FITUG e.V.
Förderverein Informationstechnik und Gesellschaft
FITUG e.V.Förderverein Informationstechnik und Gesellschaft |
|
------- Forwarded message follows ------- Date sent: Wed, 27 Feb 2002 17:09:54 -0500 To: politech@politechbot.com From: Declan McCullagh <declan@well.com> Subject: FC: John Palfrey's report from Boston in GPL MySQL license lawsuit Send reply to: declan@well.com
Previous Politech messages:
"Boston judge to hear first test of GNU license used in MySQL" http://www.politechbot.com/p-03188.html
"NuSphere replies to Politech post about FSF and Boston lawsuit" http://www.politechbot.com/p-03194.html
Background on the difference between the ordinary GPL and the Library GPL (aka Lesser GPL): http://www.gnu.org/philosophy/why-not-lgpl.html
---
Date: Wed, 27 Feb 2002 13:49:43 -0800 (PST) From: John Palfrey <jgpalfrey@yahoo.com> Subject: One perspective on Courtroom 13's Open Source story To: declan@well.com Cc: ekelly@ropesgray.com
Dear Declan:
I thought I would offer one person's take on what just transpired in Courtroom 13 at the US District Court in Boston, where Judge Patti Saris heard arguments in the complex dispute between MySQL and NuSphere/Progress, which centered in no small part on the use of open source software said to be governed by the GPL.
As previous posts to politechbot reported earlier, both sides had sued the other last summer. At this stage, Judge Saris was considering whether to grant a preliminary injunction on two issues: 1) NuSphere's use of the MySQL-related trademarks and 2) NuSphere's use of certain code developed by MySQL and governed by the GPL as part of its Gemini product. Though she did not issue any orders on the spot, Judge Saris made it plain that she intended to issue a preliminary injunction against NuSphere's use of the marks in question (likely with 90 days to stop usage of the marks) and that she was not going to issue a preliminary injunction in the more complicated matter of the use of the MySQL code and terms the GPL.
1. The trademark issue was a fairly straight-forward contract construction discussion, with no cutting edge issues involved (from my perspective). She started with the presumption that irreparable harm existed, and never moved far off that point.
2. The harder issue, and the one on which she will almost certain not rule at this time, is the issue of whether NuSphere's use of the MySQL code under the GPL, and what NuSphere provides to end users pursuant to the GPL. One interesting point: it did not appear that anyone was arguing that the GPL did not apply or was not a valid license. It sounded as though the GPL was treated as any other license would be in a software context. Judge Saris seemed to focus on the question of whether Gemini constitutes an independent or a derivative work and whether the harm caused to MySQL met the irreparable standard. Experts -- none of whom were permitted to testify today, though Eben Moglen, perhaps among others, was in the room -- had filed what the Judge called "classic book-ends," or perfectly conflicting reports, on the question of the derivative work. Much of her questioning surrounded whether Gemini could operate without MySQL (as distributed, MySQL contends, it cannot) and whether or not the two products had been "integrated". She seemed to be moved by the NuSphere argument that there was no co-mingling of the source code and that "linking" to another program did not equate to creation of a derivative work. She also pushed hard on the questions of whether the distribution clause of the GPL was violated, though little progress was made on that point by either side. Ultimately, Judge Saris seemed unconvinced that MySQL could show a likelihood of success on the question of irreparable harm.
Judge Saris amused the crowd by referring to the open source movement as "like a religious movement." She also elicited snickers from certain members of the audience by referring to "Microsoft WordPerfect."
Before concluding the proceedings, she urged the parties to attempt settlement negotiations.
Best,
John P.
_______________________________
John G. Palfrey, Jr. Ropes & Gray One International Place Boston, MA 02110 jpalfrey@ropesgray.com 617.951.7037 direct dial 617.951.7050 facsimile
***********
Date: Wed, 27 Feb 2002 13:54:26 -0800 From: Craig Dickson <crdic@yahoo.com> To: Declan McCullagh <declan@well.com> Subject: Re: FC: NuSphere replies to Politech post about FSF and Boston lawsuit In-Reply-To: <5.1.0.14.0.20020227125949.0247fa00@mail.well.com>
Mark Lorion wrote:
> MySQL AB and the Free Software Foundation contends that > NuSphere violated the GPL by simply linking proprietary software to the > MySQL system using a public API. MySQL AB is interpreting the GPL so broadly > that any commercial software that comes into contact with free software must > also become free. By that standard, a commercial email program would > violate the GPL if it downloaded mail from a GPL-compliant mail server.
As someone who is neither associated with MySQL AB, NuSphere, or the FSF, but who is familiar with both sides' statements on this issue and the FSF's traditional interpretation of the GPL, I would like to comment on this.
MySQL AB asserts that NuSphere's proprietary Gemini code is statically linked to GPL'd MySQL code. As I understand the term "statically linked" (I am a software engineer, so I expect I know what it means), this means that Gemini and MySQL code are commingled in a single executable file, and therefore together form a single program. This is clearly not the same thing as an email client communicating with an email server; in that case, two independent programs are involved. This is a distinction that I have seen the FSF make many times, and it seems to me a very sensible one, as it preserves the notion of a "software commons" without making it impossible for proprietary software to interact with free software. If it is true that Gemini code is statically linked to GPL'd MySQL code, then Mr. Lorion's analogy is wildly inapt and displays a lack of understanding of the GPL -- information about which is publicly available on the FSF's web site.
