V2 Of KDBUS Published For Linux Kernel Review

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Stellarwind
    replied
    Originally posted by dnebdal View Post
    I don't know about "not important at all". If it gets wide enough adoption, it could turn out to be like /proc - perhaps not something you think about every day, but one more useful tool (that some other neat tools depend on).
    My point was that from kernel pov it is just another IPC mechanism.
    Linus doesn't bump kernel version number every time new filesystem is accepted into mainline, why should this happen for kdbus?

    Leave a comment:


  • curaga
    replied
    Originally posted by Stellarwind View Post
    Transition from 2.4 to 2.6 was a huge step with alot of kernel subsystems changing, but since then development model changed, 2.6 to 3.0 was mostly meaningless, just because version numbers hit high values. kdbus is not important at all, it is a thing you can turn off and loose nothing and in no way major feature for a kernel.

    p.s. always entertaining: http://gentooexperimental.org/~patri...3T09_26_01.txt
    That's hilarious. Sure I knew dbus was a clusterfuck, but not that it was that bad.

    Leave a comment:


  • dnebdal
    replied
    Originally posted by Stellarwind View Post
    Transition from 2.4 to 2.6 was a huge step with alot of kernel subsystems changing, but since then development model changed, 2.6 to 3.0 was mostly meaningless, just because version numbers hit high values. kdbus is not important at all, it is a thing you can turn off and loose nothing and in no way major feature for a kernel.

    p.s. always entertaining: http://gentooexperimental.org/~patri...3T09_26_01.txt
    I don't know about "not important at all". If it gets wide enough adoption, it could turn out to be like /proc - perhaps not something you think about every day, but one more useful tool (that some other neat tools depend on).

    Leave a comment:


  • Stellarwind
    replied
    Originally posted by nilssab View Post
    Well, I have heard him say this at many events. But that doesn't mean that we cannot and shouldn't try to revise our opinion on that idea by discussing it.
    Personally I find this too seem like such a feature that would warrant a new major version, and that doing a major version bump when it gets accepted or when that code is considered stable would make sense for many users.
    but I feel I have likely missed something that would equally warrant a similar bump, so I ask what the community thinks of this. Would it demerit other large chances and are there other reasons to not do this? (save for the opinion of Linus)
    Transition from 2.4 to 2.6 was a huge step with alot of kernel subsystems changing, but since then development model changed, 2.6 to 3.0 was mostly meaningless, just because version numbers hit high values. kdbus is not important at all, it is a thing you can turn off and loose nothing and in no way major feature for a kernel.

    p.s. always entertaining: http://gentooexperimental.org/~patri...3T09_26_01.txt

    Leave a comment:


  • nilssab
    replied
    Originally posted by mark45 View Post
    Linus Torvalds said a major version bump occurs when the minor version gets up to like 40. No major version bump based on features, google for more info from Torvalds.
    Well, I have heard him say this at many events. But that doesn't mean that we cannot and shouldn't try to revise our opinion on that idea by discussing it.
    Personally I find this too seem like such a feature that would warrant a new major version, and that doing a major version bump when it gets accepted or when that code is considered stable would make sense for many users.
    but I feel I have likely missed something that would equally warrant a similar bump, so I ask what the community thinks of this. Would it demerit other large chances and are there other reasons to not do this? (save for the opinion of Linus)

    Leave a comment:


  • Apopas
    replied
    Never say never!

    Leave a comment:


  • xeekei
    replied
    Originally posted by Apopas View Post
    But from 2.4 to 2.6 or from 2.2 to 2.4 the differences were tremendous, correct?
    AFAIK, the Linux kernel's development process was different back then. Now it's much more of an evolutionary process than before. Revolutionary changes probably won't happen again.

    Leave a comment:


  • Apopas
    replied
    But from 2.4 to 2.6 or from 2.2 to 2.4 the differences were tremendous, correct?

    Leave a comment:


  • Namenlos
    replied
    Originally posted by michal View Post
    [?]so 4.0 can be released instead 3.20[?]
    In the back of my head I remember sth similar. There was also some thoughts about not adding features and focus on stability, but I am not sure. -- Well found it with this information: http://www.infoworld.com/article/261...-features.html But I think this is off the table, when there is no announcement yet and the commit window is closed in DRM next for new features and for all the other stuff it will close in a month or so.

    Leave a comment:


  • mark45
    replied
    Originally posted by michal View Post
    IIRC he said that he doesn't like high numbers like 40, so 4.0 can be released instead 3.20.
    Whih is what I said. And he also mentioned he won't do a major version bump because of any new functionality.

    Leave a comment:

Working...
X