GCC 4.2.1 has been released, the last release of the GNU Compiler Collection under the GPL v2. “GCC 4.2.1 is a bug-fix release, containing fixes for regressions in GCC 4.2.0 relative to previous GCC releases. GCC 4.2.1 will be the last release of GCC covered by version 2 of the GNU General Public License. All future releases will be released under GPL version 3.”
I’m intrigued to know how the move to GPL3 will affect those companies making use of GCC for compiling apps on their platform. Does it have any particular legal effect? Will it scare some off?
it has no real effect what so ever. Will it scare some off? well, it may, but only if they are stupid, and at that point, all i can think is… so what?
There is no restriction in either GPL v2 or in GPL v3 about use of a GPL program. In both cases you are free to use the program however you want to. This is one of the four freedoms that the GPL is all about trying to preserve for you (not restrict you in).
Restrictions apply in GPL v2 only when you distribute a GPL v2 program, and restrictions apply in GPL v3 only when you distribute or assist in distribution of a GPL v3 program. Even then, the restrictions really only amount to a requirement that you pass on to downstream recipients the same permissions that you received, and they only apply to the GPL code, not to your own code.
Compiling your own program is neither of those restricted activities. Go right ahead.
Edited 2007-07-22 14:49
There is no restriction in either GPL v2 or in GPL v3 about use of a GPL program.
Unless the program is on a server, and the guy on the client side demands your modifications. Closing the webapp “loophole” IS about restricting usage!
Unless the program is on a server, and the guy on the client side demands your modifications. Closing the webapp “loophole” IS about restricting usage!
Nonsense. The GPLv3 does not apply to anything like this. The closest that any license applies to a webapp “loophole” is the Affero GPL. The latest version is here:
http://gplv3.fsf.org/agplv3-dd1.html
Edited 2007-07-22 23:34
You can safely ignore the client’s demands. You are not distributing the software by placing it on one of your server, merely using it to format and distribute unrelated (in the context of the application binary) software.
Excuse me?
If the program you have on the server is your own code, then anyone on the client side has no claim to it at all.
If the program you have on the server is GPL code, then if you have modified it slightly then you are obliged to publish your modifications to downstream users … this is the exact same deal for GPL software as it has always been. You cannot “hide” a program on a server and thereby expect to “escape” the one and only “cost” of the GPL code you are using.
No. modified GPL software running on a server is not being redistributed. So there is no requirement to supply the modified code.
In the same way that if you modified the Open Office source code for personal use, then wrote letters using it, you would not have to publish your modifications.
(Of course, it would be nice to do so)
Meh. This is a borderline case.
You could also argue (from the opposite viewpoint) that the client users are in effect “running” the code on the server, whereas people who read letters are not running OpenOffice.org in any way.
So should Google be forced to GPL the sources to GFS?
because it is (presumably) a Linux kernel module running on publically accesible servers. Clients who search using google, or use google maps are using the GFS drivers to access data. By your argument, that makes it subject to the redistribution terms of the GPL.
It gets a little hard to judge these cases!
This is a good question, and one to which I do not have a ready answer. It is not clear from the plain language of the GPL license.
Agreed, 100%. Another borderline case. I don’t offer any judgement myself, either way, as I can’t really work out if the GPL applies here or not.
Agreed, 100%. Another borderline case. I don’t offer any judgement myself, either way, as I can’t really work out if the GPL applies here or not.
You’ve said it yourself here over and over again– the GPL is a copyright license (even though the GPL3 may have meandered into contract law, but that’s another topic), and copyright only covers distribution. The software, in this case, is not being distributed, so there is no requirement to release the source code. It’s not a borderline case at all.
Fair enough. I’m perpared to go with this reasoning.
The only caveat: GPL v3 now applies to activities beyond pure and plain distribution of the code.
But you are probably right, it doesn’t seem to apply in this instance.
Fair enough. I’m perpared to go with this reasoning.
In fact, I seem to recall that this “web app loophole” came up during some of the initial GPL3 discussions. In the end, they made a conscious decision not to close the loophole.
The only caveat: GPL v3 now applies to activities beyond pure and plain distribution of the code.
Right, and this is what I was talking about in my last post with regard to the GPL3 possibly heading into contract law. I’m certainly not qualified to speak decisively on these issues, but I did find this to be a particularly interesting read: http://www.actonline.org/library/GPLv3-Contract-or-Copyright.html
Of course, we won’t have a solid answer until the GPL3 is tested in court.
