Page 3 of 4 FirstFirst 1234 LastLast
Results 21 to 30 of 34

Thread: 2008 Linux Graphics Survey Results

  1. #21
    Join Date
    Oct 2007
    Location
    Toronto-ish
    Posts
    7,411

    Default

    Quote Originally Posted by KDesk View Post
    But... KMS needs DRI2, so, the pool was wrong. It should have been DRI2 or KMS (and DRI2) and the other options also.
    My understanding was that KMS needed memory management (GEM/TTM), DRI2 also needed memory management but KMS did not need DRI2. They will probably arrive at more or less the same time, however, since they both become possible as soon as memory management is available, and both are considered really high priority by the developers.

    EDIT - bah, ASCII diagram didn't work ;(
    Last edited by bridgman; 12-23-2008 at 12:07 AM.

  2. #22
    Join Date
    Aug 2008
    Posts
    233

    Default

    Quote Originally Posted by bridgman View Post
    My understanding was that KMS needed memory management (GEM/TTM), DRI2 also needed memory management but KMS did not need DRI2. They will probably arrive at more or less the same time, however, since they both become possible as soon as memory management is available, and both are considered really high priority by the developers.

    EDIT - bah, ASCII diagram didn't work ;(
    Oh, yes, you are complete right! The share the memory manager in common... and not DRI2.

  3. #23
    Join Date
    Dec 2008
    Posts
    1

    Default

    Quote Originally Posted by Eragon View Post
    How can people keep saying linux is not suited for "normal" users if 56% of it's users describe themselves as "mainstream users"??
    Of the four choices available in the survey, IMHO the closest choice to normal was mainstream. The survey is still available in review fashion which helps see how questions and answers can sometimes skew perception of answers. I was looking to see how the survey handled the question of hardware because I have own or use via work a combination of linux booting computers (ex: macbook pro with nvida, two dell laptops with ati, eeepc, amd tower with nvida). The multiple choice answer to that question came close to capturing my collection but is still skewed (IMHO) because it doesn't show the count per manufacturer.

  4. #24
    Join Date
    Mar 2008
    Posts
    23

    Default

    Quote Originally Posted by RobbieAB View Post
    Windows has done it for years, as has MacOS. Linux has "innovated" with eye-candy instead...
    That is the very sad truth!

  5. #25
    Join Date
    Dec 2008
    Posts
    14

    Default

    Quote Originally Posted by deanjo View Post
    That's not my theory at all.
    If that's the case, do you have an appropriate 3rd party licence?



    It's fact that despite best intentions and planning thing don't always go as well as expected. Until actual implementation is actually out there and performing as well as expected it's pure speculation.
    Geez... I was just being silly. No need to drive home an obvious point. I hope you didn't burst a vein coming up with your response.

  6. #26
    Join Date
    Aug 2007
    Posts
    6,610

    Default

    I highly doubt that somebody really needs kms. At least it does not matter for me if switching to a vt takes a few monents or if the default res is already the highest possible. Stability is much more important - also correct rendering and flickerfree videos/games when Vsync is enabled. The rest it nice to have, but not a requirement.

  7. #27
    Join Date
    Nov 2007
    Posts
    1,024

    Default

    Quote Originally Posted by Kano View Post
    I highly doubt that somebody really needs kms.
    And you're wrong. KMS is extremely important for suspend and hibernation, which is critical for laptop users. It's a huge boon for display hotplug. It's mandatory for developing alternative graphics stacks without duplicating a metric shitload of driver code.

    KMS is way more interesting to me than DRI2 is. GEM interests me most of all, of course. Without it, neither DRI2 nor KMS could really work.

  8. #28
    Join Date
    Aug 2007
    Posts
    6,610

    Default

    Ok, I do not own a laptop, so I never use suspend. Hibernation/Suspend to Disk is useless anyway because you would read in worst case more data than you would do with a normal boot when you memory is big and used.

  9. #29
    Join Date
    Oct 2007
    Location
    Toronto-ish
    Posts
    7,411

    Default

    I guess the argument for suspend/resume is that it maintains your application state as well ?

    The other advantage to KMS I see is that it puts all of the state info you need for good power management in one place. Right now the engine state is in drm (unless you're running without 3d) and the display state is in ddx, but you really need both to make on-the-fly power management decisions.

  10. #30
    Join Date
    Jan 2008
    Location
    Radoboj, Croatia
    Posts
    155

    Default

    Quote Originally Posted by Kano View Post
    Ok, I do not own a laptop, so I never use suspend. Hibernation/Suspend to Disk is useless anyway because you would read in worst case more data than you would do with a normal boot when you memory is big and used.
    This is true for hibernation/suspend to disk. But what about suspend to RAM? Most of laptop users use it a lot. And it still is at least 2 times slower than on windows (especially while using fglrx - open source driver suspends and resumes the graphics quicker, but still slower than windows!).

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •