GPL 3: An Open-Source Earthquake? (CRN) [LWN.net]

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 25, 2007 23:16 UTC (Thu) by rsidd (subscriber, #2582) [Link]

Their positions may well be different, and Perens may be closer to the FSF than ESR, but the criticism of the GP was that both ESR and Perens have proved to be good at talking but at little else. You offer nothing to rebut that criticism.

RMS is taken seriously not because of his polemical writing, but because of his actual contributions to free software, starting from emacs and gcc.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 25, 2007 23:32 UTC (Thu) by JohnNilsson (guest, #41242) [Link]

Why is it that contribution can only be measured in code? I think RMS did a great contribution by writing GPL, by forming FSF and by traveling around the world driving the politics of free software.

Similarily I think ESR has contributed a great deal through his writings and by his involvement in the Netscape/Mozilla deal.

I know to little about Bruce to comment though.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 2:52 UTC (Fri) by sbergman27 (subscriber, #10767) [Link]

"""Why is it that contribution can only be measured in code?"""

Because code is what counts. The rest is fluff.

Talk is cheap. Code is hard. It's real work.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 9:43 UTC (Fri) by bug1 (subscriber, #7097) [Link]

"Talk is cheap. Code is hard. It's real work."

You obviously dont understand teamwork.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 16:14 UTC (Fri) by sbergman27 (subscriber, #10767) [Link]

"""
You obviously dont understand teamwork.
"""

Teamwork is foremost on my mind.

Problem is... the path to fame is shorter for an advocate than for a coder.

Sure, Linus achieved fame... without really seeming like he wanted it. (Though he does, a little, I suspect.) But look at the hard work involved. And all the uncertainty.

But then we have these self-proclaimed pop icons crop up like Eric, Bruce, and Pamela.

All three being first and foremost into their own fame and notoriety. (The lady doth protest too much, methinks.)

There is a reason that Plato decided that the philosopher kings would have to be put into their positions *despite* their own wishes.

I automatically rate someone producing code higher than someone expending the same level of effort promoting themselves as advocates.

This is not to say that advocates are bad. It is simply a position that is more easily (and frequently) abused.

I hope that clarifies my position, whether I'm agreed with or not. ;-)

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 17:34 UTC (Fri) by nix (subscriber, #2304) [Link]

It seems a tad peculiar for one `into [her] own fame' to start by...
concealing her identity. In fact it seems like you've just contradicted
yourself.

(It's also interesting that you consider Groklaw to involve no work.
Obviously you imagine all that stuff just writes itself...)

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 18:22 UTC (Fri) by sbergman27 (subscriber, #10767) [Link]

Nix,

I didn't say it required no work. I said it was the shorter path to fame.

At any rate, to bicker would go against my own goals, and I think yours, too.

We might disagree upon this or that. We might trust and distrust different people.

I've expressed my suspicions. That's enough. I may be right. And I may be wrong. And I hope I'm wrong.

I'm in favor of solidarity. To argue my point further would not serve that purpose.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 20:47 UTC (Fri) by bug1 (subscriber, #7097) [Link]

"Problem is... the path to fame is shorter for an advocate than for a coder."

There is a saying, "those who make the most noise get the most done", it is an old saying, its not specific to coders, it applies to everything.

Ive read a bit of plato's republic, i would imaging them talking about this topic like this...

Would you have those naturally gifted in the abilities of an advocate spending their time coding, and those naturally gifted as a coder spending their time being an advocate ?

Fame certainly would help an advocate get their message to a larger audience, so fame lends itself to the advocate. It would be a very unsuccessful advocate that was not well known.

So let us say the best advocates need to be famous.

2nd person) And what of the coder, what abilities should they be endowed with ?

Certainly not jealously for it does not benefit anyone, i would say they should need be gifted in logic and concentration. Indeed being famous may impede the coders concentration from the constant interruptions from fans and foes.

2nd person) But has not the great coder as much right to be famous as the great athlete or the great advocate ?

True, the coder has a right to be famous, im just saying it doesnt help them produce work.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 1, 2007 9:28 UTC (Thu) by ekj (guest, #1524) [Link]

I thougth it was: "Empty barrels make the most noise."

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 14:51 UTC (Fri) by coriordan (guest, #7544) [Link]

I think you're wrong, but I think you've explained why the GPL works.

The best licence is one that developers don't have to think about when they're developing, but which they also don't have to worry about when they want to enforce it.