This statement:
sort-of contradicts this one:
Who is going to test it in court, if the GPL authors themselves deliberately chose as they did?
I keep hearing noises about “testing the GPL in court”, but I have yet to hear:
(1) any sound theory on what basis to test it, and
(2) a sound reason why anyone should test it.
If, for example, Microsoft carry on as they are, and they don’t actually take anyone to court over their alleged patents in Linux, then Microsoft is compliant with the GPL v3 in its handing out of vouchers.
If, for example, Microsoft do “take the GPL to court” and challenge the validity of the GPL itself, then all that would mean is that any permissions the GPL previously granted to Microsoft (as long as Microsoft weren’t suing anyone) would be revoked, and Microsoft would be in violation of copyright law, gining out vouchers for a copyrighted work without permission of the copyright holder.
Microsoft are actually best served by not invalidating the GPL in court, and continuing to not sue anyone over GNU/Linux use.
Edited 2007-07-23 05:40
This statement:
In fact, I seem to recall that this “web app loophole” came up during some of the initial GPL3 discussions. In the end, they made a conscious decision not to close the loophole.
sort-of contradicts this one:
Of course, we won’t have a solid answer until the GPL3 is tested in court.
I was referring to the idea that parts of the GPL3 may be considered contract law and not applicable to a copyright license. This is completely unrelated to the web app loophole.
I keep hearing noises about “testing the GPL in court”, but I have yet to hear:
(1) any sound theory on what basis to test it, and
(2) a sound reason why anyone should test it.
Like I said earlier, I think the sound theory on why it could be tested is that it seems like it may be stepping beyond the bounds of copyright law.
Though I do agree with you that I don’t see any reason to actually test it in court. Microsoft is far more interested in maintaining the status quo than they are in trying to litigate against free software.
[…] Microsoft would be in violation of copyright law, gining out vouchers for a copyrighted work without permission of the copyright holder.
I don’t agree with this. I see no reason why I can’t stand out on the street and give away vouchers for music that I don’t own. It only becomes copyright infringement when I actually distribute the copyrighted item.
Not to mention that under your hypothetical situation, if the GPL3 were to be invalidated, everyone who distributes GPL3 software would be in violation of copyright– not just Microsoft.
Microsoft are actually best served by not invalidating the GPL in court, and continuing to not sue anyone over GNU/Linux use.
Very true. I am certainly not a Microsoft basher (I make Windows/Xbox360 game middleware for a living), but their patent threats towards Linux are FUD in the truest sense– and they want to keep it that way. The dark clouds hanging over Linux in the eyes of the ignorant observer are far more valuable to Microsoft than any possible outcome of a patent suit.
GCC has been used for decades to compile all manner of proprietary software, and the GPLv3 doesn’t change this. Vendors that distribute development tools based on GCC should have their lawyers take a look. However, chances are that what they’re doing is permitted under the GPLv3 if it was permitted under the GPLv2.
In general, the only people who should be concerned about the GPLv3 are distributors who were previously circumventing the spirit, if not the letter, of the GPLv2. The Linux kernel is really the only practical exception to this rule, as the controversial restriction on restricting users from running modified works is an issue that predominantly impacts kernel code.
All in all, the GPLv3 should be rather uncontroversial among those that like the GPLv2. Almost everybody likes the final patent language except for Microsoft and its shameful proxy vendors within the Linux community. It’s compatible with the Apache license, forming a nice licensing continuum including non-copyleft (Apache), weak copyleft (LGPLv3), and strong copyleft (GPLv3) options. Together they represent a new generation of free software licenses that are boilerplate around the world and aware of the threats posed by software patents.
>>In general, the only people who should be concerned about the GPLv3 are distributors who were previously circumventing the spirit, if not the letter, of the GPLv2.<<
Not true: every project which combines GPLv2 only code (Qt, Linux kernel,…) with other code which goes from GPLv2 to GPLv3 has a problem.
>>All in all, the GPLv3 should be rather uncontroversial among those that like the GPLv2.<<
False also: as shown by Linus who likes the GPLv2 but don’t like the GPLv3 (because of its “anti-DRM/tivoization” clause).
Sure the compatibility of GPLv3 with Apache is nice, the patent provision is nice too, but this is not a reason to put the rose glass and ignore that there are dissenting voice and issues that the GPLv3 will trigger (and not only with people which tried to circumvent the GPLv2).
>Not true: every project which combines GPLv2 only code (Qt, Linux kernel,…) with other code which goes from GPLv2 to GPLv3 has a problem.
