Page 3 of 4 FirstFirst 1234 LastLast
Results 21 to 30 of 36

Thread: NVIDIA Tries To Put Fence Sync Into X Server 1.10

  1. #21
    Join Date
    Nov 2008
    Location
    Aachen, Germany
    Posts
    13

    Default

    From: Dave Airlie
    "[...]This goes for nvidia type situations as well, the whole point is
    to place the maintainer burden at the feet of the people causing the
    problems in an effort to make them change."
    Here is the whole thing: http://kerneltrap.org/mailarchive/li...4589722/thread

    Maybe "not being nice" was the wrong expression, it's more of a not giving a sh*t.

  2. #22
    Join Date
    Oct 2007
    Posts
    912

    Default

    The OSS people don't create APIs to be unusable for closed source drivers - they simply don't the burden of trying to maintain compatibility with closed source drivers for no reason other than to keep those drivers running. The kernel APIs should not be held hostage by some binary blob, and in practice there's not much effort in keeping the binary blobs working anyway.
    It's this way because the devs care about the sanity of kernel interfaces and don't want them bogged down in ancient cruft.

  3. #23
    Join Date
    May 2007
    Location
    Third Rock from the Sun
    Posts
    6,584

    Default

    Quote Originally Posted by mirv View Post
    The OSS people don't create APIs to be unusable for closed source drivers - they simply don't the burden of trying to maintain compatibility with closed source drivers for no reason other than to keep those drivers running. The kernel APIs should not be held hostage by some binary blob, and in practice there's not much effort in keeping the binary blobs working anyway.
    It's this way because the devs care about the sanity of kernel interfaces and don't want them bogged down in ancient cruft.
    Well you would like to think that was the way it worked but we have already seen in the past that Xorg, even though the patch would require virtually no maintenance at all (remember the 3 line patch nvidia submitted that simply gave priority to the blob if detected?). Also you can't say "don't want them bogged down in ancient cruft." as the kernel is full of ancient cruft.

  4. #24
    Join Date
    May 2007
    Location
    Third Rock from the Sun
    Posts
    6,584

    Default

    Quote Originally Posted by Qaridarium View Post
    what is the basement of "work together" ?? there are just enemys the best they can do is ignore the others
    That is not the best they can do, the best they can do is to attempt to work together. The worst they can do is ignore each other.

  5. #25
    Join Date
    Oct 2007
    Location
    Under the bridge
    Posts
    2,142

    Default

    Quote Originally Posted by deanjo View Post
    No, a "better way" is to work together regardless of personal opinion, politics, and ideologies for the greater good of the end user whom more then likely don't care about how it is done but the end result is providing a product that does what the end user wants it to do.
    You are confused. The end users of the kernel are its own developers, paid or unpaid. The paid ones work on the kernel for the benefit of their company. The unpaid ones work on it as a hobby.

    Why should either category bend over to accept and *maintain* code for the benefit of a closed-source company that doesn't help in Linux development (e.g. Nvidia)?

  6. #26
    Join Date
    May 2007
    Location
    Third Rock from the Sun
    Posts
    6,584

    Default

    Quote Originally Posted by BlackStar View Post
    You are confused. The end users of the kernel are its own developers, paid or unpaid. The paid ones work on the kernel for the benefit of their company. The unpaid ones work on it as a hobby.

    Why should either category bend over to accept and *maintain* code for the benefit of a closed-source company that doesn't help in Linux development (e.g. Nvidia)?
    I'm not confused. There are a hell of a lot more end users that are not devs then there are devs. Whether you like to admit it or not, nvidia does help out in linux development. In fact I would go so far as to venture that without nvidias solutions being available you would not have seen so much progress being made on the linux desktop over the last 8 years simply for the fact of them offering a full featured working solution. Then there is also items such as nvidia's contribution to the open API's such as openGL, openCL, etc. To say that nvidia doesn't contribute to development is BS, it may not offer open documentation to it's products but to say that it doesn't contribute to open projects isn't the truth at all.

  7. #27
    Join Date
    Oct 2007
    Posts
    912

    Default

    nvidia does what is best for nvidia. AMD does what is best for AMD. Intel does what is best for Intel. It's the same for any company. It's competition (in this case, usable drivers from both nvidia and amd) that have driven things forward.

  8. #28
    Join Date
    Oct 2007
    Location
    Under the bridge
    Posts
    2,142

    Default

    Quote Originally Posted by deanjo View Post
    I'm not confused. There are a hell of a lot more end users that are not devs then there are devs. Whether you like to admit it or not, nvidia does help out in linux development. In fact I would go so far as to venture that without nvidias solutions being available you would not have seen so much progress being made on the linux desktop over the last 8 years simply for the fact of them offering a full featured working solution. Then there is also items such as nvidia's contribution to the open API's such as openGL, openCL, etc. To say that nvidia doesn't contribute to development is BS, it may not offer open documentation to it's products but to say that it doesn't contribute to open projects isn't the truth at all.
    Strawmen and more strawmen.

    I said that Nvidia does not contribute to Linux (i.e. kernel) development. With that in mind, why should kernel developers be tasked with maintaining Nvidia's blob? What would e.g. Red Hat gain from that? Or independent contributors?

    Like it or not, the kernel is not developed for you and me. It's developed by Red Hat, AMD, Intel, etc *for* Red Hat, AMD, Intel, etc. The fact that we can use this kernel on our desktop is awesome, but we are certainly not the primary target.

  9. #29
    Join Date
    May 2007
    Location
    Third Rock from the Sun
    Posts
    6,584

    Default

    Quote Originally Posted by BlackStar View Post
    With that in mind, why should kernel developers be tasked with maintaining Nvidia's blob? What would e.g. Red Hat gain from that? Or independent contributors?
    Hey Redhat fixes bugs when they effect workstations with Maya on it. Why would they do that? Oh ya, probably because their customers use such applications.

  10. #30
    Join Date
    Oct 2007
    Location
    Under the bridge
    Posts
    2,142

    Default

    Quote Originally Posted by deanjo View Post
    Hey Redhat fixes bugs when they effect workstations with Maya on it. Why would they do that? Oh ya, probably because their customers use such applications.
    Since when does Maya contain kernel modules?

Tags for this Thread

Posting Permissions

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