Slashdot Mirror


Linux GPU Performance

CrzyP writes "AnandTech.com has benchmarked the most popular graphics cards from ATI and NVidia on the Linux OS (SuSE 9.1). It is interesting to see that they have also written a custom benchmarking tool which can also be downloaded from the article. Take a look at Kristopher Kubicki's "Linux 3D AGP GPU Roundup" to see how each of the mid to high end cards performed on the Penguin flavored system."

373 comments

  1. Kristopher Kubicki by TheReckoning · · Score: 5, Funny

    Get that man on the KDE team IMMEDIATELY! His parents obviously had a grasp of the KDE naming convention long ago.

    It's funny. Laugh.

    1. Re:Kristopher Kubicki by kasperd · · Score: 0, Flamebait

      Get that man on the KDE team IMMEDIATELY!

      A man who cares about performance on the KDE team? Well if it works out, the result might be something good.

      --

      Do you care about the security of your wireless mouse?
    2. Re:Kristopher Kubicki by Ford+Prefect · · Score: 2, Interesting

      A man who cares about performance on the KDE team? Well if it works out, the result might be something good.

      They seem to know something about improving performance. Try a recent KDE release, you'll be surprised. :-)

      Heading back to the topic, I've been very impressed with the Nvidia drivers when using SuSE 9.0. Fast, completely stable, and dead easy to install - and that's with the standard Nvidia installer. With 9.1, it's supposed to be even easier...

      --
      Tedious Bloggy Stuff - hooray?
    3. Re:Kristopher Kubicki by Kethinov · · Score: 2, Interesting
      A man who cares about performance on the KDE team? Well if it works out, the result might be something good.
      KDE's performance is fine. What they need is better usability, better defaults, and most importantly, a better looking QT. Even when you theme QT and GTK with Bluecurve to make them look as identical as possible, open konsole then open gnome-terminal and place them side by side. konsole, like all QT apps, has needless bordering everywhere, whereas gnome-terminal, like all GTK apps, is so much cleaner looking.

      This is why many notably amazing Linux apps (GIMP, gaim, evolution, hell even firefox) reject QT in favor of GTK. If QT was made to look cleaner and KDE's defaults/usability was more refined, it would be far more successful.
      --
      You're right, I wouldn't steal a car. But if it were possible, I sure as hell would download one!
    4. Re:Kristopher Kubicki by shawn(at)fsu · · Score: 1

      This is why many notably amazing Linux apps (GIMP, gaim, evolution, hell even firefox) reject QT in favor of GTK How much of your post is oppinion and fact I wonder?

      My oppinion is I like QT, I like the way apps that use QT look compared to GTK, which is why I use KDE over Gnome.

      --
      500 dollar reward for tip(s) leading to the arrest of the person(s) who stole my sig.
    5. Re:Kristopher Kubicki by Kethinov · · Score: 1

      There really isn't much of a difference besides the superfluous bordering everywhere in QT apps, which drives me up the wall. Seriously. Look at it it.

      --
      You're right, I wouldn't steal a car. But if it were possible, I sure as hell would download one!
    6. Re:Kristopher Kubicki by fymidos · · Score: 2, Informative

      >open konsole then open gnome-terminal and place
      >them side by side.konsole, like all QT apps, has >needless bordering everywhere,

      funny, i see many more borders is gnome-terminal ?!?!

      >This is why many notably amazing Linux apps
      >(GIMP, gaim, evolution, hell even firefox)
      >reject QT in favor of GTK.

      gimp is the origin of gtk, i would be surprised to see gimp using qt.
      gaim, evolution and firefox (and nautilus and abiword and ..) are using gtk for completely different reasons: qt is gpl'ed *not* lgpl'ed and it's only available in windows under a non-gpl license.

      believe me, if qt was lgpl'ed only gimp would use gtk. qt is just so much easier ...

      --
      Washington bullets will simply be known as the "Bulle
    7. Re:Kristopher Kubicki by Tet · · Score: 1
      believe me, if qt was lgpl'ed only gimp would use gtk. qt is just so much easier ...

      Having tried writing apps in both, I'd have to disagree with you. Given the choice, my future GUI development with be in GTK+.

      --
      "The invisible and the non-existent look very much alike." -- Delos B. McKown
    8. Re:Kristopher Kubicki by Kethinov · · Score: 1
      funny, i see many more borders is gnome-terminal ?!?!
      Um, I would encourage you to look again.

      Notice the superfluous bordering especially on the right and left of konqueror and konsole and the lack of such extra bordering on nautilus and gnome-terminal?
      --
      You're right, I wouldn't steal a car. But if it were possible, I sure as hell would download one!
    9. Re:Kristopher Kubicki by Anonymous Coward · · Score: 0

      And mine is in qt. *shrug*. To each his own.

    10. Re:Kristopher Kubicki by kasperd · · Score: 2, Informative

      KDE's performance is fine.
      Really? Every time I have compared KDE and Gnome on a lowend computer (300-500MHz 128-256MB RAM) the result have been the same. Gnome was way faster than KDE.

      What they need is better usability,
      Maybe KDE is not perfect, but I don't know any GUI which is better than KDE.

      better defaults,
      I'm sure most people (including me) will agree with that. But that is all they could agree about. Because everybody want different settings, and no default will satisfy everybody. Better just let each user configure the environment as he preffers.

      open konsole then open gnome-terminal and place them side by side.
      No, I don't want to do that. I don't like any of them. Having both of them side by side would just be too much. I only use them to start an xterm anyway, that is until I get around to replace the launcher icon with one that will launch an xterm.

      This is why many notably amazing Linux apps (GIMP, gaim, evolution, hell even firefox) reject QT in favor of GTK.
      I don't think so. A lot of people rejected QT because of license issues. AFAIK the license is no longer a problem, but there is not really much point in switching from GTK to QT. Switching would require some amount of code to be rewritten, so a very good reason would be needed.

      I don't think the GIMP people ever whined about it. They needed a toolkit, and apparantly none could satisfy their needs, so they wrote one: The GIMP ToolKit.

      --

      Do you care about the security of your wireless mouse?
    11. Re:Kristopher Kubicki by mabinogi · · Score: 1

      Disable the border then.
      It's in the settings for Konsole.

      --
      Advanced users are users too!
    12. Re:Kristopher Kubicki by OmniVector · · Score: 0

      that doesn't fix the problem with all qt apps. that also doesn't help because it should be default. windows is even remotely usable once you install about 30 apps, remove ie, remove OE, install SP2, reboot 4-5 times (due to all the 30 apps and updates). it just takes time, effort, and it isn't setup (out of the box) to be user friendly. that goes a long way towards making the whole system easier to use.

      --
      - tristan
    13. Re:Kristopher Kubicki by mabinogi · · Score: 1

      But that problem doesn't exist with other apps.
      Just look at Konqueror and Kate in your own screenshots.

      The border in Konsole is not some inherent QT flaw, it's a design decision (albeit an odd one)....

      --
      Advanced users are users too!
    14. Re:Kristopher Kubicki by Anonymous Coward · · Score: 0

      "opinion".

    15. Re:Kristopher Kubicki by stoborrobots · · Score: 1

      I don't know that it's particularly QT-related, or odd - even the gimp windows shown on the top left of the KDE screenshot show the double-border.

      Then realize that the "extra" border is actually the border of the "tab" and is there for visually describing the edge of influence of the tab button.

      It seems pretty reasonable to me...

    16. Re:Kristopher Kubicki by mattyrobinson69 · · Score: 1

      try a new version of kde (3.3) also, use a different theme (i use baghra at the moment, and there isn't "superfluous bordering everywhere"

    17. Re:Kristopher Kubicki by 10Ghz · · Score: 1
      Really? Every time I have compared KDE and Gnome on a lowend computer (300-500MHz 128-256MB RAM) the result have been the same. Gnome was way faster than KDE.


      Well, 3.2 was a big step forward when it comes to KDE. I have occasioanlly heard people claim that Gnome is faster, but I have heard even more claims saying KDE is faster. So I disagree with you assumption that "Gnome was always faster than KDE".

      FWIW: I have KDE 3.2.2 installed on my old 300Mhz P2-laptop with 320MB of RAM and it's usable just fine. Of course it's slower than something like Fluxbox or XFCE would be, but it's definitely usable. I did try Gnome on it as well, and while it booted few seconds faster than KDE did, it wasn't one bit faster than KDE was during actual usage. In fact, it was slower at times.
      --
      Lesbian Nazi Hookers Abducted by UFOs and Forced Into Weight Loss Programs - -all next week on Town Talk.
    18. Re:Kristopher Kubicki by Kethinov · · Score: 1

      Konqueror and Kate do have these problems. Look closer.

      --
      You're right, I wouldn't steal a car. But if it were possible, I sure as hell would download one!
    19. Re:Kristopher Kubicki by julesh · · Score: 1

      With [SuSE] 9.1, it's supposed to be even easier...

      Well, I went online, did an update, it downloaded them... then said I didn't have an nVidia graphics card so it wasn't going to install them.

      Not sure what went wrong. Some day, perhaps I'll figure it out.

    20. Re:Kristopher Kubicki by julesh · · Score: 1

      Every time I have compared KDE and Gnome on a lowend computer (300-500MHz 128-256MB RAM) the result have been the same. Gnome was way faster than KDE.

      Have you tried it on a recent glibc release after running prelink? KDE performance sucks primarily because of slow dynamic linking with its hundreds of libraries. Prelink solves this problem. (Or so I'm told. I haven't had a chance to try it yet.)

    21. Re:Kristopher Kubicki by julesh · · Score: 1

      Notice the superfluous bordering especially on the right and left of konqueror and konsole

      Err... that border isn't superfluous, it's there to identify (in the case of konsole) that what's inside it will be switched if you click on one of the tabs or (with konqueror) that the pane is 'inside' the draggable framework.

      I think it's better that it is there than not, although I'll agree that's probably highly subjective.

  2. Except by Anonymous Coward · · Score: 2, Interesting

    Isn't NVidia refusing to allow driver support for recent linux kernels on some of their cards?

    What good is good GPU performance if you have to run an old kernel to run the GPU at all?

    1. Re:Except by jejones · · Score: 3, Informative

      It took nVidia a while to get around to releasing a driver that could deal with 4K stacks, which more recent kernels have switched to from 8K. It was a pain in the posterior for Fedora Core 2 for a while, but no longer. Maybe that's what you're referring to?

    2. Re:Except by Anonymous Coward · · Score: 0

      So the nvidia drivers are up to speed now then?

    3. Re:Except by madrouter · · Score: 0

      I think you are thinking of their binary Nforce chipset drivers, which are for 2.4 series only (that I know of).

      Not thier graphics card drivers.

    4. Re:Except by spuzzzzzzz · · Score: 1

      Newer kernels give you the option of switching to 4k stacks. You can still use 8k stacks, however (I'm still using them).

      --

      Don't you hate meta-sigs?
  3. Penguin-flavored? by Anonymous Coward · · Score: 4, Funny

    Umm, you're supposed to be putting the Linux CDs into the CD drive, not taste-testing them... ;)

    What does a penguin taste like, anyhow?

    1. Re:Penguin-flavored? by temojen · · Score: 5, Funny

      Herring, probably.

    2. Re:Penguin-flavored? by kfg · · Score: 5, Funny

      What does a penguin taste like, anyhow?

      Fat chicken.

      KFG

    3. Re:Penguin-flavored? by Anonymous Coward · · Score: 0

      What does a penguin taste like, anyhow?

      Easy, tastes like a chicken ;-)

    4. Re:Penguin-flavored? by stratjakt · · Score: 1

      Penguins is practically chickens.

      And I just looooooove chickens!

      --
      I don't need no instructions to know how to rock!!!!
    5. Re:Penguin-flavored? by Anonymous Coward · · Score: 0

      I would imagine they taste like duck, as they're both a fishy kind of bird.

      Which I guess makes them OK for some types of vegetarian.

      Linux:-The Vegetarians choice

    6. Re:Penguin-flavored? by Anonymous Coward · · Score: 2, Funny

      So, Linux is just a little bit fishy... ? :)

    7. Re:Penguin-flavored? by Oliver+Wendell+Jones · · Score: 1

      They taste like Spotted Owl with just a hint of Snail Darter. The bigger ones taste like Bald Eagle.

      --
      A computer once beat me at chess, but it was no match for me at kick boxing -- Emo Phillips
    8. Re:Penguin-flavored? by jejones · · Score: 1

      What does a penguin taste like, anyhow?

      Maybe Lola Granola would know...?

      "Darn it, it doesn't soar!" -- Lola Granola

    9. Re:Penguin-flavored? by TheLittleJetson · · Score: 1

      FUNNY +5 gazillion

    10. Re:Penguin-flavored? by Naikrovek · · Score: 1

      That's called beta-tasting.

    11. Re:Penguin-flavored? by BHS_Turf · · Score: 1

      I once saw a japanese gameshow where the man had to eat a scorpion, and said it tastes just like penguin.

    12. Re:Penguin-flavored? by FuzzyBad-Mofo · · Score: 1

      Fat chicken.

      KFG

      Surely you meant 'KFC'? ;)

    13. Re:Penguin-flavored? by zepi · · Score: 1

      Here in Finland penguins basically taste like ice cream. http://www.valio.fi/tuotteet/jaatelot/pingviini.ht ml

      Local manufacturer produces various tastes of cones under brand 'Penguin' (Pingviini in finnish).

    14. Re:Penguin-flavored? by Anonymous Coward · · Score: 0

      No. Kentucky Fried Gnu.

    15. Re:Penguin-flavored? by glitch23 · · Score: 0

      So basically like KFC chicken then right?

      --
      this nation, under God, shall have a new birth of freedom. -- Lincoln, Gettysburg Address
    16. Re:Penguin-flavored? by barbazoo · · Score: 1
      What does a penguin taste like, anyhow?
      What flavour d'you think it is? It's bloody dead sea-bird flavour.
  4. ATI vs nVidia by Eeknay · · Score: 5, Insightful

    Wow, this is very surprising. One would expect with similar Windows benchmarks for the X800 to be matching or beating the 6800 Ultra (depending on drivers of course), so these low X800 scores in Linux really are quite a shock.

    1. Re:ATI vs nVidia by Tumbleweed · · Score: 5, Insightful

      If low-performing Linux ATI drivers are a shock to you, you haven't been paying attention for a few years. :(

    2. Re:ATI vs nVidia by Anonymous Coward · · Score: 2, Informative

      They did not use a XT or XTPE, only the Pro. The Pro never beats an Ultra and rarely bests the GT:

      http://anandtech.com/video/showdoc.aspx?i=2044

    3. Re:ATI vs nVidia by stratjakt · · Score: 0

      No shock.

      ATI is to small to spend time writing good drivers for the 3 linux gamers out there.

      nVidia is big enough to burn a few man hours doing so, although their linux drivers are nowhere near the Windows ones in terms of performance (IMO).

      Neither will release their "specs" without some serious NDA signing going on, the best hope would be (commercial) third party linux drivers for each.

      The worlds of moochers and big spenders must collide in the linux desktop of the future.

      --
      I don't need no instructions to know how to rock!!!!
    4. Re:ATI vs nVidia by tempmpi · · Score: 5, Informative

      Not really, it is common knowledge that Nvidia's linux binary drivers are much better than Ati's. Not only the performance is better in Nvidia's drivers but they are also more compatible. People often had problems getting ATI's binary drivers working, while Nvidia's drivers are working without problem in most configurations and even problems like 4k stacks were fixed withhin a reasonable time.

      --
      Jan
    5. Re:ATI vs nVidia by Anonymous Coward · · Score: 0

      http://theinquirer.net/?article=18664

    6. Re:ATI vs nVidia by stratjakt · · Score: 1

      Yeah, they say that about once a year, at least for the last decade or so.

      "Big driver push" usually means "make it compile for a recent kernel" not "make it on parity with the Windows driver".

      The OpenGL support in Windows is sorely lacking with ATI, so they need good OpenGL in the first place before they can port it to linux.

      --
      I don't need no instructions to know how to rock!!!!
    7. Re:ATI vs nVidia by HadenT · · Score: 2, Informative

      Linux Nvidia drivers perform just as good or even better than in windows.
      There wasn't Quake3 test in this review, but ut2004 was said to run faster. As for Q3, I've tested it myself: runs 5-6% faster.

    8. Re:ATI vs nVidia by wardred · · Score: 1

      When I spend from $200-$500 for a video card, $50 for a decent, modern game, and a fair amount of cash for the rest of the system, I don't see myself as a moocher.

      No, I didn't pay several hundred dollars for a single machine OS upgrade merely because support for an older OS that I spent over a hundred dollars on is dying. But what does the video card company care how much money I paid to M$? The only reason I'm not getting decent drivers from them is because their aren't enough of us gaming on Linux, not because we didn't purchase their expensive video cards...

      Heck, if they'd release the specks for their cards, said moochers would write the darned drivers for them. I'm guessing they don't do so because of all the various problems with IP laws...It might be implied that they're allowing anybody to use patent #35987322 because they released the API for accessing it....or it might be found that they've actually used somebody else's patented methods, but nobody has bothered reverse engineering things far enough to find this out...

    9. Re:ATI vs nVidia by stratjakt · · Score: 2, Interesting

      Q3 and UT2004 are both very CPU intensive, so who knows if the speed is due to the drivers or the platform.

      Maybe they're built with super-mega-optimizing flags on linux that they don't use on Windows. Maybe it's the sound or input routines. Both were developed in linux, AFAIK.

      My point is, it's true enough to say that Q3 performs better in linux, but that doesnt necessarily mean the OpenGL drivers for linux are better than Windows. I'd say that, at best, they'd be pretty much the same.

      --
      I don't need no instructions to know how to rock!!!!
    10. Re:ATI vs nVidia by arose · · Score: 2, Interesting

      Who cares about the gamers. Workstation 3D is what Linux needs.

      --
      Analogies don't equal equalities, they are merely somewhat analogous.
    11. Re:ATI vs nVidia by Anonymous Coward · · Score: 1, Informative

      Actually, it already compiles for the most recent kernel (2.6.8-rc3) so I guess "big driver push" really means something else this time. Maybe 64 bit support. :p

      Oh and they've hired some guy called Michel from the DRI project; http://marc.theaimsgroup.com/?l=dri-devel&m=109418 851616952&w=2

      *shrug*

    12. Re:ATI vs nVidia by Anonymous Coward · · Score: 1, Interesting

      ATI is to small to spend time writing good drivers for the 3 linux gamers out there.

      nVidia is big enough to burn a few man hours doing so, although their linux drivers are nowhere near the Windows ones in terms of performance (IMO).


      Are you joking?

      ATI is not the behemoth that it used to be in the 1990s, but it still has over 2500 employees worldwide. Nvidia has approximately 2000.

    13. Re:ATI vs nVidia by Justus · · Score: 1

      I'm going to go out on a limb and say that the "big driver push" will give us drivers that work on X.org 6.8.1 without crashing whenever you attempt to use the GL features.

      I predict that these drivers will be available around the time that 6.9.0 comes out.

    14. Re:ATI vs nVidia by pdxaaron · · Score: 2, Interesting

      While the Nvidia driver are most likely better, their benchmarks used 64 bit drivers for Nvidia versus 32 bit drivers from ATI. I wish they would have tested 32 bit Nvidia drivers as well to compare Apples to Apples.

    15. Re:ATI vs nVidia by Anonymous Coward · · Score: 0

      What I don't get is, why is it so hard to port the windows drivers? They're not filesystems or user interfaces, they're device drivers. It's the same hardware, the same bus, the same CPU ... the only difference is in the API, which more or less has to be public information. Porting a DirectX driver could be fairly tricky, but OpenGL should have the same entry points. What am I missing here?

    16. Re:ATI vs nVidia by 1000StonedMonkeys · · Score: 1

      In fact, if I remember correctly, they mostly are the same. NVidia uses pretty much the same code on linux and windows. It's just the interfaces that change.

    17. Re:ATI vs nVidia by Anonymous Coward · · Score: 0

      Porting a DirectX driver could be fairly tricky, but OpenGL should have the same entry points. What am I missing here?

      You're missing the fact that ATI sold out to Microsoft and primarily supports DirectX3D. ATI is notorious for having bad OpenGL support and performance. It does not matter what platform you use for benchmarking and comparison, ATI sucks for OpenGL usage.

    18. Re:ATI vs nVidia by irc.goatse.cx+troll · · Score: 0, Troll

      Theyre comparing the best NVidia has vs the best ATI has. Personally, I wish ATI had better drivers (including 64bit ones), but thats just me. I still wouldnt use them of course, as thats like buying windows 2k3-- It doesnt matter if a bad company finally makes something worth using, they're still a bad company. Knowing ATI, I'm just suprised they don't EOL their drivers for any kernel older than a week. On a related note, anyone else appreciate the performance enhancements NVidia's TNT2 cards get with the most current drivers? It's nice having a company actually support their customers.

      --
      Pain lasts, kid. Its how you know you're alive. Sometimes I think this growing up thing is just pain management-TheMaxx
    19. Re:ATI vs nVidia by Anonymous Coward · · Score: 0

      Actually, ATI has relased specs for certain cards, and the Open Source drivers have sucky performance. Fact is these companies put massive optimization work into their drivers, which the basement coders of the open source world can't hope to match until it is years too late. (Forex, Linux now has excellent 3dfx support....)

    20. Re:ATI vs nVidia by suckmysav · · Score: 2, Insightful

      "The only reason I'm not getting decent drivers from them is because their aren't enough of us gaming on Linux, not because we didn't purchase their expensive video cards."

