- Site menu:
- Latest Discussions
- Latest Newswire
- Archives
- Submit Story
- Story HOWTO
- Syndicate
- E-mail Alerts
'Cool it, Linus' - Bruce Perens
|
|
|
Comments (30)
Linus pissed off
9 earlier comments in the full thread.
It doesn't matter if he is breaching a contract or breaking a law. As a kernel developer, he has an obligation not to screw up whatever arrangement Linus has made. He was and is perfectly free to decide that his freedom to reverse engineer protocols is more important to him than being a kernel developer.
That's legit. Deciding that none of it matters and doing as he damned well pleases sounds more like a petulant child than a responsible developer.
That's legit. Deciding that none of it matters and doing as he damned well pleases sounds more like a petulant child than a responsible developer.
The most disturbing things here for me is not the bitkeeper license thing, or what Tridge did.
What disturbs me is you have Linus saying that Tridge did something wrong, in public -- Tridge not able to open his mouth, and Perens (who should know better, IMHO) calling linus "a real idiot" -- The Register is having a field day with sound bites like this. Then he goes on to tell the head of BK to "shut up".
The damage is not technical at all. BTW, Linus, in his original post called Tridge a idiot or something similar. But, reading Linus's post, I can see something fundamental going on. He's pissed because he has to write source code management software that he didn't want to write.
Not everyone can do that, by the way. I hope he decides to GPL it ;) my guess is he will.
Second, I agree with his fundamental point. here's the link to his post -- read it carefully. I think it would have gone over a lot better had he kept his diplomacy hat in check:
http://www.realworldtech.com/forums/index.cfm?action=detail&...
I'm not going to call any of the parties in question an idiot by the way. Just wish that the amplitude of the words in response to this was in the dampening direction, instead of the wave-generating kind.
--FeriCyde
What disturbs me is you have Linus saying that Tridge did something wrong, in public -- Tridge not able to open his mouth, and Perens (who should know better, IMHO) calling linus "a real idiot" -- The Register is having a field day with sound bites like this. Then he goes on to tell the head of BK to "shut up".
The damage is not technical at all. BTW, Linus, in his original post called Tridge a idiot or something similar. But, reading Linus's post, I can see something fundamental going on. He's pissed because he has to write source code management software that he didn't want to write.
Not everyone can do that, by the way. I hope he decides to GPL it ;) my guess is he will.
Second, I agree with his fundamental point. here's the link to his post -- read it carefully. I think it would have gone over a lot better had he kept his diplomacy hat in check:
http://www.realworldtech.com/forums/index.cfm?action=detail&...
I'm not going to call any of the parties in question an idiot by the way. Just wish that the amplitude of the words in response to this was in the dampening direction, instead of the wave-generating kind.
--FeriCyde
Is trigdell even a kerenel developer?
Does he write patchs for the kernel on top of Samaba? I truely don't know but I don't think so, or maybe nothing more than a few speed enhancments for the networking stacks.
Was trigdell a member bound by the bitkeeper license. no one is saying yes or no to that. If he wasn't then he can do what ever he wants. going on the assumption that tridgell was never a licensee of bitkeeper, it's absurd to think Bitmover would hav ethe authority to try and force trigdell's employer to fire him.
Now the other direction, if trigdell was a bitkeeper licensee then he was breaking a contract.
NO ONE is giving out specifics.
Does he write patchs for the kernel on top of Samaba? I truely don't know but I don't think so, or maybe nothing more than a few speed enhancments for the networking stacks.
Was trigdell a member bound by the bitkeeper license. no one is saying yes or no to that. If he wasn't then he can do what ever he wants. going on the assumption that tridgell was never a licensee of bitkeeper, it's absurd to think Bitmover would hav ethe authority to try and force trigdell's employer to fire him.
Now the other direction, if trigdell was a bitkeeper licensee then he was breaking a contract.
NO ONE is giving out specifics.
How could Tridgell reverse engineer anything about bitkeeper without access to the application? If he got access to it, one might presume he did so through the kernel development process.
