On Fri, Apr 18, 2008 at 3:42 PM, Paul Burba <ptburba_at_gmail.com> wrote:
>
> On Fri, Apr 18, 2008 at 3:19 PM, Hyrum K. Wright
> <hyrum_wright_at_mail.utexas.edu> wrote:
> > Paul Burba wrote:
> >
> > > Fixed in r30668. This change (with the fix) is again nominated for
> > backport.
> >
> > I'd like to roll RC 4 this weekend, probably tomorrow evening. I think the
> > r30644 group is the only one which lacks review, and it should go into the
> > RC.
>
> I'm still working on issue 3067, making some progress...though I just
> got back from my lunch run and found a whole slew of test failures :-(
> I'll keep at it.
>
> If I can tackle 3174 'Merge algorithm chokes on subtrees needing
> special attention' I'll do that too, as it looks very closely related
> to 3067.
>
> It would be helpful if some folks could look at 3067 and 3174 and
> understand the problem. Not necessarily understand it enough to
> propose a solution, but to offer an opinion as to whether these are
> release blockers or not (I kinda think they are but maybe I'm overly
> pessimistic).
I am testing the --record-only changes proposed for backport. They
merge cleanly to 1.5.x and the tests all pass. However, I then used
the binaries with that change to do a manual test on the 1.5.x branch:
svn merge --record-only -c30112 -c30286 https://svn.collab.net/repos/svn/trunk
I was expecting this command the elide away the explicit mergeinfo on
CHANGES. It did not, it just updated the root. I do not think this
is a bug, because a merge of the same two revisions without
--record-only also does not elide away the explicit mergeinfo. I
would like to run a command like this and see it work. What are the
other revisions I need to include for this to work correctly?
-
Thanks
Mark Phippard
http://markphip.blogspot.com/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe_at_subversion.tigris.org
For additional commands, e-mail: dev-help_at_subversion.tigris.org
Received on 2008-04-19 22:33:14 CEST