Why Torvalds is sitting out the GPLv3 process (Linux.com) [LWN.net]

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 26, 2006 23:03 UTC (Tue) by smitty_one_each (subscriber, #28989)
In reply to: Why Torvalds is sitting out the GPLv3 process (Linux.com) by mingo
Parent article: Why Torvalds is sitting out the GPLv3 process (Linux.com)

That link begs the question of whether/when glibc will go GPLv3, and what unintended consequences WRT the Linux kernel might shake out.

to post comments

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 27, 2006 0:43 UTC (Wed) by Zack (guest, #37335) [Link] (25 responses)

That disagreement was quite a while ago. There was somewhat of an aftermath and another side to the story but it got drowned out in the subsequent events.
Suffice to say that in the end the maintainer didn't fork away glibc from the GNU-project so perhaps the differences where not completely irreconcilable (although probably far from comfortable).

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 27, 2006 1:03 UTC (Wed) by mingo (subscriber, #31122) [Link] (24 responses)

Suffice to say that in the end the maintainer didn't fork away glibc [...]

as the link i gave describes it perfectly well, the Stallman "backstabbing" failed and thus Ulrich Drepper did not have to fork glibc. Thus it's not "in the end" but right at the beginning, and the "suffice to say" is misplaced at best :)

What we see here is a pattern of RMS being accused of attacking people behind their back - and i have no reason to doubt Ulrich Drepper's words. I quote:

The morale of this is that people will hopefully realize what a control freak and raging manic Stallman is. Don't trust him. As soon as something isn't in line with his view he'll stab you in the back. *NEVER* voluntarily put a project you work on under the GNU umbrella since this means in Stallman's opinion that he has the right to make decisions for the project. (link)

As i said before, this whole thing is trust-based for me. (You might have other preferences or other experience - it's your choice.)

I've never seen Linus do any "backstabbing" (or being accused of doing that) - in fact i can attest to the fact that every time Linus called me stupid it was right in front of everyone ;-)

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 27, 2006 5:53 UTC (Wed) by b3timmons (guest, #40286) [Link] (20 responses)

Breaking trust is when someone leads you to expect something and then intentionally betrays your expectations. I have never seen where FSF or RMS has ever broken anyone's trust.

Several posters have tried to explain to you that it is easy to trust RMS/FSF. They are quite predictable. After all, an alternative formulation of Drepper's complaint is that you *can* trust RMS to insist upon his views. In Drepper's case, it was a FSF project, so why should we not expect RMS to insist upon various things? It's naive to assign something to the FSF or any entity for that matter and expect it not to be subjected to their agenda.

We all know that RMS claims that people act ethically or unethically. How is that a personal attack, if he can show how it is true? You may end up disagreeing with it, but I would bet you would have to claim a different ethical standard to do so. I would suggest that if you want to counter RMS/FSF, a more constructive means would be to explain what is unethical about what they do.

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 27, 2006 14:46 UTC (Wed) by tjc (guest, #137) [Link] (19 responses)

We all know that RMS claims that people act ethically or unethically. How is that a personal attack, if he can show how it is true? You may end up disagreeing with it, but I would bet you would have to claim a different ethical standard to do so.

Most people do "claim a different ethical standard" that differs from RMS's unusual views of right and wrong. For example, what percentage of the general population thinks that non-free software is ethically wrong? Less than 1/1,000,000 of one percent?

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 27, 2006 17:15 UTC (Wed) by b3timmons (guest, #40286) [Link] (18 responses)

Most people do "claim a different ethical standard" that differs from RMS's unusual views of right and wrong. For example, what percentage of the general population thinks that non-free software is ethically wrong? Less than 1/1,000,000 of one percent?

Although most may claim differently, I have yet to find a convincing defense of their standard, say on the level of Stallman's essays. Instead of trying to show moral superiority of their standard, they seem to play down the role of ethics. Such moral relativism generally is not a constructive path toward doing the right thing.

For example, one of RMS's basic assumptions is that helping others is the basis of society. Finding fault with such an assumption could be a good starting point for an opposing argument. While I see no fault, I would gladly consider an well-expressed argument by someone who does.

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 27, 2006 17:53 UTC (Wed) by tjc (guest, #137) [Link] (17 responses)

Instead of trying to show moral superiority of their standard, they seem to play down the role of ethics. Such moral relativism generally is not a constructive path toward doing the right thing.

The fact that RMS attempts to show moral superiority does not in itself make his morality superior. His moral standard is just as relative to his own personal beliefs as is yours or mine.

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 27, 2006 18:43 UTC (Wed) by sepreece (guest, #19270) [Link] (1 responses)

In the most recent RMS interview I've seen (<http://www.redherring.com/Article.aspx?a=18757>), he comes over as one who clearly believes he is morally superior and looks at the rest of the world as being in need of his superior wisdom. He manages to be paternalistic and contemptuous at the same time that he advocates violating copyright to get access to movies outside DRM.