Obviously there's a lot of speculation involved in this thread, but if the bitkeeper license contained a provision prohibiting reverse engineering, at least the owner thought that what Tridgell was doing violated that agreement. It would appear that at least Tridgell's lawyer thinks he might have violated the terms of the bitkeeper license.
It is not particularly well settled in the law whether or not reverse engineering is automatically allowed under the fair use doctrine, nor is it settled whether the copyright fair use doctrein trumps license provisions specifically prohibiting reverse engineering. Either way, if I let you use my product without cost under the condition that you do not reverse engineer it, you accept my offer and then go ahead and reverse engineer it, how is that not unethical and/or wrong? Just because you think I should have given you "free" access in the Stallman sense? Or just because you believe you ought to be able to because of fair use?
Obviously there's a lot of speculation involved in this thread, but if the bitkeeper license contained a provision prohibiting reverse engineering, at least the owner thought that what Tridgell was doing violated that agreement. It would appear that at least Tridgell's lawyer thinks he might have violated the terms of the bitkeeper license.
It is not particularly well settled in the law whether or not reverse engineering is automatically allowed under the fair use doctrine, nor is it settled whether the copyright fair use doctrein trumps license provisions specifically prohibiting reverse engineering. Either way, if I let you use my product without cost under the condition that you do not reverse engineer it, you accept my offer and then go ahead and reverse engineer it, how is that not unethical and/or wrong? Just because you think I should have given you "free" access in the Stallman sense? Or just because you believe you ought to be able to because of fair use?
FeriCyde - my compliments, that was a very well reasoned post. [No sarcasm intended nor present!]
A couple of points:
Paulie, you are right (as usual, though I hate to admit it).
The wordstorm is inappropriate, and Mr. Perens's especially so, since it tended to exascerbate and amplify Linus's outburst. Blessed are the peacemakers and all that - Perens, for whatever reason, decided to vent rather than to heal.
Linus made his points very well, BTW, but I will add one more note. OpenOffice and Samba are, as he points out, useful products in their own right, but they are more than that. They are, in effect, open-source implementations of de-facto computing standards.
The interoperability they provide is vital to the growth of free software.
The same cannot be said bout BitKeeper. If there is a de-facto standard in commercial source-code control, ClearCase is a much better candidate for reverse-engineering than BitKeeper.
Paulie, you are right (as usual, though I hate to admit it).
The wordstorm is inappropriate, and Mr. Perens's especially so, since it tended to exascerbate and amplify Linus's outburst. Blessed are the peacemakers and all that - Perens, for whatever reason, decided to vent rather than to heal.
Linus made his points very well, BTW, but I will add one more note. OpenOffice and Samba are, as he points out, useful products in their own right, but they are more than that. They are, in effect, open-source implementations of de-facto computing standards.
The interoperability they provide is vital to the growth of free software.
The same cannot be said bout BitKeeper. If there is a de-facto standard in commercial source-code control, ClearCase is a much better candidate for reverse-engineering than BitKeeper.
dinotrac: I really wonder if you ever used that amazing product, if you can say: "... ClearCase is a much better candidate for reverse-engineering ..." If that is the better product, free software should look elsewhere.
[Caveat: I make no pretense I am or was an expert with ClearCase - though I once instructed an individual that made great claims about both the powers of and his extensive experience with ClearCase, yet copied my instructions verbatim.]
[Caveat: I make no pretense I am or was an expert with ClearCase - though I once instructed an individual that made great claims about both the powers of and his extensive experience with ClearCase, yet copied my instructions verbatim.]
TxtEdMacs --
Sorry to have been unclear.
I will tell you that most conversations I have involving ClearCase tend to replace Case with a different 4 letter word.
I meant only that ClearCase is much closer to being a de-facto standard than BitKeeper is. I never meant to imply that ClearCase is better than, as good as, or, quite frankly, even in the same league as BitKeeper.
Sorry to have been unclear.