The GPL has made development and enforcement so safe and problem-free that if someone didn't think about it, they could easily think it wasn't there or that it was just fluff.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 16:27 UTC (Fri) by landley (guest, #6789) [Link]

Neither GPLv2 nor GPLv3 are "code". Therefore, by your definition, they
are unimportant. Yet you're still trolling about them.

> Talk is cheap.

Paranoid ranting about people you've never met is apparently free.

I haven't met PJ in person, but I have spoken to her on the phone. This
is more than I've done with Al Viro or Ingo Molnar, yet I don't question
their existence or identity. Heck, I handed BusyBox off to Denis Vlasenko
and he doesn't even show up on irc. (He's in the Ukraine, we're often on
different schedules.)

I deeply disagree with PJ's advocacy of GPLv3. I think she's wrong. Then
again I disagree with Eric Raymond about lots of things, and we're still
friends. It is possible for adults to respect people they disagree with.

Grow up.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 17:25 UTC (Fri) by sbergman27 (subscriber, #10767) [Link]

Rob,

Thanks for expending the effort to maintain busybox. Your contributions are greatly appreciated. Most of your users do not even know your name. Some of us do.

My ranting is not paranoid. At least, I hope it's not! ;-) I have no problem with our agreeing to disagree.

I'm worried about the motivations of some of our OSS luminaries. If that is paranoid then so be it.

Fourteen years ago, I was naive enough to think that the Internet was going to give us all the ability to communicate with each other and make the world a wonderful place for everyone.

I didn't even consider the possibility of spam, an Internet bubble, etc.

I'm a little more suspicious than I was back in those days.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 29, 2007 17:05 UTC (Mon) by nix (subscriber, #2304) [Link]

Fourteen years ago, I was naive enough to think that the Internet was going to give us all the ability to communicate with each other and make the world a wonderful place for everyone.

Lots of people wished that. In earlier times, people thought that the balloon, dynamite, the tank, nuclear weapons and even the telephone and radio would 'naturally end war' because everyone could communicate so of course they'd all agree.

It seems to be easy for techies to forget that people have differing goals (since we're all such nice people, doncha know), and that some people have goals such as 'maximize my personal power base' that contradict everyone else's.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 30, 2007 17:48 UTC (Tue) by rgoates (guest, #3280) [Link]

Nix, well put. I can often sympathize with those who indulge in utopian thinking. I wish some of those utopias were practical. But naivete regarding human nature is troubling; worse, it can be dangerous.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 25, 2007 23:48 UTC (Thu) by johnkarp (subscriber, #39285) [Link]

I think the open source movement does need *some* people who specialize in
talking and writing. For a couple reasons:

1. Coders often don't consistently have the time, inclination, or skill to
write articles or give talks for the media.
2. Someone needs to look far ahead, and coders are usually caught up in
the moment (by necessity).

One example: Very few science fiction writers have done much science per
se, but science fiction has done a lot *for* science (by inspiration.
posing thought problems, etc.).
Another: John Locke. He never constructed or oversaw a democracy. Does
that mean we should devalue his writings, or consider him irrelevent in
comparison to actual founders of democracies?

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 2:58 UTC (Fri) by Arker (guest, #14205) [Link]

GCC and Emacs are very valuable (and all too often underestimated) contributions, but they're the smallest of RMS' work in the long run. Code is not everything.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 8:30 UTC (Fri) by beoba (guest, #16942) [Link]

"You know, I'd vote for that politician, but I just don't know how well they <i>code</i>"

Some people are good at code, some people are good at writing, and a few people are good at both. Why not let members of our community do what they like doing? If Bruce is a good writer, why should anyone care what his code looks like?

A great benefit of our community is the ability for any member to go off and do whatever they want. We should not attempt to remove that benefit by claiming control over fellow members' activities.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 1:56 UTC (Fri) by charris (subscriber, #13263) [Link]

"As such, he's shown personal growth and contributed to political discussions consistently on a progressive front."

No need to insult the man. I'm sure he's a nice person despite all that.

Although Linus doesn't like the GPL-v3, he has also said it is simply not possible to change the license anyway on account of the numerous contributors who would need to give their permission. So it's the best of all possible worlds. KISS is the word.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 15:04 UTC (Fri) by coriordan (guest, #7544) [Link]

Relicensing is possible, it could be done over a 2-year timeframe. I wrote about how before.

After writing that, I was told it's even easier than I thought. Relicensing doesn't actually require that permission be gotten from 100% of the copyright holders. About 95% is enough, and then you have to remove any of the 5% if a relevent owner complains.

