| |||||||||||||
Commitment to Open ?Commitment to Open ?Posted Mar 5, 2011 11:07 UTC (Sat) by ballombe (subscriber, #9523)In reply to: Commitment to Open ? by rahulsundaram Parent article: Commitment to Open (Red Hat News)
Every person that download the GPL linux kernel SRPMS is a legitimate user of that kernel. They should receive the preferred form for making modification. That is the whole point of the license.
It is sad to see that Oracle is finally starting to destroy RedHat as a FOSS company. RedHat should not take the bait.
What RedHat is doing is a (admittedly very limited) form of vendor locking. I doubt this will end well since they cannot do anything about leaks (which they might not even know about).
(Log in to post comments)
Commitment to Open ? Posted Mar 5, 2011 12:40 UTC (Sat) by rahulsundaram (subscriber, #21946) [Link]
"Every person that download the GPL linux kernel SRPMS is a legitimate user of that kernel. They should receive the preferred form for making modification. That is the whole point of the license."
Maybe it is. Unless a copyright holder makes the claim in court that a single tarball doesn't satisfy the license terms, it makes no difference. Arm chair lawyering in LWN serves no real purpose.
Commitment to Open ? Posted Mar 5, 2011 15:24 UTC (Sat) by ewan (subscriber, #5533) [Link]
Bearing in mind that this move seems to be aimed at Oracle, and Oracle have quite a few kernel devs of their own with an interest in the code that Red Hat ships, they certainly have the standing to take this to court if they wanted.
There is no good outcome for Red Hat here; either this move pisses off community people and doesn't inconvenience Oracle at all, or it does inconvenience Oracle and they sue.
Really, aside from anything else, can you imagine how embarrassing it would be for Red Hat to defend, or even lose, a GPL enforcement case to Oracle?
It'll be interesting indeed... Posted Mar 5, 2011 16:01 UTC (Sat) by khim (subscriber, #9252) [Link] There is no good outcome for Red Hat here; either this move pisses off community people and doesn't inconvenience Oracle at all, or it does inconvenience Oracle and they sue. Wow. Interesting theory we have here. Now, why will they sue? RedHat have not invented this "monolythic-patch-for-everyone, VCS access to selected few" scheme. This distinction belongs to other well-known organization: FSF. For quite a few years GCC, Emacs and other FSF-sponsored programs were developed behind the closed doors and only releases were dumped on FTP. Essentially what the RedHat is doing. To argue that RedHat treats GPL improperly you must argue that FSF treated GPL improperly for years - and this is not so easy to do, believe me. P.S. It's interesting to note that community was pissed off back then too - but they fixed it introducing EGCS fork which was accepted by everyone, no by using courts to sue FSF.
It'll be interesting indeed... Posted Mar 6, 2011 15:08 UTC (Sun) by ballombe (subscriber, #9523) [Link]
FSF require copyright assignement, so nobody had standing to sue anyway.
It'll be interesting indeed... Posted Mar 6, 2011 20:13 UTC (Sun) by salimma (subscriber, #34460) [Link]
That's a completely orthogonal issue. Copyright assignment only restricts the merging of contributions to the upstream tree, and has nothing to do with the form in which that tree is made available to outsiders.
It'll be interesting indeed... Posted Mar 7, 2011 19:14 UTC (Mon) by vonbrand (subscriber, #4458) [Link] No, it isn't. FSF requires copyright assignment before they accept your changes. Once you have assigned copyright to them, they can well do as they please with your changes, i.e., distributing them (or not) in whatever form they desire. Nobody has any legal standing to sue them over what they do with their own stuff. Once somebody else gets, say, GCC, from the FSF, they are bound to honor GPL. Saying that FSF usually has distributed their packages only as monolithic tarballs is sort of a red hering. But others have followed suit, and FSF didn't say a word. So, in a sense they have set an example by their own actions and what they have allowed historically. And this could presumably be important in case some court considers ruling on what GPL's meaning of "preferred form for modification" really means. Also don't forget that for the most part of its history Linux itself was distributed as tarball + monolithic patches as the only format available, and later (in the bitkeeper days) that was still the canonical format. This does weight a lot against the "git trees are the preferred form" arguments.
