login
Search
LinuxWorld recently posted an article by Richard Stallman (RMS) titled, Linux, GNU, and freedom. It begins as a reply to an earlier article by Joe Barr (The Stallman factor), explaining an incident with the SIGLINUX user group. RMS continues on to talk about the reason why the name 'GNU/Linux' is important, the influence Linus has, and his decision to utilize the BitKeeper tool.
Read on for more information, including a full mirror of the article.
Richard Stallman founded the GNU Project in 1984. He also originally authored a number of well known and highly used development tools, including the GNU Compiler Collection (GCC), the GNU symbolic debugger (GDB) and GNU Emacs.
There is a lot of history leading up to the present. The many links above make for interesting and highly recommended reading, describing events that are very much shaping the computers and general technology we now rely upon daily.
In this most recent article, RMS speaks strongly against Linus' decision to use BitKeeper. He ties this belief into the larger picture, drawing a correlation between non-free software, the CBDTPA, the Broadcast "Protection" Disucssion Group, software patents, and Microsoft's nondisclosure agreements. Referring to the 70's when UNIX systems were very much closed source, RMS says, "We can lose our freedom again just as we lost it the first time, if we don't care enough to protect it."
In closing, RMS offers the following food for thought:
"Linux, the kernel, is often thought of as the flagship of free software, yet its current version is partially non-free. How did this happen? This problem, like the decision to use Bitkeeper, reflects the attitude of the original developer of Linux, a person who thinks that "technically better" is more important than freedom.
"Value your freedom, or you will lose it, teaches history. `Don't bother us with politics,` respond those who don't want to learn."
Summary:
'Since Joe Barr's article criticized my dealings with SIGLINUX, I would like to set the record straight about what actually occurred, and state my reasons.' (1,900 words)
By Richard Stallman
(LinuxWorld) -- Since Joe Barr's article criticized my dealings with SIGLINUX, I would like to set the record straight about what actually occurred, and state my reasons.
When SIGLINUX invited me to speak, it was a "Linux User Group"; that is, a group for users of the GNU/Linux system which calls the whole system "Linux". So I replied politely that if they'd like someone from the GNU Project to give a speech for them, they ought to treat the GNU Project right, and call the system "GNU/Linux". The system is a variant of GNU, and the GNU Project is its principal developer, so social convention says to call it by the name we chose. Unless there are powerful reasons for an exception, I usually decline to give speeches for organizations that won't give GNU proper credit in this way. I respect their freedom of speech, but I also have the freedom not to give a speech.
Subsequently, Jeff Strunk of SIGLINUX tried to change the group's policy, and asked the FSF to list his group in our page of GNU/Linux user groups. Our webmaster told him that we would not list it under the name "SIGLINUX" because that name implies that the group is about Linux. Strunk proposed to change the name to "SIGFREE", and our webmaster agreed that would be fine. (Barr's article said we rejected this proposal.) However, the group ultimately decided to stay with "SIGLINUX".
At that point, the matter came to my attention again, and I suggested they consider other possible names. There are many names they could choose that would not call the system "Linux", and I hope they will come up with one they like. There the matter rests as far as I know.
Is it true, as Barr writes, that some people see these actions as an "application of force" comparable with Microsoft's monopoly power? Probably so. Declining an invitation is not coercion, but people who are determined to believe that the entire system is "Linux" sometimes develop amazingly distorted vision. To make that name appear justified, they must see molehills as mountains and mountains as molehills. If you can ignore the facts and believe that Linus Torvalds developed the whole system starting in 1991, or if you can ignore your ordinary principles of fairness and believe that Torvalds should get the sole credit even though he didn't do that, it's a small step to believe that I owe you a speech when you ask.
Just consider: the GNU Project starts developing an operating system, and years later Linus Torvalds adds one important piece. The GNU Project says, "Please give our project equal mention," but Linus says, "Don't give them a share of the credit; call the whole thing after my name alone!" Now envision the mindset of a person who can look at these events and accuse the GNU Project of egotism. It takes strong prejudice to misjudge so drastically.
A person who is that prejudiced can say all sorts of unfair things about the GNU Project and think them justified; his fellows will support him, because they want each other's support in maintaining their prejudice. Dissenters can be reviled; thus, if I decline to participate in an activity under the rubric of "Linux", they may find that inexcusable, and holds me responsible for the ill will they feel afterwards. When so many people want me to call the system "Linux", how can I, who merely launched its development, not comply? And forcibly denying them a speech is forcibly making them unhappy. That's coercion, as bad as Microsoft!
Now, you might wonder why I don't just duck the issue and avoid all this grief. When SIGLINUX invited me to speak, I could simply have said "No, sorry" and the matter would have ended there. Why didn't I do that? I'm willing to take the risk of being abused personally in order to have a chance of correcting the error that undercuts the GNU Project's efforts.
Calling this variant of the GNU system "Linux" plays into the hands of people who choose their software based only on technical advantage, not caring whether it respects their freedom. There are people like Barr, that want their software "free from ideology" and criticize anyone that says freedom matters. There are people like Torvalds that will pressure our community into use of a non-free program, and challenge anyone who complains to provide a (technically) better program immediately or shut up. There are people who say that technical decisions should not be "politicized" by consideration of their social consequences.
In the 70s, computer users lost the freedoms to redistribute and change software because they didn't value their freedom. Computer users regained these freedoms in the 80s and 90s because a group of idealists, the GNU Project, believed that freedom is what makes a program better, and were willing to work for what we believed in.
We have partial freedom today, but our freedom is not secure. It is threatened by the CBDTPA (formerly SSSCA), by the Broadcast "Protection" Discussion Group (see http://www.eff.org/ ) which proposes to prohibit free software to access digital TV broadcasts, by software patents (Europe is now considering whether to have software patents), by Microsoft nondisclosure agreements for vital protocols, and by everyone who tempts us with a non-free program that is "better" (technically) than available free programs. We can lose our freedom again just as we lost it the first time, if we don't care enough to protect it.