Of course, this attitude groups him with a lot of other leaders, some of whom we revere as visionaries, some of who we revile as tyrants or fools. Only time will tell. He's pointing at real problems; I just don't believe he's offering real solutions.

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 27, 2006 20:35 UTC (Wed) by b3timmons (guest, #40286) [Link]

Well, as you acknowledge, what matters in the end is whether he can help come up with solution, not his social manner.

Feelings of moral superiority is probably more prevalent among people than you suggest, so I do not see how the RMS interview shows anything special. It reminds me of the stereotypes of the rude East coasters who say what is on their mind versus the Californians who are afraid to say anything politically incorrect. Many, many people are holier-than-thou--RMS is just uninhibited to the point of occasional rudeness, not the worst quality IMO.

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 27, 2006 19:44 UTC (Wed) by b3timmons (guest, #40286) [Link] (10 responses)

The fact that RMS attempts to show moral superiority does not in itself make his morality superior. His moral standard is just as relative to his own personal beliefs as is yours or mine.

At least he is trying, complete with initial assumption and application of logic until conclusions. If there is no flaw found in his reasoning, then fault should be found in his assumptions. Indeed, too many counterarguments out there seem to rely on conflicting assumptions which are mistakenly portrayed as more than that.

It is entirely possible that his reasoning is incomplete, but the right thing to do would be to show how. Otherwise, it would be better if those conflicting assumptions were spelled out plainly. It seems that such assumptions are that freedom and cooperation are, at best, means to an end. Plainly disagreeing on such assumptions elevates the discussion beyond moral relativism IMHO and would be better for all concerned.

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 27, 2006 20:23 UTC (Wed) by tjc (guest, #137) [Link] (7 responses)

It seems that such assumptions are that freedom and cooperation are, at best, means to an end.

What exactly do you mean by this? If freedom and cooperation are but means to an end, then what is the end?

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 27, 2006 20:48 UTC (Wed) by b3timmons (guest, #40286) [Link] (6 responses)

What exactly do you mean by this? If freedom and cooperation are but means to an end, then what is the end?

You know--technological progress. Now we are getting into the territory that deeply divides people! No other end is ever claimed by open source advocates. Not being a social movement, they do not concern themselves with the social condition.

I suspect they feel such concern is misguided, perhaps because the social condition is less controllable than technology. Being less controllable does not justify ignoring it IMO.

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 27, 2006 21:53 UTC (Wed) by sepreece (guest, #19270) [Link] (5 responses)

Or, perhaps they [we] feel that, at least in this case, technology is not the solution to the social condition in question.

It's like the old saw, "If the only tool you have is a hammer, every problem looks like a nail."

The only tool available here is the license language, so it's tempting to try to use it to solve whatever issus you think are important. When the issue is "I want to make my code available, but I want to be able to see what changes other people make to it" that hammer works fine; when the issue is "DRM is evil", that hammer has nothing to pound on.

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 27, 2006 23:25 UTC (Wed) by b3timmons (guest, #40286) [Link] (4 responses)

The only tool available here is the license language, so it's tempting to try to use it to solve whatever issus you think are important. When the issue is "I want to make my code available, but I want to be able to see what changes other people make to it" that hammer works fine; when the issue is "DRM is evil", that hammer has nothing to pound on.

The very same concern for scope of a license has been repeatedly leveled at GPLv2 over the years; scope in and of itself does not seem to be that compelling. Again, we are brought back to the "user". One side argues for the manufacturer, the other the end owner. With Tivoization, the undeniable fact is that freedoms granted to the manufacturer are not granted to the end owner. You may be arguing that it is in line with the intentions that produced the GPLv2. If you are right here, then we must correct the FSF. However, I think your argument is that the reduction of freedoms is OK, regardless of the intentions.

Computers--embedded or not--are universal tools that embody exceptionally strong notions of freedoms. Moreover, propagation of digital information is ridiculously easy and cheap. Few other examples scenarios show the kind of artificial imbalance of freedom that we see between Tivo and a Tivo user. That few Tivo users might benefit does not subtract from the principle.

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 28, 2006 3:53 UTC (Thu) by AJWM (guest, #15888) [Link] (3 responses)

> With Tivoization, the undeniable fact is that freedoms granted to the manufacturer are not granted to the end owner.

It is perfectly deniable, because it is not a fact.

With regards to the software, the end user has _exactly_ the same rights as the manufacturer -- he is perfectly free to take and modify that software, develop some unique hardware, and to manufacture and sell that hardware along with the modified software.

