMO bogus. TT has (eg) a Linux FTP server free for download. Why? Because they wrote one, and have no need to keep it to themselves. True sharing. In fact, it meets your open-source definition perfectly.
Why is it so hard to believe that a company does not care about the community/users? The people who run TT are PEOPLE. Just because they are running a company doesn't make them evil. Two guys writing GIMP as individuals and releasing it as OpenSource is sharing, but a few guys writing a GUI tookit as a company, and releasing it as OpenSource isn't??
I don't think the cluetrain is visting your town.
--
Read more of this story at Slashdot.Read more of this story at Slashdot.Read more of this story at Slashdot.
Should we perhaps have a compatibility matrix showing which products are legally compatible on which platforms? Or how about a "license manager" where you just click on the licenses you accept?
I can imagine it: "This application has performed an illegal action and will be dragged to court for punishment".
Or how about: "License violation at 0xca85e2a - the application will be terminated".
Seriously though, this could cause problems getting Linux into the big-scale companies where everything has to conform to standards. Imagine getting a sale rejected because your product wasn't "GPL-license certified".
What is the sound of one hand clapping?
cat
I am certainly glad that Trolltech was not in this position, as they were already OSS supporters, but even they, after having showed this support for several years, still get flak from the community because there license wasn't written to the critic's specifcations, as if Trolltech was going to announce tomorrow, "Ha, fuck you! We left a nice little loophole in our license, so no more damn source code for you!"
.GIFs (everything was nice for years and years, until everybody was using it, and then the "pay us royalties or we'll sue the shit out of you" come).
Why, that is exactly what happend to
This is exactly what is happening to MP3s right now.
And they all were nice guys in the beggining.
I am a lawyer, but this isn't legal advice. If you need legal advice, see an appropriately licensed lawyer.
:)
:)
>>There is nothing illegal about writing GPLed
>>code which uses the Qt API and libraries.
>Writing a derived work from GPLed and QPLed
>source is not the problem or illegal. The
>question is wheter distributing a derived work
>based on a GPLed application and a QPLed library
>is.
No, there's a question even before that--to which the answer is that "KDE is not GPL software."
It may or may not be true that KDE has violated the GPL license of other software by including it within KDE. I don't know (and am not even interested, for that matter
However, it is *not legally possible* for the authors of software to violate their own license. The general claim that "Program XXX violates the GPL" is generally nonsense--a program may violate other licenses that provide material, but the GPL is not some abstract; either it provides the terms for that piece of code, or it does not.
In the case of KDE and several otehr purportedly GPL projects, it does not. These are a class of projects which put out code, call it GPL, but at the same time invite, implicitly or explicitly, people to modify and redistribute--in spite of relying upon (typically) libraries which are not subject to annexation by GPL software.
The action (publication) is inconsistant with the words (GPL), and the law resolves the conflict: the actions govern, and the contract/license is "reformed" in accordance. The software is not GPL, but Quasi-GPL (QGPL).
This doesn't necessarily solve the problem, though. For example, I mentioned above the possibility that GPL software was included in the QGPL software; this could be a problem (offhand, I don't see how it couldn't be). In fact, the various QGPL licenses may be mutually incompatible, depending upon which terms of the GPL must be tossed for each.
Bottom line: release software with an invitation to develop and distribute, proclaiming "GPL" from the treetops, but link to a non-annexable libray, and you land in the land of QGPL, not GPL.
Ugly? Yes. But that's life
hawk, esq.
The article seems to cleam that the distribution is legal, but that does not make it true.
>There are very few OSS projects(maybe with the
:)
>exception of KLyx) that really revolutionize
>their market segment.
Klyx??? Klyx was a one-off port of a badly out of date version of lyx. It allowed multiple windows to be open at once, and cut and paste between them, but twas otherwise *very* far behind. Unless things have changed recently, there is *no* plan for another version of klyx prior to the toolkit independence of lyx (though there was initially a plan to eventually merge the two).
however, if you s/klyx/lyx/ , I'll agree with you. I'll also point out that lyx solved the QGPL license problem
I am a lawyer, but this isn't legal advice, etc.
:)
Actually, even if you don't *say* so, but just *act* that way, the same thing happens.
*However*, whether by your disclaimer or your action, you're going to have trouble including other GPL code once you do this.
