Announcement

Collapse
No announcement yet.

Linus Torvalds Asks Kernel Developers To Write Better Git Merge Commit Messages

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

  • Linus Torvalds Asks Kernel Developers To Write Better Git Merge Commit Messages

    Phoronix: Linus Torvalds Asks Kernel Developers To Write Better Git Merge Commit Messages

    Yesterday when announcing the Linux 6.12-rc2 kernel, Linus Torvalds asked that the kernel maintainers do a better job moving forward with their commit messages...

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    Nothing controversial here, this is pretty standard and all of the OS projects that I work on/maintain (as well as my day job) uses imperative language for git commit messages.

    Comment


    • #3
      That's how we do it, too. Clear, precise, active. Leave prosa for newsblogs.

      Comment


      • #4
        completely agree

        Comment


        • #5
          English teachers all over would be proud.

          Comment


          • #6
            Isn't this potentially an actual use case for LLMs? I mean formatting and correcting grammar in a commit message to be overall consistent should be something to solve automatically, right?

            Comment


            • #7
              Originally posted by TheJackiNonster View Post
              Isn't this potentially an actual use case for LLMs? I mean formatting and correcting grammar in a commit message to be overall consistent should be something to solve automatically, right?
              If "should be" and "is" are guaranteed to overlap completely on a Venn diagram, sure. Otherwise you're creating extra, horribly annoying work for yourself.

              Comment


              • #8
                Originally posted by TheJackiNonster View Post
                Isn't this potentially an actual use case for LLMs? I mean formatting and correcting grammar in a commit message to be overall consistent should be something to solve automatically, right?
                Not really, because the LLM wouldn't have context about the code. So in the example Linus gave, it wouldn't be able to tell you what it fixes the null reference in. Sure it could tell you the function name, but not the context around it or why it needed to be fixed.

                Comment


                • #9
                  Been telling that to devs for years. Newsflash, the devs are still winning

                  Comment


                  • #10
                    Originally posted by bug77 View Post
                    Been telling that to devs for years. Newsflash, the devs are still winning
                    Those who do [the work], decide [how to do it].

                    Comment

                    Working...
                    X