Will enough of us care? That depends on many things; among them, how much influence the GNU Project has, and how much influence Linus Torvalds has. The GNU Project says, "Value your freedom!" Joe Barr says, "Choose between non-free and free programs on technical grounds alone!" If people credit Torvalds as the main developer of the GNU/Linux system, that's not just inaccurate, it also makes his message more influential--and that message says, "Non-free software is ok; I use it and develop it myself." If they recognize our role, they will listen to us more, and the message we will give them is, "This system exists because of people who care about freedom. Join us, value your freedom, and together we can preserve it." See http://www.gnu.org/gnu/the-gnu-project.html for the history.
When I ask people to call the system GNU/Linux, some of them respond with silly excuses and straw men. But we probably haven't lost anything, because they were probably unfriendly to begin with. Meanwhile, other people recognize the reasons I give, and use that name. By doing so, they help make other people aware of why the GNU/Linux system really exists, and that increases our ability to spread the idea that freedom is an important value.
This is why I keep butting my head against bias, calumny, and grief. They hurt my feelings, but when successful, this effort helps the GNU Project campaign for freedom.
Since this came up in the context of Linux (the kernel) and Bitkeeper, the non-free version control system that Linus Torvalds now uses, I'd like to address that issue as well.
Bitkeeper issue
The use of Bitkeeper for the Linux sources has a grave effect on the free software community, because anyone who wants to closely track patches to Linux can only do it by installing that non-free program. There must be dozens or even hundreds of kernel hackers who have done this. Most of them are gradually convincing themselves that it is ok to use non-free software, in order to avoid a sense of cognitive dissonance about the presence of Bitkeeper on their machines. What can be done about this?
One solution is to set up another repository for the Linux sources, using CVS or another free version control system, and arranging to load new versions into it automatically. This could use Bitkeeper to access the latest revisions, then install the new revisions into CVS. That update process could run automatically and frequently.
The FSF cannot do this, because we cannot install Bitkeeper on our machines. We have no non-free systems or applications on them now, and our principles say we must keep it that way. Operating this repository would have to be done by someone else who is willing to have Bitkeeper on his machine, unless someone can find or make a way to do it using free software.
The Linux sources themselves have an even more serious problem with non-free software: they actually contain some. Quite a few device drivers contain series of numbers that represent firmware programs to be installed in the device. These programs are not free software. A few numbers to be deposited into device registers are one thing; a substantial program in binary is another.
The presence of these binary-only programs in "source" files of Linux creates a secondary problem: it calls into question whether Linux binaries can legally be redistributed at all. The GPL requires "complete corresponding source code," and a sequence of integers is not the source code. By the same token, adding such a binary to the Linux sources violates the GPL.
The Linux developers have a plan to move these firmware programs into separate files; it will take a few years to mature, but when completed it will solve the secondary problem; we could make a "free Linux" version that doesn't have the non-free firmware files. That by itself won't do much good if most people use the non-free "official" version of Linux. That may well occur, because on many platforms the free version won't run without the non-free firmware. The "free Linux" project will have to figure out what the firmware does and write source code for it, perhaps in assembler language for whatever embedded processor it runs on. It's a daunting job. It would be less daunting if we had done it little by little over the years, rather than letting it mount up. In recruiting people to do this job, we will have to overcome the idea, spread by some Linux developers, that the job is not necessary.
Linux, the kernel, is often thought of as the flagship of free software, yet its current version is partially non-free. How did this happen? This problem, like the decision to use Bitkeeper, reflects the attitude of the original developer of Linux, a person who thinks that "technically better" is more important than freedom.
Value your freedom, or you will lose it, teaches history. "Don't bother us with politics," respond those who don't want to learn.
Copyright 2002 Richard Stallman Verbatim copying and distribution of this entire article are permitted without royalty in any medium provided this notice is preserved.
RMS has a point
"Linux, the kernel, is often thought of as the flagship of free software, yet its current version is partially non-free."
The Linux kernel is a good example of the Open Source world view. RMS should be upset when it is presented as the flagship of the free software world view.
There have been a lot of flamewars that are the direct result of lack of clairity in the Open Source world view. BitKeeper is one, but the module license symbol flamewar, and the tainted module flamewar are better examples.
Like it or not, RMS holds fast to the least complicated (rationalized) world view, even if like the Catholic Church it is wrong.
<i>I believe in Free software
I believe in Free software, and I think it's important not to bind yourself to proprietary software vendors, but I also believe in using the best tool for the job.
RMS and the other free software true beleivers say: "use the best free tool for the job"
I think you fall into the open source camp.
people do things for differen
people do things for different reasons. that does not mean that some are wrong or some are right. i personally agree with RMS on this point. that does not mean i "lose track of reality".
He doesn't value technical s
He doesn't value technical superiority. Damnit, he INVENTED bloatware when he spewed out emacs. What Stallman values is getting his name on everything - and that's why he's so pissed at Linus. Sheer jealousy!
Anti-RMS sentiment?
I was amazed to see a confused post on this page disparaging RMS's philosophy. The truth is, Free Software is the only line of defence against draconian laws like the DMCA. Once you allow even a few lines of proprietary software into your system, you become vulnerable to tools of extortion like BSA audits.
Even if you run a Debian main system and just install BitKeeper to 'keep up' with the kernel developers, you've effectively lost many of the rights the Free Software system previously afforded you. Once you install that proprietary software, complete with a non-free end user license agreement, you become just another 'software customer' and a potential 'pirate'.
I still won't call Linux GNU/Linux because what we call a Linux system comprises components from so many sources. Last time I checked, there were more lines of OpenOffice code on my box than GNU code -- does that mean I should call it a Sun/GNU/Linux box? That's just silly.
Just because it's more advanced than CVS doesn't make BitKeeper the 'right choice'. BitKeeper is a liability to developers and users; after all, you know what Voltaire said about those who give up their rights for temporary safety...
But you don't have to...
> Even if you run a Debian main system and just install BitKeeper to
> 'keep up' with the kernel developers, you've effectively lost many of
> the rights the Free Software system previously afforded you.
Do it they way people did before. In fact kernel baselines seem to be getting released faster these days. Several people on lkml have set up daily snapshot sites. If the FSF/GNU wont run Bitkeeper to sync to CVS they can just use the daily snapshots for their tree.
