The real currency of the internet is respect & reputation. Alan has those in spades, which is why he's trusted. Should he ever violate that trust, then he won't be as influential as he is now.
If he becomes sufficiently untrustworthy, then someone else that most developers trust could start managing a fork of the Linux tree. But, Red Hat's continued success depends on the general public's favorable impression of Linux. Even if a particular fork turns out to be a "good thing", the public opinion would start to turn, and Red Hat would suffer. Methinks Red Hat won't use any hidden agenda to shoot themselves in the foot deliberately.
You just don't have to claim the copyright for yourself -- for example, you can assign the copyright to FSF (as required for those contributing to egcs) or perhaps to the orignal developer.
Rereading your reply, I think some clarification is in order. The GPL & the MEULA [1] are licenses, not copyrights. If you own the copyright, then you have permission to grant people the right to copy your material. This is done with a license -- and you can use different licenses for different people, even. If you choose to use the GPL, then you are requiring developers who combine their copyrighted material with your copyrighted material to use the GPL if they distribute the combined result.
The GPL expressly does not forbid what Sun did to Blackdown. There is nothing illegal in taking a GPL'd program, making zero or more changes to it, and calling it your own, as long as you preserve the original copyright (of course, you may copyright any changes you make). The very fact that GPL does not require you to make mention of the original developers is what makes the "old" BSD license incompatible with GPL.
As to whether people will offer you good or ill will is another story, as ESR has discussed many times.
I know it's not that easy. But I'd remembered a comment from about 15-18 months ago by the CTO of some company (can't remember which) which had previously only released products for commerical Unices (IIRC: Solaris, HP-UX, SCO, AIX, and maybe IRIX). This was when "there are no apps for Linux" was pretty much true in the commercial sense, and this was one of the first commercial apps ported to Linux. One of the industry mags asked him how hard it was to port their app to Linux, and the CTO replied "I typed 'make'." Christopher A. Bohn
After I wrote the above comment (and coined "DebianBSD"), it occured to me that this effort may very well go forward... I say this without malice toward anybody. Besides the various "why"s such as "because it's there" and "to improve the portability of our software" and the like, there is a political reason to do this. If Debian makes a BSD distribution, it would almost certainly be called "Debian GNU/BSD". This would reinforce the idea that, according to the FSF, et.al., the correct name of the Linux distributions is SoAndSo GNU/Linux. By creating a GNU/BSD, the media attention would make people pay attention to the "GNU/" prefix and would also result in comparisons to GNU/Linux. Christopher A. Bohn
If the *BSD developers do not switch to GPL (and I doubt they will), then they can (obviously) keep the advertising clause. However, the advertising clause is incompatible with the GPL, which is why BSD-licensed software historically could not be relicensed under the GPL. Christopher A. Bohn
The paragraph in DWN mentions the concern that it would open the doors for unscrupulous 3d parties to take their work and sell closed-source derivatives, but I don't see that as a real problem... the FreeBSD kernel would have the BSD kernel, but much (most?) of the software that the Debian team puts in the DebianBSD distribution would still be GPL, which means FastBuck Inc. would not be able to take DebianBSD as-is and apply a closed-source license.
And now a question for the License Lawyers.
The BSD license's incompatability with GPL is centered on the BSD license's extra restrictions forbidden by GPL. Now that the advertising clause is gone (see next bullet), the question remains: does the no-endorsement clause constitute a forbidden restriction, or would the Debian team (or someone else) be permitted to relicense the BSD kernel with GPL?
Just because the Cal State Berkeley Regents withdrew the advertising clause would not, if I understand correctly, require those already holding a copy of the kernel to similarly withdraw the advertising clause (they obtained it under a previous license, and they would not be required to apply the new license; further, even if they were required to apply the new license, the new license would still permit relicensing back to the old license). So my second question is, did the FreeBSD, OpenBSD, and/or NetBSD developers withdraw the advertising clause, or are they holding onto the clause?
You realize, of course, that any license which forbids use by residents of the U.S. would not satisfy the the Open Source Definition. Of course, neither would one that would forbid use by residents of Cuba, Syria, Iraq, etc. Christopher A. Bohn
If the injuries are severe enough, the Americans with Disabilities Act takes effect. For example:
Under ADA, my organization was required to arrange for a 35" TV to be attached to a PC running in 640x480 mode so that he could use the PC (granted, this wasn't due to work-related injuries).
Someone I work with has carpal-tunnel syndrome so bad (this is a work-related injury) that our organization was required to arrange for voice-recognition SW & headset for her -- the good side is she can now prepare documents far, far faster than anyone else can.