Craig Dickson
***********
To: declan@well.com Subject: Re: FC: NuSphere replies to Politech post about FSF and Boston lawsuit From: Jake Donham <jake@bitmechanic.com> Date: 27 Feb 2002 16:40:59 -0500
Hi Declan,
You'll probably get plenty of mail about this, but I wanted to give my armchair lawyer's opinion that Mark Lorion seems to be misreading the GPL when he says:
Mark> MySQL AB and the Free Software Foundation contends that Mark> NuSphere violated the GPL by simply linking proprietary Mark> software to the MySQL system using a public API.
The applicable language in the GPL is
These requirements [that derived works be licensed under the GPL] apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.
The NuSphere code cannot be reasonably considered an independent and separate work in itself in the absence of another implementation of the MySQL APIs, because the code is useless without linking to MySQL. If there were another implementation of the MySQL APIs, NuSphere would still violate the GPL by distributing their code together with MySQL.
The Lesser GPL exists for authors who want people to be able to link non-GPL'd works to their code without infringement, but MySQL is not distributed under the LGPL.
Mark> MySQL AB is interpreting the GPL so broadly that any Mark> commercial software that comes into contact with free Mark> software must also become free. By that standard, a Mark> commercial email program would violate the GPL if it Mark> downloaded mail from a GPL-compliant mail server.
A commercial email program can be reasonably considered an independent and separate work because it also works with non-GPL'd mail servers. Also, presumably the mail client and server are not distributed together. "Coming into contact" is not a violation of the GPL.
Jake
***********
Date: Wed, 27 Feb 2002 11:23:27 -0800 (PST) From: Abraham Ingersoll <abe@dajoba.com> To: Declan McCullagh <declan@well.com> Subject: Re: FC: NuSphere replies to Politech post about FSF and Boston lawsuit
I'm a free-software publisher who has dealt with similar license issues, even so far as bringing up concerns about the GPLv2 with Bradley Kunz and in turn, Eben Moglen.
Just as Moglen's deposition concludes, I think it's pretty clear-cut that NuSphere violated the GPL by not releasing the source when they should. MysqlAB's not a happy camper from previous issues, and has now decided to stick it to NuSphere with the GPL. Too bad for NuSphere.
As far as NuSphere crying about this setting a bad example for the free software, and whining about it's implications in the business marketplace -- SHUTTUP ALREADY!
NuSphere built something off of MysqlAB's back and then didn't play by the rules (the GPL). Usually this wouldn't be too big of a problem, but they had previously pissed off Monty (MysqlAB's lead) in many other ways.
What's truly disgusting is how NuSphere's trying to muddy the waters with claims about the GPL's supposed chilling-effect on intellectual property rights.
That's just not the issue. The issue is whether they followed the GPL. They did not. Now how exactly a judge is going to sort this out, and how enforceable the GPL really is -- that's going to be fun to watch. But NuSphere crying like a baby .. all that does is simply show what a shallow corporation they really are.
NuSphere, bad. GPL, good. MysqlAB, dunno.
Abe
***********
From: Amos Satterlee <asatterlee@inta.org> To: "'declan@well.com'" <declan@well.com> Subject: RE: NuSphere replies to Politech post about FSF and Boston lawsui t Date: Wed, 27 Feb 2002 14:19:19 -0500
Declan:
I can't seem to find anything on NuSphere's original suit, just their respones to the counter-suit.
The agreement that they (NuSphere) post on their site clearly states:
a. payment UP TO 2.5 million, based on coming to a second agreement. In fact, they paid the money they said they would in the June 2000 agreement.
b. they were given right to register combination trademarks.
If their expert is correct and Gemini can act as a freestanding application apart from MySQL, then I tend to agree that it's a stretch to GPL it -- or are there other issues? It does seem to be a private spat wrapped up in the flag of Free Software, but that's because I'm not clearly what NuSphere's original beef was about.
Agreement: http://www.nusphere.com/misc_stuff/mysqlab_agreement.pdf Expert: http://www.nusphere.com/misc_stuff/declarationofbrucefwebster.pdf
Amos
***********
From: "Vincent Penquerc'h" <Vincent.Penquerch@artworks.co.uk> To: mlorion@nusphere.com Cc: "'declan@well.com'" <declan@well.com> Subject: RE: NuSphere replies to Politech post about FSF and Boston lawsuit Date: Wed, 27 Feb 2002 19:15:27 -0000
Hi,
> NuSphere violated the GPL by simply linking proprietary > software to the > MySQL system using a public API. MySQL AB is interpreting the > GPL so broadly > that any commercial software that comes into contact with > free software must > also become free. By that standard, a commercial email program would > violate the GPL if it downloaded mail from a GPL-compliant > mail server.
The two cases you mention are wildly different in scope. In the first case, the code linking against a GPL library leverages the library in order to perform some work it could not do at all were it not linked against it. In the second case, the email program you mention merely *allows* one to use it alongside a GPLed mail server. The email program would work without. Associating these two wildly different cases shows either you don't know the obvious difference between the two, or that you're trying to get people who don't know the difference to believe that what you say is true. I hope it's the former. In the case of linking against a library, the case is very specifically handled by the LGPL, which permits such linking. The very existence of the LGPL (created, as the GPL, by the FSF) shows that this distinction is not a fine one as you seem to imply.