AFAIK Linus never signed on to be an ambassador for Free (or even Open Source) software. He's just doing what he enjoys. Let the people that want to be the political activists get on with it and leave the rest who just want to write code get on with it. We are still free to make our own choices are we not?
Dissection of the dissection...
Ok. I'm probably asking for this but license threads are usually the
most "active" discussions in this community :-) And hey I've got to
answer the accusation of being a troll.
troll. v.,n
...posting on Usenet designed to attract predictable responses or
flames;...
(from Jargon File:
http://www.tuxedo.org/~esr/jargon/html/entry/troll.html)
dissection. n.
...
3. A detailed examination or analysis.
(from Dictionary.com:
http://www.dictionary.com/search?q=%20Dissection%20 )
And onwards...
>> Do it they way people did before.
> So, because people have done the wrong thing before, it's right to keep doing it?
Well I can see four non-BK choices on offer to the aspiring kernel hacker to
keep up with kernel development.
1. Track kernel changes manually. From kernel release to kernel
release with or without the help of Linus's Bitkeeper generated
Changelog
2. Import the kernel releases or daily snapshots into his/her favorite
"free" SCM tool.
3. Follow the lkml and associated special interest mailing lists to
keep their finger on the pulse of kernel development.
4. Write/Modify a distributed SCM tool under there own FSF friendly
license and convince Linus to use it in BK's place.
The aspiring kernel hacker still has all these choices, or if they can
live with themselves they could use BK. BK is not the barrier it is
put up to be. I would disagree that Linus has to change the tools he
uses just because the aspiring hacker can't keep up.
>> In fact kernel baselines seem to be getting released faster these days.
> He evades the part of the previous post which already accepts that
> BitKeeper may be better code than CVS or other Free Software source control tools.
The quote from the previous post was "Just because it's more
advanced than CVS doesn't make BitKeeper the 'right choice'. BitKeeper
is a liability to developers and users; after all, you know what
Voltaire said about those who give up their rights for temporary
safety..."
Its obviously the right choice for Linus. Its his choice to make and
its not mandated for kernel developers. IANAL but having read the
BitKeeper license I must of missed the bit that said using BitKeeper
affects the license of the source code you manipulate with it. If
anyone can point out the offending stanza in the license
(see: www.bitkeeper.com/Sales.Licensing.Free.html)
I would be most interested to know what it is.
By the way I thought it was Benjamin Franklin that said "They that can
give up essential liberty to obtain a little temporary safety deserve
neither liberty nor safety." although I believe Voltaire said
"I am very fond of truth, but not at all of martyrdom."
>>We are still free to make our own choices are we not?
> The original poster actually says "doesn't make BitKeeper the
>'right choice'" -- the use of the word choice implies that
>everyone is free to choose.
Unless the rules have changed recently I'm with you so far..
> Yet, the pro-BitKeeper poster conveniently ignores this, ironically
> trying to turn the 'freedom to choose' people against the
> anti-BitKeeper poster.
The trouble with the BitKeeper debate is that so much of it is phrased
in terms of "Linus should do this...", "We should all do that..."
terms which is inherently trying to push other people one way or the
other. I'm all for debate and putting your case but a lot of hot air
has been expended with people putting the same points over and over
again in the hope mere repetition of the same argument will win the
day. Linus has already said (See: lwn.net/2002/0425/a/ideology-sucks.php3)
he would would use a free tool if one existed. If someone
cares that much about what tool Linus uses they know what to do.
> Saying no to BitKeeper is the best way we can maintain our freedom
> to choose in the future.
I wouldn't say its the best way, in fact given the number of other
things (software related or not) going on in the world dealing with
the BitKeeper "issue" is pretty low on *my* list.
> Once you've installed it, you've opened up a legal can of worms
> -- surely if you just wanted to code, then you wouldn't want to take
> the risk of submitting yourself to arduous legal battles.
As I said before, lack of legal credentials not withstanding, if
people can point at the worms please feel free. I think the major
problem here is the difference between pragmatists and idealists. I am
an unashamed pragmatist, when I make choices I do it by balancing
risk against gain. I may have Free software leanings but it doesn't
stop me getting my job done.
> What Alex fails to understand is that Intellectual Property laws
> are only getting stronger -- the price to pay for BitKeeper is just
> too great for anyone who values their privacy and freedom to code.
I like to think I have a reasonable grasp of the state of Intellectual
Property in this day and age. However I'd choose a different battle.
Phew. This has turned out to be quite long. I might as well finish
with my option of RMS.
RMS is an unashamed idealist. I respect his point of view and he has
certainly had a big impact on Free software. However he didn't do it
on his own. The Perens's and Raymonds's of this world bought "Open
Source" to the attention of commercial organisations have also had a
big effect. I think both "Free Software" and "Open Source" have
benefited overall. While the idealists may not be happy I sometimes
wish they would choose to concentrate their energies on convincing
the great unwashed of the benefits of their way of doing things than
repeating their point of view to people that are already overly
familiar with it.
Alex
re: Dissection of the dissection...
> IANAL but having read the
> BitKeeper license I must of missed the bit that said using BitKeeper
> affects the license of the source code you manipulate with it. If
> anyone can point out the offending stanza in the license
> (see: www.bitkeeper.com/Sales.Licensing.Free.html)
> I would be most interested to know what it is
You are mistaken. BK can be used to develop Free software. (capital-f free == FSF-style Free). Using non-Free software in the toolchain you use to develop Free software can be a problem. If you write code that needs to be compiled with non-Free tools, the Free source isn't as useful to other people. (This was the case with Java, but Free java compilers and run-time environments (JVMs) are making progress.)
The BK case is marginal. BK isn't part of the toolchain, so a Free project developed with BK can drop BK and be developed using only Free software at any time by anyone. The project doesn't become dependent on BK. BK is not a building block for anything else, unlike a library, or a windowing system. For one thing, only developers of the project have any reason to have anything to do with BK, and it is possible for developers to work without using BK, or any other non-Free software. For a project as important as the Linux kernel, lots of people are interested in getting at the source code, etc.