This move is clearly intended to prevent work-related injuries from getting so bad that ADA kicks in. Christopher A. Bohn
...when ESR writes the fourth essay? IIRC, it's titled "Weaving the Web of Indra?" (or something like that), but I cannot find on his website a reference to his intent to write it, anymore. Christopher A. Bohn
Of course, if time runs backwards, we'd experience everything backwards, and we wouldn't realize time was running backwards while the universe contracted. So, to our perception, time would be running forward and the universe would be expanding. Like the Thermos-bottle question, how would you know? Christopher A. Bohn
Now the tweaked, buggy egcs has been foisted off as the ONLY upgrade to gcc. Which means sane, reliability-minded people who loved the old gcc, now have nowhere to turn to.
Unless, of course, you take the last "pure" gcc version and fork it to ACgcc. Christopher A. Bohn
At a recent presentation by SGI (Linux University Road Tour), I think it was put best (paraphrasing): "Like all evolving open-source projects, Linux will always appear to be on the verge of forking, due to the constant experimentation going on. This is a healthy thing for Linux." Christopher A. Bohn
No offense intended here; this comment is meant to be about 25 deg C, and is just a clarification for what seems to be a common misconception.
No application can make use of a parallel computer without a programmer first identifying the inherent concurrency. There are pre-processors that attempt to do this for you, such as BERT 77, but IMO that cannot replace the intuition and basic understanding of the problem that a human programmer provides (with all due respect to Doug & everyone else at Paralogic). Once that concurrency is identified, then implementing it using the MPI or PVM API, or using HPF directives, or using pthreads (depending on your platform) should be fairly straight-forward.
Computers are not magic! You cannot feed a program written for a sequential computer into a parallel computer and expect speedup, period. At run-time, the computer just doesn't have sufficient view of the application to even attempt to identify the concurrency for you, and it certainly doesn't have the cycles to spare to decide what aspects of the concurrency should or shouldn't be parallelized. The best you can do if you won't or can't parallelize your code is to submit multiple instances of the application, with different inputs, to the batch scheduler and obtain a greater throughput.
I cannot say with certainty that this article is about the same machine I'm thinking of, but... SGI recently finished a Beowulf for the Ohio Supercomputer Center, and has put it on display at SC99 before actually delivering it to OSC. 128 Xeons for computation (32 SGI 1400L's, each with 4 Xeons, with one more 1400L as an admin node; Myrinet interconnect). Since Itanium hasn't actually been released yet, I expect anything anyone debuts in the near future built around Itanium is only debuting 4-color glossies. But, that is the first step before debuting an actual product built around a brand-new processor. Christopher A. Bohn
So if you play with a x.(2y+1).z kernel while rubbing your feet on the carpet and a lightning rod attached to an ISA slot, then this is for you. If you only use a x.(2y).z kernel with z>2, then this'll probably do nothing more than occupy disk space. Christopher A. Bohn
While I was working on my master's, there were generally interesting naming schemes. The Suns in "The Zoo" were all named after animals (and configuration management was accomplished by types of animals, e.g., "All the fish will be down Tuesday to upgrade to Solaris 2.6"). The SGIs in "The Louvre" were all named after artists. There was a network with all the machines named after MASH characters, and one named after Greek mythological characters. The network used for genetic algorithms research was named after genetics terms (Codon, Allele, etc; Creation was the exception to the naming scheme, but it still fit). The Beowulf cluster I built used a less imaginative naming scheme to keep configuration easy (ABC01, ABC02, ABC03,...). The "behind the scenes" servers were all named after airplanes (sabre.xxx, mustang.xxx, etc), until someone decided to use less interesting but more suitable names (library.xxx, mail.xxx, etc). My machines at home are named after SF authors (Asimov, Heinlein, etc) I have to machines on my desk at work now. The Unix systems are all named after luxery automobiles... I got a dirty look when I replaced the Sun "Allante" with a Linux box and named it "Gremlin". The MS boxen are named according to the primary user, followed by the Windows version. For example, I started with "Bohnca95". Now that I've replaced it with an NT box, I find it amusing that my MS machine is named "Bohncant". Christopher A. Bohn
The GPL forbids combining GPL'd code with other-licenced code that would place extra restrictions beyond those in the GPL. For example, the advertising clause in the traditional BSD License keeps code with the traditional BSD License from being used with GPL'd code. RMS wants the GPL to be used instead of the LGPL precisely to force derived works to be GPL'd. As he explains it, the Lesser GPL should only be used when there's already a popular non-free library with the same functionality. In such a situation, a developer could use the non-free library to develop non-free apps, so the compromise is to use LGPL to try to get the free libraries used, even if the derived works don't end up free. Christopher A. Bohn
The problem is that the program is linked to both GPL'd code and QPL'd code. The licenses, while both open-source, are mutually exclusive (just like GPL and the traditional BSD license). Christopher A. Bohn
Slashdot Mirror
The real currency of the internet is respect & reputation. Alan has those in spades, which is why he's trusted. Should he ever violate that trust, then he won't be as influential as he is now.
If he becomes sufficiently untrustworthy, then someone else that most developers trust could start managing a fork of the Linux tree. But, Red Hat's continued success depends on the general public's favorable impression of Linux. Even if a particular fork turns out to be a "good thing", the public opinion would start to turn, and Red Hat would suffer. Methinks Red Hat won't use any hidden agenda to shoot themselves in the foot deliberately.
Christopher A. Bohn
You just don't have to claim the copyright for yourself -- for example, you can assign the copyright to FSF (as required for those contributing to egcs) or perhaps to the orignal developer.
Rereading your reply, I think some clarification is in order. The GPL & the MEULA [1] are licenses, not copyrights. If you own the copyright, then you have permission to grant people the right to copy your material. This is done with a license -- and you can use different licenses for different people, even. If you choose to use the GPL, then you are requiring developers who combine their copyrighted material with your copyrighted material to use the GPL if they distribute the combined result.
[1] is this pronounced "moola"?
Christopher A. Bohn
The GPL expressly does not forbid what Sun did to Blackdown. There is nothing illegal in taking a GPL'd program, making zero or more changes to it, and calling it your own, as long as you preserve the original copyright (of course, you may copyright any changes you make). The very fact that GPL does not require you to make mention of the original developers is what makes the "old" BSD license incompatible with GPL.
As to whether people will offer you good or ill will is another story, as ESR has discussed many times.
Christopher A. Bohn
Of course you'd mention this after I switched to the Happy Hacker Keyboard (w/o NumLock key). =)
Christopher A. Bohn
Rioters log-on to the internet and loot on-line.
(with due credit to David Letterman)
Christopher A. Bohn
I know it's not that easy. But I'd remembered a comment from about 15-18 months ago by the CTO of some company (can't remember which) which had previously only released products for commerical Unices (IIRC: Solaris, HP-UX, SCO, AIX, and maybe IRIX). This was when "there are no apps for Linux" was pretty much true in the commercial sense, and this was one of the first commercial apps ported to Linux. One of the industry mags asked him how hard it was to port their app to Linux, and the CTO replied "I typed 'make'."
Christopher A. Bohn
2a. % ./configure
2b. % make
2c. % make install
2d. debug
2e. rinse, repeat.
Christopher A. Bohn
After I wrote the above comment (and coined "DebianBSD"), it occured to me that this effort may very well go forward...
I say this without malice toward anybody.
Besides the various "why"s such as "because it's there" and "to improve the portability of our software" and the like, there is a political reason to do this.
If Debian makes a BSD distribution, it would almost certainly be called "Debian GNU/BSD". This would reinforce the idea that, according to the FSF, et.al., the correct name of the Linux distributions is SoAndSo GNU/Linux. By creating a GNU/BSD, the media attention would make people pay attention to the "GNU/" prefix and would also result in comparisons to GNU/Linux.
Christopher A. Bohn
If the *BSD developers do not switch to GPL (and I doubt they will), then they can (obviously) keep the advertising clause.
However, the advertising clause is incompatible with the GPL, which is why BSD-licensed software historically could not be relicensed under the GPL.
Christopher A. Bohn
The paragraph in DWN mentions the concern that it would open the doors for unscrupulous 3d parties to take their work and sell closed-source derivatives, but I don't see that as a real problem ... the FreeBSD kernel would have the BSD kernel, but much (most?) of the software that the Debian team puts in the DebianBSD distribution would still be GPL, which means FastBuck Inc. would not be able to take DebianBSD as-is and apply a closed-source license.
And now a question for the License Lawyers.
Christopher A. Bohn
You realize, of course, that any license which forbids use by residents of the U.S. would not satisfy the the Open Source Definition. Of course, neither would one that would forbid use by residents of Cuba, Syria, Iraq, etc.
Christopher A. Bohn
- Under ADA, my organization was required to arrange for a 35" TV to be attached to a PC running in 640x480 mode so that he could use the PC (granted, this wasn't due to work-related injuries).
- Someone I work with has carpal-tunnel syndrome so bad (this is a work-related injury) that our organization was required to arrange for voice-recognition SW & headset for her -- the good side is she can now prepare documents far, far faster than anyone else can.
This move is clearly intended to prevent work-related injuries from getting so bad that ADA kicks in.Christopher A. Bohn
...when ESR writes the fourth essay? IIRC, it's titled "Weaving the Web of Indra?" (or something like that), but I cannot find on his website a reference to his intent to write it, anymore.
