Slashdot Mirror
The Re-Unification of Linux
In the wake of the wildly successful Red Hat IPO stories mooting the possibility that Linux might `fragment' under corporate pressure seem to be proliferating. The memory of the great proprietary-Unix debacle of the 1980s and early 1990s is constantly invoked -- N different versions diverging as vendors sought to differentiate their products, but succeeded only in balkanizing their market and inviting the Windows invasion.
But amidst all this viewing-with-alarm (some of it genuine, much of it doubtless seeded by Microsoft) something ironically fascinating is happening. Unix is beginning to re-unify itself.
SGI's recent decision to drop IRIX and focus on Linux is one telling straw in the wind. Another is SCO's launch of a Linux professional-services group, clearly a trial balloon aimed at discovering whether SCO's branded-Unix business can be migrated to a Linux codebase. I visited a Hewlett-Packard R&D lab last week, and learned that many people there expect HP to deep-six its HP-UX product in favor of Linux in the fairly near future.
What's causing this phenomenon? Open source, of course. Whoever you are -- SGI, SCO, HP, or even Microsoft -- most of the smart people on the planet work somewhere else. The leverage you get from being able to use all those brains and eyeballs in addition to your own is colossal. It's a competitive advantage traditional operating-systems vendors are finding they can no longer ignore.
Playing along now and trying to defect later won't work either -- because running away from the community with your own little closed Linux fragment would just mean you didn't get to use those brains any more. You'd be swiftly out-evolved and out-competed by the vendors still able to tap the literally hundreds of thousands of open-source developers out there.
What we have now have going is a virtuous circle -- as each of the old-line Unix outfits joins the Linux crowd, the gravity it exerts on the others grows stronger. The Monterey and Tru-64 development efforts, the last-gasp attempts to produce competitive closed Unixes, can't even muster convincing majorities of support inside the vendors backing them; both IBM and Compaq are investing heavily in Linux.
Linux fragmenting? No way. Instead, it's cheerfully absorbing its competition. And the fact that it is `absorbing' rather than `destroying' is key; vendors are belatedly figuring out that the value proposition in the OS business doesn't really depend on code secrecy at all, but instead hinges on smarts and service and features and responsiveness.
These are all things the worldwide community of open-source hackers are really good at supplying. Vendors become packaging and value-add operations that never have to re-invent the wheel again. Customers get better software.
By joining the Linux community, everybody wins.
--
Eric S. Raymond
Well, I'd rather not be part of a group of people that have an exclusionary and/or elitist attitude. I have tried Linux (I installed Slackware 3.0 lats year), but I wasn't impressed enough by its technical merits to continue using it. The RMS philosophy and the GPL intrigue me, but they're counterbalanced by the rest of the crap I see in the "Linux community." On just technical reasons, ignoring any ideologies, I see no reason to switch. Sure, Windows sucks in many ways, but so does X. Windows crashes, but it has a usable, consistent interface. X doesn't crash nearly as often (though XF86Setup required me to reboot a few times), but its interface pretty much sucks, and is not consistent in the least.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
As far as I'm aware, development of IRIX may be slowing, but it's far from stopping.
Oh, despite ESR's tendency to assume all things good are a result of open source, it's a damn fine article.
I'm not sure why, but I never had to do that. make install took care of everything for me.
--
Interested in XFMail? New XFMail home page
You *did* say that you were considering FreeBSD, didn't you? :-P
But seriously, running an operating system doesn't mean you have to be part of any particular group. You just use it, and get on with your life.
I just don't get it when people say, "I won't run , because advocates are jerks. In case you haven't noticed, advocates of all OSs are mostly jerks (probably including myself). So, by your standard, you'd have to go back to paper and pencil.
Who cares what people who advocate Linux say? Measure it by its usefulness, not by its advocates.
If you prefer Windows, that's fine. I'm certainly not going to tell you that it doesn't suit your needs; it just doesn't happen to suit mine. As others will no doubt point out, X has no interface, since it is a protocol. There are various interfaces that you can use via X, including GNOME, KDE, WindowMaker, fvwm, and, of course, twm. KDE 2.0, slated for early next year, will probably be the most advanced UNIX desktop environment to date. Check it out when it's released. In the meantime, save often. ;-)
--
Interested in XFMail? New XFMail home page
+I won't run (os), because (os) advocates are jerks.
Tried to use angle brackets. Arrgh.
--
Interested in XFMail? New XFMail home page
ESR is right in that the huge number of *nix variations are slowly being abandoned. Over the years there have been hundreds of *nix variations, and it got to be ridiculous to try and support an application on more than a few of them.
Its a good thing the *nix vendors realize there is more money to be made in service and support, rather than tricky features and special proprietary hardware. As more of them are being absorbed by the OSS model, they realize exactly where the profit comes from and focus on it.
