Announcement

Collapse
No announcement yet.

The Debate Continues Over How To Transition GCC's SVN Repository To Git

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

  • The Debate Continues Over How To Transition GCC's SVN Repository To Git

    Phoronix: The Debate Continues Over How To Transition GCC's SVN Repository To Git

    Under the planned time-line for transitioning to a Git workflow for the GNU Compiler Collection that was established back at the GNU Tools cauldron conference, 16 December was to be the cut-off for deciding which Git conversion program to use for translating their massive SVN repository into Git. That puts today as the deadline in order to meet their goal of switching over to Git at the start of 2020, but it looks like it could take several more days to decide their SVN-to-Git approach...

    http://www.phoronix.com/scan.php?pag...ebate-Heads-Up

  • #2
    Michael
    sometime in January the GNU Compiler Collection could be kicking off the new decade by utilizing Git.
    Hmm, 2020 should not be the start of the new decade but 2021, for the same reason the new millenium started in 2001 and not 2000.
    Unless you're trying to say this whole thing is still another year away?

    Comment


    • #3
      Originally posted by geearf View Post
      Hmm, 2020 should not be the start of the new decade but 2021
      The silly humans insisted on an index origin of 1 (rather than 0, as it should have been). Unfortunately, the silly humans outnumber those that can count to 10, so 2020 is apparently the start of a new decade.

      Comment


      • #4
        Originally posted by phoronix View Post
        but it looks like it could take several more days to decide their SVN-to-Git approach...
        There is no excuse to delay. The deciders (who get a vote, rather than that vast numbers of others who are free to pontificate, but nothing more) are known, as was the criteria. Choose, move forward.
        Last edited by CommunityMember; 12-17-2019, 03:07 AM.

        Comment


        • #5
          For a while there, I was worried that this whole issue might finally come to a swift and quiet end. It's good to see that they felt the need to throw in one last argument over the single-use transition program.

          Comment


          • #6
            Clearly ESR thinks it's Repobludgeon or bust. And his hubris causes him to suspect anyone else's efforts that compete with his own.

            Comment


            • #7
              Originally posted by CommunityMember View Post
              The silly humans insisted on an index origin of 1 (rather than 0, as it should have been). Unfortunately, the silly humans outnumber those that can count to 10, so 2020 is apparently the start of a new decade.
              Just declare the first decade CE/AD to have been nine years long. Crazier adjustments have been made to the calendar to fix old mistakes. The switch from the Julian to Gregorian calendar happened unevenly across Europe and resulted in one to two weeks being skipped. (September 3rd through 13th, 1752 don't exist in Britain.)

              Comment


              • #8
                Keeping all of the history of GCC is extremely important, otherwise this would be impossible: find out who introduced a certain bug 30 years ago, try to ask the person in question, fail because he died 20 years ago.

                Comment


                • #9
                  Originally posted by cl333r View Post
                  Keeping all of the history of GCC is extremely important, otherwise this would be impossible: find out who introduced a certain bug 30 years ago, try to ask the person in question, fail because he died 20 years ago.
                  It's not like the svn repositories won't be available to find out information like that if necessary. It's just that it would be a lot more convenient if the history was intact in git.

                  Comment


                  • #10
                    Originally posted by cl333r View Post
                    Keeping all of the history of GCC is extremely important, otherwise this would be impossible: find out who introduced a certain bug 30 years ago, try to ask the person in question, fail because he died 20 years ago.
                    While none of the conversion approaches can convert all the scm history perfectly(*), the general consensus by people who have done this in recent past is all of the conversion solutions are good enough for all but perhaps the most exceptional spelunking trip down the longest and most dangerous historical caves and passageways.

                    (*) there are some constructs in the existing svn repo (which itself was converted from cvs imperfectly) that do not convert perfectly to git because the scms simply track things differently.

                    Comment

                    Working...
                    X