I don't agree with rms on much of anything, but, from a legal standpoint, it is *quite* usefull to have a standardized way of doing this. The only way out of this mess that I see is for the possiblity of such a clause to be included is placed in future versions of the GPL--which effectively places that clause in all GPL software, which will create new religious wars, and a fork of the GPL
Currently, there are programs that have done the same thing--I wrote the LyX disclaimer, another is mentioned above, still others do it by actions. It is fare from clear that these projects can currently share code; if they all used the same disclaimer (whether mine or rms'), they could.
hawk, esq.
How exactly is what I posted flamebait? Sure, I insulted an AC, but certainly with reason: his post was mere assertion apparently uninformed by reading the article. I pointed out that the article itself addressed his points.
I'm quite willing to take moderation based on objective facts, but being modded down solely due to having expressed an unpopular opinion is a bitter pill to swallow. I know that when I moderate I take the job seriously. Generally I only mod down people whom I agree with, but whose method of posting is indelicate, for that is the only way that I can be objective.
Insulting an AC's mental skills is hardly flamebait, esp. when his post is obviously ignorant and he has not read the article referred to.
It got me thinking... Does the API itself fall under the copyright? For a while there, the Harmony project was working (If I understand things correctly) to create a GPLed library with the same API as QT. With two libraries out there each with the same API, one free and one not, that really kind of muddies the waters. I'd think you could ship object modules (Not source) even if one of the libraries was pure GPL. I wonder if Stallman ever considered this scenerio.
I'm trying to teach myself to set people on fire with my mind... Is it hot in here?
Troll Tech's business model is basically that Qt is free for free software ("free" in the "free speech" sense), but if you want to make proprietary software with it, you pay Troll Tech. If you are serious about making proprietary software, and you don't have the money to pay Troll Tech, then I suggest that you use another toolkit that is free even for proprietary use, such as a toolkit under either the LGPL or a BSD-ish license.
All the high profile stuff is essentially a free implementation of stuff that's been done before. KDE is a CDE-workalike. Even though its better, its still nothing new. Same thing for GIMP, Apache, KDevelop, Blender, Gnumeric, Bonobo, KParts, Konqueror.
I take your point about the others, but what was the proprietary software that Apache imitated?
Also, wasn't Mosaic open source too? Okay, so it doesn't dominate the market but neither do proprietary trailblazers like Visicalc. Most open source software is derivative because, well most software in general is derivative.
. . . it's easier for everyone if Troll Tech does the changes. If Troll Tech changes the QPL, then the change propagates to every user of the QPL'd Qt, not just KDE. This also means that hundreds of contributors to KDE don't have to be contacted to get KDE licensed under a GPL+Qt exception.
Apache is a webserver, as such, it imitates other web servers already in the market. Also, Apache is very very good at what it does, but doesn't really have any features that haven't been done before. My point about OpenSource sofware being derivative still stands. I was talking more about features than the genere of the software. Stuff like document embedding which was introduced (to the masses) via OLE and later OpenDOC, or the nifty media node features of BeOS, or the array of cool features like Intellisense in VisualC++, the great integration of desktop publishing and word processing in Word Perfect, the web integration of fireworks. All these things are new, innovative, and genuinely usefull features in products that are otherwise just an image editor, or just a word processor. The problem is, that there is very little open source software that brings these kinds of new features to the table. There are very few OSS projects (maybe with the exception of KLyx) that really revolutionize their market segment.
A deep unwavering belief is a sure sign you're missing something...
The fact that Qt is well documented is in itself a big time saver.
Again, Linux is much more than a kernel. Qt is just as much a system library as Motif was for older versions of unix if and when distributions ship with Qt. That decision is entirely up to the distributions, not to GNU zealots and license police.
If you make the assumption that QT is a system library, you still fall foul of unless that component itself accompanies the executable. Sure, shipping KDE would be fine then, but you wouldn't be able to ship it with your distribution.
Arg! Okay, so, let's say you like a GPL program(let's say Program) to QT("QT"). Your binary file now has TWO licenses and TWO seperate parts. There's the Program, and there's QT. QT is not a derivitive of Program, so it doesn't have to be GPL'd. While QT functions are embedded in a statically-linked Program executable, they ARE STILL QT, and therefore are a) under the QPL, and b) not a derivative work of Program. I can distribute a statically-linked program that uses QT, so long as I say "Some parts of this binary are under the QPL, and some are under the GPL. The source for the GPL parts is available at ."
