I am the ‘host’ in this interview. I think the story of building network transparent CVS and then subversion is super interesting. The thing that got me interested in this topic was a really old question I had from when I first started using git, “why was merging so much better than in subversion”. At the time, people said it was because git was distributed but that didn’t make any sense to me. Jim answers that question towards the end of the interview, but the story Jim tells about building source control tools was even more interesting in the end.
The mistake that we made in subversion, is that we thought merges should actually mean something textually, right? That actually there should be a way to say, here is the reconciliation of these changes with that change, and I think we really just over-thought it a lot.
That mirrors well what I remember Linus writing when he came up with git. There was a lot of back and forth about merge conflicts with Mercurial, Monotone, Darcs, and others. Linus emphasized that a version control system should not care because you cannot get the general case correct anyways.
I am the ‘host’ in this interview. I think the story of building network transparent CVS and then subversion is super interesting. The thing that got me interested in this topic was a really old question I had from when I first started using git, “why was merging so much better than in subversion”. At the time, people said it was because git was distributed but that didn’t make any sense to me. Jim answers that question towards the end of the interview, but the story Jim tells about building source control tools was even more interesting in the end.
That mirrors well what I remember Linus writing when he came up with git. There was a lot of back and forth about merge conflicts with Mercurial, Monotone, Darcs, and others. Linus emphasized that a version control system should not care because you cannot get the general case correct anyways.