User: Password:
|
|
Subscribe / Log in / New account

Preferred form for whom?

Preferred form for whom?

Posted Mar 10, 2011 3:11 UTC (Thu) by branden (guest, #7029)
In reply to: Preferred form for whom? by branden
Parent article: Red Hat and the GPL

There's a typo in the above; I quoted 3b), not 2b).

But 3a) is even more squarely on point:

" a) Accompany it with the complete corresponding machine-readable
source code, which must be distributed under the terms of Sections
1 and 2 above on a medium customarily used for software interchange; or,"

That's what closes the source-printed-on-paper loophole. Not the "preferred form of the work for making modifications to it."

Put differently, if you're right about the "preferred form" wording being the cause of the elimination of the paper loophole, then the "medium customarily used for software interchange" language becomes redundant and therefore meaningless.

Ask an attorney (I'm not one), and they will tell you that there are standard rules of construction that courts use when interpreting legal documents that strive *not* to create redundancies or rob language of meaning.


(Log in to post comments)

Preferred form for whom?

Posted Mar 11, 2011 1:00 UTC (Fri) by rahvin (subscriber, #16953) [Link]

Your right, I've confused the two clauses. But that doesn't change the arguement. The GPL is a software license. It's commonly handled in the US under precedents and written law that differs materially from Contract law in only a few circumstances.

Anyway, when brought before a Judge it's the plain language of the license that matters. Where that plain language is ambiguous and the parties don't agree on the intent the court is obligated to consider extrinsic evidence including the intent of both parties at the time the agreement was executed (good luck on that). It's important that both parties know each other intent or have made available their intent in some way otherwise their intent becomes ambiguous and you move to the next step. Baring that knowledge the Judge is obligated to turn to the creators of the agreement. Baring that there are many other sources of extrinsic evidence that can be examined. Baring that it's up to the Judge to try to find an equitable solution to the disagreement.

My main point all along is that RMS has defined what every single one of the clauses in the GPL are for, what they mean and what exploit of the agreement they are meant to prohibit. Baring some exchange of intent between the licensee and the licensor those statements are the highest value source the Judge is going to find explaining the intent of the GPL license. There may be other information that adds to the information available but to argue that isn't the basis for the agreement is ignoring the entire history of how the GPL came about and why it was created. RMS is directly responsible for the GPL, through his work that agreement was created and when you sign on to use that agreement, unless you state otherwise, you are agreeing to those terms and the reasons they exist.

My secondary point is that there are probably as many work flows and preferred forms of source code as there are developers. Baring any clear statements of intent from a developer what does a licensee have to rely on other than the plain language of the agreement based on RMS's definition of what the agreement means? I'll argue that there is nothing wrong with changing that meaning, but you need to write your own agreement and call it something other than the GPL because if your intent and interpretation differs from RMS, you aren't using the General Public License.

RMS has defined preferred form, that's the definition that applies baring the use of a different license.

Preferred form for whom?

Posted Mar 11, 2011 9:51 UTC (Fri) by PaXTeam (guest, #24616) [Link]

> [...]you are agreeing to those terms and the reasons they exist.

unless those reasons are explicitly spelled out in the license, no, you can't possibly be agreeing to them. what you do agree to is what's in the license, nothing more, nothing less.

> RMS has defined preferred form, that's the definition that applies
> baring the use of a different license.

you keep mentioning his definition/interpretation, where is it?

Preferred form for whom?

Posted Mar 11, 2011 18:14 UTC (Fri) by rahvin (subscriber, #16953) [Link]

unless those reasons are explicitly spelled out in the license, no, you can't possibly be agreeing to them. what you do agree to is what's in the license, nothing more, nothing less.
If the licensor hasn't made his definition of preferred form (if it's different than the GPL definition) available to the licensee at the time the license is executed the extrinsic evidence available to the court will be considered. The best of that evidence is the statements by the FSF and RMS on what the GPL means and why they wrote it, after all they are essentially the legal team that crafted the license. To legally demonstrate any other interpretation you would have to prove that the licensee was aware of some alternative interpretation at the time of license AND prove that the GPL definition is ambiguous. Whatever your preference is (and most people will have different preferences) is not necessarily what the licensee interprets when he draws a license.

The problem I have with this discussion is that people are pulling a single sentence out of the license and trying to reinterpret that sentence however they want without regard to the agreement or it's history. There are numerous places within the GPL where the preferred form is defined implicitly although not directly.
3. B. Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or,
The reference to preferred form, in my opinion, is a direct reference back to the paper copy exploit section and the binary source exploit section. Although the language could be a little clearer I don't think there is any doubt that the preferred form is machine readable source code on a medium typically used for software interchange. How that source is arranged as long as it's not obfuscated is beyond the license agreement. It would be absolutely silly to assume that whatever the licensee interprets to be their preferred form is the form the licensor has to comply with.