      Agreed. I have two (primary use) PC's at home. Number one is an XP2400+1Gb+GF4-4200/128 and it runs Windows. Number two is a Dual PIII/933+512Mb+GF4MX400/64 and it runs Linux.

      The first one is obviously the better system for gaming, and in fact that is what I primarily use it for. It is good enough to run Far Cry with most settings on "very high" detail, believe it or not. The second one I use for all internet related stuff, as well as a central file server.

      I'm hoping that one day I won't need to have a Windows machine at all for games, but for this to happen developers will need to start developing for Linux, and for devs to do that people must start buying Linux games.

      My small contribution to the cause is to always buy the linux version of a game if at all possible, even if I intend to run it on Windows initially. I have found so far that the activation keys for most games are interchangable between the linux and windows version. So, once I have purchased my Linux game I can then obtain through more dubious channels the Windows version, and proceed to use my legally obtained key with the Wwindows binaries. I suppose that this is technically illegal, but I don't feel it is in the least bit morally wrong.

      My aim is to gradually build up a collection of Linux games, and as my systems are upgraded over time more and more of those games will become playable natively on any future linux PC I might have. In the meantime I hope I am encouraging developers to release their games natively under linux.

      --
      "You can't fight in here, this is the war room!"
    21. Re:ATI vs nVidia by swv3752 · · Score: 2, Informative

      I think it was either Icculus or Carmack that said the subsystems of Linux (primarily the kernel) were more efficient than Windows and all other things being equal, Linux would out perform Windows.

      --
      Just a Tuna in the Sea of Life
    22. Re:ATI vs nVidia by Zorilla · · Score: 1

      Or at least some fast GDI routines and preferably an alpha channel in X. Windows has had GDI acceleration since Windows 95 and Linux still needs it. If Linux is already faster than Windows, this would definitely help it feel as fast as, or faster than, Windows.

      --

      It would be cool if it didn't suck.
    23. Re:ATI vs nVidia by AstroDrabb · · Score: 2, Informative
      Did you RTFA?
      Another issue that we came across with ATI's was the lack of 64-bit Linux drivers. ATI has no 64-bit drivers for Linux, yet they have 64-bit Windows binaries. Thus, our benchmarks are limited to 32-bit binaries only.
      --
      If Tyranny and Oppression come to this land,
      it will be in the guise of fighting a foreign enemy. -James Madison
    24. Re:ATI vs nVidia by AstroDrabb · · Score: 1
      Hello troll, how are you today?

      Where did you get any of your information? It is all wrong. ATI has _more_ employees then NVidia. Also, the NVidia Linux drivers are the same as the MS Windows drivers. NVidia uses a common code base across all platforms the support. It was one of their best moves IMO.

      nVidia is big enough to burn a few man hours doing so, although their linux drivers are nowhere near the Windows ones in terms of performance (IMO).
      Did you even RTFA? The review said that _native_ Linux games that also have a native Windows version were a little faster under Linux. And from my experience this is true. I have seen a few extra FPS under Linux with native games then the MS Windows conterpart. I am not talking 30 extra FPS but around 5 FPS or so which can make the difference between 25 FPS (choppy) and 30 FPS (more smooth).
      --
      If Tyranny and Oppression come to this land,
      it will be in the guise of fighting a foreign enemy. -James Madison
    25. Re:ATI vs nVidia by strider44 · · Score: 2, Insightful

      The thing I found amazing when reading the article to the end was the very last paragraph:

      During publication of this review, we received some information from ATI about some upcoming Linux announcements which they are working on. We will keep you informed of the details as we hear them.

      Is it true? Is ATI finally getting it's act together with linux? Can linux finally become a game platform with everyone able to play?

      Looking at past experiences, don't get your hopes up methinks.

    26. Re:ATI vs nVidia by Anonymous Coward · · Score: 0

      Is ATI finally getting it's act together with linux?

      "its".

    27. Re:ATI vs nVidia by strider44 · · Score: 1

      What was the point of that post? So what if I made a minor grammatical error when hurriedly typing?

    28. Re:ATI vs nVidia by hellings · · Score: 1

      On a side note, I have been unable to get the latest release of the Nvidia drivers for my GeForce4 Ti to compile with gcc-3.4. No problems with the older versions of gcc, but the newest versions don't compile. Now, mind you, I have had no problems with performance, since Nvidia has awesome support for Linux, but there is a struggle somewhere in there to get the driver to compile.

      --
      Quidquid latine dictum sit, altum viditur. "Whatever is said in Latin, seems prfound."
    29. Re:ATI vs nVidia by timts · · Score: 1

      I just wonder how many gamers are using linux. when they cant get the linux version at the same time as pc version.

  5. ...vs. same cards with Windows? by TrollBridge · · Score: 5, Insightful

    Those numbers are all well and good, but I'd be interested in seeing them side-by-side with the same tests performed (on the same machines, of course) running Windows.

    --
    There's a Mercedes gap too. I want one and can't afford one, but it's not government's job to do anything about it.
    1. Re:...vs. same cards with Windows? by kasperd · · Score: 1

      but I'd be interested in seeing them side-by-side with the same tests performed (on the same machines, of course) running Windows.

      I'd like to see that comparision as well, but with Linux using open source drivers on documented hardware. (I know that would disqualify a lot of nVidia and ATI chips from the test.)

      --

      Do you care about the security of your wireless mouse?
    2. Re:...vs. same cards with Windows? by DeckerEgo · · Score: 5, Informative
      Here's what they noticed (more of a summary than a benchmarking):
      Although this analysis did draw some pretty strong lines as to where each card stand, we were more interested in how each game performed compared to their Windows counterparts. We drew a lot of conclusions from one of our more recent video card analyses from July. Surprisingly, most of our NVIDIA video cards scaled very similarly. Wine games like Jedi Knight took a 10% to 15% hit in performance compared to the Windows tests that we did just a few weeks ago. Other games like Unreal Tournament 2004 actually showed mild signs of an increase in frame rate on the NVIDIA graphics cards. Wolfenstein: ET generally performed with similar average FPS to our video cards from 2003. However, keep in mind that the drivers used then were almost a year old.
    3. Re:...vs. same cards with Windows? by Tet · · Score: 1
      I'd like to see that comparision as well, but with Linux using open source drivers on documented hardware. (I know that would disqualify a lot of nVidia and ATI chips from the test.)

      I was wondering if it was just me. For some of us, absolute performance isn't the ultimate goal. But I'd still like to know which card will give me the best performance without having to resort to closed source drivers. I won't buy Nvidia or newer ATI cards because they won't publish programming information for them.

      --
      "The invisible and the non-existent look very much alike." -- Delos B. McKown
    4. Re:...vs. same cards with Windows? by kasperd · · Score: 1

      But I'd still like to know which card will give me the best performance without having to resort to closed source drivers.

      That is also something I want to know, I'm just not sure where to ask about it. AskSlashdot perhaps?

      --

      Do you care about the security of your wireless mouse?
  6. My linux GPUbreakdown from independant tests: by stratjakt · · Score: 5, Funny


    NVidia: Sort-of

    ATI: Kind-of

    --
    I don't need no instructions to know how to rock!!!!
    1. Re:My linux GPUbreakdown from independant tests: by Anonymous Coward · · Score: 0

      I would add my test of integrated
      SIS GPU: Fuck-off

    2. Re:My linux GPUbreakdown from independant tests: by Anonymous Coward · · Score: 0

      Matrox: not exactly

  7. Are you saying by Anonymous Coward · · Score: 3, Funny

    We can run Linux on a GPU now?

    1. Re:Are you saying by Ingolfke · · Score: 1

      Only if they are configured into a Beowulf cluster.

    2. Re:Are you saying by Anonymous Coward · · Score: 0

      What kind of self-respecting linux geek would admit to wasting his time playing video games? You should be hacking on your GPU to give your cluster more power!

    3. Re:Are you saying by Anonymous Coward · · Score: 0

      And only in Soviet Russia.

  8. It doesn't much matter... by Anonymous Coward · · Score: 0

    is there any good use for GPUs on linux as of now? maybe so you could tell how awesome games would be if they were made?

    1. Re:It doesn't much matter... by mikael · · Score: 1

      There's Bzflag - a multiplayer networked tank game with many variations (capture-the-flag with teams, free-for-all, hunt-the-rabbit). All options can be set by the admin (jumping, gravity, ricochet, 1 to infinite missiles, tank speed, missile speed, teleporters, superflags [shockwave, superbullets, oscillation overdrive, guided missile, seer] ).
      Runs at over 60 frames/sec at 1400x1050 with full texturing on a Sony laptop.

      --
      Vintage computer adverts: http://www.vintageadbrowser.com/computers-and-software-ads
    2. Re:It doesn't much matter... by imsabbel · · Score: 1

      Yeah, but if it was anything else as 60fps is would be a shame, because the gfx look slighly better then the original battlezone 25 years ago.

      --
      HI O WISE PRINCE. WHT TOOK U SO DAM LONG?
    3. Re:It doesn't much matter... by smclean · · Score: 1
      Flightgear, Neverball/putt, Tuxracer, Tuxkart, BZFlag, Cube, Scorched 3d, GLTron, Trackballs, FooBillards, NWN, America's Army, Quake 3, UT2004, Planeshift, etc etc..

      There's enough out there to warrant purchasing a GPU... Now I must go punish myself for responding to a troll.

      --

      "'Yrch!' said Legolas, falling into his own tongue."

    4. Re:It doesn't much matter... by mikael · · Score: 1

      The 60fps limit is due to the LCD display. 'glxgears' runs at 900 frames/sec without waiting for vertical refresh. The manufacturer had the brilliant idea of sticking a GeForce FX5600 in a laptop and boosting the screen resolution to 1400x1050. Unfortunately, they had the not so bright idea of locking the refresh rate to 60Hz rather than 85Hz.

      --
      Vintage computer adverts: http://www.vintageadbrowser.com/computers-and-software-ads
  9. Better drivers and licensing please by Handbrewer · · Score: 5, Insightful

    The situation right now is quite frustrating - all distributions should be able to ship the binary drivers for the vendor kernel. It would make it so much easier, than having to get the kernel source and headers before building the module on your own. Thats an unneccesary burden only placed on our shoulders because of some paperwork. 2nd, id like some better drivers please, the ATI drivers are terrible, please stop treating me as a second rank costumer. My money is as good as anybody elses. Thanks so far NVIDIA, now we just need a better license.

    1. Re:Better drivers and licensing please by danheskett · · Score: 4, Insightful

      Your money is not as good as everyone elses.

      90%+ of the market is Windows. To support your 10% slice of the market is vastly more expensive per sale.

      Look at this way. If developing, optimizing, and supporting the drivers for a new line of cards was equally as expensive for Linux as for Windows, at say, $1M USD, and the card was expected to sell 10M units, and 90% of the customers are Windows users or OEMS who sell only Windows, then you are looking at a per-unit cost to develop that Windows driver of about $0.11 USD. For the 10% of Linux users, the cost per unit-sold is $1 USD.

      Those are assuming all things are equal. I'd wager that for any given company, developing the Windows drivers, including packaging, tweaking, etc is much easier. Add in that Microsoft through its Windows Harward Quality Labs (WHQL or whatever they call it now) basically will subsidize your development costs, I'd imagine that the cost ratio is not 1:1 but much leaner in favor of developing Windows drivers.

      So the bottom line is that, in any company, you try to maximize your profitable customers and ditch your unprotifable customers. If the Linux drivers cost only $2 per unit in volume for each sale generated where as the Windows drivers cost only pennies on that dollar, you have a big problem. Over time ATI and nVIDIA sell an awful lot of chips to OEMs and other re-branders who pay very little for hardware. If you are selling chips at $40 (my guess is that this is a high number.. your typical 800-1000 laptop doesn't cost more than 500-600 to build, and most of that goes to Intel/AMD, Micron for RAM and Samsung for LCDs) a unit that $2 cost is suddenely a huge, huge problem.

      Of course these numbers are made-up out of broad cloth. But the ratios are real. Linux drivers would have to cost 10 times less to develop to make you an "equal" customer as a Windows customer.

      In fact, you are lucky to get anything from nVIDIA or ATI.

    2. Re:Better drivers and licensing please by sloanster · · Score: 2, Informative

      The situation right now is quite frustrating - all distributions should be able to ship the binary drivers for the vendor kernel. It would make it so much easier, than having to get the kernel source and headers before building the module on your own.

      I'm not sure what distro you're using, but with suse 9.1, no such contortions are needed. I simply checked the box in yast that says "install nvidia drivers" and a message popped up saying "nvidia drivers will take effect next time X is restarted".

      The article contained some incorrect statements as well - in particular, claiming that the kernel must somehow be "recompiled" to allow the nvidia drivers to be installed. That's never been true, so seeing such a statement is mystifying.

      To recap, with suse 9.1 and nvidia drivers:

      recompile kernel? no
      edit config files? no.
      time to install drivers? about 30 seconds.

    3. Re:Better drivers and licensing please by Handbrewer · · Score: 3, Informative

      Well the NVIDIA driver installer does build a module that matches your kernel. I run Debian, and i had to apt-get the kernel source and headers, not a terrible task to do, but should not be necessary to do, just to get the newest driver. I dident say the kernel need recompilation. And you need to modify the XF86Config-4 file from "nv" to "nvidia" in the driver section and thats about it. 30 seconds is about the time yes, but only after i got the kernel headers and source, some 30-40 mbs of source, not that it matters on 10mbit, but for dialups, that would kinda suck i guess :).

    4. Re:Better drivers and licensing please by stratjakt · · Score: 2, Informative

      You're mostly right about not needing to recompile. But it depends on the distro.

      If you don't have the kernel source that was used to build the kernel you're running, you'll need to either obtain it, or in many cases, obtain the latest kernel source and build a kernel to match it.

      Wasting the space to have the kernel sources around is pretty sucky just to upgrade a video driver. I'm thinking small, set-top or embedded gaming machines, like something jammed into an arcade cabinet.

      But, what the hell, I already waste all that space for portage, just to make installing stuff somewhat less of a headache.

      --
      I don't need no instructions to know how to rock!!!!
    5. Re:Better drivers and licensing please by mrgreen4242 · · Score: 2, Insightful

      Here's a question for you, based on what you've said... do you think that Linux users would be willing to pay $2-4 for a CD with the proper drivers, as well as lifetime updates via the web? Do you thing that would that make it more likely that Linux drivers would be developed?

    6. Re:Better drivers and licensing please by sloanster · · Score: 1

      If you don't have the kernel source that was used to build the kernel you're running, you'll need to either obtain it, or in many cases, obtain the latest kernel source and build a kernel to match it.

      Or run a distro which provides a wrapper around the nvidia driver and does not require driver rebuilds when the kernel is updated. Of course, in the case of suse 9.1, that means using the 1.0-5336 driver, and not the latest and greatest. The choice is yours: Do you want simplicity? fine, stick with the out of the box suse experience. Do you want to get your hands dirty with the latest and greatest? Good, download the latest nvidia driver and perform the 3 steps:

      1. cd to kernel source and type "make cloneconfig", then "make prepare-all"
      2. type e.g. "sh NVIDIA-Linux-x86-1.0-6111-pkg1.run -q --kernel-source-path=/usr/src/linux"
      3. type "sax2 -m 0=nvidia" if this is a new driver install, otherwise skip this step.

      Of course this means you must rebuild the driver if you update the kernel, and you must have kernel sources if you update the kernel, but it's your choice, and if you don't like the idea of having to maintain the driver/kernel synchronization, just stick with the default nvidia wrapper.

    7. Re:Better drivers and licensing please by fymidos · · Score: 1

      usually all you need to get linux drivers developed is some documentation.
      it's much cheaper than developing a windows driver, that's for sure.

      >In fact, you are lucky to get anything from nVIDIA or ATI.

      they don't seem to think that now, do they?
      supporting the 10% slice in this case means that you get a 10% slice.
      Nvidia is doing a good job at that and i do recomend an nvidia card for a tight budget, but for as long as i can remember, the best card you can buy in a linux box has always been a matrox.

      --
      Washington bullets will simply be known as the "Bulle
    8. Re:Better drivers and licensing please by nusratt · · Score: 1

      "please stop treating me as a second rank costumer" [sic]

      OK, here's your big chance:
      whip up something wonderful for me to wear on Halloween, and I'll grant you recognition as a first-rank costumer. ;-)

      or did you mean customer?...

    9. Re:Better drivers and licensing please by pyros · · Score: 1

      Assuming a 2.6 kernel (not sure aobut 2.4 kernels) look at /lib/modules/`uname -r`/build and find everything you should need to build modules against your current running kernel.

    10. Re:Better drivers and licensing please by RAMMS+EIN · · Score: 1

      How can you expect them to improve their drivers if the kernel interface keeps changing?

      --
      Please correct me if I got my facts wrong.
    11. Re:Better drivers and licensing please by samadhi · · Score: 1
      But the ratios are real. Linux drivers would have to cost 10 times less to develop to make you an "equal" customer as a Windows customer.

      Hmm wouldn't Open Sourcing the drivers achieve an even lower development cost? Ooops forgot that is not on the agenda, oh well I can but dream.
    12. Re:Better drivers and licensing please by deacon · · Score: 1, Insightful
      90%+ of the market is Windows. To support your 10% slice of the market is vastly more expensive per sale.

      WTF ?

      Here, let's try your logic again:

      "90 % of otters eat clams, so the care and feeding of wombats is more expensive."

      By definition, the Linux user is "out there" on the cutting edge. Following your example, *I* will now make an unsubstantiated assertion:

      80% of the 90% of computer users that use windows don't know what a video card is, could care even less, and use their computer to sell Hummel figurines on Ebay.

      These users do not need or want a fast video card.

      Of the linux users, the vast majority are computer literate, interested in the performance of their system, and far more likely to pay top dollar to get a FAST video card.

      Makes perfect sense to cater to people who make up 60% of your market for high profit, latest generation video cards.

    13. Re:Better drivers and licensing please by Anonymous Coward · · Score: 0

      According to my information, which is not authoritative and lacks any kind of reliable source, nVidia can't release the source to its drivers because of their licensed SGI technology.

      I don't if that's the truth, but to me it makes sense.. If someone here knows better, please enlighten me.

    14. Re:Better drivers and licensing please by Fred_A · · Score: 1

      I know I'd pay for that. I'd sign right away even.

      If for example it could stop me from trying to figure what damn WiFi chipset maker X or Y put in this week's batch of PCMCIA cards and how well supported it is, you bet I'd pay $4, 4 even.

      --

      May contain traces of nut.
      Made from the freshest electrons.
    15. Re:Better drivers and licensing please by Anonymous Coward · · Score: 0

      90%+ of the market is Windows. To support your 10% slice of the market is vastly more expensive per sale.

      Why? Are you assuming that the driver must be closed-source? Would they lose any money by open-sourcing it? (I don't think ATI could just take NVidia's drivers, change the name, and get out of writing drivers for ATI boards. A driver, by definition, is only useful for interfacing with that piece of hardware!)

      Just about every type of hardware out there has either open-source drivers, or published specs or simple enough interfaces that Linus & co can write their own drivers. What's so special about graphics cards from NVidia that they have to try to keep it secret?

    16. Re:Better drivers and licensing please by Taladar · · Score: 1

      So you think all of the Kernel Module Development shall be frozen because the two (Nvidia and ATI) module maintainers with more money to throw at it than most other Kernel Developer can not keep up with people doing the same for other hardware in their spare time?

    17. Re:Better drivers and licensing please by Kiryat+Malachi · · Score: 2, Informative

      No, he was right.

      There is a market. ATI has sold X cards. These cards, hardware wise, can be used identically on Linux or Windows, its only software thats needed to adapt to the system. If 90% of those cards went to Windows users and 10% went to Linux users, here is why the Linux driver is more expensive per unit sold.

      Let's assume driver development costs a fixed number per OS, Y. Let's assume that its equal for Linux and Windows. Thus, every driver ATI develops costs them Y out of their profits.

      The cost-per-card for Windows driver Y/(0.9X). The cost for Linux drivers, similarly, is Y/(0.1X). 0.9X > 0.1X unless X is zero. So, unless ATI sells NO cards at all, its cheaper to develop for the majority of the market-share. I.E. Windows.

      Considering that desktop market share for Linux is around 3%, and Windows comes in around 90%, its significantly more expensive to support Linux than Windows, even if you assume that the ratio of ATI Linux users to Windows users is probably a little bit better (a large portion of those Windows users have cheapie integrated Intel graphics).

      --

      ---
      Mod me down, you fucking twits. Go ahead. I dare you.
      (I read with sigs off.)
    18. Re:Better drivers and licensing please by Kiryat+Malachi · · Score: 1

      Oh, as to your argument regarding the likelihood of Linux users buying higher end gear - uh-huh. Except for the number of Linux users who heavily re-use old gear and the fact that the people who want the high-end video cards are GAMERS, who are significantly less likely to run Linux.

      It makes perfect sense to cater to your market; but the market for video card makers is NOT Linux.

      --

      ---
      Mod me down, you fucking twits. Go ahead. I dare you.
      (I read with sigs off.)
    19. Re:Better drivers and licensing please by RAMMS+EIN · · Score: 1

      Look, the kernel interfaces are changed frequently. Sometimes, kernel developers deliberately break things to make the lives of module developers hard. I think broken modules is exactly what they deserve.

      Similarly, ATI deserves its reputation for having crappy and broken drivers, by neither taking the effort to produce good drivers themselves, nor allowing anyone else to do it for them.

      Users are getting what _they_ deserve by insisting on using a video card from a vendor that produces shitty drivers and a kernel that keeps breaking existing drivers.