This is how the Mozilla codebase was relicensed to MPL+GPL despite taking public contributions for years.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 29, 2007 13:58 UTC (Mon) by forthy (guest, #1525) [Link]

Please, please stop circulating this myth. Linux single-handedly "relicensed" Linux to GPLv2-only in 2.4.0-test-something, without asking any other contributor. Why on earth should he need to ask the contributors now? It either was illegal then (then Linux is still GPLv2 or later), or it is legal now for him to decide. Actually, my interpretation of Linus' comment is that he is redistributing Linux currently under GPLv2, but he just can't speak for the other contributors. They have to speak for themselves, or if they fail to, the standard assumption from the GPLv2 is true ("or any later version unless stated otherwise"). So the only part of Linux which actually is legally GPLv2 only are those parts made by Linux himself, and this should be the least problem when he changes his mind.

Linus just uses that argument since he doesn't like the GPLv3. Let's wait for two years, and then he'll forget about this argument, as well.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 29, 2007 20:25 UTC (Mon) by bronson (subscriber, #4806) [Link]

Myth?

forthy, The "or later" has never been a part of the GPL; it's only a suggestion. Therefore, if "GPLv2 or later" doesn't appear in your project's declared license, then your project is GPLv2 only.

I can confidently say that "or later" has never appeared in any kernel release ever. It's always been GPLv2 only. Linus just added language in 2.4.0test-something clarifying that fact. Nothing was relicensed.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 5, 2007 19:27 UTC (Mon) by gerv (subscriber, #3376) [Link]

How confidently?

http://www.kernel.org/pub/linux/kernel/Historic/v0.99/lin...

gerv@otter:linux$ grep -A2 "Software Foundation" kernel/blk_drv/scsi/fdomain.c
* Free Software Foundation; either version 2, or (at your option) any
* later version.

gerv@otter:linux$

The "or later" language lives (or doesn't) in the header of each individual file, not in the COPYING file. If it's there and you remove it, you are arguably removing a right granted by the previous copyright holders - the right to use their contributions under later versions of the GPL. I'm not sure if you can actually do that.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 6, 2007 14:02 UTC (Tue) by notamisfit (guest, #40886) [Link]

Something like two thirds of the kernel header files only have copyright notices, with no mention of license.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 12, 2007 10:17 UTC (Mon) by forthy (guest, #1525) [Link]

The GPL covers this lazyness: "If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation."

So for two third of the files in Linux, you can choose any version ever published, they are essential compatible with GPLv1, v2, or once v3 is released, with v3, as well. "The Program" is legally a file, not the complete kernel. Only the complete kernel, the collection of many "programs", as distributed by Linux has this v2-only tag.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 12, 2007 17:17 UTC (Mon) by zlynx (subscriber, #2285) [Link]

COPYING does specify a version. If you're thinking of the files without a GPL included and if you don't believe COPYING applies to those files, then you have no license at all to those files. Your choice is GPLv2 or nothing.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 12, 2007 18:32 UTC (Mon) by gerv (subscriber, #3376) [Link]

COPYING does not specify a version as far as I can see:
http://www.kernel.org/pub/linux/kernel/COPYING

I believe COPYING applies to all unheadered files, and the latter part of Section 9 specifically: "If the Program does not specify a version number of
this License, you may choose any version ever published by the Free Software
Foundation."

Gerv

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 12, 2007 20:15 UTC (Mon) by zlynx (subscriber, #2285) [Link]

It's right there at the top:
GNU GENERAL PUBLIC LICENSE
Version 2, June 1991

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 12, 2007 20:34 UTC (Mon) by gerv (subscriber, #3376) [Link]

Indeed. The document does contain a copy of the GNU General Public License version 2. But it also says that you should put the boilerplate on to apply the new terms to your license, and what happens if the boilerplate isn't put on (section 9).

The MPL says "Mozilla Public License version 1.1" at the top, but software under the MPL can be used under any later version of the MPL - section 6.2.

For the GPL, exactly the same thing applies unless the boilerplate says different - i.e., unlike the MPL, the author has the option of restricting it if they choose. But they must explicitly choose to do so. If they do not, then "you may choose any version ever published by the Free Software Foundation".

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 12, 2007 20:46 UTC (Mon) by zlynx (subscriber, #2285) [Link]

Section 9 says nothing about boilerplate. Section 9 applies to code that has a bit like this from kernel/cpu.c in it:

/* CPU control.
* (C) 2001, 2002, 2003, 2004 Rusty Russell
*
* This code is licenced under the GPL.
*/

I would say that listing the entire text of GPL version 2, leaving off the "or later" notice, *and* the paragraph by Linus at the top of COPYING, makes it very clear that it is version 2.