Christopher A. Bohn
Of course, if time runs backwards, we'd experience everything backwards, and we wouldn't realize time was running backwards while the universe contracted.
So, to our perception, time would be running forward and the universe would be expanding.
Like the Thermos-bottle question, how would you know?
Christopher A. Bohn
you could try reversing the polarity
If that should fail, try tesselating a recursion matrix.
Christopher A. Bohn
Christopher A. Bohn
At a recent presentation by SGI (Linux University Road Tour), I think it was put best (paraphrasing):
"Like all evolving open-source projects, Linux will always appear to be on the verge of forking, due to the constant experimentation going on. This is a healthy thing for Linux."
Christopher A. Bohn
Just took a look at the original press release
Christopher A. Bohn
No offense intended here; this comment is meant to be about 25 deg C, and is just a clarification for what seems to be a common misconception.
No application can make use of a parallel computer without a programmer first identifying the inherent concurrency. There are pre-processors that attempt to do this for you, such as BERT 77, but IMO that cannot replace the intuition and basic understanding of the problem that a human programmer provides (with all due respect to Doug & everyone else at Paralogic). Once that concurrency is identified, then implementing it using the MPI or PVM API, or using HPF directives, or using pthreads (depending on your platform) should be fairly straight-forward.
Computers are not magic! You cannot feed a program written for a sequential computer into a parallel computer and expect speedup, period. At run-time, the computer just doesn't have sufficient view of the application to even attempt to identify the concurrency for you, and it certainly doesn't have the cycles to spare to decide what aspects of the concurrency should or shouldn't be parallelized. The best you can do if you won't or can't parallelize your code is to submit multiple instances of the application, with different inputs, to the batch scheduler and obtain a greater throughput.
Christopher A. Bohn
I cannot say with certainty that this article is about the same machine I'm thinking of, but...
SGI recently finished a Beowulf for the Ohio Supercomputer Center, and has put it on display at SC99 before actually delivering it to OSC. 128 Xeons for computation (32 SGI 1400L's, each with 4 Xeons, with one more 1400L as an admin node; Myrinet interconnect).
Since Itanium hasn't actually been released yet, I expect anything anyone debuts in the near future built around Itanium is only debuting 4-color glossies. But, that is the first step before debuting an actual product built around a brand-new processor.
Christopher A. Bohn
Wasn't this in Men in Black, except about the size of a quarter, with Kay remarking "This is going to replace the CD"?
Christopher A. Bohn
So if you play with a x.(2y+1).z kernel while rubbing your feet on the carpet and a lightning rod attached to an ISA slot, then this is for you. If you only use a x.(2y).z kernel with z>2, then this'll probably do nothing more than occupy disk space.
Christopher A. Bohn
While I was working on my master's, there were generally interesting naming schemes. The Suns in "The Zoo" were all named after animals (and configuration management was accomplished by types of animals, e.g., "All the fish will be down Tuesday to upgrade to Solaris 2.6"). The SGIs in "The Louvre" were all named after artists. There was a network with all the machines named after MASH characters, and one named after Greek mythological characters. The network used for genetic algorithms research was named after genetics terms (Codon, Allele, etc; Creation was the exception to the naming scheme, but it still fit). The Beowulf cluster I built used a less imaginative naming scheme to keep configuration easy (ABC01, ABC02, ABC03,...). The "behind the scenes" servers were all named after airplanes (sabre.xxx, mustang.xxx, etc), until someone decided to use less interesting but more suitable names (library.xxx, mail.xxx, etc). ... I got a dirty look when I replaced the Sun "Allante" with a Linux box and named it "Gremlin". The MS boxen are named according to the primary user, followed by the Windows version. For example, I started with "Bohnca95". Now that I've replaced it with an NT box, I find it amusing that my MS machine is named "Bohncant".
My machines at home are named after SF authors (Asimov, Heinlein, etc)
I have to machines on my desk at work now. The Unix systems are all named after luxery automobiles
Christopher A. Bohn
The GPL forbids combining GPL'd code with other-licenced code that would place extra restrictions beyond those in the GPL. For example, the advertising clause in the traditional BSD License keeps code with the traditional BSD License from being used with GPL'd code.
RMS wants the GPL to be used instead of the LGPL precisely to force derived works to be GPL'd. As he explains it, the Lesser GPL should only be used when there's already a popular non-free library with the same functionality. In such a situation, a developer could use the non-free library to develop non-free apps, so the compromise is to use LGPL to try to get the free libraries used, even if the derived works don't end up free.
Christopher A. Bohn
The problem is that the program is linked to both GPL'd code and QPL'd code. The licenses, while both open-source, are mutually exclusive (just like GPL and the traditional BSD license).
Christopher A. Bohn