      And I get sad because people are so silly (which is what I deserve, because I care but not enough to change things).

      --
      Please correct me if I got my facts wrong.
    20. Re:Better drivers and licensing please by Anonymous Coward · · Score: 0

      Actually. All I ask for is that they open up some documentation so the community can start to write drivers. There are not many people who can do it, but they are out there and there will be more if some documentation and reference drivers starts to show up.

      Until then I will continue run the hottest with documentation given to the community, Radeon 8500.

    21. Re:Better drivers and licensing please by MrNemesis · · Score: 1

      Arguments about "XXXX should GPL their drivers!" aside...

      I may be wrong, but I was under the impression that the "core" of the nVidia driver is essentially very similar under both Windows and Linux, hence why producing Linux drivers in tandem with their Windows one should be relatively easy, since there's only one codebase. Again I may be wrong, but I believe this is not the case for ATI.

      nVidia got around the rapidly changing kernel interface by GPLing some "glue" code that bound the kernel to the binary module. Hence it's trivially easy for the kernel devs to keep the nVidia module(s) compatible with the kernel by way of altering the glue, and hence why the nVidia module works with pretty much every kernel on pretty much every x86 distro I've seen.

      OTOH, getting the ATI's to work has been a complete PITA, and frequently I give up. They too have (recently) gone down the "glue" code route, but their driver is still *very* picky as per your configuration. For nVidia all I have to do is make sure that MTRR is enabled in the kernel.

      --
      Moderation Total: -1 Troll, +3 Goat
    22. Re:Better drivers and licensing please by Anonymous Coward · · Score: 0

      >I'd wager that for any given company, developing
      the Windows drivers, including packaging,
      tweaking, etc is much easier. Add in that
      Microsoft through its Windows Harward Quality
      Labs (WHQL or whatever they call it now)
      basically will subsidize your development costs,
      I'd imagine that the cost ratio is not 1:1 but
      much leaner in favor of developing Windows
      drivers.

      Huh?!! How do you beat a well-documented and
      open kernel api with no hidden variables, or
      hidden apis, no funky bizarre interaction with
      other drivers? To quote Microsoft on writing
      drivers for Linux vs. Microsoft: (From the
      Halloween II document no less):
      http://www.opensource.org/halloween/hallow een2.php

      An important attribute to note which has led to volume drivers is the ease with which you can write drivers for linux, and the relatively powerful debugging infrastructure that linux has. Finding and installing the DDK, and trying to hook up the kernel debugger and do any sort of interaction with user-mode without tearing the NT system to bits is much more challenging than writing the simple device-drivers for linux. Any idiot could write a driver in 2 days with a book like "Linux Device Drivers" -- there is no such thing as a 2-day device-driver for NT

      And from a Microsoft developer:
      A Microsoft developer who wishes to remain anonymous comments:

      Microsoft really shoots themselves in the foot in driver space. For those of us that will willingly spend effort to improve the quality, functionality, and availability of device drivers for Microsoft OSs, the tools, etc. they provide are actually getting progressively worse. This is happening largely because Microsoft fails to grasp the basic concepts you've been talking about. In response to driver quality and availability issues, they've consistently made exactly the wrong decisions: rather than encouraging more people to help and providing better tools, they try to make the system more and more closed and dumb down the tools.

      So I'm not with you, are you implying dumbed down
      tools are easier to use to write drivers, or
      tearing apart the Microsoft kernel is easier for writing Microsoft drivers than reading a howto for Linux kernel drivers?

      Or are you implying it's better to use dumbed
      down software tools than letting dumb software
      developers write tools?

      In the case of Microsoft, Inc., does the
      experimental evidence support the hypotheses that
      they write dumbed down software FOR dumb
      software developers?

    23. Re:Better drivers and licensing please by owlstead · · Score: 1

      Yup, and don't forget to rerun the thing after updating your kernel. That screwed up my installation, and the xorg.conf file was silently using the nv driver again.

      I could have noticed that the picture was a bit to the right since it was using another refresh rate on my CRT, but you know how it goes. If your 3D screensaver crashes because of missing GLX stuff, you know you just had a kernel update. Something to keep in mind though.

    24. Re:Better drivers and licensing please by deathontwolegs · · Score: 1

      Mandrake also offers nvidia 3d drivers, not in the download editions but in powerpacks (and discovery), and mandrake move..

    25. Re:Better drivers and licensing please by Anonymous Coward · · Score: 0

      Oh, as to your argument regarding the likelihood of Linux users buying higher end gear - uh-huh. Except for the number of Linux users who heavily re-use old gear and the fact that the people who want the high-end video cards are GAMERS, who are significantly less likely to run Linux.

      It makes perfect sense to cater to your market; but the market for video card makers is NOT Linux.


      And you know this how? I'm a gamer, coder, and sysadmin.. and I just bought a 6800 series Nvidia for my machine and I love it. I run linux native games as well as Windows stuff under Cedega (WineX).

      On your other post: Indeed it may cost more to develop linux drivers per unit sold, but in that case I'd even pay a couple of bucks for a linux driver CD. As long as they're making a profit, what the hell do they care?

      Please run a bullshit filter over yourself before hitting the post button. Thanks.

    26. Re:Better drivers and licensing please by Anonymous Coward · · Score: 0

      Four dollars?!? Mac users are accustomed to paying about $100 more for niche platform support. You can barely stamp a CD for four bucks.

    27. Re:Better drivers and licensing please by Kiryat+Malachi · · Score: 1

      You fall into the familiar trap of assuming that your experience translates to the general world. I have a custom built piece of hardware designed to make a certain type of sound; does that mean everyone does? Further, I have a car. I've gone to the trouble of refitting my car audio system; this doesn't mean that the majority, or even that a significant percentage of people, have done so.

      How do I know this? I know this because I'm friends with a lot of gamers, and almost none of them run Linux. Further, I know this because the general statistics back me (where are the Linux game translaters? If 60% of the market is Linux then there's a lot of money there). The only bullshit is in seeing yourself in everyone. It's nice that you like to play games under Linux, but you are by no means the rule.

      --

      ---
      Mod me down, you fucking twits. Go ahead. I dare you.
      (I read with sigs off.)
    28. Re:Better drivers and licensing please by Advocadus+Diaboli · · Score: 1
      90%+ of the market is Windows. To support your 10% slice of the market is vastly more expensive per sale.

      But on the other way round you have 2 options:

      1. Supporting those 10% in a lousy way and probably lose them because customers get pissed of by your Linux support policy
      2. Supporting those 10% in a good way and increasing your sales because customers that got pissed off by other companies now buy your product because of the good support

      Its sometimes really annoying to see companies complain about little markets but doing not a damn to develop those markets. And they should be aware, that sales of Linux desktop PC are growing, so if they are not waking up they will lose in the long run.
    29. Re:Better drivers and licensing please by Anonymous Coward · · Score: 0

      "Add in that Microsoft through its Windows Harward Quality Labs (WHQL or whatever they call it now) basically will subsidize your development costs,..."

      WHAT? I've spent the last 2 months nearly fulltime trying to get a driver through WHQL, with no compensation from MS.

      Their instrumentation for the coverage test breaks the driver. It took about 6 hours with IDA to prove that it was their coverage probes that were causing the bluescreens. Without that proof (and barely even with the proof) they claimed it was clearly a driver bug.

      MS owes me millions in lost time and sanity for this one. Developers pay up in time and money so that MS won't scare your customers with 'UNSIGNED DRIVER BAD BAD BAD' warnings. It's extortion. Those fucking assfucks.

    30. Re:Better drivers and licensing please by Karora · · Score: 1

      There are well-founded alternative arguments as well.

      If you have two companies wrestling for market share in a particular area, but if one of them is the no-contest, hands-down winner in a 10% segment and they are then only really fighting over the 90% remainder, then that company has an advantage.

      That 10% of the market (20% of their own market share) is like a base they don't have to worry about. If the company that doesn't compete in the 10% is totally successful, they have a competitor who can still come back from that base. If the 10% one is totally successful they have wiped out the competition.

      But there are more subtle reasons for supporting the wider market - it isn't simply about 'we can make money selling video cards to 10 million Linux users'. Those 10 million users are almost by definition not "average" users. Of the people I know who run Windows, vs the people I know who run Linux, there are many many more of the Windows people who are just "users", and many many more of the Linux people who are "early adopters" and get looked to for advice by their friends. That's an important market to cultivate, if you're trying to sell the latest and greatest hardware. ATI seem to be making the mistake of thinking that only hardcore gamers influence people's purchasing decisions, but that is soo not the case.

      Also, your points about driver development costs are not necessarily true. I have seen people purport to know exactly the oppposite from their own experience, i.e. that WHQL costs $$$ to the developer. Other people have asserted that the driver architecture under Windows is somewhat baroque and difficult to write. A company that did the job right should be able to reuse an awful lot of code across platforms in any case.

      --

      ...heellpppp! I've been captured by little green penguins!
    31. Re:Better drivers and licensing please by IncohereD · · Score: 1

      usually all you need to get linux drivers developed is some documentation.
      it's much cheaper than developing a windows driver, that's for sure.


      Unless you have to pay huge sums of money to other companies, whose propriety code you licensed to use in your own driver. Sad but true.

    32. Re:Better drivers and licensing please by julesh · · Score: 1

      On my machine, /lib/modules/`uname -r`/build is a symlink to /usr/src/linux-`uname -r`.

  10. It's a driver thing by bersl2 · · Score: 1

    What else could it be?

  11. Linux Gaming, In Summary by TrollBridge · · Score: 4, Insightful
    From TFA:

    "High performance gaming on Linux certainly isn't for everyone. We spent weeks preparing for this analysis and we still ran into problems that we could not correct. So many times, we came to a solution for a problem only to find our Linux distribution had some files in a slightly different place or our file dependency tree was completely broken. These are the things that scare away people from Linux."

    That is the 100% gospel truth. I couldn't have said it better myself. How then will the Linux community and game publishers overcome this (IMHO) enormous obstacle?

    --
    There's a Mercedes gap too. I want one and can't afford one, but it's not government's job to do anything about it.
    1. Re:Linux Gaming, In Summary by danheskett · · Score: 4, Insightful

      Part of the problem is the flexibility that Linux affords.

      A highly-componentized system is great for flexibility, but a nightmare for usability and developers.

      Microsoft's big-man has a penchant for screaming "Developers Developers Developers", but's he right. Developing a graphics intensive app on the Microsoft platform is a different ball of wax completely from developing on a *nix machine. DirectX works well enough that the developer is able to focus on developing things that add value to their product. There is a lot of that coming from the SDL camp, but truth be told DirectX is a success in the Windows world because of the underlying architectural differences: DirectX being install is a boolean operation: installed or not. (Well, now actually its actually a matter of versions.. all Windows distros have it.. so its just a matter of what version you want to deal with).

      Windows has a great success at offering games and graphics heavy packages because of the broad hardware support, the straightforward development model for drivers and the resources MS provides for brand-name developers, and the unified relentless promoting and development of DirectX as a platform.

    2. Re:Linux Gaming, In Summary by Nos. · · Score: 2, Interesting

      As someone who keeps a Windows 2000 box around 95% for gaming, its nice to see however, that things have come a long way since I last looked seriously at Linux Gaming. Of course the game I play most (Half Life - well, its mods anyways) has no native support, and it doesn't look like HL2 will either. However, games like UT2004 and Doom 3 are among games that I truly enjoy playing and do have native Linux binaries. Seeing this article that says installing drivers for my Nvidia card are now simple (at least in Suse, I usually run Fedora) makes me think about going dual boot on my gaming machine just to start trying a conversion to Linux gaming.

    3. Re:Linux Gaming, In Summary by MondoMor · · Score: 1, Informative

      The way they always do: arrogantly.

      Joe Gamer will be told to fix the patch himself, or encounter condescending jerks telling him everything but a workable solution on forums across the web. He may find a HOWTO, but will be discouraged when he realizes all the HOWTOs are written for computer science majors who are familiar with every other aspect of GNU/Lunix EXCEPT the subject of the HOWTO.

      So instead of being able to use Lunix as he wishes, he'll have to deal with an annoying kluge until he either gets tired of it and gives up or the weekly version upgrade of package X and dependent package Y stumble on a fix.

    4. Re:Linux Gaming, In Summary by TrollBridge · · Score: 2, Interesting
      "A highly-componentized system is great for flexibility, but a nightmare for usability and developers."

      Which is why I believe the "desktop Linux" is doomed to fail, ultimately.

      For servers, you want all of the customizability you can get. You want users who can tweak the most minute of details, because the system's performance and security depends on it.

      On the other hand, that doesn't bode well for desktop applications, which rely on certain system assumptions in order to work properly. Microsoft's OS model makes for weak security, in this, but when you buy an application made for Windows, you can be pretty certain it's going to work out of the box.

      --
      There's a Mercedes gap too. I want one and can't afford one, but it's not government's job to do anything about it.
    5. Re:Linux Gaming, In Summary by SalsaDoom · · Score: 0, Interesting

      Well, to be fair, the article said installing the nVidia drivers was easier then it was in windows for SUSE. Thats pretty good if you ask me. Its just ATI that is acting badly when it comes to these drivers, if you read the article you'll see that basically all the problems where with ATI's crapware drivers.

      That said, I have an ATI card in my laptop, and it runs relatively slowly compared to what it would have if I went with windows, but it was easy to setup and use. I think SUSE might be involved in a lot of those problems...

      But really, yeah, gaming in linux isn't for the average person really unless he has an nVidia, then it really should not be too hard...

      And as far as game developers go... its not their problem to make distros easy. They should just make it work on their machines and go from there. nVidia's basically proved that binary drivers can be done well on linux (and fuck you to the zealots trying to ruin it for the gamers by their constant bitching) and when ATI catches up -- and word is they are going to be doing a driver push soon -- things should be better.

      We'll just wait and see what happens!

      --SD

      --
      "Computers will never truly be free until the last windows user is strangled with the entrails of the last mac user."
    6. Re:Linux Gaming, In Summary by bersl2 · · Score: 2, Interesting

      How then will the Linux community and game publishers overcome this (IMHO) enormous obstacle?

      Maybe if a site analogous to the Linux Counter were established, to count the number of Linux gamers, recording hardware, games played under Linux, games we wish we were playing under Linux, etc...

      The root of the problem is a lack (though obviously it's not as bad as it was) of communication. So maybe if we could tell more companies that Linux gamers are numerous enough to be worth pleasing, we could get some results.

    7. Re:Linux Gaming, In Summary by gosand · · Score: 1
      From TFA: "High performance gaming on Linux certainly isn't for everyone. We spent weeks preparing for this analysis and we still ran into problems that we could not correct. So many times, we came to a solution for a problem only to find our Linux distribution had some files in a slightly different place or our file dependency tree was completely broken. These are the things that scare away people from Linux." That is the 100% gospel truth. I couldn't have said it better myself. How then will the Linux community and game publishers overcome this (IMHO) enormous obstacle?

      By having static packages with everything included in the package that you need to run.

      I recently ran into a dependency (I think) problem when trying to install BitTornado. I had an older bittorrent client running fine, but my dsl modem was getting futzed up. I was just playing with bt, not doing any serious downloading or anything, just trying it out. So I wanted to download and try BitTornado because it supposedly had some built-in diagnostics. I could NOT get it working. Some dependency problem I think with wxPython. I got tired of F'ing around with it, trying to figure it out, so I fired up the old Windows box, downloaded and installed BitTornado with NO PROBLEMS!! Sometimes I have to just shake my head and wonder why it can't be this easy on Linux. OpenOffice.org seems to not have a problem including everything you need in one package. Why does EVERY other app on Linux? Why is it this still a problem? It has been a problem as long as I have been using Linux (RH5.x timeframe).

      --

      My beliefs do not require that you agree with them.

    8. Re:Linux Gaming, In Summary by theantix · · Score: 1

      How? Very simple... let the distribution manage the updates for you. If you are using free software it shouldn't be a problem -- and if you aren't a free software zealot for drivers then use a distribution that manages the binary graphics drivers for you. It's not that hard.

      "Linux" is a shitty platform for binary compatibility though, and all indications suggest that this will continue to be the case for the near future at least. Don't expect non-free games and other software binaries to work with "Linux" anytime soon, instead they will work with a subset of supported distributions.

      In the long run, there may be a LSB-alike for games, but that is a long way off still.

      --
      501 Not Implemented
    9. Re:Linux Gaming, In Summary by tuffy · · Score: 2, Interesting
      Seeing this article that says installing drivers for my Nvidia card are now simple (at least in Suse, I usually run Fedora) makes me think about going dual boot on my gaming machine just to start trying a conversion to Linux gaming.

      I'm running an nVidia card on Fedora 2 (x86-64) and the installation went flawlessly. After a few trivial X11 config changes to let the X server know about the new video card, DVI and accelerated OpenGL worked like a charm.

      --

      Ita erat quando hic adveni.

    10. Re:Linux Gaming, In Summary by southpolesammy · · Score: 1

      I thought this excerpt from the article was more indicative of the Linux/Gaming situation:

      Unfortunately, the decision to buy new hardware constantly goes hand in hand with the decision to play some new game - and if it's a gaming machine you want, then Linux isn't the operating system that you need.

      IOW, new hardware (especially graphics cards) is purchased in order to run software that can not adequately run on yesterday's gear. Even though the graphics card manufacturers do release Linux drivers, it's the game companies that drive their adoption, and overwhelmingly, this pushes gamers towards Windows.

      Of course, you really can't blame the game companies for doing this -- 99% (if not more) of their revenues will come from Windows users. It's a bad risk to put money into developing a Linux offering until you know (1) that the Windows offering is going to succeed, and (2), that there is a critical mass of willing Linux buyers to offset the cost. Otherwise, it doesn't make business sense to do so.

      --
      Rule #1 -- Politics always trumps technology.
    11. Re:Linux Gaming, In Summary by untermensch · · Score: 1

      "A highly-componentized system is great for flexibility, but a nightmare for usability and developers."

      Which is why I believe the "desktop Linux" is doomed to fail, ultimately.

      But that's exactly why we have custom distro's to handle these things. Mandrake/Suse/Gentoo/etc. simply write wrappers (RPMs/ebuilds/etc.) around apps that handle the problem of tweaking each app to fit their particular distro.

      Admittedly, they don't always do a fantastic job of this, particularly when close-source drivers are involved, but I don't think this is an insurmountable problem that is going to destroy Linux's shot at the desktop

    12. Re:Linux Gaming, In Summary by Kethinov · · Score: 0, Troll
      Which is why I believe the "desktop Linux" is doomed to fail, ultimately.
      Uh, they got these great things called standards. When people unite behind them, stuff works better. Linux just needs more and better standards and it's getting better all the time. Quit singing the *BSD^H^H^H^H Linux is dying tune cause it just hasn't happened yet and it won't any time soon so long as it remains free, open, and organized.

      I'd be more worried about its closed source corporate controlled counterparts. They could fail and/or cease development at ANY TIME should the company act accordingly. Quite frankly, using software controlled by a relatively small group of people doesn't sit well with a number of people. Using open source unrestricted software unencumbered by EULAs developed by millions worldwide just makes sense. Software as a corporate product is a flawed business model. That is, when you take anti competitive tactics out of the picture.

      Look what capitalist American has done to the computer industry. Is this what Adam Smith had in mind when he wrote The Wealth of Nations?
      --
      You're right, I wouldn't steal a car. But if it were possible, I sure as hell would download one!
    13. Re:Linux Gaming, In Summary by geomon · · Score: 1

      I couldn't have said it better myself. How then will the Linux community and game publishers overcome this (IMHO) enormous obstacle?

      They won't.

      Just throw your collective hands in the air and admit that everyone should just use Microsoft products to play games.

      Or Nintendo.....

      Or Playstation.....

      --
      "Rocky Rococo, at your cervix!"
    14. Re:Linux Gaming, In Summary by Anonymous Coward · · Score: 0

      Arguing the philosophy of OSS vs. closed source is irrelevant, from my point of view. I'm looking at what actually works. From a practical point of view, it's REAL hard to justify Linux over Windows in terms of desktop applications. Linux needs to get some consistency to allow a wide range of apps to run on it, if Linux is ever to be successful as a desktop OS.

    15. Re:Linux Gaming, In Summary by Anonymous Coward · · Score: 0

      Please read the article, and let us know how your idealogy played out in terms of real-world results.

    16. Re:Linux Gaming, In Summary by gears5665 · · Score: 1

      That is the 100% gospel truth.

      Thats funny...considering that God doesn't exist and the bible is a load of fictional stories written by men who couldn't get laid to oppress women and the poor.

      I think you need to start using a different phrase for "undisputed fact".

      but hell, what do I know?

    17. Re:Linux Gaming, In Summary by TrollBridge · · Score: 1
      "but hell, what do I know?"

      If God doesn't exist, what is this "hell" you speak of? ;-)

      --
      There's a Mercedes gap too. I want one and can't afford one, but it's not government's job to do anything about it.
    18. Re:Linux Gaming, In Summary by JonKatzIsAnIdiot · · Score: 1

      Part of the problem is the flexibility that Linux affords. A highly-componentized system is great for flexibility, but a nightmare for usability and developers.

      Very true, but these hassles can be mitigated by good organization and management schemes. Linux is way ahead of Windows here - library versioning, package management systems - but there is more work to be done.

    19. Re:Linux Gaming, In Summary by gears5665 · · Score: 1

      If God doesn't exist, what is this "hell" you speak of? ;-)

      It is a cultural literary reference which is a throwback to a time when the only book people owned was the pack of fiction called the bible and people didn't understand the science behind the world around them. It's a great set of stories and necessary reading to get along with educated people in Western Society. However, it should never ever be taken as a book of "facts"; just a fun read.

    20. Re:Linux Gaming, In Summary by Anonymous Coward · · Score: 0

      but hell, what do I know?
      Obviously, not much about the historical accuracy of the Bible.

    21. Re:Linux Gaming, In Summary by Kethinov · · Score: 1

      Isn't it obvious?

      - MS monopolizes the industry.
      - Alternative OS attempts competition but saturates the market a fraction of a percent
      - Hardware manufacturers are reluctant to release any drivers for alternative OS, or they do release drives which suck (like ATI)
      - Software authors are reluctant to write software or port to alternative OS due to its low saturation.

      Software monoculture is an extremely bad thing. Especially when the dominent OS is closed source, poorly engineered, and is actively developed in an anti competitive way. These anti competitive tactics have real-world results. The stifle competition's ability to compete.

      --
      You're right, I wouldn't steal a car. But if it were possible, I sure as hell would download one!
    22. Re:Linux Gaming, In Summary by Atzanteol · · Score: 3, Informative

      I believe that's what the Linux Standards Base people are trying to fix. I've spoken to them at Linux world. Their goal is to eventually be able to have developers say "This program can run on any distro LSB1.0 compliant."

      --
      "Ignorance more frequently begets confidence than does knowledge"

      - Charles Darwin
    23. Re:Linux Gaming, In Summary by Kethinov · · Score: 3, Insightful

      First off, a wide range of apps already runs on Linux. I don't have a problem using it as my desktop OS, nor does anyone else I know once you factor out the difficulty of installing, which is negligible. In fact, my computer illiterate anti-technology parents don't have a single issue using pre installed Linux. It's all the same to them. Click icon here, browse web. Click icon there, read email. Click icon up there, play solitaire.

      "Oh, but you maintain it for them!"
      Sure I do. But I'd be doing the same damn thing if they ran Windows. Except then I'd be scanning for viruses, removing spyware, etc. There's really no difference.

      As for consistency, that's a mute point. There already is a huge number of statically compiled apps which run in any distro. And apps that aren't are provided by your distro's package manager. Why is consistency even an issue at this point? Because you want a ubiquitous distro? Sure, that'd be nice but the world doesn't work that way atm.

      If you're talking about ease of use and installation, I'll have sympathy for you. But once installed, the Quake3 tastes just as sweet. All it needs is a little popularity and the real world performance materializes nicely. Stuff like that is proof that Linux IS ready for the desktop and that people are just unwilling to change due to some legacy nonported application or just their unwillingness to learn something new.

      To such people I say "what do you really need windows for?" In most cases the answer is nothing except proprietary games. In which case I convince them that games are not as important as running a proper and moral OS (in the sense of free and open as well as unpirated), or I at least encourage them to dual boot. (Even I do that.)

      Besides, Linux gets more games all the time. There are only TWO games left that necessitates my windows install. When I stop playing them or they are ported, I will never maintain a windows install again.

      You say Linux needs consistency to be a successful desktop OS and I say it needs time. I'm switching people. Friends of mine are switching people. Convincing people to drop the MS monopoly like the bad habit it is is a painfully one-person-at-a-time process.

      But each person is worth it.

      --
      You're right, I wouldn't steal a car. But if it were possible, I sure as hell would download one!
    24. Re:Linux Gaming, In Summary by Taladar · · Score: 1

      I know...Gentoo-Zealot-bla...but

      emerge bittornado

    25. Re:Linux Gaming, In Summary by be-fan · · Score: 1

      The biggest problem with gaming on Linux is that manufacturers don't put any effort into it. For stuff that people pay attention to, things are really nice. Eg, installing NVIDIA's drivers in SuSE is as easy as clicking a button in YaST, changing settings is as easy as using the provided nvidia-settings graphical app, and installing provided games is as easy as selecting them in YaST.

      Unfortunately, the commercial vendors (with some exceptions) don't put the kind of effort they put into their Linux products that they put into their Windows products. They write fancy graphical installers for Windows, then ship some lame buggy shell-script for Linux. ATI writes a gadget-filled graphical control panel for Windows, but makes you edit XF86Config on Linux.

      --
      A deep unwavering belief is a sure sign you're missing something...
    26. Re:Linux Gaming, In Summary by Anonymous Coward · · Score: 0
      "High performance gaming on Linux certainly isn't for everyone. We spent weeks preparing for this analysis and we still ran into problems that we could not correct. So many times, we came to a solution for a problem only to find our Linux distribution had some files in a slightly different place or our file dependency tree was completely broken. These are the things that scare away people from Linux."
      That is the 100% gospel truth. I couldn't have said it better myself. How then will the Linux community and game publishers overcome this (IMHO) enormous obstacle?
      Jeese, I dunno. I'd say capture a decent market share then implement a large scale help desk program powered by flow charts read by people who know nothing about computers... I mean really... Isn't this the biggest problem for all "power users" on all OSes? It seems to me that it's just more cost effective to have some person in India on a phone to say "Oh, [follows flow chart to] you're using [OS][ver.]. You need to [go to some OS specific setting/location] and change [some OS specific setting/location] to [something else]. Then it will work just fine!"

      Here's the deal though... If you're on the "cutting edge" in any OS including Windows(tm) You're in the same boat. You find a "solution" then modify it to meet your needs...
    27. Re:Linux Gaming, In Summary by Anonymous Coward · · Score: 0

      So if you want to install an application you to go through your distro? And what if, for a reason or another, your distro doesn't want you using the applications?

      Even Microsoft is not doing this!

    28. Re:Linux Gaming, In Summary by Anonymous Coward · · Score: 0

      You're right... Linux on the desktop is not dying. Which is not that surprising since it's not even born yet!

    29. Re:Linux Gaming, In Summary by irc.goatse.cx+troll · · Score: 1

      Or under a real distro, apt-get install bittornado.

      Why the fuck would someone remake ports using python when anyone with half a clue and some desire has had the REAL ports under linux since back in the linux 2.2 days? I don't have the link anymore (I since found out that debian is better), but you can find the ports tree in a nice tarball off sourceforge that install clean and works, goes nicely with a slack install.

      --
      Pain lasts, kid. Its how you know you're alive. Sometimes I think this growing up thing is just pain management-TheMaxx
    30. Re:Linux Gaming, In Summary by Anonymous Coward · · Score: 0

      > Uh, they got these great things called standards

      The Linux world has no such thing, and there's no movement to create such standards. Developers can't even rely on whether X is installed or not!

      The onlything that's close to a real ABI is LSB, which is so minimal to be useless for most real world applicaitons.

    31. Re:Linux Gaming, In Summary by kiddygrinder · · Score: 1

      Excuse me for being pedantic, but it's moot, not mute. Thank you for your time.

      --
      This is a joke. I am joking. Joke joke joke.
    32. Re:Linux Gaming, In Summary by Kethinov · · Score: 1

      lol
      Thanks, I'll keep that in mind next time.

      --
      You're right, I wouldn't steal a car. But if it were possible, I sure as hell would download one!
    33. Re:Linux Gaming, In Summary by gosand · · Score: 1
      Or under a real distro, apt-get install bittornado.


      You and the parent to your post make my point EXACTLY. How do you install it on Linux? Well, if you are using Gentoo, blah blah. If you are using Debian, blah blah. If you are using $distro, $blah $blah. It may work, it may not. I have never used Gentoo, but I have used a variation of Debian. Apt-get didn't work all the time.


      If you want to install something on Windows, you have basically 2 options - either put the files in some location and run them (i.e. no need to go through an installer), or run a setup program and it will install it for you. Only on rare occasions are there dependencies. On Linux it is "normal" to have problems installing things. I have been using it faithfully for 5 years, but even I can see that the Emperor has no clothes.

      --

      My beliefs do not require that you agree with them.

    34. Re:Linux Gaming, In Summary by irc.goatse.cx+troll · · Score: 1

      I was actually just trolling against gentoo, but if you want to talk linux as a whole, grab quake3's linux binaries(even just the demo) and you'll find they install just as easily as on windows-- Its a nice shell script with a tarball encoded into it that can extract/run itself, complete with a cute little gtk wizard. I'd honestly be suprised if doom3 doesnt use the same system (this is how they distributed rtcw and et also, so its sort of their defacto standard).

      The real problem with linux gaming isnt installing, its answering questions like "Why doesnt my sound work?" Explaining while in linux a single app can lock your soudncard vs windows where I regularly have aim im sounds notifying me who signs on and off while winamp plays some good music and my game has no problem. Hell, I can even switch from headphones to speakers entirely in software(TurtleBeach SantaCruz)

      --
      Pain lasts, kid. Its how you know you're alive. Sometimes I think this growing up thing is just pain management-TheMaxx
  12. It's all Mesa's fault! by Anonymous Coward · · Score: 4, Interesting

    If it weren't for the spinless Mesa developer for switching away from LGPL licence just to please the incompetent xfree crew, we would have major companies contributing open codes, instead of having them taking stuff and releasing slower, buggy, proprietary drivers.

    Remember ALSA? It sticks its gun to GPL right down to the driver, and Creative actually donate SBLive driver for it, when the company was already crushing everyone else (Aureal included) sound card market! This should be how Mesa license the code, not the lame, bogus, xfree licence.

    1. Re:It's all Mesa's fault! by Anonymous Coward · · Score: 0
    2. Re:It's all Mesa's fault! by Tim+C · · Score: 2, Insightful

      we would have major companies contributing open codes, instead of having them taking stuff and releasing slower, buggy, proprietary drivers

      Slower? Please explain to me how you've come to the conclusion that NVidia's closed source drivers are slower than the open source ones, especially given that the open source ones don't support 3D acceleration at all?

    3. Re:It's all Mesa's fault! by xenocide2 · · Score: 1

      I don't know if you're aware of this, but mesa is intended as a software rendering replacement for 3d hardware. The "code" for drawing polygons is actually a set of circuits; all the drivers do is feed the hungry beasts and put the gory remains on display.

      The two concepts are incompatible; good luck on that.

      --
      I Browse at +4 Flamebait

      Open Source Sysadmin

    4. Re:It's all Mesa's fault! by joib · · Score: 1

      Mesa started as a software renderer, but since, uh, the DRI thing in xfree86 4.0 many moons ago, Mesa has also provided hardware support.

  13. Wow by Anonymous Coward · · Score: 0

    ATI has poor drivers? What a shocker. News flash at 11.

  14. "Penguin flavored" by Anonymous Coward · · Score: 2, Funny

    Mmmm.... Penguin flavored. *drool*

  15. Totally off-topic, but need Linux advice.... by Anonymous Coward · · Score: 0, Offtopic

    I noticed this test was run on SuSE. What's the "best" Linux version to run if I'm an experienced developer, comfortable in UNIX, but never tried Linux? I don't know much about Linux, but would like to put together a system for home tinkering. Is SuSE a good choice? Sorry for the Off-Topic post, but I figure picking the OS is as important as picking the GPU.

    1. Re:Totally off-topic, but need Linux advice.... by stratjakt · · Score: 1, Insightful

      In the end, they're all the same on the inside.

      All the distro-pimping is for kids and zealots.

      --
      I don't need no instructions to know how to rock!!!!
    2. Re:Totally off-topic, but need Linux advice.... by Ingolfke · · Score: 0, Troll

      You're obviously new to /. (see I did the slashdot symbol so I'm l33t). My point is that, if you weren't new you would have known that you never have to be on topic here. Just draw some loose (and I use that word in the loosest sense) relationship between the topic you have a question about an the topic of discussion and you'll be right at home. For example if you have a question about how install bionics into your 3-month old baby using only spare parts from old Amiga you bought on e-bay, but the topic of discussion is "Microsoft evil Satan destroyer of souls or evil demonbeast destroyer of worlds?" you would lead in with something like this...

      I hate Microsoft, and cannot stand Bill Gates, and oh-my-gosh do I love Linux, and you-guys, like, I've got a poster of Linus on my wall, and I have this 3-month old child who I want to install bionics into and I need to run linux on the kid, and I only have old Amiga parts. Has anyone done this before?

      See... that's easy.

      So to your question about which Linux version is the best version. This is very easy. Knoppix.

      Many people will disagree with me, and they are all right, except for people who say:

      Gentoo - It's dead. Like BSD dead.
      Debian - It's stable, but smells funny... stale and pungent.
      Suse - It's just Redhat, with a different name.
      Redhat - Very popular, but owned by an evil corporation set on destroying the rainforests.
      Slackware - Was the best prior to kernel version .7
      Fedora - Think Redhat, but w/ a hint of the stink of Debian.
      Knoppix - Meets all your needs. Easy installation. Don't mind the difficulty in getting persistant storage setup.
      SCO Unix - It's like linux, but well I've heard there identical.

      Whatever you choose make sure you talk alot about your cooling (say it's liquid cooled even if that just means you rest your beer on the case from time to time) and also brag that your going to setup a beowulf cluster of these someday soon. Once you get the baby setup w/ the bionics.

      Don't let anyone tell you that BSD is an option. BSD is no longer being developed. The government was funding it, but quit, when they realized it was dead. Something about using old birth certificates or patents from the 50s. I don't know.

      Anyways... I hope that helps.

      Slashdot: News and commentary on par w/ CBS.

    3. Re:Totally off-topic, but need Linux advice.... by Anonymous Coward · · Score: 0

      What's the "best" Linux version to run if I'm an experienced developer, comfortable in UNIX, but never tried Linux?

      Any one you pick is fine. Just be sure to buy an SCO license for it.

      -Daryl

    4. Re:Totally off-topic, but need Linux advice.... by Sarcastic+Assassin · · Score: 0

      For the least painless install possible, try MandrakeLinux. It also is very stable once set up. However, it does many things for the user (eg, partitions), so may find you'd like to be more hands-on. SuSE is a good distro to start on too, as it also has a pretty painless install.

      Also, if you plan to dual-boot (Linux/Windows), PartitionMagic is a must.

    5. Re:Totally off-topic, but need Linux advice.... by Anonymous Coward · · Score: 0

      You friggen idi0t. You must work for Micros0ft or somethin'. Go hug a tree you John Kerry lovin' freak!

      Linux is the best 0S out there, bar none... Suse is good, Redhat is better, and Gentoo is emerging!

      And about the bionics for your child... thes guys might be able to help http://www.amuc.ab.ca/news.html.

      Nobody needs to teach me how to r0ck!

    6. Re:Totally off-topic, but need Linux advice.... by HumanTorch · · Score: 1

      In the end, they're all the same on the inside

      Not totally the same.. install the latest kernel src rpm (not the same as the kernel-source rpm) for Fedora and you will see Redhat applies a bunch of their own patches to the stock kernel.

    7. Re:Totally off-topic, but need Linux advice.... by Anonymous Coward · · Score: 0

      Sir, you are absolutely out of line. BSD is a thriving operating system. Have you never heard of FreeBSD, OpenBSD, and NetBSD? These operating systems are maintained by at least 7 different people. Why even here on Slashdot you'll see there is a seperate section dedicate entirely to BSD. This forum and these operating systems are used by at least 107 people.

      Not only is BSD the world's most secure and open operating system in the world, it is extremely easy to use. Before I started using BSD I was using Linux for about 6 years, and Solaris and HP-UX before that. Once I switched to BSD, unencumbered by GUI interfaces, web servers, TCP/IP, and all the other "inventions" so frequently touted as progess I was able to easily produce text files in almost six weeks. I had to port vim from the source, edit it for my Amiga OS, and strip out most of the featurs so it would run in the free memory I have, but man it was awesome. I felt so free. Security and portability are integrated into BSD. You can configure a firewall, router, security, and a VpN in less than 3 days using the very friendly command line interfaces, man files, and well... you don't need any gui help interfaces. It even has lynx.

      Anyways. BSD is definitely not dead. Me and 106 other people prove you wrong. If you're looking for a dead or dying OS try these on for size: option #1 and option #2.

      Oh yeah, one of the best thing about BSD is that it's not encumbered by the viral GNU license or misappropriated intellectual property. Ditch Linux, ditch Windows, ditch VMS, get yourself BSD.

    8. Re:Totally off-topic, but need Linux advice.... by Anonymous Coward · · Score: 0

      I agree BSD is dead, and Gentoo is on the way out. You mischaracterize Linux though and Linux users thoug. I bathe regularly and use Fedora, I'm actually CFO for a small financial services firm in SouthEastern Kansas (MBAs and financial types do know how to use Linux). I can't vouch for the Debian distro though. Haven't found the time to use it.

  16. I don't know by temojen · · Score: 1

    But I run Linux-2.6.8.1 with a Geforce FX 5700 and it works fine. I suspect you may have old information.

  17. You can blame it on Mesa by Anonymous Coward · · Score: 0, Troll

    If it weren't for the spinless Mesa developer for switching away from LGPL licence at version 3.1 beta 2 just to please the incompetent xfree crew, we would have major companies contributing open codes, instead of having them taking stuff and releasing slower, buggy, proprietary drivers.

    Remember ALSA? It sticks its gun to GPL right down to the driver, and Creative actually donate SBLive driver for it, when the company was already crushing everyone else (Aureal included) sound card market! This should be how Mesa license the code, not the lame, bogus, xfree licence.

    Thanks to the moronic decisions of the Mesa team, it is too late to turn back now.

  18. Things that scare away people from Linux by Anonymous Coward · · Score: 1, Insightful

    An excerpt from the article:
    "So many times, we came to a solution for a problem only to find our Linux distribution had some files in a slightly different place or our file dependency tree was completely broken. These are the things that scare away people from Linux. Although customizing our own system, contravening the Microsoft "monopoly" and roughing-it-on-our-own were refreshing and challenging, this editor immediately fired up the Tribes: Vengence demo on Windows after the Linux testing and editing were complete. Total time to install and configure: 5 minutes, 40 seconds; now that was refreshing."

    1. Re:Things that scare away people from Linux by Anonymous Coward · · Score: 0

      It took him 5 minutes to install the Tribes demo? What is he a retard who had a hard time finding the "Next" button to click?

  19. Nvidia needs DRI???? by Anonymous Coward · · Score: 0

    "NVIDIA's drivers provide Direct Rendering Infrastructure (DRI) support via SaX2 (the SuSE X configuration tool) as well. The DRI acts as somewhat of an abstraction layer between X Windows and OpenGL. We need the DRI module loaded with X Windows in order to run any OpenGL hardware acceleration. Without it, we are only running software acceleration."

    This is absolutely wrong, Nvidia uses GLX, I haven't read the entire article yet, but this is utter crap. And it also isn't true that you need to recompile the kernel for both drivers...

    Is it just me?!

    1. Re:Nvidia needs DRI???? by temojen · · Score: 1

      GLX uses DRI... DRI is how X talks to the video card, not how applications talk to X. GLX is how applications talk to X.

      Both sets of drivers have to be compiled from within the source tree of whichever kernel version you're running, but you don't nescesarily have to recompile the whole kernel.

    2. Re:Nvidia needs DRI???? by Anonymous Coward · · Score: 0

      So, you're implying that DRI is needed for acceleration of the X framebuffer?! Or implying that DRI is needed for actual hardware rendering in games?

      I can assure you that isn't correct, I'm running gentoo and I haven't compiled DRI into the kernel. In my understanding, DRI is simply an alternative to GLX.

      Please correct me if I'm wrong, but if DRI is needed for X acceleration, then I've been making a remarkable feat for some time.

    3. Re:Nvidia needs DRI???? by Anonymous Coward · · Score: 1, Interesting

      GLX is a an API for writing OpenGL applications for X. DRI is a common interface for OpenGL drivers to write directly to hardware, bypassing the X rendering layer. They're completely different things, and DRI is pretty much mandatory if you want decent OpenGL performance.

      The exception is the nVidia drivers which don't use the common DRI, instead providing their own mechanism for direct rendering. I assume this is related to the amount of common code between their Windows and Linux drivers - easier for them to maintain their own framework than to support DRI on Linux and something else on Windows.

  20. Re:Indeed by Anonymous Coward · · Score: 5, Informative

    For Windows users not familiar with the process, the kernel must be completely recompiled for ATI or NVIDIA drivers to work.

    I don't know about the ATI drivers, but this isn't true for the NVIDIA drivers. You can download an installer from NVIDIA that will create a kernel module for you and places it with the other modules. No need to recompile the kernel at all. Just load the module (if the installer doesn't do this for you) and restart your X server.

  21. Lots of good uses for GPUs by temojen · · Score: 3, Interesting

    see gpgpu.org for more information.

    1. Re:Lots of good uses for GPUs by jpvlsmv · · Score: 1
      gpgpu.org

      For a moment there, I thought you were running encryption on your video card.

      --Joe
  22. Odd by temojen · · Score: 1

    When I got a new video card (Geforce FX 5700) all I had to do was "emerge nvidia-kernel nvidia-glx" then "opengl-update nvidia" and change the chipset and videoram in my XF86Config. All of this took me about 1/2 hour, including installing the card.

    1. Re:Odd by Anonymous Coward · · Score: 0

      But what about installing [most] games and getting them to work? Not so easy, and that was the point of the exerpt from TFA.

    2. Re:Odd by stratjakt · · Score: 1

      I did the same thing.

      Now, getting the games installed and running was another task altogether.

      I managed to get Wolf-ET installed after much screwing around, I still have no sound when I try to play it though.

      --
      I don't need no instructions to know how to rock!!!!
    3. Re:Odd by BillyBlaze · · Score: 2, Informative

      Well, that's the game's problem, not Linux's. And at least UT2004 has absolutely no problem - just run the GUI installer.