If you are applying this argument to the early versions of Linux, which were less clear on licensing, then Linus could apply any version GPL he desired, and he picked GPL 2.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 12, 2007 21:18 UTC (Mon) by gerv (subscriber, #3376) [Link]

Listing the entire text is a requirement of the license. It makes no statement about versions. The paragraph by Linus at the top of COPYING (at least, the version I linked to) also says absolutely nothing about versions.

Yes, the "or later" notice is left off, but so is all the rest of the boilerplate. How much clearer can section 9 be? The Program (the file) does not specify a version number of this License (because it doesn't specify anything) and so consequentially you may choose any version ever published by the Free Software Foundation.

The boilerplate you quote also falls into the same category. But we have already agreed that if code has no boilerplate at all, then it is under the terms of GPL version 2 - as given in COPYING. And GPL version 2 clearly states that if no version number is given in boilerplate, you can use any version of the GPL ever published by the FSF.

To put it another way, the default terms in the case of no boilerplate are "any version ever published" (i.e. 1, 2 or 3). The standard default boilerplate is "this version or any later version" (i.e 2 or 3) and it can be modified to "this version only (i.e. 2). There are three different possibilities depending on the boilerplate status, and files with no boilerplate are in the first of those three buckets.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 12, 2007 22:22 UTC (Mon) by zlynx (subscriber, #2285) [Link]

I finally did look at the link you provided.

I had assumed it was a link to the latest kernel. But your link is COPYING for the kernel repositories. Each release of Linux contains its own COPYING file. Please look at COPYING in the top level of 2.6.20, for example. Then refer to my previous comment that by your argument, Linus was free to relicense to explicit version 2. I myself happen to agree with Linus that it was always version 2.

I *still* disagree with your interpretation of section 9. If I put the complete text of GPL Version 2 as the license on my program, and do not explicitly give you permission to relicense under any later version, then I *did* specify a version: version 2. I do not see how you can read "GNU GENERAL PUBLIC LICENSE Version 2" as anything other than specifying a version, and section 9 that you are so fond of says nothing about *where* the version must be specified. I believe section 9 applies only when the GPL is included by reference.

If you want *me* to agree with you, please explain and convince me why listing the version 2 GPL text does not specify version 2 GPL.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 14, 2007 21:42 UTC (Wed) by gerv (subscriber, #3376) [Link]

I've had a look at the COPYING in the latest kernel (2.6.20). Linus's statement there is interesting because I don't think it's immediately obvious that he has the legal authority to make that statement about any contribution other than his own - certainly not for contributions made before he added the statement.

If I contribute some code to the pre-Linus-announcement Linux kernel, and don't put a header on it, section 9 says that recipients can choose any version of the GPL ever published by the FSF. (You may still dispute that; but see below for further comments.) Linus can not take away that right I have given, because it's imposing further restrictions on the rights I am granting - something section 6 of the GPL forbids.

If I contribute unboilerplated code to the Linux kernel post-Linus-announcement, there is a conflict. The license supposedly applying to the kernel says recipients can use any version; Linus says it's only version 2. Which wins? Well, Linus doesn't own the copyright on my contributions, so I'd say the text of the GPL wins. But I agree it's a different and more complex situation.

Your interpretation of section 9 is wrong because it says "If _the Program_ specifies...". In other words, it *does* say where the version must be specified - it must be specified on the Program. The copy of the included license is not the Program.

If the mere inclusion of the licensing text of one particular GPL version were sufficient to fix the licensing of the accompanying software at that version, then that part of section 9 would not need to exist. Why would you ever have a clause beginning "If the Program does not specify a version number of this License", if it were impossible to not specify a version number, because the inclusion of the license text always meant that you had specified one?

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 6, 2007 20:14 UTC (Tue) by bronson (subscriber, #4806) [Link]

This is becoming a FAQ...

Obviously, the files that make up the Linux kernel of have all sorts of different copyright holders and licenses (many even have multiple licenses). There is one thing that holds them all together: each is somehow compatible with the GPLv2.

So, what license applies to the entire kernel as a whole? COPYING in the root directory, of course. Not coincidentally, this also happens to be the intersection of each license exception found in kernel. The "or later" clause is just one example.

And has /COPYING ever included the "or later" clause? Never. That's why I said "kernel release" instead of "kernel file": I'm referring to the kernel as a whole.