Yes, but this problem was created by the “GPLv2 only” users and not by the GPLv3. The FSF has known from the beginning that a new GPL will be incompatible with old GPLs. Therefore they have introduced the “or any later” language. You could call this a “bugfix”. Than people used the license and removed the “bugfix” and now they start complaining about this “bug”. Sound strange, doesn’t it?
The problem isn’t created by GPLv2 or GPLv3 because there was a solution for it. The problem was created by people who decided to not use the GPL 1:1 like released by the FSF.
Edited 2007-07-22 20:26
Sure. Why would I hand the future of my code over to some fundamentalists who could go berserk against the Great Satan of closed source and add such limitations in using/combining/distributing my code in commercial environment which I could never accept.
Edited 2007-07-22 20:45
Sure. Why would I hand the future of my code over to…
You can do what you want. But if the original license manage compatibility to new versions and you decide to remove it than you can’t blame the FSF or the GPL for missing compatibility.
You have decided to remove compatibility so now you have to life with it.
Edited 2007-07-22 20:49
But I’m not the one who has removed compatibility. It is those who have changed from 2 and later to 3 and later. They are the ones who intentionally and explicitly have broken compatibility with 2 only software.
Edited 2007-07-22 21:08
>It is those who have changed from 2 and later to 3 and later.
No, because every “GPLv2 or later” will be compatible with “GPLv3 or later”
Compatibility was broken by people who haven’t used GPLv2 like released by the FSF but instead without the “or any later” language. So the people who have modified GPLv2 has explicitly stated that they don’t want to be compatible with new versions of the GPL.
It’s their right to decide not to be compatible with new versions of the GPL. But than they also have to take responsibility for their decision and not complain about others.
Everyone who has used the GPL like released by the FSF has absolutely no problem to link against GPLv1, GPLv2 or GPLv3 code.
EDIT: fixed typos
Edited 2007-07-22 21:22
But now you’re getting nihilistic. We live in a world where there are large code base licensed under v.2 exclusively. These include such as Linux kernel and Qt and organisations behind these projects have very clear reason to license their stuff this way (see my earlier post).
Now we get to the point. Those who change from 2 or later to 3 or later explicitly break compatibility with those vast code bases licensed under 2 only. And this is the problem. People forget the value of cooperation and helping each other and instead try to force feed their new license-of-choice.
This effectively kills some projects as there is no way for a poor little programmer working on a mash-up of Qt and lets say Samba to ever get this thing sorted out. Even he wants to he can’t force Trolltech to switch to 2 and later or 3 only. And now this poor bastard finds himself in a situation where he is unable continue doing his thing only because some found something in the current world immoral.
But don’t get me wrong. I’m not against GPL 3 per se but I think it foolish to cut compatibility with some of the greatest open source projects there are. And you have to admit that whose who break the compatibility here and now are those who change to 3 and later only.
Edited 2007-07-22 22:02
But don’t get me wrong. I’m not against GPL 3 per se but I think it foolish to cut compatibility with some of the greatest open source projects there are.
I am having trouble seeing the fairness of your comments. You start off by characterizing the FSF as “fundamentalists” who could “go berserk” and add use limitations to future versions of licenses that have only ever had distribution limitations, including GPLv3. The “or later” language is natural and logical to ensure compatibility among copyleft licenses. Why should the unwillingness of those of a relatively few projects who refuse to use such language count for more than the wish of those of relatively many projects who want to restore protection (among many other benefits) for their software that was compromised by loopholes in an old license amid a business climate of patent threats? Moreover, you implicitly assume that not upgrading the license would lead to lesser likelihood of forking within the project from those who wish to upgrade than the more typical converse case. Why should cooperation among projects ever matter to any one project as much as cooperation within the project itself?!
I am having trouble understanding how you are so quick to attribute malice or stupidity to those who simply seek the project integrity and kind of protection that the GPL was always supposed to encourage and provide. Oh, and yes, the GPL was always motivated by what is moral, so it is strange that you should be bothered by that motivation as people upgrade to simply another version.
Edited 2007-07-22 23:23
People who move to v3 or later can use code from v2. It is people who insist on v2 only who are denying themselves access to V3 the code.
For what purpose would people want to stay with v2 only? Are you saying it is worthwhile to stick with v2 only, and deny yourself all that v3 code, just in order to allow Tivo to rip people off?
I just don’t get this attitude at all.
>We live in a world where there are large code base licensed under v.2 exclusively.
