Slashdot Mirror
RMS Accused Of Attempting Glibc Hostile Takeover
Bram Stolk sent a bit in thats been floating around lately where Ulrich Drepper, glibc maintainer announces the new version, and sidetracks to discuss an an RMS takeover attempt and how he feels about it. He raises several good points and I tend to agree with him. The FSF has done, and continues to do so much good, but more and more tension continues to grow between the extreme free speech faction and the more moderate folks. People have asked my opinion, and I'll just leave it by saying I don't prefix "Linux" with those 3 little letters and a slash even tho I've been asked.
Adding GNU to the front of Linux seems to me a lot like adding "FedEx" to the front of "Super Bowl" or something like that. It just seems like somebody wanting to get their name in lights for doing background work. Maybe I'm strange, but when I do background work, I usually enjoy being credited in the background someplace, not out in front of the people who pull things together in the end...
Posted from the wireless couch.
I will probably be the only here to defend RMS, but I totally agree with Linux being called GNU/Linux when talking about an entire distribution instead of just the kernel. You are ignoring the work of a lot of developers by just calling it Linux. And I know people will say "We should call it Xfree86/BSD/GNU/Linux then." Well Xfree86 and a lot of the BSD code isn't needed at all to get an operating system up and running. GNU and Linux represent the core that is needed.
I'm pretty sure those of us who have met Stallman in person would agree wholeheartedly.
;)]
Despite the fact the cause has some degree of validity, the extremes which he takes it to regularly stomps on people's toes, and is generally antisocial.
I had the (mis?)fortune to meet him during one of his visits to Canberra, Australia - which, over lunch, he proceded to argue that our local Linux Users Group (CLUG) should rename itself to the Canberra GNU/Linux Users Group. This did not go down well.
Even though there are some fairly valid reasons as to why, its still fairly egotistical of him - did he ask for a consensus of all the developers releasing "GNU Software"? Does his own technical work make up a large slice of the GNU works used by linux? [No, Emacs does not count as a large slice, despite its footprint.
Just consider RMS as what he really is, a politican.
The core that is needed to whom? YOU? What if I need Xfree86, BSD, and perhaps other commercial applications as part of my operating system? Then do I refer to all of them?
I mean, surely when I tell people what OS I use, I say Windows 98, not Windows98/Office2000/Winamp/AOL/ATI Drivers/Creative Labs Drivers/Winzip/Acrobat Reader...
I approve of different vendors calling their distributions whatever they want, based on Linux or not. Let Red Hat Linux simply be Red Hat... let them call it Red Hat Linux if they have a Red Hat Windows Compatible OS too. Maybe there's good reasons not to do that either, but I see no reason why Linux should be called GNU/Linux. GNU does not own Linux. And I would laugh if Linus sold the rights off to the kernel one day, as Stallman would be very very screwed...
Okay, lets set this straight once and for all.
What is part of the operating system?
- kernel
- libraries necessary to run C programs
- the most basic interface possible
What is not part of the operating system?
- GUI
- web browser
- office suite
- your mom
Okay... so, you should call linux "GNU/Linux", because GNU tools are a larger percentage of the Operating System itself than even the Linux kernel.
You should not call windows "Windows98/Acrobat Reader" because Acrobat Reader in no way qualifies as a "part of the Operating System".
I have an idea. GNU/Linux is a bit of a mouthful, so let's use an abbreviation. I recommend "Linux".
Patrick Doyle
I mod down every jackass who puts his moderation policy in his sig. Oh, wait a sec....
I use some Gnu tools and many more non-gnu-licensed tools with my linux kernels. I'm not gonna say Gnu/Apache/Perl/BSD/etc/Linux, and neither should anyone else. Yeah, the system would be less useful without gnu tools, it'd also not be what it is without all the other pieces of software on it.
The point isn't that the GNU tools are a major part of a standard Linux distribution.
In the early 1990s, the GNU project had everything you needed for a baseline operating system. Compiler, assembler, linker, C library, shell. Everything except a kernel. Linus took those tools and added the final piece, the kernel. Linus didn't need X-Windows or Perl. Apache didn't exist. Linus needed a compiler, a linker, an assembler, a C library, and a shell. He used the GNU project's tools. Linux is built upon a foundation of GNU tools.
That's why the Stallman can claim the GNU project has a valid claim to share the Linux title. Why bother? Politics. Stallman is pushing a political and ethical agenda. Free Software or nothing. Part of his job is to spread the word, and getting the GNU name used is a great way to do it. Every user who says "What's the GNU thing in front of Linux?" is an opportunity to spread the word.