    4. Re:Odd by Enucite · · Score: 1

      In the article they did say installing the nVidia drivers in Linux was easier than installing them in Windows. So they aren't talking about installing nVidia drivers when they say they ran into problems.

      Read the article for details on the problems they ran into; most of the problems were with ATi's drivers and specific bugs/glitches in games.

    5. Re:Odd by nvrrobx · · Score: 1

      When I bought a new video card for my PC, I put the card in, inserted the CD and booted my machine. Windows installed a driver and told me to reboot. Upon reboot, everything was fine.

      As a Linux user, I have to tweak configuration files. Yeah, that's fine for me because I'm a geek. Try to get your mother to do that.

      Why should you have to put your videoram in your X config anyhow? It can figure it out from the card!

    6. Re:Odd by Anonymous Coward · · Score: 0

      You mean Linux does sound now? I've been running Fedora Core 2 without a peep from my speakers no matter what the app is. It just can't seem to find the sound card and I can't get it going no matter what I try.

    7. Re:Odd by ozric99 · · Score: 1

      Funny, I had exactly the opposite problem trying to get my Radeon8500 working properly in Gentoo. After multiple installations following all manner of differing opinions and help on the official gentoo forums I gave up and put win2k back on it.

    8. Re:Odd by DogDude · · Score: 1

      When I got a new video card (Geforce FX 5700) all I had to do was "emerge nvidia-kernel nvidia-glx" then "opengl-update nvidia" and change the chipset and videoram in my XF86Config. All of this took me about 1/2 hour, including installing the card.

      And how'd you figure out how to do all of this? I would have *never* figured this out. I can't imagine a whole lot of other people would be able to figure this out, either.

      --
      I don't respond to AC's.
    9. Re:Odd by O · · Score: 2, Insightful

      Ha! Try to get your mother to open up a computer and install a new video card first....

      --

      1, 1, 2, 3, 5, 8, 13, 21 -- Mathematics is the Language of Nature.
    10. Re:Odd by Googo · · Score: 1

      Well, based off the instructions that come with the card, I'm pretty sure most people could install the card. They just need to be brave and not be afraid that they will mess up the computer. It's usually a simple pull out old card and install new card. Then follow the instructions. Though I guess you could get stuff like driver conflicts and stuff, the card would still be installed.

    11. Re:Odd by jburroug · · Score: 1

      Well he's a Gentoo user, as the listing of emerge commands would suggest, and all of what he just wrote is in the Gentoo install documention. In fact it's a required step when you install Gentoo. Installing Gentoo can be a pain in the ass if you're in a hurry to get a working system but the end result of going through it's install process is that you know how to do shit like this by the time you have a working install :)

      --
      "Listen: We are here on Earth to fart around. Don't let anybody tell you any different!" - Kurt Vonnegut
    12. Re:Odd by Synn · · Score: 3, Informative

      That's the difference between a vendors that supports you and one that doesn't.

      NVidia isn't a problem under Linux because they actually put out decent drivers for it. But ATI support is horrid, because the company barely puts any effort into Linux drivers.

    13. Re:Odd by russint · · Score: 2, Insightful

      What? My (very non-technical) mother managed to install both a cd-writer and a video card the first time she opened up a computer

      Thats what manuals and common sense are for.

      --
      ^^
    14. Re:Odd by Aeiri · · Score: 1

      I'm a Linux user, and I wouldn't have figured that out, those are Gentoo specific commands. On Slackware, for nVidia drivers, you run the file you download, wait about 10 seconds, and it says "Finished". Restart X and you are done...

    15. Re:Odd by jovlinger · · Score: 1

      by searching the web for install nvidia geforce fx 5700 linux, and voila.

      Perhaps if you're lucky (and google somehow figures out the "install" bit from context) you'd end up on this comment.

    16. Re:Odd by Anonymous Coward · · Score: 0

      Strait, the problem is probably artsd is using the sound system and not letting the game get a look-in. Either disable artsd or ue the setting that makes artsd drop access to the soundcard if it is idle.

      Worked for me when I had problems with Wolfenstein.

    17. Re:Odd by gears5665 · · Score: 1

      funny, in windows, all I had to do was install the card and boot the system. It sucks that I lost my 5 days of uptime though.

    18. Re:Odd by be-fan · · Score: 1

      And got to www.nvidia.com, download the drivers, find where you put them, click on the executable, click next a half a dozen times, then reboot. I seriously doubt your way was quicker...

      --
      A deep unwavering belief is a sure sign you're missing something...
    19. Re:Odd by be-fan · · Score: 1

      Okay. She can follow a manual to open a case, plug in the IDE cable, and make sure the jumpers are right, but she couldn't follow a manual that told her to "download and run this installer" then open up a file and change "nv" to "nvidia"?

      --
      A deep unwavering belief is a sure sign you're missing something...
    20. Re:Odd by nvrrobx · · Score: 1

      With my mother, that's not an issue. But, that's beside the point.

      The same idea applies for plugging in a random USB device, however. I'm betting most people's parents would be likely to plug in a new digital camera via USB to download their images...

    21. Re:Odd by jdhutchins · · Score: 1

      I have to tweak configuration files

      There are several utilities to help you configure your video card, and they're getting better all the time. They may not be perfect yet, but they still work. Knoppix can auto-detect almost any video card, and it's only a matter of time (it's already happening, iirc) before that support gets ported to other distros. There are plenty of GUI configuration tools for linux, don't complain becuase you don't about them.

    22. Re:Odd by gears5665 · · Score: 1

      In windows, it was plug and play. The grandparent's description of a linux installation with the words "All I had to do:" is misleading because no sane human being (i.e. not an operating system hobbiest) would consider what he did trivial and lying to potential users about the ease of using linux is not the way to prepare people for working with Linux or maintain their trust for lifelong use.

      In legal terms, it is considered fraud and negligence to imply something is one thing when in fact it is another.

      These costs in time and learning are not trivial and setting up a driver requires higher skilled labor than in that other opterating system.

      Some of us are working on ways to make this easier, but the grandparent does not make the world a better place when their fanboy attitude does not convey the true difficulty of using and configuring a linux system. We can't improve a system if we don't identify the places where it is hard. If we lie to ourselves about what is hard, then it will never get better.

      Hence, in driver installation, linux sucks and windows rocks, but linux's potential is still blowing Windows out of the water as the future OS of the world.

    23. Re:Odd by nvrrobx · · Score: 1

      I certainly know about them.

      Changing to init level 3, logging in as root, installing the nvidia driver, running sax -n 0 nvidia, changing to init level 5 is not intuitive.

      And yes, I still end up tweaking configuration files because of things like this:

      Option "TwinView"
      Option "MetaModes" "1280x1024, 1280x1024; 1024x768, 1024x768; 800x600, 800x600; 640x480, 640x480"
      Option "NoLogo" "true"
      Option "RenderAccel" "true"

      Show me the GUI for all of that, and how to apply those changes without restarting X, and I'll happily eat crow.

      I fully understand it's all a work in progress, but it's not there yet. Until Linux is up to par with Windows for things like this, it can't be expected to overtake the desktop. Period.

    24. Re:Odd by be-fan · · Score: 1

      In windows, it was plug and play.
      Except it isn't. The NVIDIA drivers don't come standard with either Windows or Linux. You have to install them manually in both cases. You can use whatever comes with XP or XFree86 (without having to do anything), but nobody who uses their cards to actually play games does that. Everyone goes to nvidia.com to download the drivers.

      "All I had to do:" is misleading because no sane human being (i.e. not an operating system hobbiest)
      I think that's a bit of an exaggeration. "No sane human being?" We're talking about gamers here. They know how to edit text files (at least historically, many games could be tweeked this way), and anybody who uses Gentoo knows how to use emerge (just like anyone who runs Windows knows how to run a driver installer). Hell, there are even graphical tools to do what he did! As long as the driver is known to Portage (or APT or YaST, depending on your distro), installation is a highly automated process, unlike in Windows. On SuSE, his steps are even unnecessary, YaST *asks* you to install the driver, downloads and unpacks it automatically, and configures X for you.

      Hence, in driver installation, linux sucks and windows rocks
      What you fail to understand is that all the fancy packaging you see in driver installers is provided by the vendor. If your driver vendor takes the care to provide a good installer (NVIDIA), or your distro takes the care to repackage the driver with a good installer, the process is transparent. If they don't, it can be painful, but the exact same thing is true of Windows drivers. There is no way to fix this on the Linux end, because it's the vendors' crappy installer.

      --
      A deep unwavering belief is a sure sign you're missing something...
  23. No Americas Army? by planckscale · · Score: 3, Informative
    I've only played 2 games on Linux, America's Army and Postal2. Postal2 looked pretty good but torching people gets boring.

    On an nVidia MX 400 card, AA is playable and actually pretty fun online, but shadows are mostly chunks of squares on the ground. Otherwise, rpg's and smoke grenades look fantastic. I wonder why they didn't do comparisons of at least AA? I would think that's one of the first games people download for Linux especially because it's free.

    Oh yeah, I had some original difficulty installing the nvidia drivers on a knoppix hd install with the 2.6 kernel, but I finally got it running well and documented the installation here: http://www.knoppix.net/forum/viewtopic.php?t=10314 &highlight=

    --
    Namaste
    1. Re:No Americas Army? by _Sprocket_ · · Score: 1


      I wonder why they didn't do comparisons of at least AA? I would think that's one of the first games people download for Linux especially because it's free.


      That's where Enemy Territory (which they tested) comes in. Free download. Pretty strong player community with plenty of servers. Fun gameplay that veers closer to arcadeish than AA's attention to reality (such as it is).
  24. What about DRI? by lspd · · Score: 4, Informative

    You can find my benchmarks of DRI compatible cards here. They're a first attempt at benchmarking DRI and still need some tweaking.

    Eric Anholt's benchmarks of DRI on FreeBSD are here.

    Roland Scheidegger's comparison of the three drivers available for the Radeon 9000 (DRI, FGLRX, XIG) is here.

    It's a bit surprising that the Radeon 8500 series is completly absent from this comparison. The 8500 and FireGL 8800 are still remarkable video cards.

    1. Re:What about DRI? by ebrandsberg · · Score: 1

      Nice benchmark, what surprised me was how well the Voodoo 5 ran with DRI given the age of the product at this point.

    2. Re:What about DRI? by Anonymous Coward · · Score: 0

      "It's a bit surprising that the Radeon 8500 series is completly absent from this comparison."

      The 9700 was only getting an unplayable 24fps in UT2004. What would the last-gen get? 10fps?

  25. mod the parent up by poohsuntzu · · Score: 1

    This is an incredible correction that needs to be known. Anonymous is correct, as the Nvidia installer uses the kernel headers to create it's own module which can be loaded by a simple xorg.conf change and reloading the X server.

    There is -no- need to recompile the kernel unless they are referring to the old 2.6 kernel bug with the old nvidia drivers(the stacks bug). But that was a GLITCH that only lasted a week or two, and something quickly fixed (not to mention only really experienced by Fecora Core 2 users) by nvidia.

    --
    "We're breaking out the ramen noodles. . . "
    "Really? Is it someone's birthday?"
  26. Proprietary driver hell by freelunch · · Score: 5, Interesting

    I have an Nforce2 based MB with built-in video..

    For a few months I ran Nvidia's proprietary driver but found that their support was poor. Countless people would report the same problem and Nvidia would basically just shrug and not even reply to the postings on their website. Stuff like "not our problem". They were very slow to support 2.6.

    And as a gentoo user, I hated the binary installation program.

    I finally dumped their stuff and went to the OSS driver. It is much slower, even when just opening new browser windows or xterms. But not having to mess with nvidia installer hell each time I gen a new kernel (which is pretty rare, actually) makes it worth it.

    This was a great article, however, because it shows just how much chance and luck there is in getting these drivers to work. Buying the latest and greatest MB and CPU for use with Linux is still a huge unknown for the novice and experienced Linux user alike. And then there is the very real fear of whether it will work after you upgrade your kernel, etc.

    Sad to see that Nvidia is the most Linux friendly vendor??

    1. Re:Proprietary driver hell by Neil+Watson · · Score: 1

      I still don't understand why people bash Nvidia because of the binary drivers they provide. In spite of Linux being low on the desktop count, Nvidia is nice enought to provide current drivers for all of their latest cards. They were quick to support the 2.6 kernel. If the driver works and works well why does it matter whether it's open or not?

    2. Re:Proprietary driver hell by Anonymous Coward · · Score: 0

      "And as a gentoo user, I hated the binary installation program."

      Why would you use the binary installation program? I just do "emerge nvdia-kernel".

    3. Re:Proprietary driver hell by Anonymous Coward · · Score: 0

      Gentoo has had nVidia binary driver ebuilds for a very long time. You should never have had to mess with the binary installation program (unless you really wanted to get the driver like half a day early). Or are you referring to nVidia's nforce chipset drivers?

      Their video drivers have never given me problems. Then again, I don't use built-in video.

      nVidia also isn't exactly the most Linux-friendly vendor, depending on what you mean. I believe Intel and Via have released full specs for their cards so that the Xorg people can implement open-source drivers for their cards. Granted, nVidia and ATI probably can't legally do that, but it probably results in better drivers for the Intel and Via camps (and it most certainly would in the ATI case).

    4. Re:Proprietary driver hell by Bull999999 · · Score: 1

      You can use the "-x" argument to extract the source code and also use "-n" to prevent the installer from using precompiled modules.

      --
      1f u c4n r34d th1s u r34lly n33d t0 g37 l41d
    5. Re:Proprietary driver hell by cowbutt · · Score: 3, Insightful
      I still don't understand why people bash Nvidia because of the binary drivers they provide. In spite of Linux being low on the desktop count, Nvidia is nice enought to provide current drivers for all of their latest cards. They were quick to support the 2.6 kernel. If the driver works and works well why does it matter whether it's open or not?

      Oh, little things, like the reassurance of being able to continue to use the hardware I've paid for even if nVidia don't feel like continuing to develop the drivers if-and-when the kernel API changes - like with the recent 4k stacks issue. That, and Free drivers are more stable that proprietary drivers in my experience, and when they aren't, you can look at the code to try to figure out why, rather than crossing your fingers and waiting for a driver update that may never come.

      --

    6. Re:Proprietary driver hell by Anonymous Coward · · Score: 0

      Sounds like a problem with your kernel, not the drivers. Maybe you should file a support request on the lkml.

    7. Re:Proprietary driver hell by brunes69 · · Score: 3, Insightful

      Oh, little things, like the reassurance of being able to continue to use the hardware I've paid for even if nVidia don't feel like continuing to develop the drivers if-and-when the kernel API changes

      Yeah. You mean like my Zoran TV tuner card that hasnt worked since the 2.2 series, despite their being public specs and sources for it? Drivers are only maintained as long as the developer is around. And unless you have the skills to write your own drivers (and most of us, including large numbers of application developers, *don't*), the having specs/source or not is irrelevant.

      ...like with the recent 4k stacks issue...

      You mean the "issue" that nvidia had *working drivers* for within weeks after it was *even an option* in the kernel? You mean the "issue" that "open" drivers like *ahem* ATI have and NVidia does not?

      That, and Free drivers are more stable that proprietary drivers in my experience, and when they aren't, you can look at the code to try to figure out why

      Again. you are in the vast minority in being able to do this. So don't bash NVidia for catering to the rest of us.

    8. Re:Proprietary driver hell by cowbutt · · Score: 1
      The lkml doesn't help debug problems in proprietary drivers for Windows. Funny thing, that.

      If you'd have taken some effort to comprehend my point before attempting to flame me, you'd have realised I was talking about proprietary drivers in general not nVidia's or ATI's specifically.

      --

    9. Re:Proprietary driver hell by freelunch · · Score: 1

      Sounds like a problem with your kernel, not the drivers. Maybe you should file a support request on the lkml.

      Actually, that is part of the challenge.

      The kernel guys generally don't want to hear about closed source problems because they don't have code access. It is one thing to not have the hardware but quite another to not have the code.

      Also, those types of problems tend to be extremely tedious to debug and support (aka "not fun"), so I don't blame them. Wouldn't you rather spend your donated time working on things that will have lasting value, like fixing the OSS driver?

    10. Re:Proprietary driver hell by Anonymous Coward · · Score: 0

      With the TV tuner, the specs are out there, so port it to the newst kernel. Take a look at "Writing Device Drivers for the Linux Kernel" by Alesandro.

      Alternatively pay someone to port it or ask someone nicely.

    11. Re:Proprietary driver hell by MrNemesis · · Score: 1

      "I have an Nforce2 based MB with built-in video"

      I've never liked onboard much, so have never used it. Some people have problems with the nForce built-in both under Windows and Linux; there's also a fair few people happily using them for MythTV. YMMV, I guess!

      "They were very slow to support 2.6"

      Sorta true, but 'slow' isn't an objective measurement; I believe they were faster than everyone else (not that I paid much attention to it at the time - I didn't try 2.6 until 2.6.5). As an aside, they supported x86-64 before the chips were even available, something ATI *still* hasn't done a year on - so if you want an accelerated ATI card on an AMD64 mobo, you have to stick with 32bit Linux. Yuk.

      "And as a gentoo user, I hated the binary installation program."

      As the AC pointed out, gentoo has had ebuilds of this since forever (also heavily mentioned in the installation howto), since nVidia doesn't mind other distros supplying the drivers (read their license), and packages for the drivers are available in both gentoo and debian (the only two distros I commonly use). Yes, those packages do contain binary modules, but both nVidia's and ATI's hands are tied in this respect - there is no way they can open source their drivers.

      "This was a great article, however, because it shows just how much chance and luck there is in getting these drivers to work"

      Heh, this is obviously gonna be one of those "YMMV" cases ;) Installation for me has always been a snap (easier and quicker than in windows at any rate), and I have personally never had the nVidia driver obviously b0rk on me (except on a dodgy gigabyte mobo with a dicky AGP slot), and their ease of installation, performance and stability has done nothng but increase with each successive upgrade.

      "And then there is the very real fear of whether it will work after you upgrade your kernel"

      This has pretty much been eradicated, at least with nVidia, since the module uses GPL'd glue code to bond to your new kernel. It's only if you [en|dis]able some new function that nVidia haven't caught up with yet (such as the infamous 4/8k stacks issue) that you may run into problems.

      I use nVidia cards in the three systems I have that need acceleration; namely my desktop and two MythTV boxes. XvMC and OpenGL are used heavily on both, and I have more problems with other things crapping out (like ivtv and mythbackend, grr!) than I do my nVidia cards.

      Personally, I've had more problems with my motherboard (AMD64 with 64bit gentoo) than I have with my GFX card, and I go through about a kernel a week. After the initial installation, upgrades have been totally seamless (there's even several scripts around that will check your kernel-dependent external drivers - nvidia, ivtv, lirc, etc - against your running kernel, and recompile and restart services as neccesary, which makes upgrades seamless *and* pretty much transparent) and the driver has been rock solid (within reason) - for a binary driver without the kernel devs sanction (albeit using GPL glue code), I find this impressive.

      "Sad to see that Nvidia is the most Linux friendly vendor?"

      Try an ATI card next time then ;)

      Some might call me an nVidia fanboy, some might call me lucky, but in my experience nVidia's efforts under Linux are to be applauded. They're supporting a very niche OS, and (in my experience at least) are doing a very good job of it. Their drivers are worlds apart compared to those crummy closed drivers that require a specific kernel version on a specific distro, and then require a specific ambient temperature, moon phase alignment and virgin sacrifices to work (yes, I once tried to get a Promise RAID card working in Linux too).

      In an industry dominated by secrecy, competition and closed sources, nVidia are considerably further along in penetrating and supporting the Linux market than anyone else I've had the misfortune to deal with.

      Disclaimer: I don't work in any way for nVidia, and I am not trying to discredit the parent, just pointing out my side of the coin

      --
      Moderation Total: -1 Troll, +3 Goat
    12. Re:Proprietary driver hell by Dr.+Photo · · Score: 1

      You mean like my Zoran TV tuner card that hasnt worked since the 2.2 series [...] And unless you have the skills to write your own drivers (and most of us, including large numbers of application developers, *don't*)

      No TV since kernel 2.2 means you've had years to learn driver development! Why aren't you a guru yet? ;-) (*ducks and runs*)

    13. Re:Proprietary driver hell by Anonymous Coward · · Score: 0

      i run slackware.

      the longest part about installing the nvidia driver?

      downloading it. it took 27 seconds to download.

      it took me 2 seconds to drop out of X and run the installer script. it took the script about 10 seconds to do it's thing. i made one change in the x config file (3 more seconds)...

      i restarted X, started Tux Racer. yea.

      total time? ~50 seconds.

      i'm really restraining myself now...but you show your complete ignorance in a way that begs someone to bash your face in for stupidity.

      maybe it's all those "optimizations" you are dicking around with in your precious gentoo.

    14. Re:Proprietary driver hell by cowbutt · · Score: 1
      You mean like my Zoran TV tuner card that hasnt worked since the 2.2 series, despite their being public specs and sources for it?

      Does this help? (found using google.

      Drivers are only maintained as long as the developer is around. And unless you have the skills to write your own drivers (and most of us, including large numbers of application developers, *don't*)

      I do get frustrated with this "poor helpless me" attitude. I didn't know what I was doing when I first ported a piece of kernel code from one kernel API to another. But I learnt by doing it. It's not like the process is shrouded in secrecy; you have the source of a driver that used to work, the specs, and mailing lists that'll help if you can post coherent and well-researched queries. What more do you want, for Pete's sake?! And, as other posters point out, there's also nothing stopping you from finding a developer who'll port the driver for you for money - perhaps raised by putting up a donations webpage where other owners of the same hardware can contribute.