There is absolutely no doubt that the vanilla with patches and comments method is far more useful. But the license only requires that you be provided the source code in a manner that is editable and machine readable.

As far as providing quotes of RMS's speeches I have better things to do with my time. I've seen him discuss this and the methodology they went through when they created the agreement (basically trying to figure out how a naughty licensee will try to exploit the license to prevent free use of the source). I'm sure a paralegal can dig up plenty of references to RMS's and the FSF's intent when the GPL was drafted. Maybe at some point he'll weigh in on the issue but I doubt he will. If you want to pursue this line though you need to consider that there are probably more projects using the unified tarball than there are projects that are providing vanilla, patches and comments.

I think RedHat's move will complicate development of the kernel. But I also believe it's a necessary move if RedHat has any intention of responding to Oracle's attempts to destroy them. And I also believe RedHat could address the concerns by giving the concerned developers access to the information that's now missing. And I believe that if they fail to address this issue it could hurt them more than it helps. But I don't believe this is a GPL violation and I think it's a dramatic situational reinterpretation of the GPL and is something that could scare a lot of companies away from GPL. Just imagine for a moment that anyone that uses the software can redefine "preferred form" however they would like and what that would do to the community.

Preferred form for whom?

Posted Mar 11, 2011 22:35 UTC (Fri) by PaXTeam (guest, #24616) [Link]

> If the licensor hasn't made his definition of preferred form (if it's different than the GPL definition) [...]

ok, i think i see where your thinking went astray. you keep mentioning this 'GPL definition' of 'preferred form'. thing is, it does *not* exist. really, look at the GPLv2 (the license of linux) and see for yourself. what the GPL does define there is the 'source code', in terms of 'preferred form for modification'. it does *not* further elaborate on just what this 'preferred form' is. this is the entire reason for these discussions here because different people have quite different (or even contradictory) ideas about 'preferred form'.

> The reference to preferred form, in my opinion[...]

see, that's your own opinion only, noone seems to share it. and for good reason, that sentence in section 3.b does not define 'preferred form for modification'. really, if the license wanted to further elaborate on that term, it should and would have done so explicitly, like it already does for other terms where the license authors saw it necessary to be precise (whether they achieved that or not is another question but let's not digress). as to what your highlighted sentence was about see this thread somewhere else where others explained it. it's most definitely not about the 'preferred form for modification' term.

> Although the language could be a little clearer I don't think there is
> any doubt that the preferred form is machine readable source code on a
> medium typically used for software interchange. How that source is
> arranged as long as it's not obfuscated is beyond the license agreement.

the GPL does *not* contain the word 'obfuscation'. where did you pull that requirement from? and what does it mean? what is considered 'obfuscation'?

> It would be absolutely silly to assume that whatever the licensee
> interprets to be their preferred form is the form the licensor has to
> comply with.

i've seen people argue the exact opposite view in defense of RH here, but let see, if i take your view then it's bad news for RH as their *own* 'preferred form' is clearly sending patches around among their own developers, not entire tarballs as in the RHEL6 srpm. you know, not unlike what they currently distribute in the RHEL5 kernel srpm. how do you explain that two very different forms can both be preferred at the same time?

> But the license only requires that you be provided the source code in a
> manner that is editable and machine readable.

somewhere above you said it must not be 'obfuscated'. now you say it does not matter as long as the 'source code' is 'editable' (another word not present in the GPL, i don't know where you're getting these from ;). what does that mean? does it have to be in ASCII? 'cos your editor may not do EBCDIC, or whatever i invent for my own 'preferred form'. what if i translate everything into some intermediary representation and do my modification on that, what am i supposed to distribute then? certainly my own tools are mine and not affected by the GPL, so what will you do with that form then? really, instead of inventing new terms not present in the GPL, you should stick to its language and acknowledge where it is ambiguous.

> As far as providing quotes of RMS's speeches I have better things to do with my time.

it's just that you seemed very confident that he had precisely explained 'preferred form'. but it's not important at all, what really matters is the copyright holders' opinion, not his.

> But I also believe it's a necessary move if RedHat has any intention of
> responding to Oracle's attempts to destroy them.

paying customers of RH have access to the patches in question (one wonders why they exist if they're not the preferred form for modification, but let's not digress) and so does Oracle (how do you think Sun servers are certified to run RHEL?).

> Just imagine for a moment that anyone that uses the software can
> redefine "preferred form" however they would like and what that would do
> to the community.

i don't need to imagine anything, i just need to look at the past decades and see what people preferred and did not complain about. i'm sure you know it as well and apparently it didn't scare anyone.


Copyright © 2017, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds