Thanks, Will understand and vote for r34547 and this group tomorrow.
I observed 'CHANGES' file not getting svn:mergeinfo set when I merge r30257 using 1.5.4 client, but merging r33024 and r33026 sets the mergeinfo, will dig that too tomorrow.
From: Paul Burba [mailto:ptburba_at_gmail.com]
Sent: Tue 12/9/2008 9:56 PM
To: Kamesh Jayachandran
Subject: Re: [Don't try to filter natural history from added mergeinfo during reverse merges] ?Complete?
> Hi Paul,
> I was trying to understand [Don't try to filter natural history from
> added mergeinfo during reverse merges] block of revisions to vote for 1.5.x.
> I somehow could not see it seemed to fix the reversal of r32625 issue
> with the revision numbers recommended in STATUS.
> Will explore on it bit, in case you have thoughts let me know.
> I tried the following
> $/home/kamesh/1.5.x-build/bin/svn co -r32625
> $cd subversion-svnpatch-diff-r32625
> $/home/kamesh/1.5.x-build/bin/svn merge -c-32625
> With and without patch above merge does not delete 'svn:mergeinfo' on
> 'build.conf' which got added via r32625.
> With trunk binary this property gets deleted.
> With regards
> Kamesh Jayachandran
Sorry about that Kamesh, the dependencies between all these
merge-related backports can be a bit much at times...
For the specific example in
to be fully solved requires that the fix for issue #3323 'Mergeinfo
deleted by a merge should disappear' be present. I have nominated
that fix (r34547) and made notes about the relationship among the
r33026, r34547, and r34432* groups.
Thanks for spotting this,
* Which is unrelated to these other two, except that r34547 conflicts
if it hasn't been applied.
Received on 2008-12-09 18:06:23 CET