      ...like with the recent 4k stacks issue...

      You mean the "issue" that nvidia had *working drivers* for within weeks after it was *even an option* in the kernel? You mean the "issue" that "open" drivers like *ahem* ATI have and NVidia does not?

      Have you got a reference for that assertion? I'm still using 2.4, so I can't provide experience to the contrary, but it would surprise me if that's still true...

      --

  27. Standardization by acherrington · · Score: 4, Interesting

    In my opinion this is the best thing to happen to linux in some time. Any time that you can develop standards for an industry, you can finally give a target for competitors to aim at (e.g. each other). This will drive competition and really drive the market forward. I would consider this a first step forward.

    After both ATI and nVidia clobber each other with better framerates and better overall performance, I think that a new competitive advantage will develop... perhaps this may be better graphics quality or easier installs.

    --


    Victory is gained, not in knowing your opponents next move, but in preempting them.
    1. Re:Standardization by Anonymous Coward · · Score: 0

      What we needs is standardization in the developement.

      You see mesa rocks, not because it's high performance, but because it's cross platform and open source.

      You see right now Linux drivers are a bitch because of 2 things:
      1. Liscencing issues. (which is a good thing, without GPL there would probably be no point to even trying)
      2. dual nature of Linux video drivers...

      You see with X Windows (not specific to Linux BTW), and unlike any other GUI you need 2 seperate sets of drivers.

      You need a set for X11 and then another for OpenGL. One 2-d driver and another 3-d one.

      So that's twice the work for companies to do + they need to open up their drivers...

      Which sucks for Nvidia and ATI because both have some considurable magic associated with them, and lots of very secretive optimizations.

      So it's a double whammy against Linux drivers.

      So what X Windows guys need to do to make it much better for drivers developers is to make a purely OpenGL accelerated enviroment.

      That way you only have to develope 3-d drivers and developers can concintrate on that alone and make thigs much snappier and stable as a result. No more duplication of effort, no more dividing resources.

      Yep, that's the solution. No more caring about 2-d acceleration. You simply use the texturing handling and vector handling that's already built into OpenGL.

      Legacy no-3d cards or cards without all the features that newer OpenGL specifications require they will have software acceleration.

      That's one of the nice things about Mesa and DRI.

      DRI uses mesa even for it's hardware acceleration. So when working with older cards you simply use what hardware resources they handle, and then emulate the rest.

      I know it's not Windows, but DirectX versioning is well known...

      It would be like you can upgrade your drivers on your older cards and take a DirectX 7 or DirectX 8 card, and turn it into a DirectX 9 card. Just with a driver upgrade.

      Of course it will be much slower then a "real" DirectX 9 card, but it will be completely compatable with all the newest games.

      The future of X is OpenGL with vector and texture acclerated in the 3-d hardware portion of the card.

      If you look at the hardware sections of a modern GPU card you'll notice that the 2-d accelerated portions are the same as they were back in the Geforce 1 and the original Geforce 2 days.

      2-d is "fast enough", so all the work goes into 3-d. While the 2-d has remained the same as it was 4-5 years ago, 3-d portions of the cards have increased exponentially in complexity and performance.

      3-d is the way of the future, baby.

  28. Re:Indeed by prisen · · Score: 3, Informative

    No need to recompile the kernel, true - but you'll have to have your kernel's source installed in order for NVIDIA's installer to compile a custom module on the spot.

    All in all, it does work really well...until you upgrade or replace your kernel, and then X of course won't work. Many times you can run the installer again, though. Simple enough!

  29. Re:Indeed by poohsuntzu · · Score: 2, Informative

    Anytime you upgrade or replace the kernel, just run the binary again before loading X (you aren't booting straight into x, are you?). Nvidia updates the module, reinstalls the new module while removing the old one, and bam. Now boot up X.

    --
    "We're breaking out the ramen noodles. . . "
    "Really? Is it someone's birthday?"
  30. The real question is by IceFox · · Score: 4, Interesting

    Planning to plunk down some money soon and what I want to know is: What is the best video card you can get that works in Linux that *doesn't* require binary drivers? I don't perticularly care to be locked into one kernel if given the option. -Benjamin Meyer

    --
    Do you changes clothes while making the "chee-chee-cha-cha-choh" transformation sound?
    1. Re:The real question is by untermensch · · Score: 1

      If you want good 3D support and you don't want a binary driver, then you're simply out of luck. The best open-source drivers available for even the fastest of cards, is still completely incapable of keeping up with the closed-source binaries.

      If you don't care about 3D support then, it doesn't matter much which card you get. Basically any card from a major vendor will do just fine playing movies, displaying your desktop apps, etc. I recommend just buying something nice and cheap.

    2. Re:The real question is by arose · · Score: 1

      As far as I know it's the ATI FireGL 8800.

      --
      Analogies don't equal equalities, they are merely somewhat analogous.
    3. Re:The real question is by Bull999999 · · Score: 1

      The Nvidia driver package also comes with the source so that it can compile a module for you if it doesn't already have a binary driver for your kernel.

      --
      1f u c4n r34d th1s u r34lly n33d t0 g37 l41d
    4. Re:The real question is by Anonymous Coward · · Score: 3, Informative

      OEM Radeon 8500 based cards are cheap and are among the fastest cards with open source 3D drivers.

    5. Re:The real question is by fymidos · · Score: 1

      the best choice is matrox, and it's a good choice, indeed...

      however, you are not "locked into one kernel" with the nvidia drivers either ... and ATIs' *opensource* drivers are somewhat better than nvidia's *opensource* drivers.

      --
      Washington bullets will simply be known as the "Bulle
    6. Re:The real question is by orkysoft · · Score: 1

      Actually, on my TNT2 card, if I use the open source nv driver, I get a crap framerate when playing movies fullscreen, but if I use the closed source nvidia driver, I get a good framerate.

      So the difference is not just 3D support.

      --

      I suffer from attention surplus disorder.
    7. Re:The real question is by untermensch · · Score: 1

      hmm, If you're not using Xfree 4.4 or later I would think about upgrading as the 2D acceleration for TNT cards was greatly improved as compared to older versions. I'm not 100% sure but I could've sworn I had a TNT2 playing fullscreen video.

      But in any case, when I said a cheap card I didn't really mean TNT2 cheap. Even in the TNT2 really can't do it with the nv driver, older GeForce's shouldn't have a problem (or ATI from the same era).

    8. Re:The real question is by damiam · · Score: 1

      "Best" is subjective, but if you mean "fastest 3D performance at a consumer price", you want the Radeon 9000. Any Radeon or GeForce will work superbly in 2D with open-source drivers, but only Radeons 9000 and below are 3D-accelerated.

      --
      It's hard to be religious when certain people are never incinerated by bolts of lightning.
    9. Re:The real question is by lspd · · Score: 1

      The FireGL 8800 is the fastest DRI compatible video card, but it has problems with projected textures (Descent3, HeavyGear2, Shogo) and there's a problem with UT2K3/UT2K4 that causes X to lock up when the shock rifle is fired.

      As far as reliability goes, the Matrox G400Max and Voodoo 5 - 5500 seem to be the best options. You may have problems with AGP slot compatability with those though.

      Another good option is the Radeon 9000 or 9200. They still have the projected textures problem, but UT2K3/4 doesn't lock up.

      The DRI team seems to be closing in on a solution to the problems with projected textures on r200 cards. The shock rifle bug doesn't seem to be going anywhere.

    10. Re:The real question is by hemanman · · Score: 1

      Use Fedora and check www.fedorafaq.org, installing Nvidia drivers(When having updated your yum.conf) is as easy as writing this in the terminal:

      yum install nvidia-glx kernel-module-nvidia-

      It can't be much easier :-)

      -H

    11. Re:The real question is by Anonymous Coward · · Score: 0

      ATi Radeon 9200 is (IMHO) the best supported non-binary card. Works perfectly with open-source only drivers. Full support for multihead and rather good 3D acceleration. The only part that doesnt work is tv-out. I have been using such a card with two monitors for a year without video-related crashes. (I usually reboot every 14 days) Earlier experience with nvidia-binary drivers made non-free drivers a non option. (system became very flaky, perhaps good for gaming if you can stand rebooting a lot, not usable in a workstation)

    12. Re:The real question is by Anonymous Coward · · Score: 0

      I believe it's Radeon 9200 and below.

    13. Re:The real question is by glitchvern · · Score: 2, Informative

      As others have mentioned, the best cards with open drivers are based on ATI's R200 chipset. These cards include the Radeon 8500 and 9100, and FireGL 8700 and 8800. While the FireGL 8800 is probably fastest, it is also crazy expensive. The 9100 is a rebadged 8500 with different core and memory clock speeds. I have been told the 8500 should be faster, but have never seen any benchmark proving one to be faster than the other.

      I am not sure how fast the ViaCLE266 is, but it does not matter since it is a chipset for motherboards, and I do not think it is available as a card. I read a review somewhere claiming it was decent, which given the time since the 8500's release may mean it is the roughly the same, worse, or better. It has totally open drivers.

      S3's DeltaChrome (S4/S8) is suppose to get an open source driver released from S3. Also it has been claimed people who are not S3 have received the specs necessary to write drivers for the card. Via is S3's parent corporation, and these announcements were made at approximately the same time Via opensourced the CLE266 driver and the driver for their hardware mpeg decoder. DeltaChrome cards are not yet available in the United States. They were suppose to be available quite awhile back. They are available in Asia and Europe and have been for a few months now. Any DeltaChrome card (even the budget S4) would smoke an anceint Radeon 8500, but I do not know that I would wait forever for DeltaChrome boards and linux drivers to appear.

  31. ATI multi-monitor support a shocker by hgiddens · · Score: 2, Informative

    Even with my shiny new (-ish) Radeon 9800, I can only get around 80 fps in glxgears - because ATI's drivers don't support Xinerama, I'm stuck with the functional, but much slower, open-source drivers. The framerates AnandTech are be getting single-headed are a dream for me.
    However, having read the article, ATI claim to have some Linux announcements in the pipeline - with any luck, maybe these drivers will allow me to use both my monitors with some decent 3d acceleration.
    Anyway, to anyone thinking of getting an ATI card for use with multiple monitors under linux: caveat emptor.

    1. Re:ATI multi-monitor support a shocker by cowbutt · · Score: 2, Informative
      Even with my shiny new (-ish) Radeon 9800, I can only get around 80 fps in glxgears - because ATI's drivers don't support Xinerama, I'm stuck with the functional, but much slower, open-source drivers.

      That's still poor; I'm using the Free ATI drivers from the 2.4 kernel and XFree86 4.3.0, and I get about 800fps with glxgears (lousy benchmark, BTW). On my two year old P4 2.4 and Radeon 7500.

      --

    2. Re:ATI multi-monitor support a shocker by mczak · · Score: 1
      That's still poor; I'm using the Free ATI drivers from the 2.4 kernel and XFree86 4.3.0, and I get about 800fps with glxgears (lousy benchmark, BTW). On my two year old P4 2.4 and Radeon 7500.
      That's not surprsing, since you do get 3d acceleration for the radeon 7500, but none at all for the radeon 9800, for which no documentation is available unfortunately to the dri developers. (This project here, http://r300.sourceforge.net/ has a driver for r300 which is being developed by reverse-engeneering ATI's driver, but it's not usable currently.) (and btw, 80fps seems to be way too slow even for software rendering, maybe the software renderer is synced to screen refresh...)
    3. Re:ATI multi-monitor support a shocker by quan74 · · Score: 1

      Even with my shiny new (-ish) Radeon 9800, I can only get around 80 fps in glxgears - because ATI's drivers don't support Xinerama, I'm stuck with the functional, but much slower, open-source drivers. The framerates AnandTech are be getting single-headed are a dream for me.

      The ATI (fglrx) drivers do not support xinerama, but they DO support multi head display WITH acceleration, either seperate sessions on each screen, or one big screen across both monitors. My 9800 pro does this just fine. The options to do so are in the fglrxconfig utility. There are some problems with the ATI dual head setup though, since it doesn't support xinerama properly, some things are a PITA, for example maximizing a window makes the window take up both screens instead of just one (in big desktop mode) and any program that uses xinerama to draw to both screens (for example xscreensaver) will only cover the primary screen (the dvi connector). I haven't found a workaround for the fullscreen problem, but using xlock as a screensaver instead of xscreensaver covers both screens. YMMV.

    4. Re:ATI multi-monitor support a shocker by Fnord · · Score: 1

      Don't count on it. It's actually a flaw in XFree, xinerama and dri (which the ati drivers do depend on) don't function simultaneously. And the reason you only get 80 fps in glxgears is because its pure software rendering. The XFree drivers for ATI cards past the 8500 have absolutely no 3d acceleration.

    5. Re:ATI multi-monitor support a shocker by Anonymous Coward · · Score: 0

      6200fps on glxgears. Tweaked, whipped and accelerated dual DVI on a geforce 5700 on Debian.

      I think i've gotten every last tweak squeezed out of it. runs ut2004 just great (on one of the LCDs, heh)

    6. Re:ATI multi-monitor support a shocker by joib · · Score: 1

      Uh, that's pretty bad. With my Radeon 9200 i get 1900 fps in glxgears with the opensource dri driver. And the cpu is no screamer either, 800 MHz duron.

  32. Depends... by temojen · · Score: 0

    Gentoo is very flexible and configurable, and can keep you on the bleeding edge of software versions. It also takes a long time to set up.

    Debian Stable and RedHat Advanced Server are very slow changing, but behind the times. The stability is good in some situations.

    Suse (reportedly) has a slick installer. I haven't used it.

    I prefer Gentoo, but I've got a herd of machines to handle and 9 years of experience with Linux.

  33. What's the beef with rebooting? by poohsuntzu · · Score: 3, Insightful

    I don't get why some users complain about rebooting, linux and windows alike. In a company/mission critical server type situation I could see it, but for home use? My Windows XP machine takes a total of 20 seconds to shut down, pass BIOS, reboot, and hit the desktop ready to work. In that twenty seconds (which mind you, isn't very long to begin with) I can actually do that thing we forget to do, and stretch my legs and arms. Grab a cup of water, hell.. even look out the window.

    The same with my slackware machine. About 25-30 seconds for a reboot. None of that bothers me because I -know- 30 seconds on my home machine doesn't mean a damn thing. I'll enjoy that time to rub my eyes, refresh myself with maybe spending that 30 seconds taking all the dishes out of the room back up to the kitchen.

    Don't treat 30 seconds as a long and unbearable time unless you want to start complaining about having a manually flushing toliet in your home, followed by hands that can't wash themselves.

    --
    "We're breaking out the ramen noodles. . . "
    "Really? Is it someone's birthday?"
    1. Re:What's the beef with rebooting? by mark1348 · · Score: 2, Funny

      30 seconds!! Time enough to flame another windoze user group!!!

    2. Re:What's the beef with rebooting? by JoScherl · · Score: 2, Interesting

      It's not only the booting process, it's stopping and starting the applications,too. Closing my web-brwoser including every application and if you testdifferent settings these 30 seconds sumup plus you need timeto reopen the wizard...

    3. Re:What's the beef with rebooting? by poohsuntzu · · Score: 1

      wizard? browser? programs? I have windows boot on startup, that's it. Not Windows+trillian+msword+spyscanners+supersound configuration+something else in the systray.

      Let me introduce you to msconfig.exe and from there, I'm sure you can google to learn how to correctly keep your system optimized. If you are having problems with booting time, it isn't windows, it's your plethora of programs loading on bootup that don't need to be there.

      Especially in windows since clicking Shutdown will close all your programs for you and use the Save featre built into them.. there is no reason for you to complain about 30 seconds. If clicking that little X box to close a program (takes, what? one second.. maybe two if you use the keyboard to ctrl+s and save?_ really upsets you, then should you really even bother being on a computer in the first place?

      --
      "We're breaking out the ramen noodles. . . "
      "Really? Is it someone's birthday?"
    4. Re:What's the beef with rebooting? by Hobadee · · Score: 2, Insightful

      It's not the length of the reboot time that has me (and I think lots of people) up in arms. It's the fact that you have to reboot period. Except for updating VERY central things (such as the kernel), you should never have to reboot. Period, end of story. Programmers like those at Microsoft have made people content with constantly rebooting thier machine because it either crashed, or they installed a new program.

      --
      ...Had this been an actual emergency, we would have fled in terror, and you would not have been informed.
    5. Re:What's the beef with rebooting? by JoScherl · · Score: 2, Insightful

      First I'm using Linux not Windows but to the point: I wrote bout thefact that it isn't just the time for booting, if you change your display settings you either want to test it, so you test different settings and reboot many times or you want to run a special application so you set the display settings (which takes some time any way you're doingthis), wait 30 seconds, start the special application, work close it, (these three steps need to be done anyways) wait30 seconds and open you're usual workdesktop (browser, maillcient, ...) if you have to do this more often it's really annoying.

    6. Re:What's the beef with rebooting? by poohsuntzu · · Score: 1

      Wait, what on earth are you talking about?

      NEVER any point in time on windows OR linux do you have to reboot your computer for the display settings, unless you don't have your video drivers installed and need to reboot to install them. I don't care if you go from 640x480 to 1024x768 at 32 bit, you -wont- need to reset. Windows 98 had the option to reset, but on that same menu it had the option to not reset and have the video card adjust the resolution anyways. Everyone noticed that there wasn't a reason to reboot for video resolution changing, and thus why it was never default past windows 98.

      Get your facts straight. Screen and color depth does -not- create a machine reboot, nor should it ever need it.

      --
      "We're breaking out the ramen noodles. . . "
      "Really? Is it someone's birthday?"
    7. Re:What's the beef with rebooting? by poohsuntzu · · Score: 1

      Which brings us back to step one.

      WHY is it so bad to reboot? Your "Period, end of story" doesn't make any sense because you didn't once state why it was bad. So? You are without an OS for 30 seconds, get over it.

      If it was a mission critical server or likewise I can understand it completely, but not for normal home usage. Since when is your "Period, end of story" reasoning going to state what should and shouldn't be acceptable in the computer industry world? Rebooting does nothing more than require you to have 30 seconds away from the computer after system changes (including installing programs if we want to talk about win32 enviroments due to dll installs).

      In fact, you -dont'- have to reboot many of the times it says reboot. It's just that it may be safer on certain systems, and thus they have to account for all.

      Don't be a zealot if you don't have backed up reasoning besides "Period, end of story, my opinion rulz all4eva".

      --
      "We're breaking out the ramen noodles. . . "
      "Really? Is it someone's birthday?"
    8. Re:What's the beef with rebooting? by thinkninja · · Score: 1
      Just in case you didn't actually know:

      uptime : n.

      Technically, a machine's time since last reboot; jargonically, how long a hacker has gone without sleep. "What's your uptime?" "Oh, about 28 hours so far, but I think I can probably do another 12." This is, of course, a reference to the uptime command and the pride with which most Unix types note how long their computers go without reboots. Uptime is a testament to the stability of the OS and the stamina of the hacker.


      Emphasis mine.
      --
      "The number of Unix installations has grown to ten, with more expected." (Unix Programmer's Manual, 2nd ed.; june 1972)
    9. Re:What's the beef with rebooting? by timmyd · · Score: 1
      to reboot and get back into the state that i'm currently in, i have to:
      • login again
      • startx again
      • open firefox and open the 9 pages that i have open right now in tabs, again
      • open a gnome-terminal and start irssi and connect to two different irc servers, again.
      • open gaim and arrange my chat window so it is in the bottom left of my left monitor, again.
      • open two xterms-one for each window and set the sticky bit, again
      • open gedit in another desktop so that i can input different languages as a scratchpad for looking up words in stardict, which i have to start again
      • open up some xterms in another desktop and ssh connect to my server so i can record television programs, again.
      • open up xterms for which i can play music videos with mplayer
      • open emacs with a terminal so i can work on a cs project using smlnj
      • open another instance of emacs and a xterm with a sql prompt where i can enter commands to postgresql and work on my cs database project
      • open some xterms so i can use bittorrent
      • open an another emacs instance and start gnus so i can check email
      • open some more xterms and log in as root so i can keep my computer up to date and emerge new programs
      • do the same as the previous but be ssh'd into my server box
      • another emacs instance and a gv instance to read a ml manual so that i can work on my cs project using both monitors in my dualhead setup
      • leave my volume control program open in another desktop


      altogether i have 24 desktops i can access with the F1-F12 keys that are spread between two different monitors. with a total of 279 processes and 1280MB of ram, you can see i like having my programs open for me so i don't have to open a new program everytime i change what i'm doing at the computer. if i have to reboot, i have to gradually set this state up again which is very annoying.

    10. Re:What's the beef with rebooting? by puddpunk · · Score: 1

      Here's what you do:

      Dump your lame window manager and get one that has "Remember" support.

      Get a session manager (GNOME and KDE have one, but my guess is your not using either) like rox-session.

      Problem solved.

    11. Re:What's the beef with rebooting? by Taladar · · Score: 2, Informative

      I don't know about you but I tend to do several things in parallel on my PC most of the time which accounts for several open webpages in my browser, an open MP3-Player, open Movie-Player, SSH-Sessions to other machines, IRC-Client, ICQ-Client,...
      Now why the hell should I have to close all that and open it again (which takes at least several minutes) for some minor change like installing new Software?

    12. Re:What's the beef with rebooting? by Anonymous Coward · · Score: 0

      taking all the dishes out of the room back up to the kitchen.

