perl-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Joe Orton <>
Subject Re: syncing branches with trunk, how?
Date Fri, 13 May 2005 08:37:51 GMT
On Thu, May 12, 2005 at 04:06:06PM -0400, Stas Bekman wrote:
> Joe Orton wrote:
> >You can use a semi-automated technique by recording the last-merge-point 
> >as a property; I've been using this script:
> >
> >
> >
> >for one of my projects, it seems to work quite well.  To begin, set the
> >"merge-point" property on the branch to the revision number of the last
> >merge from the trunk, and just run the script to merge again from a wc
> >of the branch.
> >
> >(it assumes you have the standard repos structure with a "trunk/"
> >alongside "branches/)
> Thanks a lot, Joe. So since I have tagged the trunk before doing the 
> branch, can I now go back in time and add a merge-point? I guess I just 
> need to checkout the tag and add a merge-point? is that right? if not how 
> do I do it?

(I just uploaded a new version of that script which should work properly
with the ASF SVN repos)

The "merge-point" property used by that script is a property set on the
root directory of the branch, used to record the *last* revision which
was merged into that branch from the trunk.

I presume you are dealing with the "clone-skip-unstable" branch, right? 
It looks like that branch has not been merged from the trunk ever, so
far, yes?

Looking at the commit logs, that branch was created from revision 165203
of the trunk - so that is the last "merge point".  To use my little
script, you do:

$ svn co \
$ cd my-working-copy
$ svn propset merge-point 165203 .
$ svn.remerge

the script does the merge from between 165203 and HEAD of the trunk into
that working copy, and resets the merge-point property to the revision
number of the HEAD, ready for next time.  So finally you just commit the
working copy.

Does that make sense?


To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message