Now, if QT was to link against a GPL'd program(NOT the same as a program linking against QT), then we'd have a problem. But QT only links against libraries that it's allowed to(ie: either LGPL or some other license is applied to the libraries).
Barclay family motto:
Aut agere aut mori.
(Either action or death.)
I have been trying to make this point for ages -- that Qt can't be used for internal/research projects w/o paying $1500, or more, per workstation, to TrollTech.
Very well written. I agree that it would be better to use any of the others (even give up and go MS) before using anything that depends on Qt.
The fee for the commercial license is ridiculous. I can get a great OS, window system, window manager, thousands of programs, distribute them thoughout a company and develop on them internally for free, without releasing any code (as long as GPL licensed code is not modified).
Add TrollTech's Qt libs and all of that grinds to a halt. The MS solution is far cheaper in such a case.
Treatment, not tyranny. End the drug war and free our American POWs.
See my user info for links.
I was just thinking the same thing! You missed a couple of examples: WindowsNT vs. Linux/*BSD, MS Office/WordPerfect/Lotus SmartSuite vs. just about any non-office editor/LaTeX, IIS vs. Apache. Oh, wait, you covered Apache.
A student of mine told me that he thought the GIMP was actually better than Photoshop. Superiority is in the eye of the beholder.
They tend to innovate more. I've yet to see Open Source software that really redifines a market segment or brings something new to the table.
You know, you're right here, too. Good thing proprietary companies developed closed-source versions of BIND, sendmail, Mosaic, perl, and python, so that others could imitate them and release their imitations as open source.
Okay, so we haven't seen a huge push of innovation from open source lately. We haven't seen a huge push of innovation from closed-source lately, either. Products take time to develop, and to say that companies innovate more than open-source projects, or produce better software than open-source, is a gross overgeneralization. To say the other way around is also an overgeneralization: each software package needs to be evaluated on its merits.
However, whenever *I* do such an evaluation, true open source projects *always* get an extra point or two, because I can trust that if the software gains a solid developer/user base, it will continue to improve, and will continue to be free. I can't ever say the same for closed-source, proprietary software.
--keith
To quote from the editorial:
The QPL version 2.0 will hopefully end the license discussions once and for all and get us all back to coding again.
Not likely, Eirik. Sorry to bear bad tidings. Developers such as these try to do the right thing, but they aren't lawyers, and open source licenses still have yet to be put to the test in the courts, so nobody is really sure how they are going to stand up anyhow. Hell, GPL supporters spend gigabytes arguing over the fine points of the GPL, nevertheless other licenses. So there are no truly effective guidelines to creating an open source license, and "just use the GPL" doesn't work for everyone, nor should it have to.
It seems as though there is a decidedly hostile atmosphere surrounding open source software; when a company decides to open the source to their code, they seem to be opening themselves up to attacks against their business practices by the OSS legions, even when they had no problems when the source was closed! Admittedly, it is a small portion of the open source supporters that do this, but they are quite vocal, and the last thing that a company needs when considering a controversial (within the business community) move like opening the source code to their software, needs is bad press associated with the decision. I am certainly glad that Trolltech was not in this position, as they were already OSS supporters, but even they, after having showed this support for several years, still get flak from the community because there license wasn't written to the critic's specifcations, as if Trolltech was going to announce tomorrow, "Ha, fuck you! We left a nice little loophole in our license, so no more damn source code for you!" Its a wonder that traditionally closed source businesses ever decide to open their source, considering the immense risk, not from hax0rs or competitiors, but from the bad press that can result from the crowing of OSS supporters! It never ceases to amaze me how a bunch of computer geeks suddenly become Harvard law professors just because the law happens to involve software. Give these folks a break, already; sure, their licenses may not be ironclad bastions of open source might, but damn, try to be cool about it! After all, they were cool enough to release the source code to their products, and they sure as hell didn't have to do that, as closed source software seems to sell just fine.
Deo
Qt is a high quality GUI toolkit, well documented, and easy to use. That is a given.