It's perfectly possible to include code with an optional exception into a project that doesn't include the exception. In fact, the GPLv3 makes this explicit.

As you say, if for some oddball reason you make the exception mandatory ("you MUST distribute this file under the most recent version of the GPL"), then that would be most problematic. Thankfully, that's not the case. Since all exceptions in the kernel are optional ("either version 2, or (at your option) any later version") then there is no problem when the entire kernel is distributed without them.

Does that make sense?

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 7, 2007 22:54 UTC (Wed) by gerv (subscriber, #3376) [Link]

If every file which has a GPL header has the "or later" language, then that entire codebase can be used under GPLv3 terms as well as GPLv2. (Any code under other licences which is compatible with GPLv2 will be compatible with GPLv3.)

COPYING defines the distribution terms of the aggregate body, yes. It also serves as the copy of the GPL that the GPL headers say must be present. However, the question is not "what are the terms attached to Linus's release of kernel x.y.z", but "can I change those terms to GPLv3?". And the answer to that question lies solely in the GPL headers on each file.

If a file has no header at all, I'm not sure what the legal position is. I would suggest that it's under the terms of the code surrounding it. But I don't know if a court has ever ruled on this.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 8, 2007 6:33 UTC (Thu) by bronson (subscriber, #4806) [Link]

It seems clear to me but of course IANAL...

If a file was contributed without any license information, I can see only one way to interpret this... We can't assume it was intended to be public domain since basically everything is under copyright unless those rights are expressly waived (in the US since 1970 anyway). The court can only guess that the contributing author(s) intended to use the project's umbrella license. But, you're right, this hasn't been ruled on yet.

"Any code under other licences which is compatible with GPLv2 will be compatible with GPLv3."

That's not true is it? I could write a license that's compatible with the GPLv2 and also requires Tivoization ("you must keep your signing key private"). That hypothetical license would be compatible with the GPLv2 and yet incompatible with the last draft of the GPLv3.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 8, 2007 9:51 UTC (Thu) by gerv (subscriber, #3376) [Link]

Pedant :-) Yes, you can imagine a licence for which that is not true. But all the licences used in the Linux kernel in this universe (BSD variants and such) fit that profile.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 2:09 UTC (Fri) by bronson (subscriber, #4806) [Link]

I don't know, Don. I would hardly call Bruce Perens's voice "sane and thoughtful." I think shrill and vitriolic would describe many of his LWN posts from the past two months. I hope he cringes at his choice of language when he re-reads some of them now.

Personally, I was solidly in the anti-Novell-deal crowd until Bruce jumped in the middle and started trying to whip everybody into a frenzy. At that point I decided that I couldn't agree with anybody in the room and ignored anything Novell for the next few weeks.

I'm glad Bruce failed, I just wish he hadn't have gone so far over the top in doing so. He turned what could have been very pro-Linux news stories ("Microsoft claims it owns you, here are alternatives") into a solid black eye ("hey look, the Linux guys are slinging mud at each other over yet another petty issue").

If Bruce would just tone it down a bit, he might win me back over (same goes for ESR). But I wish he would quit trying to wrestle control of a project he has no interest in coding for and stop using outrageous hyperbole and personal attacks, calling Nat and Miguel amoral, in his public discourse. It just makes everybody look bad.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 3:17 UTC (Fri) by b3timmons (guest, #40286) [Link]

What exactly is meant by Bruce's "failure"?

It seems to me that we often lose sight of one of the fundamental differences between the community and corporations. Is transparency not worth the price of potential embarrassment?

I think it goes beyond transparency to freedom -- just a little more freedom to express oneself publicly than, say, in a corporate context. Failing to appreciate this freedom can easily lead to, among other things, rash choices of words such as a few that Bruce made. Who among us has not done this?

Many forget that long ago RMS and the FSF sought to focus more and more on helping education about free software* to catch up with its production and use. Bruce has helped in many ways in this area, an area that many open source folks seem to be hostile toward.

(*) Sadly even just the four freedoms seem to be too much to accept for many in the FOSS community.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 7:06 UTC (Fri) by bronson (subscriber, #4806) [Link]

Failure to influence the deal in any way. I mostly agree with what you say. If you (or Bruce) wants the freedom to shoot your mouth off, that's fine by me. However, if you do, don't expect businessmen or journalists to take you seriously. And please don't try to present yourself as the voice of an industry; that role absolutely requires responsibility and decorum. Crying wolf too much can make the entire industry look bad.