Really? I would say most GPL code is licensed under the GPL like released by the FSF -> with “or any later” language.
Large examples are:
(1) Linux, which doesn’t really matter because nobody who writes Apps have to link directly against Linux. Linux is a quite monolithic program which hasn’t large side effect (e.g. like libs) so it’s no problem to stay with GPLv2 only.
(2) Software where one entity holds the copyright (e.g Qt). This is also no problem because the code can easily be relicensed.
So (1) is not really a problem and (2) can easily solved by the copyright holder otherwise the copyright holder from this code (e.g. Trolltech) have to explain their customers why they want to deny them to write GPLv3 code. It is their decision and so they have to take responsibility for it.
You can’t cut of compatibility and than complain about others because you don’t want to be compatible. It was your decision so now you are in charge of it.
Why would Trolltech possibly object to GPL v3? Trolltech comply with the provisions of the GPL, and Trolltech don’t have any desire to sue Linux users … so why on earth would Trolltech not simply go along with GPL v3?
BTW, the largest single piece of any GNU/Linux distribution is the GNU software. GNU software is about 28% of the code in a typical GNU/Linux distribution, where the Linux kernel is about 3%.
The GNU foundation wrote the GPL v3 license text, so all of GNU software will go to GPL v3.
GNU GCC, which this article is about, was originally written by Stallman himself.
http://en.wikipedia.org/wiki/GNU_Compiler_Collection
Edited 2007-07-23 13:19
>Why would Trolltech possibly object to GPL v3? Trolltech comply with the provisions of the GPL, and Trolltech don’t have any desire to sue Linux users … so why on earth would Trolltech not simply go along with GPL v3?
We will see. At the end Trolltech has to decide but i expect that they will move to “GPLv2 or later” or “GPLv3 only” to allow KDE and other Free Software to update their license.
Your post is ‘Yes, but <a totally unrelated topic>’: I don’t like straw man arguments: the GP was wrong in saying that only ‘evil guys’ need to be concerned by the GPLv2/GPLv3 split, and I pointed this out.
Being wrong is still being wrong whoever created the problem.
So who created the problem is quite unrelated to the topic and a matter of opinion: had the FSF made a GPLv3 without the anti-DRM/tivoization clause, it’s quite likely that the shift from GPLv2 to GPLv3 would have happened without a blip.
Your post is ‘Yes, but <a totally unrelated topic>’: I don’t like straw man arguments: the GP was wrong in saying that only ‘evil guys’ need to be concerned by the GPLv2/GPLv3 split, and I pointed this out.
And I don’t like the quoting out of context and selective reading upon which your most of your reply to the GP relied.
The GP acknowledged the Linux kernel as an exception to the isolated statement that you quoted first, and yet you ignored it by mentioning the kernel–as if he never acknowledged it. Moreover, you ignored his “in general” [1] qualification. After all, how many out of thousands of projects use “GPLv2 only” besides the Linux kernel, git, busybox, Qt, and MySQL?
In addition, the GP acknowledged that the GPLv3 was controversial among kernel developers because of its anti-Tivoization language, and yet you–again quoting out of context–chose to ignore this and tried to play up how Torvalds does not like the anti-Tivoization language. Is he not a kernel developer? You also ignored the GP’s “rather” qualification of “uncontroversial.” Fortunately, he considered sentiments at large and not merely from some of the few that the media likes to hype for who knows what purposes.
Finally, nothing specifically about the GPLv3 matters to any other cases that you cite, such as Qt. Everyone knew years ago that a new copyleft license would be incompatible as well as knowing the long-recommended “or later” language around it. Failing to deal with Tivoization would not have changed the incompatibility, of course. Trolltech, MySQL AB, etc. have a generic concern about copyleft licenses. Again, the GP identified the specific concerns about the GPLv3: the patent deal blocking and the anti-Tivoization language.
To be fair to your concerns, perhaps the GP should have mentioned the incompatibility that naturally results between copyleft licenses and the usefulness of the “or later” language. Hopefully, such elementary considerations will soon be common knowledge among developers.
[1]:
in general
adj : considered altogether; “the country at large”; “I enjoyed the play as a whole though I thought the acting could have been better” [syn: {as a whole(ip)}, {at large(ip)}, {in general(ip)}]
Edited 2007-07-23 20:30
Do you realize that the GPLv3 license is both a license for FSF’s code and a template license text for others to use if they so desire? Being a mere template, it means that other parties can simply remove whatever clauses they don’t want, and apply the remainder of the text of the template as their license.