Trolltech wants everyone to use Qt instead of Motif on the pure commercial side or Gtk+ on the free software side. That is also a given.
Trolltech wants to make money. That is also a given.
The problem is, these goals work against each other.
And now, for the first time, they are scared! They thought they had a big advantage over GNOME/Gtk+ five years ago (GNOME doesn't exist yet!). That advantage was in the process of being eliminated two years ago, when they released Qt under the QPL.
The QPL was meant to kill the copycat project called the Harmony. They succeded in that.
The QPL was also meant to somewhat blunt the rapidly maturing GNOME project. Thay failed at that miserably.
Now that GNOME 1.2 is out and getting rave reviews, it woun't be long before more of the distributions make GNOME their default desktop.
If that happens, Qt will lose out completely. And that is a sure thing now. They can see it. They can sense it. They have no way of slowing the trend down.
At this stage, not even releasing Qt under the GPL will save them.
The GPL has won another battle.
--
IIO
-- Weiqi Gao weiqigao@speakeasy.net
He compares Qt with Motif:
"When we first released Qt there was already de-facto acceptance in the community that one could make GPLed programs using Motif, even before Lesstiff came along. Motif was a proprietary and non-free library with a much stricter license than Qt."
GPL allows linking (dynamic and static) to libraries that are not GPL compliant if they are considered a part of the OS. Hence, an application in windows that uses WIN32 API has no problems. Qt is not part of the OS on any system, not even Corel Linux.
He also seems to confuse GPL with LGPL:
"If the GPL effectively protected a GPLed library from being used to develop proprietary software, we would allow relicensing Qt under the GPL. But, as I have said, it is not our belief that using a library is making a derivative work. "
The LGPL license was developed to give non GPL compatible software access to GPL libraries. If a library is GPL:ed there is no way of linking (dynamic or static) with that library. E.g. the cygwin library is GPL:ed in Windows, so you can only create GPL:ed software with Cygwin.
Having said all this, Debian could include KDE if they added a clause in KDE that allowed linking to Qt.
This at least is my interpretation of how GPL works.
Sounds like the NPL.
Point taken. I had misread a crucial sentence.
However, it has been generally agreed that a binary that is prepared for use with a library is a derivative work. That's why GNU readline, for example, is GPLed, and (IIRC) the reasoning behind why NeXT had to release their Objective C compiler under the GPL. And that's the reasoning behind Sleepycat software's licensing of the Berkeley DBM libraries.
Now either copyright law permits the copyright holder on a library to restrict distribution of binaries that dynamically link to that library, or it doesn't.
If it does, then all we have is a semantic quibble over the meaning of "derivative work", since it is clearly the intention of the GPL to place restrictions on such distribution. In that case, then the conflict could be solved by distributing Qt under the GPL, with a footnote to the effect that dynamically linked binaries are considered to be a "derivative work". This would clearly be compatible with the GPL.
If it doesn't, then no license whatsoever will protect TrollTech from people who want to distribute proprietary binaries which dynamically link with Qt code. (As long as they don't also want to distribute Qt as part of the same package.)
--
The main sticking point with the QPL is that it requires all software, even software written written for internal or research purposes, to be released, and to be released under the open source model. Troll Tech's intent is clearly to get every potential commercial user to pay for each of their developers, presumably under the assumption that commercial institutions have lots of money to spend. And at between $1500 and $2400/developer, Qt is (IMO) uncompetitively expensive.
Unlike many other libraries, GUI libraries are complex and require a significant amount of time on the part of the programmer to learn how to use them. So, you should ask yourself: if I invest this much time in the library, what do I get back? With Qt, you commit to either contributing code to Troll Tech (under the QPL) or money (under the commercial license).
If Qt were the only game in town, theremight be some justification for that kind of deal. But there are lots of good, free toolkits and there is no need to make such a tradeoff.
Incidentally, similar considerations apply to GPL'ed GUI libraries. If you invest a lot of time learning a GPL'ed GUI library, you severly limit your options later with what kind of GUI programs you can write. So, I think even placing Qt under the GPL wouldn't help much given the available alternatives. At least the GPL allows internal and research use, however.
Altogether, I doubt Qt is going to make it in the door where I work, and I wouldn't view knowledge of Qt as a big advantage when hiring. My recommendation is: learn GTK/Gnome, Tcl/Tk, wxWindows, FLTK, Java/Swing, even Motif or Win32/MFC.