      That's what I always intend to do, but I always end up drooling at my home made silent bootsplash. ;-)

    13. Re:What's the beef with rebooting? by Tore+S+B · · Score: 1

      It did on all versions of W95 except OSR2, which made it optional, which carried through to 98.

      --
      toresbe
    14. Re:What's the beef with rebooting? by timmyd · · Score: 1

      session managers don't remember half of what is going on(that i'm ssh'd into a box, what files i have open in emacs, what tabs are open in mozilla, who i'm talking to on gaim, what irc servers i'm connected to and my nicknames, etc). they remember like what X program is running and where the window is, but they don't do more than that. i'd need some sort of OS support to have the kind of persistancy that i want. maybe you should have read my post before replying.

    15. Re:What's the beef with rebooting? by Anonymous Coward · · Score: 0

      I'm still running Win2K and unfortunately my reboot times are around 140 seconds.

    16. Re:What's the beef with rebooting? by Anonymous Coward · · Score: 0

      Personally, at this moment, I have 22 application
      windows open. (And 53 different websites in tabs
      in Mozilla.) I guess it's just a matter of
      familiarity with what's open and where it is,
      after the machine has been up a significant
      amount of time. (For me, anyway.) I just checked,
      and this machine claims it booted up last
      December.

    17. Re:What's the beef with rebooting? by poohsuntzu · · Score: 1

      Hey, that sounds like the default window manager for Windows XP.

      --
      "We're breaking out the ramen noodles. . . "
      "Really? Is it someone's birthday?"
    18. Re:What's the beef with rebooting? by julesh · · Score: 1

      Yeah, that's the best reason for not liking rebooting. I have Mozilla open with about 30 tabs on various web sites, scrolled to information that I might need to refer to. I have 6 terminal sessions open to different machines, applications running in each of them. I have e-mail windows open in the background that have information I intend to use. If I need to reboot, I'll have to make notes of which e-mails these are because my mail client doesn't support remembering about open windows. I have several documents open in Acrobat that I'll need to refer to in the future. I have a text editor with about 30 open files, most of which I'll need to edit soon and don't want to have to find again in a hurry. I have a file sharing application open that is stuck in the middle of some very long download queues; rebooting at the moment will probably put me back by at least a day in getting those files I'm downloading.

      This is why I don't like rebooting. It interrupts my workflow.

    19. Re:What's the beef with rebooting? by grumbel · · Score: 1

      The problem is not 20sec vs 1min, but 20sec vs. almost instantanious. Of course this is not a problem when I plan to work for a few hours on the PC, then it its a non-issue.
      Its however a big problem for all those times where I do not plan to work on the PC for hours, but where I just want to quickly look up something on the net or such. Such an action itself might take only a minute, so 20sec boot + 20sec shutdown becomes a long time compared to the action itself and often enough its to long to even bother to start the PC at all.

      Last not least 20sec is not the boottime, it might be the time till the GUI is first visible, until all application are started and fully working can take a hell of a lot longer.

  34. Why Linux sucks by Animats · · Score: 4, Insightful
    Enabling 3D acceleration (DRI) still needs to be done manually by editing the /etc/X11/XF86Config file after running the SaX2 utility. Enabling FSAA must be done by editing the XF86Config file by hand as well (see our AA/AF section for details). After a little more than 8 hours of playing with configurations, we hit paydirt.

    Any questions?

    This binary driver thing has got to go. As Linux gains desktop market share, pressure will increase to open up the hardware interface to the driver. It's not like hooking OpenGL to the card involves any technology that isn't well known in the industry.

    1. Re:Why Linux sucks by Stevyn · · Score: 1

      As said a million times here, Linux is a kernel, XFree86 is an implementation of the X11 standard. Linus would love to see binary drivers to go.

      NVIDIA's example: It's NVIDIA that's releasing the closed source drivers. The kernel does ship with an open nvidia driver called "nv." You can use it, but you won't get 3D acceleration. NVIDIA doesn't release their code for several reasons. One is industry secrets they want to keep intact. The other could be licenses from other companies they're not allowed to release. Either way, NVIDIA is pretty good to the community.

    2. Re:Why Linux sucks by be-fan · · Score: 1

      Um, this is mainly a reason why ATI's drivers suck. NVIDIA's drivers are absolutely painless in this area. Many mainstream distros today handle installing the NVIDIA drivers automatically. If you're on SuSE, for example, all you have to do is say "yes" when YaST update prompts you to install the driver. Recent versions of the NVIDIA driver come with a nice GUI tool that let's you do stuff like change anti-aliasing settings, digital vibrance control, etc.

      The situation on Windows and Linux is the same here. You think that NVIDIA control panel in Windows comes with the OS? The vendors supply the tool, just like they have to supply it for Windows. NVIDIA allows you to set the parameters through the GUI or through XF86Config, while ATI only allows you to do it through XF86Config. The fact that you have to dig around in XF86Config is ATI's fault, not XFree's or Linux's.

      --
      A deep unwavering belief is a sure sign you're missing something...
    3. Re:Why Linux sucks by MrNemesis · · Score: 1

      As others have pointed out, the bitching was entirely about ATI's appallingly awful Linux drivers. They are exceptionally picky about what they will run on; nVidia's just seem to need MTRR in the kernel to work.

      For me at least, installing and upgrading my gentoo systems' nVidia drivers is quick and painless, and in fact much easier than windows (unless you use the WHQL drivers from windows update).

      Granted, nVidia have snafu'd up a coupla times in the past when they couldn't keep up with kernel development (slowness WRT 2.6 and 4/8k stacks for example), but it's always been remedied fairly speedily. Contrast to ATI, who don't even support x86-64 yet, which nVidia have supported since before the chips came out.

      Back to your point, we'd *all* love to see nVidia and ATI GPL their drivers and bung 'em in the kernel, but since they both want to keep alot of secrets from the other and they don't actually have rights to alot of the code, their hands are tied.

      I do however share your optimistic belief that the code will become progressively more open in the future; the high end GFX workstation will hopefully prove instrumental in this respect.

      --
      Moderation Total: -1 Troll, +3 Goat
    4. Re:Why Linux sucks by Tore+S+B · · Score: 1

      "SNAFU" can not be used as an exception!

      It means "Situation Normal, All Fucked Up", and thus would imply that it can't have been an exception.

      --
      toresbe
  35. installing games by temojen · · Score: 1

    Most of the popular ones available for Linux have ebuilds. Just pop in the CD, mount it, and emerge the game.

  36. We need... by SaDan · · Score: 2, Insightful

    A dedicated distro for gaming on Linux. End of story.

    Until EVERYONE adheres to some sort of guidelines (HA! Yeah, RIGHT!), people are going to be dealing with oddball dependancies, kernel/driver issues, and filesystem layout annoyances.

    Documentation all around needs to improve too, for both the Linux distros and the game makers.

  37. nvidia by poohsuntzu · · Score: 3, Informative

    There are OpenSource drivers for Nvidia video cards, however they are no where near as fast as the official binaries and can't preform 3d worth a damn.

    You have to realise what you are asking here. The binary drivers that you are mentioning (which, by the way, never lock you to a certain kernel) are using the code made by ATI/NVIDIA to take advantage of their hardware's features. PixelShading, 3d processing. Each have their own way for their hardware to preform 3d functioning. This is not something they are going to disclose (they are a buisness too, remember?) and thus the open source drivers for video cards are always going to be horrid compared to the avalaible binaries.

    Seriously though. Get a card that works great on linux despite the binary packaging. And I still don't see how it locks you into one kernel? Could you explain further in depth what you mean for me? Because upgrading nvidia (which must be done each time you update/replace your kernel) is as simple as shutting down the X server, rerunning the nvidia binary, and then rebooting the X server.

    --
    "We're breaking out the ramen noodles. . . "
    "Really? Is it someone's birthday?"
    1. Re:nvidia by erikharrison · · Score: 4, Informative

      Have you tried to run multiple kernels with the Nvidia drivers? Everytime I booted into a different kernel, I had to uninstall and reinstall the driver. And what about 4k stacks?

      Besides, you didn't answer his question - he said "What's a good card with solid open source drivers?" You said "Nvidia has open source drivers but they suck, you shouldn't care about the binary only drivers."

      I'd still advocate a Nvidia or ATI card. ATI makes regular code drops to the DRI and Mesa projects, and the open source drivers are of reasonable quality, and the nv drivers are high profile, with lots of work going into them. These cards are the most likely to see solid render acceleration in the future as XAA is replaced with a new acceleration architecture, so even with the Open Source drivers you'll see best performance with stuff like Composite (the basis of much of the X11 6.8.0 eye candy) with these cards.

      Of the two, ATI and Nvidia, the open source drivers seem to be of roughly the same quality in my experience, but the Nvidia binary driver is far superiour to the ATI binary driver. ATI has got more bang for your buck, the GATOS project is working to support a lot of ATI's extra features, and ATI seems minimally more involved in the community with an eye to becoming moreso.

      I think that pushes things solidly in ATI's favor if you're absolutely commited to the open source driver. If you're willing to use the binary driver, things become more even - it's ATI's price versus Nvidia's better support for the card under Linux/BSD

    2. Re:nvidia by Anonymous Coward · · Score: 0

      Each have their own way for their hardware to preform 3d functioning. This is not something they are going to disclose (they are a buisness too, remember?) and thus the open source drivers for video cards are always going to be horrid compared to the avalaible binaries.

      This comes up all the time but I don't think it's really true. It's not that difficult for your competitor to reverse engineer your driver to an extend where he understands how it works. That's not a big problem. The problem is for the open source developers to write a driver that always works and also takes into account some implementation bugs on the cards side. It would be very helpful if the vendors would disclose the basic acceleration functionalities to the public. They can keep the very new and "secret" stuff for themselves I don't care to much. But give the community something that works reasonably fast and stable.

    3. Re:nvidia by poohsuntzu · · Score: 2, Informative

      >>Have you tried to run multiple kernels with the Nvidia drivers? Everytime I booted into a different kernel, I had to uninstall and reinstall the driver.

      Correct, as it should apply to any kernel specific module. This isn't something to whine about, as each nvidia vinary wants to use your latest kernel headers or configurations. That shouldn't be a big deal, especially when the binary process takes about ten seconds.

      >>And what about 4k stacks?

      This was already fixed, oh a few months ago. About two weeks after this problem was brought up, Nvidia released newer drivers to solve the incompatability.

      --
      "We're breaking out the ramen noodles. . . "
      "Really? Is it someone's birthday?"
    4. Re:nvidia by Anonymous Coward · · Score: 0

      Hey, the oss nvidia drivers aren't bad.
      I don't find them slower for general desktop stuff... If the OP finds xterms opening slow, that's probably something else... the only place I've found them slower on the desktop is watching dvds with -x11 in mplayer. With xv they are the same as far as I can tell...

      and a few more dots .....

    5. Re:nvidia by Anonymous Coward · · Score: 0

      Normal kernel modules don't need to be uninstalled when switching between different drivers, as they are located in different directories, and the system will pick the module from the directory matching the current kernel version.

  38. Re:Pronounciation by jesuscash · · Score: 1

    Soo-See

  39. plea for help.... by B5_geek · · Score: 0, Offtopic

    This is similar to the problems that I have faced (and many others too I am sure) in getting linny & nvidia to play nice together.

    Hardware:
    AMD AthlonXP 2600+
    1.5GB ram
    Nvidia Ti4600

    1st attempt: Mandrake v10.0
    I ran the Nvidia driver installer and it would state: "Nvidia driver installed..."
    In Mandrake's config files it did report that I was using the correct card.
    No splash screen and GL games ran like a hog (software emulation it felt like).

    I played around with the xfree86.conf (I think that's the name) file, switching "nv" with "nvidia" and back again.

    No splash screen.

    I then loaded Suse v9.1 It detected the correct video card at installation AND I got the Nvidia splash-screen at boot.

    but

    the DVI output on my card wouldn't work.

    Any suggestions?

    --
    "The price good men pay for indifference to public affairs is to be ruled by evil men." ~Plato (427-347 BC)
    1. Re:plea for help.... by plcurechax · · Score: 3, Informative


      I played around with the xfree86.conf (I think that's the name) file, switching "nv" with "nvidia" and back again.


      The binary driver is called nvidia. So switch it to that, and leave it.

      Read the log file (/var/log/XFree86.0.log), look for lines with (WW) and (EE). This will go a long way to track down your problems.


      the DVI output on my card wouldn't work.


      Do you mean a second video port? Under Linux the second port (video output) is independent of the first in the XFree86 configuration, so you have to configure it to use it explicitly. Something like 'Screen 1' or 'Option "ConnectedMonitor" "DFP" ' in the Section 'Device' should do it.

  40. Re:Pronounciation by Anonymous Coward · · Score: 0

    Sue Say

  41. Re:...vs. same cards with Windows? and X.org? by Anonymous Coward · · Score: 0

    And much better, with a stock X.org server so we can see how the cards match up when they use the open software. For many of us, it's more important to have the stability of a full open source system.

  42. Re:Pronounciation by Anonymous Coward · · Score: 0

    Soo-suh

  43. Not quite. by temojen · · Score: 3, Informative

    DRI lets X communicate with the hardware faster. X acceleration works without it, but not as well.

    The Direct Rendering Infrastructure (dri.sourceforge.net), also known as the DRI, is a framework for allowing direct access to graphics hardware in a safe and efficient manner. It includes changes to the X server, to several client libraries and to the kernel. The first major use for the DRI is to create fast OpenGL implementations.

    from Gentoo Hardware 3D Acceleration Guide.

  44. Little quirks by tempfile · · Score: 2, Informative

    My 9500 Pro still doesn't work and just crashes the system hard when 3D acceleration is enabled. Nobody knows this problem and nobody can explain. :(

    1. Re:Little quirks by spinlocked · · Score: 1

      Try fiddling with

      Option ChipId

      in the driver section of the X config file. My card (a 9800SE) was detected as a 9800 proper and the driver tried to use the two extra pipelines (faulty in my case, sadly - that would have been a bargain) causing the box to lock up hard.

      Using the ChipId of a less capable card (I went for a 9600 I think), solved the problem for me. ATI's linux drivers suck really badly, but their support is even worse. I fill their support form out with each new driver release (politely!), but I never get a reply, the problems go unheeded and my solutions are ignored. I'll be going back to NVidia when I can justify a new card. :)

      --
      # init 5
      Connection closed.


      Oh... ...bugger.
  45. mod parent down, incorrect information by poohsuntzu · · Score: 1

    Not only are you acting biased, but you are giving incorrect information. Let's start on the BSD part.

    You say its not an option because it is no longer being developed. I say you are wrong. In fact, the website of OpenBSD says you are wrong too:

    OpenBSD is freely available from our FTP sites, and also available in an inexpensive 3-CD set. The current release is OpenBSD 3.5 which started shipping May 1, 2004.

    OpenBSD 3.6, our next release, will be made available for FTP on November 1, 2004. It can be ordered now (CDs are already shipping!)


    In fact, Freebsd has already released their 5.3 beta and is having people test it out as we speak (look at their news page).

    So, your views on BSD and where it is at in the world are completely inaccurate and misguided. So, because of just that alone, I don't see why I should trust -any- of your other advice above.

    --
    "We're breaking out the ramen noodles. . . "
    "Really? Is it someone's birthday?"
  46. love of chickens by chickens · · Score: 1

    Me too!

  47. One or the other monitor by Anonymous Coward · · Score: 1, Informative

    Are you sure your card supports dual output? Some only support one or the other but not both at once; try plugging in your DVI monitor on boot and making sure the CRT is unplugged from the card (or it will take precedence).

    If both should work simultaneously, look at TwinView settings in the NVidia driver README and edit your XF86Config.

    Option "TwinView"
    Option "SecondMonitorHorizSync" "31-60"
    Option "SecondMonitorVertRefresh" "50-75"
    Option "MetaModes" "1024x768, 1024x768"
    #Option "TwinViewOrientation" "RightOf"
    #Option "TwinViewOrientation" "LeftOf"
    Option "TwinViewOrientation" "Clone"
    #Option "NvAGP" "1"

    1. Re:One or the other monitor by B5_geek · · Score: 1

      Thank-you for your quick reply.

      Yes, multi-monitor does work. I use it in XP all the time.

      --
      "The price good men pay for indifference to public affairs is to be ruled by evil men." ~Plato (427-347 BC)
  48. No... by Greyfox · · Score: 2, Insightful

    That's why ATI's crappy drivers suck. That entire page seems to be bitching exclusively about ATI's drivers. Apparently the commentary on the Nvidia drivers was on the previous page and went something like "We told it to install in yast and went init 3 and it worked." I paraphrase. In fact, only a paragraph or two on the previous page talk about Nvidia's driver, in glowing terms. The rest of it is complaining about ATI driver configuration. Then the next page (Where you got that quote) is talking about ATI driver configuration.

    --

    I'm trying to teach myself to set people on fire with my mind... Is it hot in here?

  49. Chip Spec's would be better by johnjones · · Score: 2, Insightful

    frankly its a joke

    nvidia has always fscked around trying to get mangled source into XFree

    ATI where unable to understand what this linux thing was but people who ati trusted wrote drivers...

    solution - provide the chip spec's
    frankly with CG and other GPU compilers the hardware should be patented if not then what harm is there ?

    giving out spec with COPYRIGHT stamped all over it only supports you

    for example those XGI people could capture the market by opens sourceing the drivers (they failed under SIS because of the drivers so let people hunt bugs WITH you for a better product and respect...)

    regards

    John Jones

    1. Re:Chip Spec's would be better by runderwo · · Score: 1
      for example those XGI people could capture the market by opens sourceing the drivers
      XGI released a driver package that has an open source DRM and framebuffer driver. Unfortunately, the XAA and DRI components are binary only. Get it here
  50. nvidia source by poptones · · Score: 1
    I think that's quite wrong. The nvidia driver package comes with object modules and enough source to link it into your kernel IF you have the kernel source installed on your machine. At least, that's the way it worked yesterday when I built the network and sound modules for mdk10.1 community on my shuttle mn31n.

    Are the drivers for older ATi cards oss? I put together a machine from spare parts and it has an old Rage128 card that works like gangbusters in openGL. If this isn't open source then I would say the best truly open 3d performance will be had from an SiS based setup. I have an SiS7xx based motherboard that has served me well for about two years and it uses all open source parts. The driver doesn't install 3d support ootb in either suse or mdk, but they're available from the website of the fellow who maintains the driver packages for linux.

    And for the record: the Sis based board worked damn near flawlessly from the moment of install. The MN31N doesn't even run mdk10 ootb (it crashes randomly and usually can't even complete the installer) but it kinda sorta works OK with suse91. - if you can tolerate all the other crap that's broken (like the entire gnome desktop) in Suse. Basically, I got this mbd because I got a good price and I wanted to try out the MCP-T sound support (which so far completely sucks, but that's another story) and because it has built in dualhead support. I managed to get things working ok with mdk10.1 community, but I don't believe I will ever buy another board that requires proprietary binary drivers in order to access all those wonderful "features." In so many ways, it just ain't worth the trouble.

    1. Re:nvidia source by Bull999999 · · Score: 1

      I think that's quite wrong. The nvidia driver package comes with object modules and enough source to link it into your kernel IF you have the kernel source installed on your machine.

      When I extracted the contents of the package using "-x" argument, it seemed to contain more than "engouh source to link". Besides, why would their display drive package come with modules for network and sound? I think that we may be talking about two different packages.

      --
      1f u c4n r34d th1s u r34lly n33d t0 g37 l41d
  51. That's only needed for Debian. by brunes69 · · Score: 1

    NVidia provides binary drivers precompiled for the most common distors - RedHat, Fedora, Suse, Mandrake. They can't possibly provide binaries for all of them.

    Compiling your own is the price you pay for using a less popular distro. I have to compile my own in Gentoo as well. Not that I mind, it takes all of 10 seconds.

    1. Re:That's only needed for Debian. by Timothy+Brownawell · · Score: 0, Redundant
      NVidia provides binary drivers precompiled for the most common distors - RedHat, Fedora, Suse, Mandrake. They can't possibly provide binaries for all of them.
      Debian provides compiled nvidia kernel modules (in the non-free section, not in the main distribution) to match the compiled kernels they provide.

      Tim

  52. NVidia delayed NForce GARTpatch to screw ATI users by Anonymous Coward · · Score: 0

    NVidia previously withheld their Nforce GART drivers, forcing those without Nforce boards to be unable to use their non-NVidia AGP cards. The NVidia graphics card drivers had the Nforce GART in them, so that NVidia graphics cards users could still use them on Nforce boards. People wanting to run ATI cards on Nforce boards were screwed for many months until NVidia finally released their drivers officially. I got around this only by obtaining a leaked kernel patch with Nforce GART.

  53. UT2004 benchmark matches my experience by AvantLegion · · Score: 1
    I have a Radeon 9600 XT. Not today's super-stud graphics card, but I can play UT2004 in Windows at 1280x1024 (my LCD's native res) with all graphical settings on "Normal" or "On" (for the on-off selectable ones). The game runs very smoothly throughout.

    Fire up my Gentoo install, and the game is terrible at the same res, with all graphical settings at "Lowest" or "Off". I could probably play at like 800x600, but I like running things on my LCD's native resolution.

    Of course, that's why I still have a Windows partition in the first place.

    1. Re:UT2004 benchmark matches my experience by Norgus · · Score: 1

      Not a gentoo problem, more a problem with grasping the installation and use of the latest drivers. I run UT2004 in gentoo 1024x768 with every setting maxed out and its a beauty.

    2. Re:UT2004 benchmark matches my experience by AvantLegion · · Score: 1
      The drivers are installed and used correctly.

