Announcement

Collapse
No announcement yet.

GCC Soars Past 14.5 Million Lines Of Code & I'm Real Excited For GCC 5

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

  • GCC Soars Past 14.5 Million Lines Of Code & I'm Real Excited For GCC 5

    Phoronix: GCC Soars Past 14.5 Million Lines Of Code & I'm Real Excited For GCC 5

    If you thought LLVM/Clang with just under four million lines was a huge code-base for a compiler as the entire Linux kernel is over 19 million lines, just wait until you see the current size of GCC...

    http://www.phoronix.com/vr.php?view=MTg3OTQ

  • #2
    Contributions sorted by company

    It would be great to see again which company contributed how much.

    It is always interesting to see, that e.g. RedHat is heavily involved in most important projects whereas Canonical usually doesn't give back a whole lot to the community as they usually fork stuff, maintain their own patch-sets or simply prefer to do their own thing instead of building something great together (Unity, Mir, Upstart, ...).

    Comment


    • #3
      Hm, no wonder why GCC compilation takes forever and doesn't even fit into 4 GiB tmpfs. But yes, optimisations are always nice.

      I also hope GDC could be more readily available than it currently is...

      Comment


      • #4
        Originally posted by Linuxhippy View Post
        It would be great to see again which company contributed how much.

        It is always interesting to see, that e.g. RedHat is heavily involved in most important projects whereas Canonical usually doesn't give back a whole lot to the community as they usually fork stuff, maintain their own patch-sets or simply prefer to do their own thing instead of building something great together (Unity, Mir, Upstart, ...).
        Usually I will agree with you. Upstart, Mir are unnecessary/double work, but I really like Unity. It is the most lean environment for me. I tried Fedora 21 and Gnomes Hell for a while, but I got frustrated with the ugly fonts and so much vertical space wasted. I would like to ditch off Ubuntu, but the only viable alternative is OSX unfortunately.

        Comment


        • #5
          However, this is behind LLVM seeing contributions from 272 individuals for 2014.
          Which doesn't surprise anyone given the much cleaner and easier to understand (thus facilitating contributions) LLVM codebase.

          Comment


          • #6
            Originally posted by dom0 View Post
            Which doesn't surprise anyone given the much cleaner and easier to understand (thus facilitating contributions) LLVM codebase.
            One day while I was relaxing on the couch with my laptop I accidentally contributed a gorillion KLOC to LLVM! Who would've thought that the only thing inhibiting everyone to contribute to incredibly complex software projects was after all just a vague notion of "clarity" and not, I don't know, excellent education, enormous motivation, being paid to do so, concrete reasons to contribute, and so on. Thank god we have individuals such as you who make otherwise totally unnecessary and meaningless comments, so LLVM just continues growing devs ad infinitum!

            /snark

            Comment


            • #7
              Originally posted by ClosureSpin View Post
              One day while I was relaxing on the couch with my laptop I accidentally contributed a gorillion KLOC to LLVM! Who would've thought that the only thing inhibiting everyone to contribute to incredibly complex software projects was after all just a vague notion of "clarity" and not, I don't know, excellent education, enormous motivation, being paid to do so, concrete reasons to contribute, and so on. Thank god we have individuals such as you who make otherwise totally unnecessary and meaningless comments, so LLVM just continues growing devs ad infinitum!

              /snark
              Hmm? Actually, it really does encourage a lot of contributions from academic types - PHD students, etc. who are experimenting with a compiler, creating a new language, etc., but don't want to spend all the time it takes to get up to speed with GCC. Contributing to LLVM is much simpler because of it's design, and the lack of tons of old legacy code interacting in various "interesting" ways. I don't think that's a controversial point of view.

              Comment


              • #8
                Originally posted by smitty3268 View Post
                Hmm? Actually, it really does encourage a lot of contributions from academic types - PHD students, etc. who are experimenting with a compiler, creating a new language, etc., but don't want to spend all the time it takes to get up to speed with GCC. Contributing to LLVM is much simpler because of it's design, and the lack of tons of old legacy code interacting in various "interesting" ways. I don't think that's a controversial point of view.

                Yes, we had for instance a compiler course, and the home project was to make a Ruby compiler that targets the LLVM IR.

                Comment


                • #9
                  Originally posted by Linuxhippy View Post
                  It would be great to see again which company contributed how much.

                  It is always interesting to see, that e.g. RedHat is heavily involved in most important projects whereas Canonical usually doesn't give back a whole lot to the community as they usually fork stuff, maintain their own patch-sets or simply prefer to do their own thing instead of building something great together (Unity, Mir, Upstart, ...).
                  Originally posted by Drago View Post
                  Usually I will agree with you. Upstart, Mir are unnecessary/double work, but I really like Unity. It is the most lean environment for me. I tried Fedora 21 and Gnomes Hell for a while, but I got frustrated with the ugly fonts and so much vertical space wasted. I would like to ditch off Ubuntu, but the only viable alternative is OSX unfortunately.
                  Except that Upstart wasn't unnecessary or "double work". It was created for a reason, and it was designed to be used across all distros. Just because systemd came along later and happened to be better at what Upstart was trying to do doesn't make it horrible.

                  Similarly, just because you hate Ubuntu/Canonical, it doesn't make everything they've done horrible.

                  Comment


                  • #10
                    Usually I will agree with you. Upstart, Mir are unnecessary/double work, but I really like Unity. It is the most lean environment for me. I tried Fedora 21 and Gnomes Hell for a while, but I got frustrated with the ugly fonts and so much vertical space wasted. I would like to ditch off Ubuntu, but the only viable alternative is OSX unfortunately.
                    I didn't say Unity is bad in any way - it is however licensed and maintained in a way no sane distributor would ship it as anything more than an optional add-on user-interface.
                    In my opinion this is a questionable way of doing linux bussiness (taking whats already there and replacing certain parts with self-written stuff licensed in a way others won't benefit - simply to differentiate).

                    And its not just Unity, Canonical is almost always absent when it comes to upstream development.
                    I've seen quite a few OSS project stats and for almost all important projects the usual suspects are main constributors (RedHat, SuSE, Intel, .....).

                    Best regards, Clemens

                    Comment

                    Working...
                    X