May 21, 2011
6:02 a.m.
It would be better, IMO, if there was a single developer who would migrate changes to svn, or to have some semi-automatic procedure for that.
Yep. It's rather a pain to determine what those changes are though if the best you can do is a recursive diff on the two trees.
You can do better than that. If the hg and svn changes are in the same order, and have the same commit messages (as they should), you can look at the log of the hg branch to find out what the youngest change is that has been copied; all changes that are more recent then still have to be applied.
Regards, Martin