I don't know if the GPL implies that source be provided in a way that doesn't require the use of non-Free tools, but that is something to be concerned about. You can always write to the authors and ask them to send you the source. As long as you can get your hands on a Free tarball, you can do anything you like.
AFAICT, the only real arguments against using BK for managing the Linux kernel is that it encourages the kernel developers to use BK, and that it is an endorsement of non-Free software.
The kernel developers can endorse whoever they like. IIRC, the source for BK itself used to be available, and the license was GPL + you can't remove the openlogging stuff. That seems to have changed, however, so now you can't run BK on platforms they don't provide binaries for. That's a big problem, and it makes BK a lot less Free than it was when BK itself was GPL+logging.
--
Peter Cordes; peter@llama.nslug.ns.ca
Installing BitKeeper does not bring the apocalypse
Once you allow even a few lines of proprietary software into your system, you become vulnerable to tools of extortion like BSA audits.
You are never "vulnerable to tools of extortion like BSA audits". The BSA has no authority. You can refuse to allow them to audit your licenses. You don't need to waste your time with them. You can completely ignore their threats. They are not a law enforcement agency.
Even if you run a Debian main system and just install BitKeeper to 'keep up' with the kernel developers, you've effectively lost many of the rights the Free Software system previously afforded you. Once you install that proprietary software, complete with a non-free end user license agreement, you become just another 'software customer' and a potential 'pirate'.
No. You lose no rights. Any program's license specifies your rights to use and distribute it. Not other programs on your system.
Its not a GNU/Linux system...
I can see where RMS is comming from in terms of crediting GNU for what makes up most modern distributions but where do you stop? To credit everyone I should probably call my box:
My GNU/MIT/BSD/X11/PD/Linux Kernel powered system put together by MandrakeSoft.
But that is a bit of a mouthfull. That probably explains why "Linux" the distribution (as opposed to "Linux" the kernel which is all I've seen Linus day on the matter) has become handy for headlines. When people ask me about Linux I can then go into more details about its mixed roots :-)
Alex
Indeed, it isn't (completely)
Someone should try communicating that to RMS. Credit where credit is due but I think the name of the OS is just the wrong place.
Credit for what?
rms has never claimed that Linux should be GNU/Linux because of lines of code. His point is that the GNU project was a vision of an operating system completely made up of free software, back before the idea was popular (or believable). The intention was never that the software would be written specifically for the project. It was always the intention that existing bits of code would be used when that made sense.
So yes, a Debian GNU/Linux machine might only have a small proportion of GNU software, but the whole is what the GNU project was about: a complete, useful system that's made up of software that anyone can share and modify.
Credit for what?
I'm sorry, but this still sounds like a bunch of sour grapes from rms. The "vision" that rms had of an operating system made up of free software contained a kernel called "Hurd", which was eternally late/broken and is even today some way short of prime time. The fact that Linus was able to write a kernel and to get other people to contribute to it in a (more-or-less) orderly way is tribute to his technical and project management skills. It was very convenient for rms that somebody was able to write a Unix-like operating system that was released under the GPL. But the fact that this code happened to fill a hole in the GNU "vision" does not give rms any sort of moral right to force the GNU brand name onto it.
Debian has chosen to brand their offering as "GNU/Linux" and it is certainly their right to do so. But rms is not satisfied with people making up their own minds whether or not to go down this path -- he holds his breath and stomps his feet whenever anyone uses the term "Linux" to refer to anything other than the kernel. He wants to force the Linux project to exist under the GNU umbrella and to force people to refer to it using the terms that he wants them to use. Which you have to admit is a bit rich coming from somebody who likes to talk about freedom so much.
The ignorance here is severe.
The ignorance here is severe. HURD is not the point. The C compiler and the rest of the tool chain, the shell, all the command line utilities, as well as init, getty, login, -- these were developed by the GNU project. *Everything* in the system, back in the days when the Linux kernel was developed, came from GNU except for the kernel. The objection, justifiably, is to use the name "linux" to refer to the system that the linux kernel was *added to*.
As for RMS forcing anyone to do anything -- that claim is stupid, moronic, and transparently intellectually backrupt.
What's the issue with the drivers?
What drivers are using binary-only code distributed with the Linux source? I've followed the discussions about flagging a tainted kernel for debugging purposes, and there was the debate about binary modules linking to free code, but neither of those issues deserve to be FUDified like this.
Are there any binary-ony drivers included the official Linux source?
Some Clarity Would Be Good
What drivers are using binary-only code distributed with the Linux source? I've followed the discussions about flagging a tainted kernel for debugging purposes, and there was the debate about binary modules linking to free code, but neither of those issues deserve to be FUDified like this.
I don't think RMS is "FUDding" this at all ... from his standpoint it is a very real and important issue, making Linux (the kernel, thus no "GNU" in front of it) non-free.
If I recall correctly, my old SCSI drivers used to load firmware into the device during the boot sequence. Perhaps this was proprietary code, or he is referring to something like this?
It would be nice to know (a) what modules have binary firmware or BIOS code in them [or, for that matter, binary code of any kind] and (b) what the copyright status and licensing terms of that binary code is. Until someone like Alan, Macedo, or Linus can definitively assure us that the code (and modules) are clean and unencumbered, or alternatively specify what their constraints are, I'll keep an open mind vis-a-vis RMS's comments and certainly not dismiss them out of hand.
He was one of the people who, after all, pointed out the early licensing problems that Trolltech and KDE had, which ultimately led to Trolltech modifying their license into something compatible with the GPL and fixing the problem before it ever became acute (i.e. resulted in someone ... a free software developer perhaps ... being dragged into court). We ignore warnings like this at our own peril, and RMS, for all his lack of diplomacy, has a pretty good track record with respect to licenses and licensing issues.
GNU-Free Linux?
RMS has valid points, coming from his ideological position. Unfortunately, I do find one gaping hole in the overall "GNU/Linux" debate- it presumes that all users of Linux (the kernel) will have reason to couple it with GNU (the Not-UNIX UNIX-like suite).