That said, I'm not sure I agree that it should be called GNU/Linux. It seems a bit pushy to me. But don't make the mistake that he wants it called GNU/Linux just because the GNU tools are part of the typical package. He wants it added to help spread the Free Software word. His claim is that the GNU tools where the foundation.
Search 2010 Gen Con events
A more complete version of the tale can be found in the Contributors file in the ispell distribution. That narration bends over backwards to avoid starting a flame war, so it is quite generous in describing Stallman's actions. But I haven't forgotten his attempts to trick the general public into doing what he wanted (which continue to this day), nor the generally rude way in which he behaved.
Essentially some wording that postures Linux as
an *alternative* to what the GNU project considers
it's 'main' thrust, Hurd? And this fellow didn't
manage to win over enough support within the
glibc project to stop a wording change?
Admittedly, RMS's obsession with this wordplay
seems to me to be unwise (it'd be better if he'd
focus on keeping the movement ideologically pure
rather than focusing on diction), but this fellow
doesn't strike me as being any more wise.
Posturing and replies to it are a waste of time --
where there are no actual effects on the way
things are run, it's better to just ignore such
things and spend that time coding, fighting
intellectual property, and other worthwhile causes.
For every problem, there is at least one solution that is simple, neat, and wrong.
I really admire RMS, but I have to say, he does go off the deep end to the point that he may be doing more to HURT the FSF than help it.
I understand that free software is as much a political movement as it is an idea for better software. However, RMS seems to be HOSTILE to those who don't make the same choices he does. Freedom to me, means, that, freedom. It's about having the freedom to make good or bad choices.
The KDE controversy, and this takeover attempt on GLIBC etc, makes him look more like a raving lunatic, and by extension, makes ALL of us who support the principle of the GPL and open source look the same. Why? Because Stallman proclaims himself the leader of the whole movement whenever asked, or not asked.
While I have tremendous respect for the man, and his philospohy, his despotic style runs contrary to the whole anarchistic nature of free software. RMS needs to realize that not EVERYTHING needs to be called "GNU/".
=== The price of freedom is eternal vigilance
Never mind that when I purchase or download a Linux version 70% or more of the included software is GNU. Right?
Being under the GPL is not the same as being GNU. I've written stuff and released it under the GPL and I'll be damned if anyone is going to tell me that the FSF deserves naming credit for my software.
NO CARRIER
But Linus Torvalds did not take an existing unfinished GNU operating system and merely add the missing piece. Yes, GNU could have taken the Linux kernel and completed its OS, but they did not. And it's not what Linus did.
The real, unrevised, history is very different. Linus started with the goal of creating a complete operating system. Once he got the kernel and a few bits of infrastructure done, he and his collaborators chose to use off-the-shelf parts already available to complete it. Some of those parts were from GNU, but many others from elsewhere. And many of the crucial components were written *specifically* for Linux.
To use an analogy, imagine that RMS set out to create an automobile. He was all finished except for the engine. Now Linus comes along and builds an engine. He goes and grabs a drive train and chassis from GNU Autoparts Store, and an electrical system from BSD. He and many friends contribute to the miscellaneous components. Voila! It's an ugly car, but it works.
GNU does not get to name this automobile. They did not build it. They only supplied some critical parts.
A Government Is a Body of People, Usually Notably Ungoverned
Why define an OS that way? It's just dumb. The OS at its most basic form is a command interpreter of some sort, which can be just a kernel. Why specify that an OS must contain libraries that can run C programs? Don't people use languages other than C? And how about straight assembly? If someone writes something that functions as an OS, but it doesn't have C libraries and must be coded for by other means, it's still an OS. Or are we going to start arbitrarily defining things by what languages and libraries they use? Doesn't a piece of software serve the same function, whether it's written in C or Java or Python or whatever? Then stop defining an OS by a compiler or a library. If you want to see an OS at its most basic, just put an ancient DOS command.com on an empty drive, along with whatever text config files that version of DOS will require to load itself. Sure, DOS usually has other files for "external commands" and for access to upper memory, etc.,--but they aren't necessary to do the absolute basics.
It's just stupid, and besides many people use OSes who wouldn't use a compiler and wouldn't know or care what libraries their software is linked to. An OS, like it or not, is defined by its kernel.
Let me lather, rinse, and repeat: an OS is defined by its kernel. And here's where I prove it: If I run a Solaris box and install and link to a bunch of GNU stuff, does that magically transform my OS into GNU/Solaris? NO.
This is why I think Stallman should be largely ignored now that he has already made his historic contribution of the GNU tools. He will go down in history for that accomplishment. But at the moment he's a hindrance, not a help. He has passed his prime, made his contribution, and is now being a petty bitch who squabbles about naming an OS he didn't write. He is actively trying to harm Linux, what with his devotion to the HURD. Anyone who doubts this, should read the post referenced in this story where the Linux glibc porter/maintainer states that Stallman tried to push him into working on glibc for Hurd instead of for Linux.
