Slashdot Mirror
The Stallman Factor
An anonymous reader sent us linkage to a LinuxWorld story about Stallman's Position
in the Linux World. Talks a lot about RMS's tacticts for getting his
acronym included with the kernel's name. This has been a long-running debate,
but personally I just don't care. I respect the GNU Project's involvement. But
I'm not gonna spit out extra syllables and keystrokes just to appease anyone.
Philosophically, Stallman is as far on the extreme left of the software world as PETA is in animal rights or the CP-USA is in politics. While not outright advocating total software anarchy now, he certainly wouldn't object to the idea if it happened. The problem is that while I somewhat understand his desire for credit for GNU, he's gone about it the wrong way, attempting to coopt an essentially non-political project (at least to its creator) for his own agenda.
Like it or not, Linus ain't in it for the politics; that's just a collateral benefit of having a free, high-quality kernel. Stallman is just another extremist with a useful ideal but no practical value.
No matter how controversial, the community needs RMS. Where Linus has openly admitted Linux is not the most important part of his life, RMS remains the prophet while Linus settles down with family life. Stallman is the perfect example of a person who is unwavering in his beliefs that all software should be free - to the point where some believe he defies common sense (e.g. coders should work as waiters to pay their bills). But he is genuinely passionate about open source, and for that we can all learn something from him. I know I am not looking forward to the day RMS is unable to continue his mission with the open source movement.
If Bill Gates had a nickel for every time Windows crashed... Oh wait, he does.
Every major free work can trace its roots to GCC (which Stallman originally wrote): Linux kernel, KDE, GNOME, XFree86. All the modern free *BSDs would be crippled without GCC as well.
He fought for ideals that may be fashionable now - but were hardly so 20 years ago.
He is a visionary. His work will benefit rich and poor alike, large corporations and third world nations.
We are all in Richard Stallman's debt.
Of course, Bitkeeper is a relatively minor issue. If, fortune forbid, the Linux project were to face a substantial crisis in of some sort, hopefully he'll have the balls to act the same way.
ESR is just some sort of leeching gasbag.
Where does it end, and why there? Please be as precise in your answer as Mr. Stallman prefers to be in his.
sPh
Hard-driving people tend not to be like that!
They promote their causes, their organizations, even themselves.
Because if they don't do it, they tend to get run-over by others who are doing it.
Stallman is the CEO of a foundation. Compare him to other CEO's of foundations, and see how he ranks then. But it's not an easy job.
Sig: What Happened To The Censorware Project (censorware.org)
I think there is a group that fits in between somehow. The group that uses it because it is free (as in beer) and because it does what they need as well as Windows. This is the typical business person, someone weighing the costs and benefits of each decision. Like it or not, that is the person that Linux, and OpenSource in general must be targeted toward. If we keep this "two camps" mentality, we will alienate the middle group and drive away potential users. We must avoid polarizing the OpenSource community.
THIS SPACE FOR RENT
Weird. By about a third of the way through that paragraph, by brain had adapted, so that I was up to normal reading speed halfway.
Kind of like any loud, annoying, repetitive sound (or person), you get used to it... or ignore it.
I watched C-beams glitter in the dark near the Tannhauser gate.
This is one of the most intelligent comments on the issue I've seen in a long while. Nobody's ever asked me my position on "Free Software versus Open Source", but the answer I've had prepared for a long time in case anyone ever did was along these lines:
Free software advocates claim that this kind of software is morally superior. Open source advocates claim that this kind of software will likely (in fact, inevitably) end up better quality and provide more benefits to both the user and the developer. I say that both of these are true, and therefore I support both. I don't feel a need to identify with one of these exclusively or even to prioritize one over the other, since both ends are achieved by the same means. My position is just to be happy that there are two important benefits from this kind of software!
Now, if the goal of getting the most benefits from my software contradicted the goal of taking the moral high ground, I honestly can't say which I would choose. But I had a hard time even phrasing that last sentence, because it's almost inherently contradictory - the moral goal of freedom is inherently what produces the benefits that I consider important. So it's impossible to separate the two and even more impossible for them to ever contradict each other.