It would be a bad thing if there were too few *nix variations, as many knowledgeable slashdotters point out whenever there is a melissa style virus sweeping thru the media. If there were only 10 or so variations of *nix just like there are only 10 variations of Windoze, then an exploit could hurt many more people with less effort.
I doubt there will ever be only 1 version of unix in the future, but it would be nice to see no more than 20 or 30, with most of them touting their adherence to a common standard for libraries and structure.
the AC
Hemos is like...sci-fi fans;he thinks technology is cool, but he hasn't bothered to understand the science it's based on
I am amazed at all the Microsoft FUD. Microsoft has what, now, 30 people who do nothing but read Slashdot and spread FUD?
I work for a software vendor. We make commercial software for Linux. It works on all distributions. It ain't pretty to make it work on all distributions (we basically have to distribute a statically linked copy, along with a dynamically linked copy in order to comply with the LGPL), but so it goes. We run our entire internal infrastructure off of Linux, and our developers have various Linux distributions (heck, my desktop is FreeBSD!). WordPerfect and Applix are our internal word processor and office suite, and both work on every machine in our office, even on my FreeBSD box.
In other words, we're talking pure FUD. Yes, it takes a bit of care to make your software work on all Linux distributions, but a commercial vendor can do it without much problem. WordPerfect does it. Applix does it. BRU does it. If vendor X doesn't do it, that's vendor X's problem, not Linux's.
-E
Send mail here if you want to reach me.
ESR's comments here truly outdo the negative attitude seen following the RHAT IPO. In fact while taken in comparison with suck.com's article a few days back, it shows how the community is (or at least should be) reacting to the IPO vs. how the rest of the world views what occurred in the community during the past 2 weeks.
He concisely addresses the whole "shareholder demands" argument by showing that these publicly owned companies are seeing that the advantage in adding to the unix codebase via the linux community.
I argued the other day (in response to the suck article) that shareholders outside of the community don't mean squat in the matter of development. And this is precisely due to the way that linux evolves. However, I do believe that shareholders within the community now realize the importance of their contributions since it breaks down monetarily.
Finally, I believe that the end result, once we've looked past the IPO, will be more of the same. And this is good. The group that did not get the letter will still (hopefully) continue to contribute. Some naysayers say the contributions will be due to the promise of tomorrow's IPO and this may very well be the case for some. But I say the contributions will continue since people enjoy contributing.
If RedHat or any of the other companies must develop something to meet the demands of shareholders, then the product must also meet the demands of the community for two reasons.
1. It must be useful for the community for our own reasons or adoption will not occur, causing the doomed fragment to be weeded from the standard Linux distribution.
2. It must be well developed within the community or someone will be compelled to develop something else to compete. And the competing project may indeed have an advantage simply due to the "anti-establishment" vibes that are prevalent within our group.
Well, I'm glad to see another article in which I can agree with ESR. Sometimes they seem far and few between.
And that's my whole take on things.
ALL HAIL BRAK!!!
You did okay on your FUD Method #1 (exaggerating weaknesses) with the "mediocre device support". But you need to apply some FUD Distraction Methods for the FUD #2 (outright lies) where you state "no meaningful GUI", since Linux has at least two meaningful GUI's (GNOME and KDE). I suggest that next time you try more extensive "Sandwiching" (distraction method #1), preferably by using FUD Method #1 to attack various attributes of those GUI's. Same goes with the 'ages of cruft' and 'so-so performance', you really need to use some distraction methods to make your FUD stick. If you're stuck with how to do that, go to Microsoft's very own "Linux is a poor value proposition" page, which is a masterful blend of FUD#1 (exaggerating weaknesses), FUD#2 (outright fabrication), and FUD#3 ("spinning" a strength as a weakness).
Sheesh, how much is Wagged paying you anyhow? Whatever it is, it's too much, 'cause you're doing a LOUSY job of FUD! I know you can do better, after all, your firm did an excellent job on the "Linux is a poor value proposition" page...
-E
Send mail here if you want to reach me.
Go to http://www.estinc.com
Yes, our shrink-wrapped commercial software product will run on every commercially available Linux.
I am happily running my 1997 vintage Applix Office on SuSE, Red Hat 6.0, and FreeBSD with no problems. So it's not just BRU that runs pretty much everywhere.
Next FUD, please!
-E
Send mail here if you want to reach me.
Linux is fragmented moreso than any commercial flavor of Unix ever will. (snip) We have GNOME and KDE (need I say more?). umm.. Yes. You need say more, as I don't understand what you're getting at. KDE and Gnome, while different beasts, interoperate pretty damn well. I recently installed Gnome on my (previously) KDE system, and had no problems at all. KDE software ran very well in Gnome (including the KWM!) If you prefer to code for KDE, use QT, if you prefer Gnome, use GTK. Software from one will run in the other.