It should be obvious that Linux is RMS's "bastard child"--it's the first OS born from the GNU tools, and it has made the Free Software movement what it is today as well as helping spawn Open Source. Without Linux, Free Software would still be a tiny little movemwent instead of being on so many desktops and servers. Yet Stallman doesn't care about Linux, he cares about finally building the kernel for his GNU/HURD dream and eventually putting Linux out to pasture. And that's fine. But don't be a schmuck and think Stallman cares about Linux or should be listened to about a damned thing that has to do with Linux. If it were up to him, all Linux developers would drop their work and start on the Hurd. Things like the attempted coup mentioned in this story just go to show that RMS is slowly sabotaging Linux, in order to promote his Hurd. And before marking this as flamebait, at least read the account linked in the story.
Chasing Amy
(We all chase Amy...)
"The more corrupt the state, the more numerous the laws"-Tacitus
That's a contraction due to length constaints of RMS's actually sig in a letter to The Register.
Sincerely,
Richard Stallman
Principal developer of the operating system often inaccurately called "Linux"
http://www.theregister.co.uk/content/4/18291.html
No, no glory grabbing at all, nothing to see here, move along...
Tastes Like Chicken
RMS may like control, but look at this case we can see Drepper is the one with a bigger ego than RMS. Let's look at the facts:
1. RMS is accused of taking over the control of a GNU project. Not mentioning that RMS probably started the glibc project and contributed code in earlier years, how has RMS tried to control glibc? Does RMS decide, say, how glibc should be written? I don't see that. Drepper is in full technical control.
2. The only place where Drepper is unhappy about seems to be the "GNU/Linux" mentioning in LGPL 2.1. Otherwise LGPL 2.1 and 2.0 are about the same. The licenses give the same rights to users. Drepper makes a big deal out of a naming issue which is not even part of the actual license requirements. And glibc being a GNU Project, switching to LGPL 2.1 seems ony natural. Just a routine step.
3. Drepper seems unhappy about the creation of a SC. He accuses the SC was an attempt to steal the project. From him. Now, who is the one wanting control here? The SC is a more democratic way to run a project than a single maintainer. At least the other contributors have more say than letting Drepper decide everything.
4. Drepper wants control, which can be seen by his handling of the gcc 3 issue. Drepper disagreed with gcc developers (many of them) on certain technical issues over gcc 3. He once declared he would never accept patches to make glibc capable of being built with gcc 3. Despite other glibc contributors' attempts to find a solution, he just says, "NO, I won't accept any patches". This issue does not involve RMS at all, and Drepper just goes against many gcc developers, who are perhaps some of the smartest compiler people in the world. It is hard to say that Drepper is right and all these gcc people are wrong.
RMS may like control, but in this story Drepper is more of a control freak and has a bigger ego.
Free Software: the software by the people, of the people and for the people. Develop! Share! Enhance! Enjoy!
Think about what RMS does, and why -- he has a passionate belief "software should be Free", and expresses it in a fairly consistant manner. He fights a somewhat unpopular fight with little reward -- outside of recognition within of a small, tight-knit community, which isn't much.
RMS has been fighting this fight longer than some GNU/Linux nerds have been alive. He had the vision to kick the thing off in the first place. His reward? A string of Slashdot readers questioning his relevency, sanity and parentage.
While I may disagree with some of RMS's views (I get the occasional whiff of Unreconstructed Socialist from some of his writing, and nobody hates a commie (or a socialist) more than me), I have the utmost respect for his work, and I'm thankful for it.
While Ulrich may have a genuine beef with RMS, waving it about in public (and Slashdot posting the story) is not very professional, nor productive.
Potato chips are a by-yourself food.
Nah, RMS only developed the bits that suck. The cool stuff was developed by some Finnish guy.
If Stallman wants proper credit given to GNU/Linux, and he contributed so much to GNU, I say we combine Stallman and Linux to give both central figures credit to the OS.
We will now call it "Stallinux".
D'OH!
The basic idealogical dispute is that previously it was illegal to link glibc with proprietary software linked by non-GNU compilers due to a special "modified GPL" in the libio section of the GNU C Library. The change that the steering committee (who are developers like Roland McGrath, not just "Stallman") made was primiarily to convert that code to LGPL. Ulrich was the one being an idealogue about it. In this case, the steering committee was the group that was actually trying to get the right thing done for the users.
The glibc-2.2.4 announcement advised everyone to switch to it. What the announcement did not mention is that if you try to configure glibc-2.2.4, you discover that it does not want to build under gcc-3. The steering committee is pushing for a fast release of glibc-2.2.5 which will not have this problem.