We need someone who understands the art of marketing - the art of diplomacy, rhetoric, whatever you want to call it. Part of this includes choosing good names for projects and concepts. "GNU/Linux" is a horrible name. "Free Software" is even worse. His insistance on spreading the MIT-centric idea of what the word "hacker" is supposed to mean is incredibly counter-productive. Changing the way people think involves changing the way they speak, but attempting to redefine new words does not work, especially in the most critical first stages of a movement.
"Hacker" means criminal to everyone outside our movement and associating yourself with this word is not the way to spread your ideology. The operating system that runs Slashdot and Google is known as "Linux" to everyone outside our movement, and changing that only spreads confusion.
The one that really bothers me is "Free Software." Here, the capitalization makes the difference between respected projects such as GCC and the crap shareware you'll find on tucows.com. I completely understand the whole "gratuit" vs. "libre" thing, being an amateur linguist and having read most of the FSF's philosophy. I can understand that "Free" and "free" are supposed to mean different things, but I don't like it.
What if I'm talking to someone? Much business is done over the telephone or through face-to-face meetings, not through email. With email/usenet/etc. it's easy to spot the difference between "Free" and "free", but what if I want to tell my boss that the new proxy server I installed is "Free"? Do I say:
(a) "It's Free software, as in free-with-a-capital-eff.", or
(b) "It's Open Source."
(A) will lead only to confusion because my boss doesn't care about the tiny differences between copylefted and Free-but-copyrighted software. Don't fool yourself into thinking that (a) will lead to a discussion about the nature of Free software; the business world is not academia, and has little patience for discussions which (seemingly or in reality) do not affect the bottom line.
Look at DivX versus DivX ;-). Look it up on Google if you don't
know the difference between the two (hint, the punctuation makes
all the difference). Now that DivX ;-) has become popular and DivX
is dead, the DivX ;-) people are trying very hard to break away
from their old name. What was initially a bad pun has become the
bane of this company attempting to sell their codec. Don't look
at this example from a Free software standpoint; the example is
just meant to show how a bad name can hinder the acceptance of a
codec/idea.
We (computer professionals or hackers if you insist) linguistically think much differently from most people. To me, puncuation and capitalization are as important as words. Compare:
(c) "DivX" to ;-)", and
(d) "DivX
(e) "find name copying and print" to
(f) "find / -name COPYING\* -print"
You and I know that (e) is gibberish and (f) is correct/useful, but Joe Schmoe only knows that (e) is "easier" than (f). Joe Schmoe has to "remember" (f) in its entirety in order to use it; he has difficulty deconstructing it into its component pieces and then reconstructing the pieces together again the next time the command must be used. Those of us linguistically gifted (all good programmers) do this instinctually, so we have little difficulty figuring out "DivX" versus "DivX ;-)" and "Free" versus "free". Most people
don't have the time or patience to do this.
This has nothing to do with ideas; it's all about marketing. Microsoft has some of the best marketing people, so we would do well to take an example from Redmond. Why did Microsoft change Windows NT 5.0 to Windows 2000? Joe Schmoe thinks like this:
"Oh, 5.0, but I'm already running 98, so is that an upgrade?"
"Oh, they came out with Windows 2000, and I'm only running 98."
"Oh, I can get IE for free, so it doesn't mean anything that Netscape is now free software."
"Oh, Netscape is now Open Source, so I'll download it because it gives me more freedom than IE."
In this case, Microsoft was deliberately trying to create confusion to increase the acceptance of NT 5.0. We have no interest in deliberately creating confusion (right?), so we should have a name for our software that immediately and succintly distinguishes it from gratis-but-encumbered software. Go to tucows.com and you'll see that most people (most people run windows) interact with gratis-but-encumbered software much more often than Free software.
Let's look at Microsoft's marketing strategy with NT 5.0 versus 2000 again. Syllabically, we have:
win-dows en-tee five-point-oh
win-dows two thou-sand
The renaming cut out three syllables. Coincidence?