If the kernel devs truly do want Tivo to be able to rip end users off, and they truly don’t want the anti-tivoisation clause, then why don’t they simply rip out that clause from the GPLv3 license template and release the kernel under that modified license, calling it GPLv3(b) or somesuch? That way the kernel devs would gain all of the benefiits and improvements and protections of GPLv3 without the anti-tivisation clause to which they apparently (and somewhat mysteriously) object.
Yes, that’s obviously their primary motivation.
I challenge you to come up with a widely applicable license that will be applied to 1,000s of public products, that is universally liked.
I challenge you to come up with a widely applicable license that will be applied to 1,000s of public products, that is universally liked.
BSD and MIT licenses have done very well, thank you very much.
But they aren’t universally liked.
BSD and MIT licenses are playing in a completely different space. How many big commercial products have been released under BSD/MIT? How many big companies are paying people to work on BSD/MIT-licensed software? There are examples, sure, but they’re much rarer than for GPL’ed software.
The GPL has some fairly unique characteristics that allow for a surprisingly pain-free cooperation between corporate and community users. There is a merit in it that BSD/MIT proponents really fail to see.
GCC has been used for decades to compile all manner of proprietary software, and the GPLv3 doesn’t change this. Vendors that distribute development tools based on GCC should have their lawyers take a look. However, chances are that what they’re doing is permitted under the GPLv3 if it was permitted under the GPLv2.
Only if the GCC release has something akin to this again:
The source code of libstdc++ is distributed under version 2 of the GNU General Public License, with the so-called “runtime exception,” as follows (or see any header or implementation file):
As a special exception, you may use this file as part of a free software library without restriction. Specifically, if other files instantiate templates or use macros or inline functions from this file, or you compile this file and link it with other files to produce an executable, this file does not by itself cause the resulting executable to be covered by the GNU General Public License. This exception does not however invalidate any other reasons why the executable file might be covered by the GNU General Public License.
The GPL is a distribution licence and not a usage licence. You do not have to agree to anything within the GPL to use GPL software. You have to follow the terms of the GPL if you wish to *distribute GPL software*.
Edited 2007-07-23 11:15 UTC
You have to follow the terms of the GPL v3 if you wish to *convey GPL v3 software*.
“Convey” includes doing things like “give someone vouchers so they can obtain”.
Even though there are no practical immediate consequences, it’ll be interesting to see what will Microsoft’s reaction be once Novell includes GPLv3 code in SuSe (Samba, GCC, and counting). According to GPLv3, the current patent deal between Novell and Microsoft would be infringing, and the only way for it to comply is by delivering the patent protections Novell currently enjoys not only to its customers, but to the entire community.
RMS explains it here: http://gplv3.fsf.org/static/release/rms_gplv3_launch_high_quality.o…
That it is be infringing is the FSF’s party line, of course. However, they’re on shaky legal ground with it, and pretty much trying to go hundreds of years of western law.
It’s doubtful that they will try to ‘acquire’ Microsoft’s stuff this way. If they did try to take this to court, they have a 50/50 chance (or greater) of having the GPL v3 declared legally void.
The GPL “license” is effectively a conditionaly waiver of the owner’s funadmental rights under copyright. If the GPL were ever declared invalid, the situation would revert to standard copyright, where no-one would have the right to redistribute the product.
This simple construction makes the GPL far stronger than the EULAs used in traditional proprietary software, which rely on licensing law, and are hobbled by the fact that you often have to buy the product before you can view the license, which breaches the basics of contract law, regardless of refund schemes.
Consequently it is unlikely any court will declare the GPL invalid, or that any organisation will try to do so. The basic simplicity of its construction lends it enormous legal strength.
Furthermore, the FSF is in no way trying to “take over” Microsoft: frankly that’s a lot of astonishingly paranoid nonsense. The additional restriction to the GPL is that if you copy and paste GPL code when creating your own product, and you patent your additions, you must distribute the result under the GPL, and waive all rights to your patents on those additions, so that other users can copy and paste your code just as easily as you copy and pasted from the original author’s code[1].
The issue for Novell really only comes into play if they knowingly mix GPL code and patented code together. Of course, if Microsoft has informed them of the patents, and they’ve accepted those patents’ validity as part of the contract, then they will find themselves in an awkward position, effectively unable to redistribute GPLv3 products… until they create patches to work around the patent issues. If anything, this should help open-source software work around patents, insofar as possible in the current, uncertain patent climate
[1] Note there is always the possibility that you could ask the original author to relicense his code under a different license (e.g. commercial one), should you wish to avoid these constraints. And if they refuse, well then you just write everything from scratch yourself, instead of copy and pasting other peoples’ work.