One huge advantage of Qt is that you can target Windows and *NIX with the same source. We use it in our lab at NASA, and have no regrets at all. I do agree that Qt is a bit pricy, but we've found that the platform flexibility allowed by Qt is worth it. I evaluated wxWindows and FLTK before commiting to Qt, and although they are both quality toolkits, Qt offered more functionality, flexibility and has a (IMO) cleaner coding style.
Unlike many other libraries, GUI libraries are complex and require a significant amount of time on the part of the programmer to learn how to use them.This is another big advantage of Qt: it's really, really easy to learn and use. The coding style is very similar to Java, and it leads to clean, easy to read, easy to maintain code.
Java, unfortunately, just isn't there yet. It's not stable or fast enough for many of the things we need to do (although we do use Java for some projects). Because of the similarity of Java and Qt, many of our APIs and some of our apps will work with both toolkits, and some developers bounce between the two.
I'm not a laywer, and really don't care to become one, but why do you think the QPL is any different to the GPL for internal use? Section 6 has this line: These items, when distributed, are subject to the following requirements: It seems to me that Qt is find to use for internal research, as long as you don't distrubute it - same as the GPL.
In my opinion, Qt is a superior technical solution than any of the other toolkits available (free or commercial). It is expensive, but we've found it to be worth the cost on more than one occasion.
Cheers, Mark Allan http://ic-www.arc.nasa.gov/ic/projects/neuro/ifc/"Good people drink good beer"
Troll Tech is a business, and releasing Qt under the QPL isn't an act of charity, it's a business strategy. In particular, Troll Tech wants two things. First, they want bug fixes and code contributions from users. Second, they want lots of people (in particular, students) to get started using their toolkit so that when they move into companies, they'll buy the toolkit. I think it's valid to ask the question whether this is a good deal.
While your opinion may differ, I don't think it is. There are several high quality toolkits out there without such restrictions, and where bug fixes and code benefit everybody. And Qt is so expensive that I would have serious concerns about using it even as a purely commercial toolkit.
Just because a company dumps some source code on the world under some license doesn't make them immune from analysis. But to be clear, my view isn't that Troll Tech should change the license or that they have done anything "wrong". My view is that people should stop using Qt because it simply isn't a good deal and because there are good alternatives available.
That said, Troll Tech was in a perfect position to fix them (3-4 clauses changed in their current license would do it) but they chose not to. As soon as Red Hat agreed to ship KDE, they didn't have to worry anymore in their thoughts. The license issues aren't their problem, but they certainly do profit from nobody beliving KDE has a problem.
Read the actual claims made by myself and some of the other Debian people. Don't rely on Slashdot histeria and rumors as to what we think the problem is. Find out what the problems really are - they're a lot simpler in nature and really very clear.
Borland/Inprise has a long history of bad business decisions, so I guess we shall see?
This isn't a hard task, read the license for yourself.
It is a far more subtle document than one would think just reading it, and sometimes it has surprising consequences. For instance here is a summary of Debian's issues with Qt. Can these be addressed? Yes. Are they somehow claiming that Qt is a derivative work of KDE? Nope. They are claiming that KDE is a work that cannot be distributed by the terms of its own license.
But I think that there is a way to solve this problem for once and for all which is not discussed. Produce a GPL-friendly interface to Qt with a GPL-friendly implementation behind it. Have it support the full programming interface, but don't worry too much about it "looking nice" or usability for the end-user. Ship KDE with this and with some easy way to switch from this (bad) default to Qt. Then KDE can ship in compliance with the GPL completely separately from any license issues that Qt may have.
Cheers,
Ben
My usual seat in the cluetrain is at A HREF="http://pub4.ezboard.com/biwethey.ht
1) If people are complaining to TrollTech about making Qt free, they (the complainers) are idiots. The problem is not Qt, the problem is KDE. Qt is not in any danger, KDE is the project is that has (potentially) linked GPL with non-GPL.
2) You are right that a->b != b->a. But the GPL also talks about "derivative works". So, for instance, I take work "myprog.o" and add "libqt.a" and create work "myprog". "myprog" is a derivative work of a GPL'd work "myprog.o" and a non-GPL'd work "libqt.a".