Actually they do... Posted Mar 8, 2011 8:11 UTC (Tue) by khim (subscriber, #9252) [Link] Nobody has any legal standing to sue them over what they do with their own stuff. Actually they do. FSF's copyright assignment includes promises to the contributor, but yes, it's not GPL, it's separate assignment. It includes words like "machine-readable source code", but does not include words like "preferred form", so it's more relevant to spirit discussion and less relevant to legal discussion. But others have followed suit, and FSF didn't say a word. Actually they said lots of words about Cygnus (which had roughly similar policy to today's RedHat) - and most of them were bad. But again, they were about moral responsibility, etc - and not about legality...
Commitment to Open ? Posted Mar 5, 2011 23:36 UTC (Sat) by zonker (subscriber, #7867) [Link]
"or it does inconvenience Oracle and they sue."
Er, on what grounds? RHAT is perhaps doing the bare minimum required by the GPL - but they are doing it.
What this does does not prevent Oracle from recompiling the kernel and running it, nor does it thwart studying the kernel, etc. It just means that they have to do detective work to figure out what's changed and where. Unless I'm mistaken about the GPL, there is no requirement whatsoever to provide any commentary on changes or break changes out as patches from the original source - only that you have to supply the changes, which they've done, but as part of the overall kernel.
I can't see Oracle having any cause to sue, even if they are (I hope) inconvenienced by this.
Commitment to Open ? Posted Mar 6, 2011 0:42 UTC (Sun) by vonbrand (subscriber, #4458) [Link] Sorry, no. GPL talks about passing source code if you distribute a changed version, not original source code, changes, and running commentary. So here Red Hat has historically done much more than was asked for (distributing to all comers, distributing upstream source and individual patches), now they are still doing more than strictly required. Again the ages old discussion: If you distribute your stuff under $LICENSE, you can't then complain if somebody does what said $LICENSE allows.
Commitment to Open ? Posted Mar 6, 2011 2:08 UTC (Sun) by zonker (subscriber, #7867) [Link]
"If you distribute your stuff under $LICENSE, you can't then complain if somebody does what said $LICENSE allows."
You can complain, you just can't sue.
I compare this to the recent Supreme Court decision about the WBC - the "church" has the right to do what they do under the First Amendment. They should not be jailed for it, nor barred from it. That doesn't mean that the rest of the society/community can't or shouldn't shun them.
I'd much rather see companies kept in compliance with the spirit of the license by people complaining than by trying to write licenses that dictate every community standard.
Commitment to Open ? Posted Mar 6, 2011 2:35 UTC (Sun) by DOT (subscriber, #58786) [Link]
You can sue, you just can't win. ;) I also don't think complaining is going to win a whole lot. Companies are used to complainers.
Commitment to Open ? Posted Mar 6, 2011 11:07 UTC (Sun) by jch (guest, #51929) [Link]
> Unless a copyright holder makes the claim in court...
What has that to do with anything?
RedHat are clearly not distributing the "preferred form of the work for making modifications to it". They're therefore violating the spirit of the license.
The community ought to work with them to help them amend that mistake. This has nothing to do with the legal system.
--jch
Commitment to Open ? Posted Mar 6, 2011 11:37 UTC (Sun) by rahulsundaram (subscriber, #21946) [Link]
You claim a violation of "spirit". Some people claim it is a violation of the license. I was addressing the latter and not the former.
Commitment to Open ? Posted Mar 6, 2011 13:05 UTC (Sun) by jch (guest, #51929) [Link]
Let me correct myself, then.
RedHat are clearly not distributing the "preferred form of the work for making modifications to it". By failing to do so, they are violating both the spirit and the letter of the GPL.
Whether or not anyone intends to sue them is irrelevant -- they should fix what many of us see as a serious mistake, lest they loose much of the community's goodwill they have earned over the years.
Better?
--jch
Commitment to Open ? Posted Mar 6, 2011 14:19 UTC (Sun) by rahulsundaram (subscriber, #21946) [Link]
It expresses your opinion and belief better however a claim of license violation is a legal one and considering that Red Hat has a well qualified legal team that would have certainly reviewed the legal validity of such actions before Red Hat made them, I am inclined to trust their judgement instead of a layperson interpretation of it. In the case of a alleged infringement, whether or not one is willing to take it to court is certainly relevant and the absence of it, I consider such claims spurious ones.
I have no qualms about people arguing about the "spirit" of the license since that is open to interpretation and one doesn't need to be a copyright expert to do so.
|
|||||||||||||