I will tell you that most conversations I have involving ClearCase tend to replace Case with a different 4 letter word.
I meant only that ClearCase is much closer to being a de-facto standard than BitKeeper is. I never meant to imply that ClearCase is better than, as good as, or, quite frankly, even in the same league as BitKeeper.
McVoy is still the bad guy. He is punishing everyone but Andrew Tridgell for whatever "transgression" he committed. That's lame and irrational. McVoy is who Linus should be mad at, not Tridgell. And it's far from clear what crime Tridgell committed. By some accounts he was doing what he always does- decoding the protocols as they came over the wire, and figuring out how to interoperate with a closed, restrictive, proprietary product.
Linus can preach all day about how BitKeeper has advanced the state of the art (see Fericyde's link), but it hasn't- it has only advanced itself. McVoy has exerted tremendous effort to keep BitKeeper's advances locked away, even to the point of complaining about anyone who might try to copy the concepts and ideas behind it. That is not advancing anything but his own interests. This beautifully demonstrates why GPL software is needed, and how quickly a closed, proprietary product can turn around and bite back.
And it also demonstrates what a self-serving hypocrite McVoy is. He received tremendous support and priceless PR from the kernel team, and he built on the work of others. And now he wants to keep it all to himself, and not even acknowledge that anyone else contributed to BitKeeper's success, like he invented it in a vacuum and every idea and line of code is completely original. That's the kind of baloney that the GPL aims to protect us from, and that's why Linus was wrong to make a deal with McVoy in the first place. Especially when he demonstrated how undependable he was by changing the terms of the BitKeeper license practically every day- what kind of crap is that? You don't do business with people you can't trust. Linus blew it. It's not the end of the world, but he should own up to making a bad deal, instead of blaming the wrong people for the deal blowing up in his face.
Linus can preach all day about how BitKeeper has advanced the state of the art (see Fericyde's link), but it hasn't- it has only advanced itself. McVoy has exerted tremendous effort to keep BitKeeper's advances locked away, even to the point of complaining about anyone who might try to copy the concepts and ideas behind it. That is not advancing anything but his own interests. This beautifully demonstrates why GPL software is needed, and how quickly a closed, proprietary product can turn around and bite back.
And it also demonstrates what a self-serving hypocrite McVoy is. He received tremendous support and priceless PR from the kernel team, and he built on the work of others. And now he wants to keep it all to himself, and not even acknowledge that anyone else contributed to BitKeeper's success, like he invented it in a vacuum and every idea and line of code is completely original. That's the kind of baloney that the GPL aims to protect us from, and that's why Linus was wrong to make a deal with McVoy in the first place. Especially when he demonstrated how undependable he was by changing the terms of the BitKeeper license practically every day- what kind of crap is that? You don't do business with people you can't trust. Linus blew it. It's not the end of the world, but he should own up to making a bad deal, instead of blaming the wrong people for the deal blowing up in his face.
I vote to rename this thread "tuxchick pissed off" ;-)
Wow. Lot's of animosity here. Well, FWIW, here's my take on the matter. The whole bitkeeper thing was one of those rare examples of Linus making a real mistake, and then stubbornly refusing to see it. That's unusual. I usually find myself in agreement with him. (Not that I always feel qualified to have an opinion. The guy's is in a totally different league than me.)
Linux seems to feel that Tridge ignored the 'inevitable' problems that his actions were sure to cause. I submit that the current situation became inevitable the day that Linux decided to use a proprietary SCM system for the kernel. Note that I am intentionally *NOT* arguing either way on whether using a proprietary tool for the kernel is morally right or wrong. I have my opinion on the matter, but it is totally irrelevant to the point that I am trying to make.
Linus' famed sense of "pragmatism" failed him completely in this instance. You see, introducing a proprietary tool into Open Source's flagship, "poster boy" project was the single *LEAST* pragmatic thing that Linus could possibly have done. And from the day that he did so, the only real question was how long it would take before the whole thing blew up in his face. People told him this over and over. (Though perhaps not in those exact words.) Linus refused to listen.
