On Fri, Apr 27, 2012 at 10:55, Bert Huijben <b...@vmoo.com> wrote: >... > This doesn't document the behavior of replacements within copies, nor > with multi layer operations. And I would avoid the wc1 'revert base' as > that can relate to op depth 0 or higher.
I think it does. It says local-add (whether that is after deleting a copy/move-here node or not) will return INVALID_REVNUM/NULL. For any copy/move-here (no matter how deep), it returns the original of that node. For simple changes or a local-delete it returns BASE. For deletions within a copy/move-here, it returns the origin again. > The behavior talks about where it will be committed to and about where > things are copied from for op-depth > 0, so personally I would have > left the 'badly defined' comment as it sounds like the old URL field in > svn_wc_entry Hmm? I see it talking about "what is being used as the pristine for this node?" (which might be a better name) That's why I used it in client/ra.c: get the pristine info for a given path. Cheers, -g