So, I have to wonder- can anyone provide an example of a GNU-free Linux system? While a project to create one would be a largely academic (read: useless) pursuit, there should be nothing preventing the creation of one, and if there is, I would consider it a design flaw in the kernel. (Okay, here's a counterpoint: glibc is probably a serious dependency, and as such, every Linux-the-kernel needs a little GNU in it.)
But, ignoring that glibc note for the sake of argument:
Look at it this way; Linux is Linus, not RMS. Linus doesn't have anything against Bitkeeper (even if he should), a closed product that runs atop Linux. Therefore, he shouldn't have a problem even if Microsoft decides to use Linux as the kernel for Windows: Linux Edition, so long as they don't violate his (GPL) license on the kernel code itself. Closed source atop GPL is not necessarily a GPL violation, at least by the interpretation that allows products like Bitkeeper, Loki ports, etc.
So really, it all comes down to semantics, and whether or not you're willing to ignore glibc. Personally, I would think a group called SIGLINUX would have a focus on applications of the Linux kernel, with or without GNU. If their focus was more distro-level, and is willing to include BSD as well, "SIGOPEN" or "SIGOSS" might be a more appropriate name. Only if the group's mission was to promote the FSF's goals (which it obviously wasn't, at least not by consensus), would "SIGGNU" or "SIGFREE" be appropriate. RMS's ruleset is a little skewed; he should certainly ask that he be represented properly (if he's coming to speak about the combination of the GNU toolchain with the Linux kernel, promoting the speech as about 'GNU/Linux' is appropriate), but requesting namechanges of his *hosts* limits him to preaching to the converted.
The real *problem* that keeps the naming-convention war going is the perception that "Linux" can refer to anything other than the kernel itself. It makes a handy shorthand, but it would be much more useful to specify things by distribution, or as LSB (GLSB?) compliant in the case of software.
I think the problem is that there aren't enough LSB-*noncompliant* distributions out there; it'll take more products like Cosmoe (or Windows: Linux Edition ;)) to make people rethink things. After all, the *LSB* is what requires most of the GNU 'dependencies.' ... Groups like SIGLINUX may care more about Linux-the-kernel than GNU/Linux, but the present market gives them only GNU/Linux distributions to talk about.
Imagine how much worse the naming conventions will be when 'RedHat HURD' comes to market...
Examples...
> So, I have to wonder- can anyone provide an example of a GNU-free Linux system?
I don't know about GNU-free but there are certainly distributions that only have minor GNU components. I'm mainly thinking of single disk distributions which replace most of the GNU tools with something like Busybox (which I don't think is official GNU despite being GPL'ed).
Glibc isn't magic. Pure embedded apps may just make syscalls directly into the kernel. Failing that there are alternatives to glibc such as dietlibc or the elks libc.
Of course to be fair to the GNU crowd your need the GNU tool chain to build these things (or at least I *think* so as a result of the number of gcc'isms in the kernel).
Alex
Can't agree with you more. ;)
Can't agree with you more. ;)
Don't rename LINUX
A much simpler solution would be for RMS to rename GNU to GNL.
Simpler solution: Change "GNU" to "Linux"
RMS made the GPL license, but as one of the posters pointed out, where is the Hurd? The Hurd has been a failure. And since most of the GNU software are a bunch of UTILITIES that have been ported over to MS-DOS, does that make MS-DOS part of GNU? Should it be called GNU/MS-DOS?
Linus Torvalds never received funding from the Free Software Foundation, right? He wrote the kernel. The kernel is non-trivial. Therefore he can call it whatever he wants. The utilities that currently run under the Linux kernel can be rewritten and outmoded.
RMS is a brilliant man, no doubt. But his vision is like communism. It only works if "everybody were exactly alike." Sorry, RMS, not all of us have a Physics degree from Harvard or could be accepted to graduate school at MIT. Not all of us could charge $250 PER HOUR like you did to develop the GNU software. Not all of us have YOUR CONNECTIONS to get companies to pay us that kind of money (under your terms of "Free Software" )to PAY OUR LANDLORDS for our RENT, UTILITIES, and FOOD. Not all of us could convince our universities/companies to live in our offices; MIT was nice to let you do that. Not all of us can get a MacArthur Genius Grant to pay us a quarter of a million dollars so that we can invest in mutual funds and live off the interest. Not all of us are fortunate to live in cities where we don't need a car and gas like you do.
Do you really think "society" is less better off because companies retain an edge over their competition by "hoarding" source code? Then why aren't all companies using Emacs for their word processing? Don't you think WYSIWIG is the way to go? And isn't that how you got carpel tunnel syndrome -- because you refused to use the mouse and stick to your Emacs macros? Now your body is paying for this ("Don't fool with Mother Nature") , and you haven't programmed any code in what, almost a decarde, according to some interviews?
"Contemporize, man."
Or get GNU Hurd working, and you won't have to care, right?
: RMS made the GPL license, b
: RMS made the GPL license, but as one of the posters pointed out, where is the Hurd? The Hurd has been a failure.
Here it is:
http://www.gnu.org/hurd
It has been in development for a long time, but it's now almost in the stage of being usable. It seems the long development period has to do with a couple of things: the long delay into making free the mach microkernel - upon which Hurd relies; the rise of Linux (what to do?? continue with our kernel or use an already built one? as it became appearent that Linux was a project on its own other than GNU, the FSF rightfully ended up sticking with its original purpose of building a complete free OS on its own); the technical hurdles themselves (hey, building a kernel is no easy job! if an obsolete monolithic kernel like Linux got so many man-hour steem via descentralized development over the web, just try to picture the much fewer FSF devotees making their best to pull out a much more modern kernel architecture); and the fact that all device drivers must be written from scratch, since using proprietary ones (like in Linux) would invalidate FSF's original goal.
It is alive and while still not kicking, it's certainly all but a failure. It currently runs moderately only in intel machines. You gotta admit that getting all possible device drivers rewritten from scratch - unlike getting Linux to run with proprietary ones - is no easy task. It probably won't ever see the kind of general acceptance as "Linux" gets, after all, it's aimed at being the _true_ GNU system, which means: no non-free parts. That means no proprietary device drivers, no Netscape browser, no Sun's StarOffice suite and the like.