The original manufacturer's right to create a box that will only run software that he approves has _nothing_ to do with software, it's inherent in his right to design the hardware any way he wants to, and sell that hardware if he so chooses and if anyone else is stupid enough to buy it. (So long as he meets the license terms of any included software).

The fact that Moglen and RMS don't seem to have a problem with distributing GPL software embedded in ROM indicates that they fundamentally agree with this position, but they're trying to pretend that "Tivoization" is something different, to the detriment of the v3 license.

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 28, 2006 4:48 UTC (Thu) by b3timmons (guest, #40286) [Link] (2 responses)

It is perfectly deniable, because it is not a fact.

It is too a fact, because you know very well that the freedoms under discussion are the four freedoms. Tivo has freedom one--the freedom to study how the program works, and adapt it to their needs--and you don't, since it is impossible to run any possible modified version of yours.

The original manufacturer's right to create a box that will only run software that he approves has _nothing_ to do with software, it's inherent in his right to design the hardware any way he wants to, and sell that hardware if he so chooses and if anyone else is stupid enough to buy it. (So long as he meets the license terms of any included software).

The issue is that is that this right should not damage the four freedoms. GPLv3 restores protection for them, and a manufacturer, unable to exercise his right to Tivoize with software under the GPLv3, should choose software under a license permitting Tivoization. The FSF has the same right as anyone else to set their own terms for their license. GPLv3 looks more like a bug-fix than anything else.

The fact that Moglen and RMS don't seem to have a problem with distributing GPL software embedded in ROM indicates that they fundamentally agree with this position, but they're trying to pretend that "Tivoization" is something different, to the detriment of the v3 license.

There is no pretending at all: with ROM freedom one is moot.

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 28, 2006 22:07 UTC (Thu) by vonbrand (guest, #4458) [Link] (1 responses)

It is too a fact, because you know very well that the freedoms under discussion are the four freedoms. Tivo has freedom one--the freedom to study how the program works, and adapt it to their needs--and you don't, since it is impossible to run any possible modified version of yours.

How is it impossible? Modified it won't run on your TiVo, but it runs here on my PC just fine.

And what about the same TiVo, just with the software in ROM? That is OK under GPLv3, but is the same situation from the user's point.

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 29, 2006 0:34 UTC (Fri) by b3timmons (guest, #40286) [Link]

How is it impossible? Modified it won't run on your TiVo, but it runs here on my PC just fine.

Modified it must be possible to run it on my Tivo because that is what this whole thing is about: my ability to implement the same functionality in the same range of circumstances as what Tivo does by exploiting GPLed code for the device that I bought from them and now own. Fortunately, the GPLv3 restores my ability lost from Tivo exploiting a loophole.

And what about the same TiVo, just with the software in ROM? That is OK under GPLv3, but is the same situation from the user's point.

No--not the same situation, because now Tivo and I are now on an equal footing--Freedom One applying to neither of us, unlike the Tivoization case above.

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 27, 2006 22:52 UTC (Wed) by smitty_one_each (subscriber, #28989) [Link] (1 responses)

>At least he is trying, complete with initial assumption and application of logic until conclusions.

I agree with you here. RMS is an archetype. We need that 'crazy uncle' figure to keep us from complacency.

I start to taper off from the gentleman when he asserts that his neat logical conclusions preclude the possibility of other conclusions. Only a Supreme Being would tote such potency. The FSF's goals are one set of legitimate goals among many. Labeling other goals 'unethical' does not, of itself, render them so.

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 27, 2006 23:33 UTC (Wed) by b3timmons (guest, #40286) [Link]

I seem to have the most trouble with some of his assumptions. If you could offer an alternative conclusion given his assumptions, that would be quite notable and noble IMO.

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 28, 2006 15:05 UTC (Thu) by lysse (guest, #3190) [Link] (3 responses)

The difference is that RMS sticks absolutely to the morals he advocates. Too many people confuse "moral relativism" with violating one's own moral code, and the presence of someone who makes a point of not doing so makes them feel like a hypocrite. Likewise, most people can't tell the difference between explaining and supporting one's own moral standards and passing judgement on those of others.

That, I suspect, is the source of most people's anger and defensiveness in reacting to RMS and his statements. It's the collective outrage of "good people" being made to feel a little less so by their own consciences.

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 30, 2006 12:51 UTC (Sat) by jstAusr (guest, #27224) [Link] (2 responses)

...But you like the idea of Linus being able to force his own set of morals and ethics on others? If I buy a TiVo, TiVo can change the software on that specific device without my permission or knowledge. As the owner of the device I am not allowed to change the software on that specific device. Under what definition of tit-for-tac does that fit?

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Oct 5, 2006 11:02 UTC (Thu) by lysse (guest, #3190) [Link] (1 responses)

I can see no part of what I said that you're actually replying to...

