> In this case, I think that a single patch would still fit this
> definition.

Ok, Issue #2851 addresses this patch, which can be found here:
http://subversion.tigris.org/nonav/issues/showattachment.cgi/681/svnmerge_ignoreInitializedRevisions-2.patch

I've written the test into the existing test that exposed the changed
behaviour since all the setup is already in place.

The only area of concern I have with this patch at the moment is having 
to force property checking for all revisions being considered, which was 
previously only done when --bidirectional was specified.  This means 
that svnmerge is making a lot more calls to svn and with Raman's recent 
patch to resolve transitive merge property conflicts the extra work 
seems a little pointless.

-- 
Michael Willmott
I.T. Support
LeadGen Marketing Solutions
Tel: 0870 160 7871

Helping our clients create the BEST call centres
through technology and smart marketing ideas

_______________________________________________
Svnmerge mailing list
[email protected]
http://www.orcaware.com/mailman/listinfo/svnmerge

Reply via email to