So, if you want to be in a professional setting, I think it's a good idea to behave in a professional manner.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 8:34 UTC (Fri) by beoba (guest, #16942) [Link]

Do Novell and MS execs read LWN? Should I be wearing a tie right now?

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 17:22 UTC (Fri) by vonbrand (subscriber, #4458) [Link]

If they did demand that, I'd proudly only read LWM in a T-shirt

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 18:18 UTC (Fri) by bronson (subscriber, #4806) [Link]

That depends, beoba. Are you producing press releases and sound bytes intended to be picked up in trade mags? Do you claim to be speaking for the entire community?

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 29, 2007 17:07 UTC (Mon) by nix (subscriber, #2304) [Link]

Do you expect to turn up in a court action describing a 'file of people to be killed'?

(Mind you that was definitely not Bruce's fault, and there's no way he could have anticipated anything that crazed... ;) )

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 29, 2007 20:37 UTC (Mon) by bronson (subscriber, #4806) [Link]

Hey, yeah! I'd forgotten about that. This is Bruce's comment:

http://lwn.net/Articles/107571/

I thought it was a great post. Merkey just failed th grasp the difference between "email killfile" and "file of people to kill".

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 29, 2007 22:25 UTC (Mon) by Arker (guest, #14205) [Link]

I personally would say it's entirely beyond the limits of credibility to suppose that a man with his background didn't know what a killfile is. I'd bet he has one with entries going back to 1990 or so, at least. I see it as a cynical ploy based on the belief that a *judge* wouldn't be likely to understand the reference.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 16:17 UTC (Fri) by landley (guest, #6789) [Link]

> Raymond may have become a one-trick pony who has drifted away from
> current relevance,

Depends on your definition of "relevance". The Art of Unix Programming
was a new york times bestseller around 2003/2004, is still in print 3
years later, and he was asked to do a teaching supplement so it could be
used as a textbook. He may no longer maintain the Emacs Lisp library, the
sunsite infrastructure, or fetchmail (which he handed off to other
maintainers 4-5 years ago and people are _still_ blaming him about for the
exploit du jour), but his current coding project is fairly well used (he
maintains gpsd, the global positioning system daemon). Before that he did
a tree comparison thing (comparator/shredcompare) which gets used in court
from time to time, and probably other stuff I haven't noticed. He
regularly starts things and hands them off to other people: he could have
stayed president of OSI for life (after all he founded it) but he handed
it off (_twice_, it came back the first time) because that's not what he
wanted.

> but Bruce has certainly taken a different path -- one which has brought
> him far closer to the FSF's camp. As such, he's shown personal growth
> and contributed to political discussions consistently on a progressive
> front.

What's the last remotely relevant thing Bruce did beyond "give
interviews"? Other than argue in favor of a position you like?

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 1:14 UTC (Fri) by sbergman27 (subscriber, #10767) [Link]

"""
Bruce is a former Debian Project Leader. He is the primary author of the Debian Free Software Guidelines (on which the Open Source definition is based) and the Social Contract. He also did lots of other things for Debian.
"""

If only the death penalty were allowed in this case! Sigh...

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 1:18 UTC (Fri) by drag (subscriber, #31333) [Link]

That and things like the Holloween documents and attempting to make sure that free software developers were aware of the threat that patents are to them.

But he does have his dipshit moments, like everybody else. Getting upitty with busybox when he was uninvolved for years, and Novell overraction are two big ones.

I know I'd like to take back some things I've said online.

Anybody who is passionate about a subject and is very outspoken is going to get their judgement clouded time to time and be straight wrong other times.

As far as ESR goes... people like to degrade him for his non-politically correct view points (of some I share). He is not, never was that relevent, and he isn't a profitant developer and he was only kinda-sorta right with the Cathedral vs Bazzar. However a couple things he has worked on I like a lot.

For instance he hit the nail on the head with the design of GpsD, for instance. It's simple, correct, flexible, easy to deal with, easy to write software that is comaptable with it, and easy to troubleshoot. Networkable also. Classic Unix design. What it does is the sort of things that most people tend to realy realy over engineer and make application-specific manner. Lots of developers trying to do things could learn from that. Simple is better.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 3:44 UTC (Fri) by Thalience (subscriber, #4217) [Link]

I would also point to "The Art of Unix Programming" as a positive, semi-recent contribution from Mr. Raymond. Its the first Unix book I would give any programmer who only had experience in the Windows or corporate Java universe.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 1, 2007 13:12 UTC (Thu) by sean.hunter (guest, #7920) [Link]

Read "The Unix Programming Environment" by Kernighan and Pike to understand what a really good programming book can be.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 1:25 UTC (Fri) by JoeBuck (subscriber, #2330) [Link]

Too bad this comment thread just seems to be about name-calling.