--
Linux MAPI Server!
http://www.openone.com/software/MailOne/
(Exchange Migration HOWTO coming soon)
If you read on, you will see that this section is actually ambiguous, since later it states that Troll Tech can always request software from you that you developed under the QPL. Furthermore, Troll Tech states in their QPL FAQ that it is the intent of their license that you must buy a commercial license even if you develop only for internal purposes.
Companies don't like ambiguity and they won't generally risk getting dragged into some legal spat, so making the QPL ambiguous and stating their intent in the FAQ essentially ensures that most companies who want to use Qt will license it, yet to the open source community, Troll Tech will appear as if their license is "almost like" the GPL. You have to decide for yourself whether this is legal incompetence or a Machiavellian plan. The fact remains that the QPL is much more restrictive than all of the other widely used, free or open source toolkits.
As for the quality, if you think Qt is worth the license fee, good for you. My point is that most people (including consultants, corporate developers and students) who pick Qt for development should view it as a commercial toolkit that costs between $1500 and $2400/developer, since most likely they will have to pay. Qt needs to be evaluated and compared to other toolkits on that basis. I suspect for the kind of money you are going to pay Troll Tech for all your developers, you could often already have some consultant or in-house developer do a significant amount of custom development and fix whatever shortcomings you see in one of the other toolkits.
Gosh. Can't people learn?
:)
<p>First off, GNOME is not GPL'd. GNOME libs for example are lgpl'd, as is gtk+, a few parts are BSD licensed. KDE isn't GPL'd either. kdelibs are lgpl'd just like GNOME. kdebase is mainly artistic licensed, although it contains some other licenses. koffice is completely artistic licensed. Other apps are under a variety of free licenses, from BSD to LGPL to GPL to QPL to Artistic. Just like GNOME.
<p>The 30 million dollar question comes down to -- what about those few pesky apps that are under the GPL?
<p>Well, some of the most active developers (like the author of ksysv) have changed there license (ksysv author added a gpl exception). But there are a few KDE apps that the authors have no current way to contact them (their email address are invalid) or they are no longer avalible (dead, moved away, etc.).
<p>What do we do with these programs? 1) Discard them? 2) Move them to kdenonbeta? 3) Change the license without the author's permission? Obviously, if you were one of those authors and had not been notified about one of the above happening, you would be likely pissed, to say the least. It's not fair to discard somebody's hardwork, just because there was a new licensing problem discovered.
<p>The KDE Qt Free Foundation are doing their best to resolve this matter, but it certainly isn't easy. No solution is ideal for either party.
<p>Lets all try to get along now
Of course there is a browsable kernel source tree, and it is wicked cool. You can click on any function/variable/etc and get the definition. The page is http://lxr.linux.no/
The GPL seems to clearly define what it takes to make a derived work:
"You may modify your copy or copies of the Program or any portion of it, thus forming a work based on the Program"
Dynamic linking to an unchanged library would not seem to be a derived work then, it clearly specifies modification. What am I missing?
Get permission from Troll to use their interface files (literally).
Put in useless stubs for everything you don't want.
Turn it into a scriptable toolkit, usable for writing automated tests against. Basically do for Qt-based applications what Expect does for interactive command-line applications.
Start working on an automated test suite, similar to the one that (for instance) Perl uses.
Cheers,
Ben
My usual seat in the cluetrain is at A HREF="http://pub4.ezboard.com/biwethey.ht
When Erik Eng tries to explain why they didn't put the QT libraries under the GPL, he perpetuates the biggest myth about the GPL: that the GPL forbids making proprietary derivative works.
It doesn't. It just doesn't permit making proprietary derivative works. Copyright law forbids making derivative works without permission; licenses, such as the GPL, give that permission.
If TrollTech GPLed QT, they could still make money by selling it to people who wanted to use it for developing proprietary software. This is exactly what Sleepycat Software does. They make the ubiquitous Berkeley DB libraries, which are used for free software development under a very GPL-like license. (Essentially the GPL, without the long rant at the beginning.) But if you want to use their libraries to make proprietary software you pay them for a different license.
Maybe QT has other good reasons for not GPLing QT, but claiming that the GPLing it would preclude it's use for proprietary software development is just plain false.
--