Stephan Witt wrote: > But in general owner_->fileName() should be used instead of file_.
yes > Another question: when the working copy has no diff to repository > the check in operation isn't sensible. But it's expensive to test > this. So I'd like to test it before prompting for the log message > and then doing the noop check in. Does this sound sensible? this is already bug #chrrm (bugzilla is down). i was about to fix it few months back but there was some catch (hard to remember, maybe because checkin can release locks only or something like that), so i postponed it ;) i agree this should be fixed but dont have time recheck that all other rcs/svn usecases are not affected, so this testing would be up to you. pavel