At least, the FSF will have fulfilled its goal of making a completely free operating system. Now they can go on to the more noble goals of completely banishing non-free software from the face of Earth and making love, not war. ;)
: And since most of the GNU software are a bunch of UTILITIES that have been ported over to MS-DOS, does that make MS-DOS part of GNU? Should it be called GNU/MS-DOS?
MS-DOS the OS don't depent upon the GNU OS utilities to run, unlike Linux the kernel.
: Linus Torvalds never received funding from the Free Software Foundation, right?
What this has to do with anything?
: He wrote the kernel. The kernel is non-trivial. Therefore he can call it whatever he wants.
That's right. It's the Linux kernel from the GNU OS.
: The utilities that currently run under the Linux kernel can be rewritten and outmoded.
That can be done. Not without huge efforts and possibly splitting the Linux community and bringing in many compatibility problems. And until then, it's possible that GNU/Hurd will be a far better alternative. :)
: RMS is a brilliant man, no doubt. But his vision is like communism. It only works if "everybody were exactly alike."
Hmm... that sounds more like Microsoft.
: Sorry, RMS, not all of us have a Physics degree from Harvard or could be accepted to graduate school at MIT. Not all of us could charge $250 PER HOUR like you did to develop the GNU software. Not all of us have YOUR CONNECTIONS to get companies to pay us that kind of money (under your terms of "Free Software" )to PAY OUR LANDLORDS for our RENT, UTILITIES, and FOOD. Not all of us could convince our universities/companies to live in our offices; MIT was nice to let you do that. Not all of us can get a MacArthur Genius Grant to pay us a quarter of a million dollars so that we can invest in mutual funds and live off the interest. Not all of us are fortunate to live in cities where we don't need a car and gas like you do.
That's right! Not everyone has the background of Mr. Stallman and that's exactly why he is the right man to pull it off! He has the history, the skills, the conscience, the moral. That's why you can play with your very own Unix system at home today, that's why you can learn from source code today, that's why you can download software for free or even update your system off the internet, that's why you can share software with your friends today, like once in the 60s like RMS well remembers.
: Do you really think "society" is less better off because companies retain an edge over their competition by "hoarding" source code?
Yes.
: Then why aren't all companies using Emacs for their word processing? Don't you think WYSIWIG is the way to go? And isn't that how you got carpel tunnel syndrome -- because you refused to use the mouse and stick to your Emacs macros? Now your body is paying for this ("Don't fool with Mother Nature")
haha, that was actually fun. But it is very much known for anyone with the skills to do it that hitting some key combos is much faster for everyday jobs than moving the mouse pointer around and pulling down menus like the novice do. Even in WYSIWYG environments, like say, MSWord, you are way better hitting combo keys to perform operations than working with a mouse. And you said anything bad about macros?? Macros are of such importance that MS even has their very own macro recorders and even a VB macro editor embedded. Just shut up and don't look like a dumb pc newbie.
, and you haven't programmed any code in what, almost a decarde, according to some interviews?
give the man some rest, will ya? After actively working so many years on projects as distinct as Emacs, GCC, Make, Autoconf and many others and still having the time to pull off a completely free OS, a political movement, and a highly influencial software license, the guy surely needs a well deserved rest... though he isn't likely to accept it.
By the way, does anyone knows of any software Bill Gates has actually developed with his bare fingers?
: Or get GNU Hurd working, and you won't have to care, right?
Seems like i'm not the only one to wanna try it out... :)
whining
(hey, building a kernel is no easy job! if an obsolete monolithic kernel like Linux got so many man-hour steem via descentralized development over the web, just try to picture the much fewer FSF devotees making their best to pull out a much more modern kernel architecture)
Why did the Linux kernel project attract so "many man-hour steem" instead of the FSF/GNU/Stallman? What pushed people away from FSF/GNU/Stallman?
It probably won't ever see the kind of general acceptance as "Linux" gets, after all, it's aimed at being the _true_ GNU system,...
So the Hurd is largely an academic exercise? Why bother since the mach kernel already existed? Sounds like some additional utilities, such as an improved SCM tool, would be a better use of ones time. Or, finding ways to make the existing code base completely free, such as rewriting all of the supposed non-free device drivers. Why not join the party and contribute to the code base in a way that is both pragmatic and 'moral': rewrite non-free stuff.
that's why you can share software with your friends today, like once in the 60s like RMS well remembers
Ouch! Is that as in remember the good old days when only the annointed could see so much as use a computer and they were mostly used to design weapons or eliminate labor in factories? I guess that's pretty much still true today, at least the labor part, but computers as a whole are much more accessible in spite of the for-profit, closed-source software model. Advances in semiconductor scaling and packaging technology are probably the biggest reason for the spread of computers, not the software license, Windows, or X/Gnu/Linux.
I heard somewhere years ago that Apple paid Novel to port the Mac OS to the IBM PC platform. They succeeded but Apple canned the project. Just think how different todays landscape might be if Apple had allowed that project to go forward. There might have been real competition in the desktop OS market, prices may have been reasonable, and Gnu/Linux would still be someones academic project.
yes, whinning
: Why did the Linux kernel project attract so "many man-hour steem" instead of the FSF/GNU/Stallman? What pushed people away from FSF/GNU/Stallman?
Many factors, such as objection to strong political and ideological ideas, objection to RMS himself and the fact that most people just want to code without getting involved in political movements. Also, open-source people like the Microsoft business model, where all programmers are vendors of prepackaged software, as opposed to the more insecure, but more accurate, FSF's view of a programmer like what he really is: a liberal professional, selling his services much like a doctor or a lawyer, rather than selling a product.
: So the Hurd is largely an academic exercise?
No. I was just stating that it'll develop much slower than Linux and will suffer from not readily supporting all kinds of hardware - device drivers - like Linux. hey, there's a price for freedom...
: Why bother since the mach kernel already existed?
A microkernel is not a complete kernel. And it seems Carnegie Mellon University - Mach creators - was reluctant into making it free.
