Julian Foad wrote on Sat, 14 Apr 2018 18:21 +0100: > Daniel Shahaf wrote on 2018-04-14: > > > It's because it uses "svnmucc -r <REV> put <...>" but STATUS has already > > > been modified between r<REV> and HEAD. > > > > Agreed. It's perfectly possible for the branch to have changed between the > > magic revision and HEAD. When editing STATUS, the base revision (argument > > to > > the -r option) should be HEAD resolved to a number. > > > > Something like this, perhaps? > > I agree the semantics you propose is/are correct. The patch looks right, > inspecting the diff; I haven't tested it.
Neither have I. Thanks for the review. I've gone ahead and committed the patch. (It's probably correct and unlikely to worsen the situation.) Cheers, Daniel