At any rate, that's all milk spilled over the dam, now.
The good news is that:
1. It's starting to look like the recovery time from this mistake might not be that great. GIT is coming along nicely, and since it's not an SCM, but a filesystem suitable *for use by* and SCM, I predict a marriage between it and one or more of the existing SCM projects. (This seems a pretty safe bet since it is *already* happening. ;-)
2. If this situation had to happen (and it did) this was a good time for it to happen. Kernel development is *ahead* of where anyone would ever have predicted that it would be at this point. (And yes, in part that is due to Bitkeeper.)
3. The main benefits derived from bitkeeper by the kernel development process had less to do with Bitkeeper, per se, and more to do with what Linus and the other developers learned from using Bitkeeper. These are lessons that can be taken with them. They will continue to apply them regardless of what direction they go with kernel SCM.
4. This last point is either good, bad, or irrelevant, depending upon your point of view. But I'll let it slip that I don't much care for Larry's attitude. (Hi, Tuxchick! :-) Mark my words. (And keep in mind the marriage between GIT and Monotone, Arch, DARCS, whatever.) At some point in the future, with the benefit of hindsight, we will all look back on these recent events and say: "That, right there, was where McVoy made his mistake. In trying to avoid competition from the Open Source community, he inadvertently conjured up a monster out of his worst nightmares, and sowed the seeds of his own company's destruction. Isn't poetic justice cool?
(OK. Perhaps a more likely outcome is that one day Bitover will have to embrace OSS in order to compete. But, hey, saying that would have been sooooooooo much less dramatic. ;-)
Linux seems to feel that Tridge ignored the 'inevitable' problems that his actions were sure to cause. I submit that the current situation became inevitable the day that Linux decided to use a proprietary SCM system for the kernel. Note that I am intentionally *NOT* arguing either way on whether using a proprietary tool for the kernel is morally right or wrong. I have my opinion on the matter, but it is totally irrelevant to the point that I am trying to make.
Linus' famed sense of "pragmatism" failed him completely in this instance. You see, introducing a proprietary tool into Open Source's flagship, "poster boy" project was the single *LEAST* pragmatic thing that Linus could possibly have done. And from the day that he did so, the only real question was how long it would take before the whole thing blew up in his face. People told him this over and over. (Though perhaps not in those exact words.) Linus refused to listen.
At any rate, that's all milk spilled over the dam, now.
The good news is that:
1. It's starting to look like the recovery time from this mistake might not be that great. GIT is coming along nicely, and since it's not an SCM, but a filesystem suitable *for use by* and SCM, I predict a marriage between it and one or more of the existing SCM projects. (This seems a pretty safe bet since it is *already* happening. ;-)
2. If this situation had to happen (and it did) this was a good time for it to happen. Kernel development is *ahead* of where anyone would ever have predicted that it would be at this point. (And yes, in part that is due to Bitkeeper.)
3. The main benefits derived from bitkeeper by the kernel development process had less to do with Bitkeeper, per se, and more to do with what Linus and the other developers learned from using Bitkeeper. These are lessons that can be taken with them. They will continue to apply them regardless of what direction they go with kernel SCM.
4. This last point is either good, bad, or irrelevant, depending upon your point of view. But I'll let it slip that I don't much care for Larry's attitude. (Hi, Tuxchick! :-) Mark my words. (And keep in mind the marriage between GIT and Monotone, Arch, DARCS, whatever.) At some point in the future, with the benefit of hindsight, we will all look back on these recent events and say: "That, right there, was where McVoy made his mistake. In trying to avoid competition from the Open Source community, he inadvertently conjured up a monster out of his worst nightmares, and sowed the seeds of his own company's destruction. Isn't poetic justice cool?
(OK. Perhaps a more likely outcome is that one day Bitover will have to embrace OSS in order to compete. But, hey, saying that would have been sooooooooo much less dramatic. ;-)
I don't think Linus failed at all...he chose something that was free and that worked great. Then the license changed.