Also, look at the components of the words:
2000
NT 5.0
The first is a simple year (easy to remember), while the second is an obscure acronym combined with an obscure version number.
When my mother first took some computer classes, she came back home to ask me what version of Windows her PC was running. I replied "Windows 3.1." (I cut my teeth programming with Borland C++ 3.0 on MS-DOS 5.0/Windows 3.1.)
She said, "No, that's not right. What version of Windows is my PC running?"
Me: "3.1."
Mom: "No, that's not a version of Windows."
Me: "Yes it is."
Mom: "Then what version of Windows is the PC at work running?"
Me: "Some are Windows 98 and some are NT 4.0."
Mom: "???"
Me: "Microsoft's versioning schemes divulged with the release of 'Chicago', aka. Windows 95. Whereas versions of Windows prior to Windows 95 were simply given a version number, Windows 95 and later were given a common name which refers to the year of intended release in addition to a normal version number. Thus, Windows 95 is in actuality Windows 4.0 and Windows 98 is Windows 4.5. Windows NT, on the other hand, is a completely separate product line. Windows NT 3.5 was the contemporary of Windows 95 and Windows NT 4.0 is the contemporary of Windows 98."
Mom: "So what version of Windows is my PC running?"
Me: "Umm...Windows 93."
The point of this who's-on-second? People want brevity and clarity, not philosophy or linguo-technical mumbo-jumbo.
In short, we need a name for our movement that:
ESR is just some sort of leeching gasbag.
ESR still writes code to this day. Ever browse Sourceforge's Trove? He created it. Compile a recent kernel? The piece that figures out the complex kernel configuration dependency logic is his. Then there's fetchmail. of course (not that it was really worth writing a software engineering theory paper over). He maintains the Jargon File. And he probably has more elisp contributions to emacs than RMS. Just about everywhere I go, I see something with ESR as a contributor.
But you weren't really interested in the truth, were you?
I've finally had it: until slashdot gets article moderation, I am not coming back.
RMS wrote GCC. From scratch. By himself. GDB too. That's not the same as what Hopper did (which was paid research). It's not easy to appreciate what RMS went through to do this, especially given when it was done (mid 1980s).
Linus Torvalds wrote a kernel by himself, with very little usefulness (but heaps of promise) in the early 1990s. He was able to use the tools created by Stallman.
John Carmack (think Commander Keen, Wolfenstien, Doom, and Quake) claimed that he wouldn't have been a programmer had it not been for the tools created by Stallman. Once after Carmack won a large jackpot in Vegas (I don't know how often he does this =-), he donated the whole thing (>$10,000 I believe) to the Free Software Foundation (i.e. Stallman's group).
In the battle of the lisp machines, Stallman was afraid the highly-non-free side was going to win. To provide balance, he recreated their features and donated his code to the more free alternative company. He did this in real-time, by himself, unapid. His output equalled the output of a collection of commercial programmers hired from Stallman's lab (and others).
So GCC starts at Stallman. Free Software starts at Stallman. GNU starts at Stallman. The Open Source definition came about because of Stallman's work (and to some degree because of his contrariness and Tim O'Reilly's stupid decision not to invite RMS to his west coast summit that settled on the Open Source name). Stallman is where GCC started. Not Hopper, not Lovelace or Babbage, not Boole or Aristotle.
We're talking about running code that is still in widespread use after nearly 20 years, not paid research that was eventually perfected by others to resemble what we today think of as a compiler. Stallman is *the* person who started all of this, by himself, on his own time, taking part time jobs to survive until sufficient donations came in (for instance, the MacArthur Genius Award).
Stallman did this because of his ideology. Linus Torvalds' comments about the world being better with less ideology really seem stupid in this context, don't they? Torvalds' comments about only idiots or freaks or something choosing Free tools over superior proprietary tools really says something about Torvald's view of Stallman, doesn't it? Does Torvalds recognize that quality Free Software would not exist if everyone thought like that? Of course Torvalds has a right to his own opinions, but I wish he'd keep his mouth shut instead of revealing how shallow he is.
-Paul Komarek