      I can run it fine in lower resolutions too. But lower resolutions are shit.

  54. The ATI drivers are bad, but not *that* bad by gotan · · Score: 3, Informative

    I gathered some experience with the ATI drivers on Suse 9.1 recently and i too think that they're bad, but it got a little better lately.

    While there *is* an "auto-installing" driver-package from ATI you'd better avoid that (unless they fixed a good number of bugs). Just running the package resulted in an error for me, googling around i found some hints and managed to install them: run the package in extract-mode, make manually, ignore error, make install accompanied by some messing with /usr/src/linux/.config.

    there is also (for Suse 9.1) an rpm-package. Following the README in that path closely will get the video driver installed. Like nvidia ATIs driver combo too consists of a kernel driver and a n X-driver, and as usual the kernel-driver is a little fiddly to install. There is *no* (longer?) need to compile a custom kernel, you need to install the kernel source though (and really, read the README!).

    Be careful though when configuring the XF86config. fglrxconfig is *not* a good idea since it asks you about mouse settings, monitor modes and whatnot, things that are running perfectly well and shouldn't be touched anyway. NVIDIA does a much better job just telling you the few lines you have to change in the config, fglrxconfig produces an XF86config-4 that is mostly useless and contains heaps of garbage.

    To make the kernelmodule load automatically add two lines to the "modprobe.conf.local" (i think the first is unnecessary):

    install fglrx /sbin/modprobe --ignore-install fglrx && { /sbin /modprobe nvidia_agp; /bin/true; } alias char-major-226-0 fglrx

    In XF86config load "glx" and "dri" in the Modules sections and put
    Driver "fglrx"
    Option "UseInternalAGPGART" "no"
    in the device section. If you've got access problems put:

    Section "DRI"
    Group "video"
    Mode 0666
    EndSection

    After restarting the X-server (twice to be sure, and check if the kernel module loaded) "fglrxinfo" should tell you something about ATI (and not Mesa), if that works do a "sync" for good measure and try tuxracer.

    In my experience the nvidia-drivers are definitely easier to install, but it's really not impossible to get the ATI-stuff running.

    --
    "By the way if anyone here is in advertising or marketing... kill yourself." -- Bill Hicks
  55. MOD UP TRUTH by Ayanami+Rei · · Score: 0, Troll

    Seriously.

    Why post anon?

    --
    THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
  56. [nt] thank god for The Weather Channel! by Anonymous Coward · · Score: 0
  57. Also... by Ayanami+Rei · · Score: 1

    Try making sure that 1) you're using ALSA drivers 2) Your OSS-Free emulation driver allows for multiplexing (ala Direct X). So that way artsd and Wolfenstein can play nice together.

    --
    THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
  58. Re:Linux is G-R-E-A-T by Anonymous Coward · · Score: 1, Funny

    Fortresses don't have windows.

  59. ATI driver Info for Debian users by grolschie · · Score: 1

    Debian users might find it easier to use Flavio Stanchina's ATI debian packages and HOWTO. I found these much easier to install.

  60. Coresponding Windows Benches by Deliveranc3 · · Score: 1

    Here they are.

    Mobo, Vid and cpu = same.

    If linux was faster I would actually switch... I spend so much time tweaking windows maybe linux would be easier...

    1. Re:Coresponding Windows Benches by oddfox · · Score: 1

      If you use an NVidia card you don't have these sorts of issues, just FYI. I kinda wish I had one but my 9800 Pro/XT was a big investment lol.

      --
      "We invented personal computing." - Bill Gates
  61. Well... by Kiryat+Malachi · · Score: 0, Offtopic

    If you're SCO, it's pronounced "Sues".

    --

    ---
    Mod me down, you fucking twits. Go ahead. I dare you.
    (I read with sigs off.)
  62. Racer? by stu42j · · Score: 1

    What is the Racer game they use as one of the benchmarks? At first I assumed they meant TuxRacer but then it describes driving a car, not a penguin.

    Freshmeat gives two other possibilites, neither of which seem quite right.

    1. Re:Racer? by Anonymous Coward · · Score: 0

      I believe they are talking about this.

    2. Re:Racer? by big+ben+bullet · · Score: 1

      It's one of the games that comes prepackaged with SuSe 9.1

      You can find more info here. It actualy is a nice car simulation with good physics.

  63. Here is the Beef by Anonymous Coward · · Score: 0

    Nunber one Why.

    Linux systems normally no need to reboot(I have a key linked to serial port to force init 3 fixs a dead X11 interface so I can init 5 to reset it one day I will add a second button ie init 3 and init 5).

    Just because the graphic interface dies why do I have to unloaded all drivers?

    I install a new kernel most cases no reboot until normal shutdown and startup the next day. Video drivers most offten can be installed in init 3 mode.

    The reason my machine shares files people get very upset when the files they want are off line ie a windows driver update would force me to kick them out. A linux driver update no they keep on working (the slowdown of a kernel being build a root still gets a few yells)

    Basicly you cannot get away with rebooting servers. Note windows also uses a dirty trick Called after loading XP not afterloading takes twice the time of suse default install starting up. ie 30 secs not 15. 15 secs are been hidden from normal users. Start playing a game quake2 or bigger of the start line and ask why its it slower to load than when I let the machine sit for a min ie if fully loaded there should not still be stuff loading slowing stuff down.

    Also XP does a non safe I will kill you and I hope you can sort it out. Linux says program I am killing you please get your affiars in order I will wait if asked.

    Allso try playing Quake 2 and other games only to have the computer blue screen of death while you are winning.

  64. 6600GT by Trejkaz · · Score: 1

    With all the news on the GeForce 6600GT lately, I would have hoped that they showed its placing in the benchmarks here... particularly since I've had my eyes on buying this card ever since Slashdot mentioned it.

    What a rip-off.

    --
    Karma: It's all a bunch of tree-huggin' hippy crap!
    1. Re:6600GT by Mildog · · Score: 1

      The 6600GT is PCI express only at the moment. And, there are currently no mobo's that have both socket 939 and PCI-e

      mildog

  65. Open your wallet by anti-NAT · · Score: 4, Insightful

    You mean like my Zoran TV tuner card that hasnt worked since the 2.2 series, despite their being public specs and sources for it? Drivers are only maintained as long as the developer is around. And unless you have the skills to write your own drivers (and most of us, including large numbers of application developers, *don't*), the having specs/source or not is irrelevant.

    Why not offer to pay an open source developer to update it ? At least you have that option, independent of the manufacturer's support for doing so.

    You going to have the same problem with Linux kernel version 4.0, when Nvidia don't provide a driver for their XYZ card in 2010. Problem is, at that time, you won't have access to the specs, so you won't even have the option of paying an open source developer to update the driver for you.

    --
    The Internet's nature is peer to peer - 20050301_cs_profs.pdf
    1. Re:Open your wallet by tyrantnine · · Score: 1

      Do you really expect an individual is going to be able to afford this? I have no real idea what kind of timeline it'd take to write a driver from scratch, nor how much an update from 2.2->2.6 would take. Anyone with any background in writing a driver, please inject.

      I think its safe to assume the time commitment for someone with the skills wouldn't be complete peanuts. I think its safe to assume the cost of paying someone or a bounty would eclipse the cost of the hardware by several orders of magnitude.

      In that case, I suppose you might try and form a group of like-minded owners of the same hardware to spread the burden. I'd imagine pulling this off would be nigh impossible (and quite a bit of work on the part of the organizers too, just to get a bit of old hardware functioning), but if theres a story out there of something like this sucessfully happening at a grassroots level (IE - no company involved), I'd like to hear about it.

      I also have to take issue with the other implication you made regarding open specs. I'm all for them, but do you really expect this guy is going to worry about drivers in 2010 for his ancient TV tuner card? Whatever consumer level PC hardware you have today is going to be worthless junk in that sort of timeframe (6+ years). Its not much of an argument for open specs.

      An individual funding this sort of development/bounty is not within reality.

    2. Re:Open your wallet by anti-NAT · · Score: 1

      I have no real idea what kind of timeline it'd take to write a driver from scratch, nor how much an update from 2.2->2.6 would take. Anyone with any background in writing a driver, please inject.

      I wouldn't expect all that much effort to move a 2.2 driver to 2.6 (which would apply to the original poster's TV card) - LDD 2nd edition goes into what is involved in a 2.2/2.4 compatible driver, and kernel developers are regularly co-developing 2.4/2.6 compatible versions of their drivers. I'd suggest it would be probably easily only a days work (probably less, maybe 3 to 4 hours) for a competent linux kernel developer. That may still be too expensive for an individual, however, the point is it could happen, because the only resource lacking is time and / or money, not the essential programming information required. Time and / or money is an "unlimited" resource, in the sense that "money" can always be made by working for it, and time can be taken from something else - switching the TV off magically produces more of it.

      I'm all for them, but do you really expect this guy is going to worry about drivers in 2010 for his ancient TV tuner card? Whatever consumer level PC hardware you have today is going to be worthless junk in that sort of timeframe (6+ years). Its not much of an argument for open specs.

      The four ISA 10Mbps ethernet cards I have in an old 486DX66, manufactured in 1991/92, aren't "worthless junk", and are currently supported in the latest 2.6.8.1 kernel, which this machine is capable of running (actually, it is running 2.6.7, I haven't got around to upgrading it).

      Of course, if those cards had "binary" drivers, it's unlikely I would have been able to use them in recent kernels (quoting your "6+" year figure, that would probably mean the binary drivers stopped being updated around 1998), irrespective of them still working perfectly, and would have had to buy both new network cards and a new machine to fit them. I have no need to spend any money on "better" performance for what I'm using this machine for (IP routing if you're interested, not throughput based, just an experiment "lab" running routing protocols such as OSPF, BGP etc.), I would be being forced to spend money needlessly just because a hardware manufacturer won't tell me or anybody else how to get their hardware to do what it is designed to do.

      I'm sure there is a lot of perfectly functional and good enough performing hardware lying around that can be useful to people - not only people like me, but also people in less priviledged countries. Open hardware programming specs ensures that those people don't end up being given "obselete" hardware that can't actually be used, just because the vendor has decided to abandon it, "driver development"-wise.

      --
      The Internet's nature is peer to peer - 20050301_cs_profs.pdf
    3. Re:Open your wallet by tyrantnine · · Score: 1

      I wouldn't expect all that much effort to move a 2.2 driver to 2.6 (which would apply to the original poster's TV card)

      Sounds reasonable, though in thinking about it a little more it'd also obviously depend on how much functionality the driver needs, and whether there's optimizations to be made. I'd guess writing or even updating a driver for a video card would be quite a bit more work than for a ethernet card.

      f course, if those cards had "binary" drivers, it's unlikely I would have been able to use them in recent kernels (quoting your "6+" year figure, that would probably mean the binary drivers stopped being updated around 1998), irrespective of them still working perfectly, and would have had to buy both new network cards and a new machine to fit them. I have no need to spend any money on "better" performance for what I'm using this machine for (IP routing if you're interested, not throughput based, just an experiment "lab" running routing protocols such as OSPF, BGP etc.), I would be being forced to spend money needlessly just because a hardware manufacturer won't tell me or anybody else how to get their hardware to do what it is designed to do.

      I didn't mean worthless junk in terms of "non-functional". Let's assume those cards werent supported, but have an open spec (ala the original posters TV card. Your situation is different - someone has done the work already). You have three possibilities -- 1) You write a driver yourself 2) You go purchase 4 brand new, cheapy (but faster) ethernet adapters for $25 brand new that are supported 3) You pay someone else to write the driver.

      That is what I meant by worthless junk. If no one has developed a driver yet, paying for someone else to do it (or doing it yourself) is not cheaper than simply replacing it. If you can think of a single piece of commodity PC hardware you bought 6+ years ago where this wouldn't be the case, I'm all ears.

      I'm not arguing against the value of open specs. However "at least you could do it yourself" or "at least you could pay someone to update it" are both are weak arguments. The original posters point:

      "And unless you have the skills to write your own drivers (and most of us, including large numbers of application developers, *don't*), the having specs/source or not is irrelevant.

      Is the pragmatic reality. If no one happened to do the work previously, its really of no use to you.

    4. Re:Open your wallet by brunes69 · · Score: 1

      The four ISA 10Mbps ethernet cards I have in an old 486DX66, manufactured in 1991/92, aren't "worthless junk", and are currently supported in the latest 2.6.8.1 kernel, which this machine is capable of running (actually, it is running 2.6.7, I haven't got around to upgrading it).

      Actually, considering I could buy a PII or newer with two PCI nics, with full system specs that would thrash that sustem, for under 50 bucks, yes I would call that whole system "worthless junk".

      The savings on your power bill alone yb upgrading to a modern mor epower efficient system/PSU would probably pay for the price of the upgrade in a few months.

    5. Re:Open your wallet by jdschulteis · · Score: 1

      Do you really expect an individual is going to be able to afford this?

      Put in a request for bids at RentACoder and find out.

    6. Re:Open your wallet by anti-NAT · · Score: 1

      Actually, considering I could buy a PII or newer with two PCI nics, with full system specs that would thrash that sustem, for under 50 bucks, yes I would call that whole system "worthless junk".

      Why spend money when you don't need to ?

      The savings on your power bill alone yb upgrading to a modern mor epower efficient system/PSU would probably pay for the price of the upgrade in a few months.

      You are assuming I run it often. I probably run it less than an average of 3 hours a month. The electricity savings you suggest are probably not going to pay for a new $50 machine within a decade, based on my usage patterns.

      --
      The Internet's nature is peer to peer - 20050301_cs_profs.pdf
    7. Re:Open your wallet by brunes69 · · Score: 1

      While I cannot possibly comprehend why you would need to run such a machine for a few hours a month (if I needed an extra machine for testing or something running a local UML kernel on my desktop would be both simpler and *faster* than running an old, old 486), if that is the case, then I guess you are right.

  66. ATI 9200 / 8500 by anti-NAT · · Score: 1
    --
    The Internet's nature is peer to peer - 20050301_cs_profs.pdf
  67. Re:Indeed by MrNemesis · · Score: 1

    I imagine it's true for most popular distros, in that it's recommended to use the distros package rather than the nvidia installer.

    On gentoo at least, there are a multitude of scripts available that will check your foreign modules against the running kernel on bootup. If they find they were compiled against a different one, they will recompile pointing at your new kernel tree, and then restart the respective services (X, lircd, etc), making upgrades stupidly easy. I imagine people have made similar scripts or utils for pretty much every major mainstream distro.

    I haven't used it since the RH9 days, but I am told if you run the nvidia installer, it also includes binaries for specific kernel versions from major distros, so if you're a n00b you don't even need the kernel source - it'll sense you're running the latest version of Fedora or whatever and plug the modules in without needing to compile anything.

    --
    Moderation Total: -1 Troll, +3 Goat
  68. I figure its http://www.racer.nl/ by Anonymous Coward · · Score: 0

    Unfortunately, its not open source

    There have been misunderstandings about the source code in the past time. This project is NOT OpenSource. The source code IS and will REMAIN for the oncoming time copyright (c) of Ruud van Gaal / Dolphinity BV. The source code is provided for general interest, and to build platform-specific versions (for Linux mainly) in case the provided binaries don't work.

    All suggested changes to the source code will be reviewed by me, but before you do that, realise that the source code will remain copyrighted. Fixes are appreciated ofcourse, but I would advise against putting much into bigger things.


    Link to statement

  69. ATI sucks worse than described by Anonymous Coward · · Score: 0

    All the NVidia cards are, more or less, supported with Linux drivers when they ship. I've had my ATI Radeon Mobility 9700 since March. Still no ATI-accelerated drivers. This is a common trend -- ATI very frequently supports only their older cards under GNU/Linux. Now I see why -- apparently, their newer cards underperform older NVidia cards, so no clueful GNU/Linux user would buy them anyways.

  70. Better drivers and licensing please-Euphoria. by Anonymous Coward · · Score: 0

    "In fact, you are lucky to get anything from nVIDIA or ATI."

    Gee! And we talk about how powerful we are. Guess the caffine wore off.

  71. did you actually look at any of it? by poptones · · Score: 1
    Mkay, so you opened the package... uh, did you actually look at any of the contents?

    Since the Linux kernel does not support a binary driver interface, we provide for rebuilding these files on the target machine (or distribution) and then linking with the binary version of the NV kernel driver.

    And yes, there are two packages: a display package and a sound/networking package.

    Both contain proprietary, closed source modules.

    1. Re:did you actually look at any of it? by Bull999999 · · Score: 1

      I'm not a programmer but won't you need the source to recompile the module? I could be wrong so enlighten me if I'm. I did take a look that the contents in detail in order to install the older versions of the drivers when I was using Slackware 9. I did see precompiled modules for the more "main stream" distros, but had to compile the module from the sources for slackware.

      People who aren't afraid of recompiling the nvidia kernel module or even reinstalling the nvidia driver each time the kernel has been updated and want or need to use the latest and greatest nvidia driver can use the following steps 1-3.

      --
      1f u c4n r34d th1s u r34lly n33d t0 g37 l41d
    2. Re:did you actually look at any of it? by Anonymous Coward · · Score: 0

      Yes, you need the source to recompile the driver. No, you don't get the source for the driver.

      You only get the binary objects (*.o) and some glue code (source), and the only thing that is recompiled is the glue code, which is then linked with the binary object files.

    3. Re:did you actually look at any of it? by Bull999999 · · Score: 1

      Thanks for the info, I stand corrected.

      --
      1f u c4n r34d th1s u r34lly n33d t0 g37 l41d
  72. That's not why NVidia is closed source by Animats · · Score: 1
    The main reason NVidia still has closed source drivers is the GeForce/Quadro distinction. Quadro cards use the same chip as GeForce cards. There's a software switch that cripples the driver when running on a GeForce card. Some GeForce cards clock a bit faster. But the distinction lets NVidia charge 3x as much for the "pro" card.

    The big difference seems to be that in GEForce mode, some extra bugs that mess up high-end animation applications are turned on.

    That probably has more to do with the issue than any "proprietary technology".

    1. Re:That's not why NVidia is closed source by Stevyn · · Score: 1

      Yeah I've read about that theory people have. It probably makes sense since it's cheaper to produce fewer different GPUs. Car manufacturers do a similar tactic by throttling their engines at high speeds. My toyota corolla will go above the speed limit of 112 mph, but it's limited.

  73. Now this is what I like...pure data by carlmenezes · · Score: 1

    They have actually put up a downloadable CSV of the results for us techies:
    http://images.anandtech.com/reviews/linu x/gpu_roun dup/fps.csv

    Now that's really nice. If they would make this a regular feature, it would enable regular readers of anandtech to stock up on CSVs of hardware they might want to buy in the future.
    if any of you AnandTech chaps are reading this,
    Excellent idea guys! keep it up.

    --
    Find a job you like and you will never work a day in your life.
  74. Before dual OS processors... by Anonymous Coward · · Score: 0

    Can we compile linux to run on the GPU?

  75. Porting cheaper than new development by mdfst13 · · Score: 1

    If you already have a working driver on one platform, it is cheaper to rewrite that driver than it is to write a new one. If you do this over several design iterations of your product, you can develop a design system where platform dependent code is separated from platform independent code in such a way that you have to make few changes to the platform dependent code per product. I.e. platform dependent code will be most product independent and product dependent code will be most platform independent.

    To respond using your numbers as a base: if it costs $1M USD to develop a MS Windows driver, then I would expect it to cost less to develop a Linux driver from the MS Windows driver, perhaps $100K. Thus, even though your Linux numbers are wildly optimistic (Linux is closer to 3% than 10%), it is not nearly as bad as it seems at first.

    Another issue is that video cards are a competitive market. Even if the overall market is only 3%, it is worth noting that by having the better driver, that company gets most of the 3%. They have to split the 90% from MS Windows.

    It's also worth noting that MS Windows sales are largely through OEMs and get the OEM discount. Linux sales are largely singletons and get the retail premium. Thus, Linux users as a group are somewhat more profitable than MS Windows users as a group.

    Finally, nVidia makes motherboard chipsets as well as video chipsets. By producing the better video drivers, they encourage people to buy their motherboards as well. ATI otoh, produces Mac and Sun video cards; thus, they already need to maintain platform independent code. This makes it much easier to produce lousy drivers, which they do so as to keep a toe in the market.

  76. Uhh.... by warrax_666 · · Score: 1
    I'm betting most people's parents would be likely to plug in a new digital camera via USB to download their images...
    ... and would work fine with any desktop-oriented distro. I seem to recall that Mandrake (or was it SuSE?) pops up a new camera icon on the desktop when you plug in a USB camera. You almost pulled of that bait-and-switch you did there.
    --
    HAND.
    1. Re:Uhh.... by nvrrobx · · Score: 1

      And that almost works reliabily in SuSE. And yes, I am speaking from experience.

      Funny, your ego almost pulled through there, but not quite.

      If you were going to flame me on that, you should have verified which distro does that first.

  77. Please learn how to make links. by Anonymous Coward · · Score: 0
    Please learn how to make links.
    <a href="http://images.anandtech.com/reviews/linux/gp u_roundup/fps.csv">CSV file</a>
    (without any spaces put there by Slashdot) yields: CSV file

    If that's too much typing for you,
    <URL:http://images.anandtech.com/reviews/linux/gpu _roundup/fps.csv>
    (without any spaces put there by Slashdot) yields: http://images.anandtech.com/reviews/linux/gpu_roun dup/fps.csv
    1. Re:Please learn how to make links. by carlmenezes · · Score: 1

      being rude will get you nowhere.

      --
      Find a job you like and you will never work a day in your life.