Community technical support mailing list was retired 2010 and replaced with a professional technical support team. For assistance please contact: Pre-sales Technical support via email to firstname.lastname@example.org.
On Mon, Aug 11, 2008 at 2:41 PM, Tony Hoyle <tony.hoyle at march-hare.com>wrote: > Bryce Schober wrote: > > > > > I've had better luck with polarion's svn-importer, regarding file > re-naming > > that I've done. It does have a problem with file creation on branches, > and > > how CVSNT marks them as "deleted" on the trunk at the same time they're > > added on the branch. I've done some experiments with modifying the ,v > files > > to remove that one "deleted" mark, and it seems to be fine. So I'll > probably > > be writing a perl script to pre-process a copy of my ,v files to strip > out > > that deletion mark, but I haven't gotten around to it yet. > > > > That's not CVSNT specific - it's required by RCS that a file always > exists in some form on the trunk.. branches are held as deltas from that > point. An importer that can't handle that isn't handling the basic > RCS file format properly. > Interesting. Goodness knows I'm only fumbling around in the dark here. I had to get away from cvs2svn because it couldn't handle the file renames that I've done in CVSNT. Removing the dead state from that revision just leaves you with an extra > file on the trunk, that is probably not what you want. > As long as the branch is merged to the trunk, why would I care? It seems like most cases would turn out just fine, though clearly any ham-fisted mucking about in the RCS files by someone who has no idea what they're doing (me) could be risky. -- Bryce Schober