: Sounds like some additional utilities, such as an improved SCM tool, would be a better use of ones time.
Really? And running it upon some non-free OS? Come on, when you have a goal in life and strong ideological reasonings, you better working on it to make it real. What did you want them to do? Just give up on everything they did to this day?
: Or, finding ways to make the existing code base completely free, such as rewriting all of the supposed non-free device drivers.
What do you think it's taking so long for Hurd to be ready? Did you actually read what i wrote previously?
: Why not join the party and contribute to the code base in a way that is both pragmatic and 'moral': rewrite non-free stuff.
That certainly would be a good idea, except... it would mean scrapping years of work worth of Hurd and begin working on the more modest Linux kernel. And i don't think RMS and his strong views would permit him to bow to Linus. ;)
: Advances in semiconductor scaling and packaging technology are probably the biggest reason for the spread of computers, not the software license, Windows, or X/Gnu/Linux.
I agree for the widespread use of computers. But those computers would be running proprietary software only today wasn't it for RMS and his stronghold on a moral use of computing. I guess wasn't for it and everybody would be happily phoning to Microsoft support every hour or so to get permission to print a file...
: I heard somewhere years ago that Apple paid Novel to port the Mac OS to the IBM PC platform. They succeeded but Apple canned the project. Just think how different todays landscape might be if Apple had allowed that project to go forward. There might have been real competition in the desktop OS market, prices may have been reasonable, and Gnu/Linux would still be someones academic project.
I don't think RMS would just give up on his ideals. He would pull it off no matter how many time it would take. And freedom, my friend, doesn't mean low prices, as you seem to sugest.
Re: Whining
[Note: I submitted this reply about 20 minutes ago and I don't see it updated. So advanced apologies if this appears twice on the forum]
:But those computers would be running proprietary software only today wasn't it for RMS and his stronghold on a moral use of computing. I guess wasn't for it and everybody would be happily phoning to Microsoft support every hour or so to get permission to print a file...
Certainly RMS deserves a big share of credit. However, he's not the only one: Remember the history of Linux. Linus learned how to create kernel's from Andrew Tanenbaum's (BS MIT, PhD Berkeley) book on Tanenbaum's Minix, which was created precisely in the spirit of TEACHING students the nuts and bolts of how to write OS Kernels. All the source code was there and fully documented, with an accompanying textbook written in an extraordinarily lucid style. Yes, Minix wasn't free, but it was Open, and Linux is a logical evolution whose roots can be traced directly from Minix, which can be directly traced from a professor who gave a darn about actually transferring his insight and knowledge to another generation (unlike many so-called "professors" who are employed by research universities and screw their paying students by giving lousy lectures, letting amateur Teaching Assistants do all the grunt work, and refusing to advise and mentor...yes spoken from experience). To clarify, Linus was NOT a direct student of Tanenbaum, and even Tanenbaum and Linus have traded seemingly hostile barbs over design considerations in Linux.
In fact, RMS knew about Minix and contacted Tanenbaum early on before Linus was even in high school, but Tanenbaum was so offended by RMS's aggressive ("obnoxious") demeanor, that Tanenbaum decided not to participate. So in a way, it's ironic that RMS's favored kernel (Hurd) would be overtaken in popularity by the son-of-Minix.
And once again, RMS never seems to learn from his tactics. RMS's current fighting style is turning people off to his cause. If he is worried about losing "credit" , then he is going about it in an ungracious manner by attacking the name of "Linux." "Let it be" -- it's released under the GPL anhyhow, so everybody KNOWS it's governed by GNU. On the other hand, if he portrayed himself as the silently overlooked "saint" of the Free Software movement, in the style of a Mother Theresea, he would draw a more sympathetic and ardently supporting crowd.
: I heard somewhere years ago that Apple paid Novel to port the Mac OS to the IBM PC platform. They succeeded but Apple canned the project. Just think how different todays landscape might be if Apple had allowed that project to go forward. There might have been real competition in the desktop OS market, prices may have been reasonable, and Gnu/Linux would still be someones academic project.
And finally don't forget FreeBSD, as an alternative to Linux, which was developed independently of FSF and GNU, if memory serves correctly (and do correct me if I'm wrong). Now that Apple is going Open Source with a BSD variant, the landscape is already changing, even for Linux and Sun Microsystems. Another interesting idea was Sun's attempt in the early 90's to purchase Apple -- aka "Snapple". Thus, Unix-on-the-desktop could have been a lot earlier, and now Sun is starting to scramble, having lost business from Oracle, which is swtiching from Sun to Linux servers.
:: RMS is a brilliant man, no doubt. But his vision is like communism. It only works if "everybody were exactly alike."
:Hmm... that sounds more like Microsoft.
...
:By the way, does anyone knows of any software Bill Gates has actually developed with his bare fingers?
Gates was known as an excellent "code bummer" with assembly language. Along with Paul Allen and Monte Davidoff, he developed the BASIC interpreter for the Altair and contributed programming articles in the early Dr. Dobb's Journals. He had a big part in the development of the IBM PC architecture with Don Estridge. That doesn't mean Gates was a computer scientist or anything, but he was technically astute.
I actually reserve my disdain for the non-programmers Tim O' Reilly (O'Reilly and Associates) and Robert Young, CEO and Founder of Red Hat. These guys are like car salesmen, and strike me as being charlatans who have seen where the parade is going, and rushed to get in front of it, waving the right banners, yet financially exploiting the work of programmers. It makes me sick that these guys contributed nothing technically , but have benefited financially in a way that makes Bill Gates look like a saint. After all, at least Gate's programmers get a salary and stock options, and Microsoft has made many millionaires from among its employees.
And no, I don't work for Micro$oft. I hear they have a very combative style among their own peers, and I would be easily crushed into dust! Witness the recent resignation of Seamus Blackley, who looks like a football quarterback.
:as opposed to the more insecure, but more accurate, FSF's view of a programmer like what he really is: a liberal professional, selling his services much like a doctor or a lawyer, rather than selling a product.
