Announcement

Collapse
No announcement yet.

GCC's Conversion To Git Is Still Turning Out To Be A Massive Headache

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

  • #11
    Originally posted by FireBurn View Post

    I don't think that's comparing like for like, KDE was mostly small repositories in comparison to GCC, the feature work was also much bigger branch merges than would have been seen in GCC
    Actually it was huge repositories, but they were split up into smaller ones during the transition.

    Comment


    • #12
      Originally posted by anarki2 View Post
      It's rather surprising that to this day, noone has come up with a reliable and scalable conversion tool between 2 such mainstream VCSes.
      It's rather suprising that to this day, you have never heard of svn2git

      What's suprising to me is that there are still some projects (not GCC) using CVS :/
      Last edited by Vistaus; 30 July 2018, 12:39 PM.

      Comment


      • #13
        If even the gurus have problems that is a real issue!
        Last edited by Danielsan; 30 July 2018, 01:48 PM.

        Comment


        • #14
          Originally posted by Danielsan View Post
          If even the gurus have problem that is a real issue!
          This. I was really shocked to read the last thread on this and see all the armchair experts telling the gurus how it's done. Found that quite amusing.

          Having said that, I enjoyed reading the thread because as a non-guru (hehe), you learn a lot by people's suggestions. So I think tone matters, but still a good discussion.

          Comment


          • #15
            Tek Syndicate on how/why they built "The Great Beast" for ESR: https://www.youtube.com/watch?v=875NbdL39A0

            This was shared in the last thread but I wanted to bump it back up. Really fun watch.

            Comment


            • #16
              Originally posted by torsionbar28 View Post
              I still don't understand why this is a performance problem. Is he trying to do some kind of live conversion without taking the repository offline? I.e. still accepting new commits while the conversion is in process?
              Pretty much... my understanding is they require as little downtime as possible. And it doesn't need to be done now, just needs to be done quickly when it does happen.

              Comment


              • #17
                Seriously? It just....

                Comment


                • #18
                  Originally posted by perpetually high View Post

                  This. I was really shocked to read the last thread on this and see all the armchair experts telling the gurus how it's done. Found that quite amusing.

                  Having said that, I enjoyed reading the thread because as a non-guru (hehe), you learn a lot by people's suggestions. So I think tone matters, but still a good discussion.
                  Really? Because all those "armchair experts" were saying that throwing more memory at the issue was the wrong approach, because his work shouldn't need 64GB of history loaded simultaneously. Myself included. And ESR of course found out that throwing GB at the problem didn't help after 3 weeks of wasting time thinking about that.

                  Originally posted by ESR
                  My tools need to get a lot faster, like more than an order of magnitude faster, before digging out of the bad situation the conversion is now in will be practical. Hardware improvements won't do that.
                  Now he's looking at porting his code to a new language, but unless he changes his core approach, it's still going to be a turd polished by high-end hardware. But who knows, maybe he'll get it to work by yak-shaving random things instead of tackling his design issues head-on.

                  BTW, half the reason we're so brutal to ESR is that he's known for being absolutely tone-deaf to anyone who isn't sucking up to him. If people are going to dish it out, they'd better be able to take it. And ESR, who is famous more for getting the right opportunities at the right time than for impressive skills and abilities, is a pretty easy target. He's extremely egotistical and considers himself a field-transcending expert, but it's pretty clear that he gets in over his head a lot and relies on his reputation to lend weight to his words and save him from embarrassing questions. Many developers are like fine wines, which just get better with age. ESR isn't a fine wine. He's a decades-old gallon jug of milk. He's that guy at your job the guy who spends more time talking up the difficulty of his tasks and making them sound like sorcery than he does actually solving them. He's the guy who drinks his own kool-aid.

                  It's fine when he just wants to be a figurehead representing history, but when he gets up and start spouting nonsense, it's best to put him in the corner in timeout.

                  Comment


                  • #19
                    Originally posted by Vistaus View Post

                    What's suprising to me is that there are still some projects (not GCC) using CVS :/
                    The GCC wwwdocs repo is still in CVS. Granted, that is quite small and insignificant compared to the main repo.

                    Comment


                    • #20
                      Originally posted by Terrablit View Post

                      Really? Because all those "armchair experts" were saying that throwing more memory at the issue was the wrong approach, because his work shouldn't need 64GB of history loaded simultaneously. Myself included. And ESR of course found out that throwing GB at the problem didn't help after 3 weeks of wasting time thinking about that.



                      Now he's looking at porting his code to a new language, but unless he changes his core approach, it's still going to be a turd polished by high-end hardware. But who knows, maybe he'll get it to work by yak-shaving random things instead of tackling his design issues head-on.

                      BTW, half the reason we're so brutal to ESR is that he's known for being absolutely tone-deaf to anyone who isn't sucking up to him. If people are going to dish it out, they'd better be able to take it. And ESR, who is famous more for getting the right opportunities at the right time than for impressive skills and abilities, is a pretty easy target. He's extremely egotistical and considers himself a field-transcending expert, but it's pretty clear that he gets in over his head a lot and relies on his reputation to lend weight to his words and save him from embarrassing questions. Many developers are like fine wines, which just get better with age. ESR isn't a fine wine. He's a decades-old gallon jug of milk. He's that guy at your job the guy who spends more time talking up the difficulty of his tasks and making them sound like sorcery than he does actually solving them. He's the guy who drinks his own kool-aid.

                      It's fine when he just wants to be a figurehead representing history, but when he gets up and start spouting nonsense, it's best to put him in the corner in timeout.
                      I understand but don't you think he's confided and spoken to domain experts on this before typing out each one of his emails? Come on, don't be naive. I'm sure he's not looking to be difficult here, he does these conversions often; this isn't new to him. And if it's so easy, you do it.

                      Comment

                      Working...
                      X