I did not say that Linus failed. I said that his famed sense of pragmatism failed him completely in this instance. Linus "making a mistake" is quite a different thing than Linus "failing", which is something that I never expect to say. It does demonstrate that he is human.
You are correct. The license changed. Many times. But that was not the core of the problem. BK was proprietary at the beginning, and was still proprietary at the end. Now without passing judgement on that issue myself, there is no doubt in my mind that the situation could never have been contained. The BK issue has been a ticking time bomb since the day Linus made the decision to use it.
I'm not complaining, though. I'm quite happy with where we are today. They kernel is in great shape. Things are getting worked out. OSS SCM systems are about to get a big boost. And I do like the fact that kernel development no longer depends upon proprietary tools.
You are correct. The license changed. Many times. But that was not the core of the problem. BK was proprietary at the beginning, and was still proprietary at the end. Now without passing judgement on that issue myself, there is no doubt in my mind that the situation could never have been contained. The BK issue has been a ticking time bomb since the day Linus made the decision to use it.
I'm not complaining, though. I'm quite happy with where we are today. They kernel is in great shape. Things are getting worked out. OSS SCM systems are about to get a big boost. And I do like the fact that kernel development no longer depends upon proprietary tools.
Found a better article on this. Many new sites it seems were blasting things out of proportion...go figure. This one actually reports the NEWS and doesn't try to incite a flame war. Wow...what a concept.:
http://www.computerworld.com/softwaretopics/os/linux/story/0...
http://www.computerworld.com/softwaretopics/os/linux/story/0...
devnet: The last line of the article was a gem. And I thought 'Git' was slang for 'Get Out of Here!', something spoken to horses in grade B cowboy movies ...
GIT is among other UK slang for a worthless person.
We own Linus a great big thank you!
THANK YOU LINUS!!!
THANK YOU LINUS!!!
robT: this is how the reverse engineering proceeded:
telnet hostname.of.bk.server 5000 help ...think... ^]q echo clone|nc bk.svr 5000 >file.sccs
Complicated, huh? And immoral? Who was the help intended for?
telnet hostname.of.bk.server 5000 help ...think... ^]q echo clone|nc bk.svr 5000 >file.sccs
Complicated, huh? And immoral? Who was the help intended for?
AnonymousCoward: Yeah, thanks for the post. You must have realized that I unable to read the other 1/2-dozen links on this. It is so nice of you to be so helpful.
Complicated? No. Is this all that was going on? I'm not convinced, but I'm a cynic.
Immoral? I can't answer that without a lot of speculation, but if this is all that was going on a lot of people need to consider self medication.
Who was the help for? You probably have your own ideas for that.
Complicated? No. Is this all that was going on? I'm not convinced, but I'm a cynic.
Immoral? I can't answer that without a lot of speculation, but if this is all that was going on a lot of people need to consider self medication.
Who was the help for? You probably have your own ideas for that.
Cool it, Bruce
Calling somebody an idiot just because you don't agree with him is not the most constructive way to deal with controversy. Perens ought to apologize to Linus for this ad hominem.
Post a comment
A free account is needed to post. Write your comment here first: your text is kept in this browser while you sign up or log in, and is waiting for you when you come back.
Your message:Markdown: **bold**, *italic*, [text](https://link), - lists, > quotes, $x^2$ for math. Formatting help
Sign up and post (it is free) or Log in and post
Secondly, the person creating the trouble is McVoy. He's self-serving, as most humans are, and Linus was foolish to make a deal with him. McVoy is the one to be mad at, not Tridgell. Yanking everyone's license because of Tridgell's actions is absurd. McVoy already proved he could not be trusted by continually modifying the BitKeeper license and placing more and more restrictions on its use.
If Tridgell is breaching a contract, then shame on him. But so far there is no indication of any such contract. And even if that is the case, McVoy still looks like a dork for "punishing" everyone but Tridgell.