So far, the steering committee seems to be a very positive influence. In the past, people were giving up hope on glibc due to its bloat, arcaneness, and legal issues. The SC seems much more focused on what users want.
By the way, let me say that Ulrich Drepper has made many contributions to glibc and I hope he will continue to be involved as a contributor.
Speaking of intelligent discourse... Am I the only person who finds it ironic that the primary reason the BSD license was incompatible with the GPL was its advertising clause? (You know, that clause that says that people who derive their work from the BSD-license-covered source must advertise that fact by saying "Contains code developed by so-and-so"...)
And yet, isn't that what RMS is asking of the Linux community? That is, for us to slap "GNU inside" on our Linux boxes?
Oh, the irony...
--JoeProgram Intellivision!
How about this:
"The glibc situation is even more frightening if one realizes the story
behind it. When I started porting glibc 1.09 to Linux (which
eventually became glibc 2.0) Stallman threatened me and tried to force
me to contribute rather to the work on the Hurd."
That's how it's supposed to be bad. If you look at ESR's recent article, he says that developers should have the freedom to do what they want. If i'm not reading this wrong, Drepper is the maintainer of glibc, and so should decide what goes on - if he has a plan for how it will work and evolve, and it's his project, then he should have the right to have the project follow his plan, and not be taken out of his control.
He also says:
"I find this completely unacceptable and can assure
everybody that I consider none of the code I contributed to glibc
(which is quite a lot) to be as part of the GNU project and so a major
part of what Stallman claims credit for is simply going away."
If he's an important contributor, and the project maintainer, I think he has every right to control the project. He is not a control freak just because he wants his project to be given to someone else! This doesn't apply to all situations, but for some projects it's good to have one person in control who decides how things will work, and controls the overall architecture and the project in general.
They that quote Benjamin Franklin on liberty and safety deserve neither.
You (I assume you are Geoff Kuenning) leave out a lot from that story...
1. The version wasn't merely not "GPL-compatible", it wasn't open software either. Specifically, it did not allow for-profit distribution alone.
2. People suggested removing these restrictions to you was vicioucly flamed, you wouldn't even accept that these restriction existed. This might be the cause of the "misunderstanding".
3. ispell 4.0 was not derived from your code. It was derived from the code of _original_ ispell author (i.e. not you), who had assigned his code to the FSF. Specifically, it lacked all the i18n features you had added.
It is true that FSF withdraw[1] ispell 4.0 as soon as ispell 3.x was released under a free software license. I think that makes it pretty clear that the action was in defence for free software, not an attempt to increase their control.
[1] As far as one can withdraw alreeady released free software -- ispell 4.0 still have DOS/Windows users as version 3 was much harder to port to DOS. This, b.t.w. is still a cause of confusion about what version is newer. Something that could easily be solved by releasing a version 3 derivative as version 5. That would require someone to be more pragmatic and less determined about whose fault it is, though.
Read the licenses carefully and rip out parts which give Stallman any possibility to influence your future. Phrases like
just invites him to screw you when it pleases him. Rip out the "any later version" part [...]
And sure enough, it wasn't part of the License itself, but of the (although suggestive) part on how to apply the license to your source code. In the License it says:
This clause seems a little strange at first, and note, that you can restrict the licensing of a library to a specific version of the LGPL (although it's not explicitly said so you can do so by specifying the LGPL-Version). I think there is a good reason for using that option though, as long as one assumes, that the LGPL and the GPL will stay the same in spirit (the [...] part in above quote): what if you merge two libraries or use part of one library with part of another, soon you'll probably find all versions of the LGPL applicable to different parts of the code. Also an upgraded Version might close some loopholes of previous ones, so if you trust the FSF to do the right thing with the LGPL it's probably a good thing to leave the option of a License upgrade open to later developers. And anyway, as long as one person or group of persons keep control of a project (in the sense of being responsible for it) it's their choice, what specific licence the actual code ships with.
I also noted, that (3) allows to elevate LGPLd code to GPLd code. Again this makes sense, in the case that you want to use LGPLd code in a GPLd project (but not vice versa, which wouldn't make sense anyway, since that would 'degrade' the GPL to the LGPL). I think these paragraphs are in there for convenience's sake and not to give RMS total control over anything GPLd. Anyway, ripping the first quoted snippet out of context and using it to picture RMS as a controlfreak is, in my opinion, bad style. RMS often enough comes through, well, overenthusiastic, to say the least. The "GNU/Linux" vs. "Linux" debate doesn't help that either, but let's be honest, Linux wouldn't be what it is today, hadn't the GNU Software and the free software idea already been in place.
"By the way if anyone here is in advertising or marketing... kill yourself." -- Bill Hicks