My guess is that if the FSF produces a license that is acceptable to enough of the big players (e.g. Sun, for Java and Solaris; developers of other interesting projects), we'll probably wind up seeing an evolutionary approach that will work things out for those who still have problems, and we'll see "GPL v2 or any later version" code more and more, possibly motivated by non-ideological folks who just want to share code between Linux and Solaris kernels. As GPLv3 code becomes more common and the world doesn't end, the battle will cool off.

And maybe we'll wind up seeing the Linux kernel have GPLv3 parts with a special exception striking the anti-DRM language. Many of the corporate contributors might like that a whole lot, since they generally prefer the draft GPLv3 patent language to the GPLv2 patent language.

Or maybe we won't and the Linux kernel will stay GPLv2. It's not really all that big a deal.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 1:32 UTC (Fri) by JoeF (subscriber, #4486) [Link]

Or maybe we won't and the Linux kernel will stay GPLv2. It's not really all that big a deal.

As has been stated lots of times before, Linus doesn't have the power to change the kernel license. He could of course change the license on the parts of the code that he wrote, like everybody else who contributed to the kernel. But for the whole kernel to change licenses, every contributor would have to agree. And that's just unlikely to happen.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 18:05 UTC (Fri) by sbergman27 (subscriber, #10767) [Link]

"""But for the whole kernel to change licenses, every contributor would have to agree. And that's just unlikely to happen."""

Some of the contributors are dead. Their benefactors control the license... if they had a proper will. If they didn't... then what?

It's not so much a matter of unlikely. It's a matter of how long it would take to sort out all the legal mess.

And a license change would permanently weaken Linux's legal position.

With SCO, badly as their case has gone, it has taken *years*, with no evidence(!), and the thing hasn't even made it to trial yet.

Do you really want to open the door to another "SCO Case", or even several?

Except that in these future "Linux license change" cases, the opposition might really *have* a case.

Changing the Linux Kernel license would almost certainly precipitate such an ugly situation. I, for one, do not want to see it.

Whether GPLv2 or GPLv3 is superior, and I will refrain from giving my opinion here... a license change for the Linux kernel is a really bad idea.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 1:36 UTC (Fri) by drag (subscriber, #31333) [Link]

I've said it before, but I'll say it again because I think it's a cool idea.

If a project is GPLv2, but would like to upgade to GPLv3 then one of the big things they could would be to either dual license it GPLv2 and GPLv3 or add a exception to the GPLv3 for backward compatability.

This would be to show people that if your moving to GPLv3 that your not doing it to force anybody and it wouldn't require other projects to move to GPLv3 in order to remain compatable. This should avoid a lot of fights.

Also if the GPLv3 turns out to be a flop in practice then this enables you to go back to GPLv2 if you have to without headaches.

Then if/when the GPLv3 becomes widely adopted and there is no longer any sort of compatability problems with other people's software and such then you can move to pure GPLv3 and then get all the benifits of the updates to the license.

I think that this would allow a nice transition from one to another. It may take a few years to get it all the way done, but I think that will be worth it to avoid project forking and infighting.

Much in the same way that my computer is 64bit, but retains 32bit compatability. There is a few issues that cause me to need to have some 32bit applications but that is getting less and less important everyday and eventually it will just stop mattering.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 11:20 UTC (Fri) by pinky0x51 (guest, #40742) [Link]

>If a project is GPLv2, but would like to upgade to GPLv3 then one of the big things they could would be to either dual license it GPLv2 and GPLv3 or add a exception to the GPLv3 for backward compatability.

It is much easier, just release your code as "GPLv2 or later" so everyone can choose to use GPLv2 or GPLv3 and you don't need special execptions etc.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 12:47 UTC (Fri) by drag (subscriber, #31333) [Link]

Well the point is to move gradually to the GPLv3 license.

This way you can make sure that everybody involved in the project would be onboard for it because you know would know that anybody contributing code will understand that it will probably be under the GPLv3 license eventually.

This will give time for people to mull it over and give time for the real-world effect of the GPLv3 to sink in instead of jumping full bore into it and pressuring people to make a desicion right away.

Plus you would have people who contribute code to your project do it under the same dual license so that the patent language goes into effect.