This is like saying "breast cancer" is the same as "brain cancer". In other words, the word "cancer" is a generic term that is describing two extremely different diseases. (Please don't read anything into my choice of a word that could have a negative connotation.) Certainly SOME software is service-oriented and can support its developers via a subscription service, other software can be a "loss leader" for a manufacturing enterprise that makes money off the hardware (eg device drivers, or software for programming embedded microcode). But not ALL software can fit under RMS's narrow and restricted vision ala the "GNU Manifesto".
Look at all the shareware being written today. These programmers can (barely) make a decent living only if you treat it as a "product". The software is small and functional. It gets the job done. But shareware authors certainly can't charge the same rates as lawyers or doctors. And shareware authors are HUMAN BEINGS, they need to eat, have shelter, and pay health and auto insurance. Where does shareware fit in the GNU Manifesto?
Software by Bill
According to a biography of Bill, if I recall correctly, he wrote the
"donkey" game that shipped with somethingorother (an early MS-DOS,
perhaps?). And I think did a chunk of the coding for Microsoft Basic.
Egads, what an ass.
Egads, what an ass.
I find it fascinating that St
I find it fascinating that Stallman believes the rest of the Linux community is worried about whether a software package they use is "free" or not. Free software is fine and good, but I'll be damned if I'm going to make myself into some kind of ideological zealot who says that ALL software must be free. Stallman can rant and rave all day long if he wants to. The legions of wet behind the ears "hackers" who regurgitate everything he says can rant and rave too. Those who are actually writing code and doing real development work are very unlikely to care. They are too busy getting the job done to worry about politics.
My 2 cents...
Just my 2 cents.
>>Linus Torvalds, having just completed the kernel back in 1991:
- Great! Now what? Well, let's try to make it run, right?... hmm.. But run what? Oh, lucky me! There seems to be a load of free OS tools already available: compilers, linkers, shells, editors. Every possible tools and servers for make up for a complete OS.
- It's just like they been lying around just waiting for the moment for my kernel to rise and give birth to a complete UNIX-like system! Hmm, it seems all those tools are copyrighted under an odd software license by some organization named FSF... it seems in order to redistribute my kernel and the tools together to make up for a complete OS my kernel will have to be licensed under that very same funky license: the GPL.
- Oh, well... guess i'll have to do it! After all, my kernel alone is nothing more than a technical showpiece without an OS to make use of it and where else can i get all those tools to build my OS?
- I look everywhere around and all free Unix reimplementations i find seem to be from this very same FSF! Heck, who the hell are these guys?!! It's just like they were already trying to buid an UNIX-like system before me!! I feel spoiled...**sob**<<
By the way, yes i am a free software advocate and while acknowledging that the POSIX implementation most people run today is mostly a modified GNU system and that without the efforts of RMS and his FSF, there would be no free, nor open source, no Linux, no reasonto make freely distributable source code and, mind you, an even greater
Micro$oft monopoly today, thanks to no Apache, Sendmail or other competition in the server wars, still, i think while it's valid for RMS to claim for his due credits, calling the system GNU/Linux is a real pain in the ass. Sorry, Richie...
RMS: You're right, but...
...you are overlooking a significant (if generally unrealized) point. In naming something as "Linux" instead of "GNU/Linux", the namer may be simply taking the natural approach - "Linux" is infinitely more practical from a speech/memory perspective than "GNU/Linux". People are syntactically lazy and will simplify to the description they feel is as small as possible yet still adequate (to them) to describe the item being named.
Certainly credit should be heaped on GNU and FSF for creating the mountain supporting the Linux kernel. But a simpler name is needed and LONG overdue. Surely such creative people as are present in these groups can come together with Linus to agree on a mutually beneficial name that is easy to pronounce and remember. Or was that the intention when names like Red Hat, SuSE, and Debian became synonymous to GNU/Linux in the public mind?
The argument begins to feel petty when we have bigger concerns like the DMCA, CBDTPA, and Microsoft's unlawful acts. Work together and find a solution to this sore point, or just drop it. It's not worth the baggage you are adding to the effort.
You were right, RMS
http://kerneltrap.org/node/4966
thanks...
You are mistaken. BK can be used to develop Free software. (capital-f free == FSF-style Free). Using non-Free software in the toolchain you use to develop Free software can be a problem. If you write code that needs to be compiled with non-Free tools, the Free source isn't as useful to other people. (This was the case with Java, but Free java compilers and run-time environments (JVMs) are making progress.)
The BK case is marginal. BK isn't part of the toolchain, so a Free project developed with BK can drop BK and be developed using only Free software at any time by anyone. The project doesn't become dependent on BK. BK is not a building block for anything else, unlike a library, or a windowing system. For one thing, only developers of the project have any reason to have anything to do with BK, and it is possible for developers to work without using BK, or any other non-Free software. For a project as important as the Linux kernel, lots of people are interested in getting at the source code, etc.
I don't know if the GPL implies that source be provided in a way that doesn't require the use of non-Free tools, but that is something to be concerned about. You can always write to the authors and ask them to send you the source. As long as you can get your hands on a Free tarball, you can do anything you like.
AFAICT, the only real arguments against using BK for managing the Linux kernel is that it encourages the kernel developers to use BK, and that it is an endorsement of non-Free software.
The kernel developers can endorse whoever they like. IIRC, the source for BK itself used to be available, and the license was GPL + you can't remove the openlogging stuff. That seems to have changed, however, so now you can't run BK on platforms they don't provide binaries for. That's a big problem, and it makes BK a lot less Free than it was when BK itself was GPL+logging.
Thanks.
don't know if the GPL implies that source be provided in a way that doesn't require the use of non-Free tools, but that is something to be concerned about. You can always write to the authors and ask them to send you the source. As long as you can get your hands on a Free tarball, you can do anything you like.
AFAICT, the only real arguments against using BK for managing the Linux kernel is that it encourages the kernel developers to use BK, and that it is an endorsement of non-Free software.
The kernel developers can endorse whoever they like. IIRC, the source for BK itself used to be available, and the license was GPL + you can't remove the openlogging stuff. That seems to have changed, however, so now you can't run BK on platforms they don't provide binaries for. That's a big problem, and it makes BK a lot less Free than it was when BK itself was GPL+logging.