Announcement

Collapse
No announcement yet.

Apache Kills Off Its C++ Standard Library

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

  • Apache Kills Off Its C++ Standard Library

    Phoronix: Apache Kills Off Its C++ Standard Library

    The Apache Software Foundation has decided to end its Apache C++ Standard Library (stdcxx) project...

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

  • #2
    I'm all for competition..but why was this needed? The project i mean. GCC has a C++ implementation, LLVM has a C++ Implementation, there's probably a c++ implementation out there for embedded. It doesnt seem like it filled any niche except for those who hated both GCC AND LLVM.

    Comment


    • #3
      Originally posted by Ericg View Post
      I'm all for competition..but why was this needed? The project i mean. GCC has a C++ implementation, LLVM has a C++ Implementation, there's probably a c++ implementation out there for embedded. It doesnt seem like it filled any niche except for those who hated both GCC AND LLVM.
      Yep, that sounds like wasted efforts.

      Comment


      • #4
        Open-Source Projects Rely on Popularity

        I hate it when celebrity programmers will act like their own workflow equals one thousand people's workflows. If a project has very few supporters, a lot of competition, and nothing really innovative coming out of it then the project doesn't have a use for a big organization like Apache. I'm sure Bergström will try and drum up support for something that barely anyone cared about.

        Comment


        • #5
          Originally posted by caryhartline View Post
          I hate it when celebrity programmers will act like their own workflow equals one thousand people's workflows. If a project has very few supporters, a lot of competition, and nothing really innovative coming out of it then the project doesn't have a use for a big organization like Apache. I'm sure Bergström will try and drum up support for something that barely anyone cared about.
          Well, as long as he works on it, he's free to scratch his own itch. But he shouldn't expect others to support it.

          Comment


          • #6
            Proper c++11 support will require a bit of work. It's not exactly trivial to retool a library for it. And c++11 is extremely popular. There are TONS of follks out there very upset with MS about their poor c++11 support, more than a few have dumped msvc for mingw.

            Comment


            • #7
              Originally posted by bnolsen View Post
              more than a few have dumped msvc for mingw.
              I'm not seeing that happen, especially since MS has declared that full c++11 and then 14 support are high priority for them.

              Comment


              • #8
                Originally posted by Ericg View Post
                I'm all for competition..but why was this needed? The project i mean. GCC has a C++ implementation, LLVM has a C++ Implementation, there's probably a c++ implementation out there for embedded. It doesnt seem like it filled any niche except for those who hated both GCC AND LLVM.
                LLVM probably didn't exist when this was written. Even when Apache took it over as an open-source project in 2005, LLVM wasn't the big deal it is now, not really catching on until after Apple adopted it. And 2005 was simply the date the code was released - presumably the project was created long before...

                Comment


                • #9
                  Apache isn't stupid: The future is libc++/libc++abi.

                  Comment


                  • #10
                    Originally posted by bnolsen View Post
                    Proper c++11 support will require a bit of work. It's not exactly trivial to retool a library for it. And c++11 is extremely popular. There are TONS of follks out there very upset with MS about their poor c++11 support, more than a few have dumped msvc for mingw.
                    There was no releases since 2008, I doubt there is any C++11 support on the killed library.
                    Also, want to know who are upset with MS? The ones who didn't get support for C99. Now that's shitty support.

                    Originally posted by Delgarde View Post
                    LLVM probably didn't exist when this was written. Even when Apache took it over as an open-source project in 2005, LLVM wasn't the big deal it is now, not really catching on until after Apple adopted it. And 2005 was simply the date the code was released - presumably the project was created long before...
                    Correct, but that doesn't change the fact that there is no need for it now, so maintaining it would be a wasted effort.

                    Originally posted by Marc Driftmeyer View Post
                    Apache isn't stupid: The future is libc++/libc++abi.
                    I don't really understand your post, but if I'm correct, you misread the article: Apache is killing their implementation, and IMO they're right doing so, because even when I agree that the future is C/C++ based the same way it was for years, we already have two good implementations of it, there's no need for another one. We probably have even more from the several smaller compilers.
                    Last edited by mrugiero; 07-18-2013, 09:08 PM.

                    Comment


                    • #11
                      Originally posted by mrugiero View Post
                      Correct, but that doesn't change the fact that there is no need for it now, so maintaining it would be a wasted effort.
                      Agreed, but I was addressing the question "why *was* this needed?". And citing the existence of LLVM as a reason kind of misses the point that this isn't a new project written to compete with LLVM - it's an old one that predates it.

                      Certainly, it has no value today, which, despite the hostile response the announcement received on the mailing list, is no doubt why the project hasn't had any activity for years.

                      Comment


                      • #12
                        If it is open source maybe PathScale should pick it up.

                        Frankly this is one example of a company saying why aren't you doing this for me! People complain about Apple and open source but at least they are piloting several well regarded and heavily used pieces of software. So instead of whining like a stuck pig maybe PathScale should sponsor the continued development of this library.

                        Comment


                        • #13
                          Originally posted by Ericg View Post
                          I'm all for competition..but why was this needed? The project i mean. GCC has a C++ implementation, LLVM has a C++ Implementation, there's probably a c++ implementation out there for embedded. It doesnt seem like it filled any niche except for those who hated both GCC AND LLVM.
                          I wouldn't call it a wasted effort like some have but this package was last updated in 2008 so it looks like abandon ware to me. The people complaining are the worst example of open source users I can think of, all give me but no offers of help or support.

                          I look at it this way, if the software is valuable somebody will pick up maintenance. The software doesn't really need to fill a niche as much as it needs to used and supported.

                          Comment


                          • #14
                            Originally posted by smitty3268 View Post
                            I'm not seeing that happen, especially since MS has declared that full c++11 and then 14 support are high priority for them.
                            What they say and what they do appear to be two separate things.

                            Comment


                            • #15
                              Originally posted by wizard69 View Post
                              What they say and what they do appear to be two separate things.
                              There was a post on the forums some time ago which had a link to a page comparing compiler support for C++11, namely GCC, Clang, MSVC, Intel's compiler and a couple other proprietary compilers.

                              And of all the proprietary compilers in the list, Microsoft's compiler offered the best support for C++11, implementing features that Intel's own compiler still lacked support for. Explain how that equates to Microsoft not taking C++11 seriously.

                              Comment

                              Working...
                              X