Another variation of the idea would be to release the project tarballs and binaries under the GPLv3, but have a linking exception with GPLv2 code as well as make it project policy that if a GPLv2 licensed project would like to use source code then that would be fine.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 14:09 UTC (Fri) by pinky0x51 (guest, #40742) [Link]

>This way you can make sure that everybody involved in the project would be onboard for it because you know would know that anybody contributing code will understand that it will probably be under the GPLv3 license eventually. [..] Plus you would have people who contribute code to your project do it under the same dual license

I don't see the difference if you say to a contributor: "You have to dual-license your code under GPLv2 and GPLv3", "You have to license you code under GPLv3 plus my special backward clause" or "You have to license you code under GPLv2 or later".

You always have to ask contributors for a special license and you always achieve the same but using "GPLv2 or later" is imho the easiest and most common way.

GPL 3: What is the fuss all about?

Posted Jan 26, 2007 15:46 UTC (Fri) by vonbrand (subscriber, #4458) [Link]

don't see the difference if you say to a contributor: "You have to dual-license your code under GPLv2 and GPLv3", "You have to license you code under GPLv3 plus my special backward clause" or "You have to license you code under GPLv2 or later".

There are certainly differences... "GPLv2 or v3" is not the same as "GPLv2 or later" (this would include GPLv4!), "GPLv3 + backward clause" is not the same as "GPLv2" either. Etc.

And please don't lose sight of the fact that the anti-DRM provisions in GPLv2 are the problem here. I agree with Linus that if I get Linux on, say, a wireless router or any other type of appliance, I should not be automatically able to run any other junk on that machine. It is quite reasonable for the manufacturer of certain devices to restrict what they can do (it might be even mandated by law), be it by physical limits or software-enforced ones. And in the current world, where it is already much cheaper (and better) to control a washing machine or elevator by software and not mechanically, this later case will be ever more prevalent. I still can take the source code to Linux I get with such a device and run it elsewhere.

Also, please do remember that GPLed software (and open source in general) is a tiny minority in all software that is out there, even more so in the areas where GPLv3 anti-DRM provisions could make some difference. Changing the license of a piece of software to prohibit some uses will make no difference whatsoever, except for making certain people feel self-important.

GPL 3: What is the fuss all about?

Posted Jan 27, 2007 0:04 UTC (Sat) by drag (subscriber, #31333) [Link]

Well.. the 'Anti-DRM' clause isn't designed to prevent the use of DRM or discourage it.

Just like the clause stating that you have to distribute source code if your redistribute GPL'd software is not designed to try to make other people (say microsoft) open source their programs.

It's designed to protect the '4 freedoms'. It is essentially a way to attempt to enforce the idea that you can't make GPL software non-Free (by FSF/GNU's definition). Either refusal to release code or using encryption to enforce specific software versions the effect on end users is the same.

In all actuality if somebody (Sun has a project to do just this) releases open source software that is able to succcessfully impliment DRM, even if you try to modify the code, then you could license that under the GPLv3.

Also if you would like to use GPLv3 software to play back DRM content or control hardware that impliments DRM, then that is fine also.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Feb 10, 2007 18:40 UTC (Sat) by RareCactus (guest, #41198) [Link]

This whole article is a blatant troll.

The Linux kernel has always been under the GPLv2-- check COPYING at the root of the tree. After another two years, there will have been another two years' worth of contributions to the Linux kernel by various people, explicitly tagged as "GPLv2 only," because that's what Linus is now requiring.

Whatever your opinions are about the GPLv3 or any other license, it's irrelevant. The Linux kernel is under the GPLv2.

GPL 3: An Open-Source Earthquake? (CRN)

Posted Jan 26, 2007 16:54 UTC (Fri) by dhess (guest, #7827) [Link]

It's off-topic, since the article was written about the GPL v3 and not Bruce Perens' merits, but I want to second what jeroen said about Bruce's contributions to free software. Bruce has done a lot for the free software community, much of it without any financial reward.

I've worked with Bruce in the past, and I think he's a very reasonable guy. When it comes to GNU/Linux and free software issues, he's opinionated but not uncompromising. He's spent a lot of time thinking about this stuff and about how to make it palatable to people who don't understand free software. He also keeps his opinions on unrelated issues to himself, unlike certain other members of "our tribe."

Bruce has earned the right to be contacted for stories like this one. I'm glad to have him speaking up on free software issues, and I hope he continues to do so. If you disagree with him, that's fine, but it's no reason to attack him or to question his impact on the free software movement.