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Oct 10, 2006 2:12 UTC (Tue) by jstAusr (guest, #27224) [Link]

Interesting, I can't either.

Why Torvalds is sitting out the GPLv3 process (Linux.com)

Posted Sep 27, 2006 8:42 UTC (Wed) by forthy (guest, #1525) [Link]

Ulrich Drepper: NEVER voluntarily put a project you work on under the GNU umbrella since this means in Stallman's opinion that he has the right to make decisions for the project.

I don't know if Ulrich Drepper ever read the FSF copyright transfer agreement. It's not an exclusive copyright transfer, and puts a lot of trust on you - the original author - rather than on the FSF; the FSF promises to hold up the four freedoms forever. It's basically a way to allow the FSF to sue third parties which violate the license, rather than to have the original author to sue himself. He can, nonetheless, if he likes to.

I'm maintainer of a GNU project (gforth), and I've never ever been stabbed into the back by RMS. Agreed, this is not such an important project as glibc, and it's sufficiently agnostic to the host OS (though it is not completely agnostic WRT GNU/non-GNU environments: we slightly treat GNU ones better). RMS helped us once to clarify the license conditions, which are a bit odd on a Forth system compared e.g. to a C compiler (the further is interactive, and contains an incremental compiler).

I have seen Torvalds being accused of backstabbing, too

Posted Sep 27, 2006 12:09 UTC (Wed) by robilad (guest, #27163) [Link] (1 responses)

http://www.theregister.co.uk/2005/04/14/torvalds_attacks_...

So you should stop trusting Linus now, based on this single source, that cobbles together a bunch of e-mails and posts to some web sites, I guess, if you are to go by the same standard for your vilification of RMS.

Linus and RMS are the same type of person.

I have seen Torvalds being accused of backstabbing, too

Posted Sep 28, 2006 20:24 UTC (Thu) by h2 (guest, #27965) [Link]

Exactly. The person I trust less after reading through most of the threads on the GPL 3 debate is Linus, who clearly shows himself willing to do several things that do not help generate trust: be inconsistent in his tone, to a massive degree, jumping from near hysterical screaming on groklaw to near rational elsewhere, why he does this I don't really care, what I do see is that he does it.

The Tridgell example is a very good one to bring up, and shows clearly that he is a bit too willing to start on the downhill slope of compromise, without being willing to consider the long term consequences, or to listen to people who warned him about this potential.

And his refusal to address the potential issues of that particular decision until the situation blew up in his face does little to make me trust his ability to do any clearsighted long term thinking on the possible results of his actions and decisions. This alone should make people give some serious thought to just how much you can trust Linus to really watch out for even the restricted long term interests of Open Source software.

It is very fortunate for us that the GPL2 proved sufficiently strong to prevent Linus from making who knows how many other potential compromises he might have wanted to make over the past years, but which the license prevented him from making. And since from what I gather from the more clear headed comments here, the GPL 3 proposed text really only tightens that long term protection. For me it's very easy to see that the spirit, and implication, of GPL 2, definitely does not preclude the fact that all software must be able to run on some hardware by definition, so changes should not be able to deny real world running of the code, it just didn't make that point explicit enough.

The only thing I could accuse Stallman of is absolute consistency in his outlook, and more importantly, in his long term behavior. He believes what he believes, and he lives his life by what he believes. Personally, I respect that, since it's an amazingly rare quality in people. Others may choose to deride this is zealotry or whatever, but I'd like to take a very close look at those people's lives and decisions before I'd take what they have to say very seriously. Luckily we have that opportunity with Stallman, Moglen, etc, so we can all decide for ourselves whether we agree or disagree with their positions.

If people think living your life within a consistent set of ethical principles is a negative that says I think more about them than Stallman. Now you can disagree with Stallman and what he believes, that's quite easy to do, but accusing him of departing from core values over time is not something you can do unless you ignore reality. So I'd say, if you trusted Stallman with GPL 2, very little has changed in him or his outlook when it comes to GPL 3, except that he now can see areas that he could not anticipate in GPL 2.

By the way, I'd like to thank Mingo for at least admitting that he is merely 'trying' to communicate the reasons for the disagreement. In my case, I have to admit that the effort has failed, I have not seen any particularly convincing arguments put forth by the kernel developers. But I do applaud the effort to at least try to be rational, which is more than I have seen from Linus, whose opposition seems to stem from personal dislike of the FSF and Stallman more than any other single factor, no matter how he tries to whitewash this.

I still trust Linus and the core developers to do the very best job they can do on the kernel code, and I trust that it will continue to improve over time, that trust hasn't changed at all, and I hope they continue with their great work. What has changed is that I no longer trust them to foresee events that might end up severely compromising even the freedoms given by their open source vs/ free software beliefs.