That’s exactly why the FSF won’t be pursuing this; the chance of destroying the GPL v3 is too great. As well as losing their license, the risk of destroying so many OSS products — especially since their sponsors rely on using GPL code instead of hiring developers and creating their own products — is too great for them to legally pursue this.
Excellent point. That’s the superior solution in virtually all cases anyways.
Congratualtions gentlemen, I think we have a confirmed new NotParker.
If it was, your ad hominem response would certainly make him feel right at home.
Seriously, where do you get this drivel from?
The FSF do not have to “persue” anything.
If you stick to the terms of the license under which FSF code is diseminated, then you are licensed for your activity.
If you do not, then you are in violation of copyright law.
Therefore, the GPL v3 license is good only for the defense of GPL software itself … to stop other parties from attacking GPL software. It has no purpose other than that.
Finally, once again … (how many times must this be repeated to you before you get it?) … if the GPL license is somehow decalred void, then the copyrighted software formerly released under the GPL reverts to copyrighted software of the FSF. The FSF will simply fix whatever legal problem is found in the text of the GPL, and then re-release their software under an amended version of the GPL. Voiding the GPL will simply NOT give ownership rights over GPL software to Microsoft.
Edited 2007-07-23 00:05
That’s the superior solution in virtually all cases anyways.
The hell it is. We use Linux/GCC at our company for custom-designed hardware. The odds of us coming up with a better combo from scratch on our own is statistically insignificant…
There is no way that anyone in the OSS community would, or could acquire Microsoft Closed-source technologies in this manner. The problem is that the Microsoft patent-deal does not, cover the GPLv3. As more projects switch to the v3, the patent-deal has less and less effect.
THIS IS NOT ABOUT THE OSS COMMUNITY STEALING IP!
There is no way that GPL v3 is an attack on Microsoft, or that there is any intent whatsoever to take Microsoft to court.
If, however, Microsoft try to take GNU/Linux to court, then GPL v3 affords a legal defense against such bullying by Microsoft. Microsoft are giving out vouchers for other people’s copyrighted works … specifically, Microsoft are giving out vouchers fro GNU software … and FSF holds the copyright for GNU software.
Under copyright law, Microsoft need permission from the copyright holder (not Novell, but the actual copyright holder) to give out vouchers for GNU software.
The copyright holders of GNU software (namely, the FSF) have given such permission to anyone who abides by the terms of the GPL license. Outside of abiding by the terms of the GPL license, there is simply no other permission given.
Either Microsoft abide by the terms of the GPL license (in which case they have the required-by-law permissions for their vouchers), or they do not (in which case they have no permissions). If Microsoft have no permission to give out vouchers for someone else’s copyrighted work, then Microsoft are in violation of copyright law. It really is that simple.
Going to court and somehow getting the GPL declared legally void will not somehow magically provide Microsoft with permission to violate copyright law. If the GPL is somehow decalred legally void, Microsoft will still have no permissions for their vouchers. What Microsoft would have to do is get copyright law decalred legally void … do you really believe Microsoft can do that or would want to do that? Get real.
Edited 2007-07-22 23:49
But the vouchers does NOT entitle you to any GPLv3 code.
They entitle you to a support agreement for SLED, exluding any and all parts that are under GPLv3.
Novell is free to include additional parts if they want to, but that is entirely up to them.
There is no need to take it all to court, unless you think MS can be seen as a distributor of GPLv3 code through Novell adding parts that MS explicitly exclude from the offer.
Sorry, but this is not what the vouchers themselves say.
http://www.groklaw.net/article.php?story=20061107194320461
From Novell’s 8K
It says “combined offering”, it says “expires January 1, 2012” and it says “SLES and a subscription for support”. It also says
… upgrades and updates are both software. It says absolutely nothing about any limitation to “GPL v2 only”.
The vouchers do indeed entitle you to GPL v3 code.
Edited 2007-07-23 12:13
Well, having GPL v3 tested in court right off wouldn’t hurt it. I presume this could be one of it’s advantages. It may be cynical but if FSF wants to quickly promote it they should start some GPL3 related lawsuits in number of EU countries and USA. The more high profile they are, the better.
Edited 2007-07-23 08:13
cool finally i can use the fstack-protector for all functions on a vanilla kernel:-)