On Sat, May 25, 2013 at 10:22:35AM +0100, Philip Martin wrote: > Bert Huijben <b...@qqmail.nl> writes: > > The diff command should never change the wc(no wclock, and no > > db/pristine changes), so I really hope the problem that corrupts is > > part of the merge on update, or we should check for another problem. > > In my case it was merge that corrupted the working file that was being > altered, the diff problem was just a SEGV.
I believe I also ran 'svn update' using the broken client. It might not have been 'svn diff' that corrupted the files.