Why the G3, anyway? I've never quite figured that out... I thought it was called 'Mesa' or 'Mesa3D'. It's v3.x right now. Maybe that has somethng to do with it.
---
"'Is not a quine' is not a quine" is a quine.
"'Is not a quine' is not a quine" is a quine.
Quine "quine?
I'd have to disagree with that. The main allure GNU/Linux has for me is that it's free software. On its technical merits alone, I don't like it. I've used it before (Slackware 3.0), and I was not impressed. X especially is pretty shitty for a windowing system, unless you need to do networked stuff (which I don't). The whole "configure everything by editing textfiles" thing doesn't impress me either.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
Well, the main reason I was going to use Linux in the first place was because of it's advocates. The GNU project and free software seem like worthwhile causes. If I ignore the advocacy from both sides, then I really don't have a need to switch. I don't need extensive uptime, and this box isn't a server.
As for X being only a protocol, yes, but it's also the foundation of the whole windowing system. Many of the problems are caused by X itself, and the various window managers try kludges that sometimes work around them, but usually only partially. Even something as basic as cut and paste proves problematic in X, while a 1984 Macintosh can cut and paste between apps without any problems.
Lots of other things seem strange, nonintuitive, or just downright dumb to us Windows users. Why can't you configure X within X itself? What's with the separate XF86Setup? Why do you have to run XF86Setup to run X? Why doesn't X have good auto-detection routines and some decent defaults so you only need to run XF86Setup if you wish to further customize X? Why is installing a new kernel an 8-step process? Why isn't there a decent archiver (one that lets you extract a single needed file, like zip, rar, arj, zoo, ace, etc., rather than tar.gzip which requires you to extract the whole thing)? Why isn't there a decent simple text editor? (no, pico doesn't count, and if you consider vim "simple" you're insane) Why is ppp so damn hard to configure?
I can think of a few more, but that's enough for now.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
All Linux distros use the same kernel.
No they don't. Red Hat Linux distros use a custom Red Hat kernel. Usually, it's an older kernel than the current latest "stable" kernel, but with some of the newer features and bugfixes added in, to make for a truly stable kernel (the "stable" kernel tree itself is somewhat of a misnomer).
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
I'd say ESR is one of the main reasons I'm still using Windows rather than Linux. Articles like this, filled with half-truths, omissions, and outright lies are what's kept me away. Of course, ESR isn't the only guilty party, much of the Linux "community" behaves likewise.
IRIX is not being dropped, nor is it being replaced with Linux. IRIX is still being supported and developed for SGI's high-end servers, which Linux cannot, and most likely will not, run on. Linux is for low to mid end computers, not enterprise-class servers. That's what IRIX is, and will continue to be, for.
Linux is not "re-unifying" UNIX. There are still many different fragments of UNIX, ranging from Linux to FreeBSD to Solaris. The various BSDs seem to mess up ESR's arguments, so he just omits them. Typical.
Anyway, RMS's writings had almost convinced me to switch to Linux. Bruce Perens has done a good job as well. Unfortunately, the rest of the Linux community, along with ESR, has done the opposite. That, and the fact that I REALLY dislike X, is going to keep me in Windows, at least until I get some spare time to install FreeBSD.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
ESR really needs to check his facts before he goes spouting off.
That's what I've thought after nearly every single article of his I've read. Apparently he'd rather generate good PR than be accurate and truthful.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
I wasn't implying that the Red Hat kernel was different because it was a different version. It's not even part of the Linux kernel development tree. It's not *any* kernel that you can find on ftp.kernel.org, but a custom kernel of their own, that differs not just in how it was configured, but in the code itself.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
Eventually, the Linux kernel will be kept alive by corporations who has an interest in the kernel because they can make money off it. These companies might be working together to reunify Unix, but we'll see some fragmentation between companies and the bleeding-edge hackers. And I think we'll see this very soon.
Open Source is anything but central control. It means nobody can force a single vision on the market, anyone can branch out at any time. Control is totally given over to market forces (i.e. the users).
However, as long as the individual need of a majority of the users are better served by options in a single development tree, that is what most users will get. When the users are better served by divergent trees, that will become more widespread.
That is the difference between free software and proprietary systems. With free software, control is in the hand of the users. Including control over when to fork the project. With proprietary software, control is in the hand of the company owning the software.
Thus, your response had nothing to do with the message you were responding to. :)
---
"'Is not a quine' is not a quine" is a quine.
"'Is not a quine' is not a quine" is a quine.
Quine "quine?
I suspect suspect the smarter the technology press journalist follows /. closely. So memes originating here might end up in the press. Especially if your name is ESR.
This is what worked for me:
make xconfig /etc/lilo.conf&
make bzImage
make modules
make install
make modules_install
xemacs
lilo
YMMV. HTH. HAND.
--
Interested in XFMail? New XFMail home page