On Fri, Nov 21, 2008 at 10:33 AM, Julian Foad
<julianfoad_at_btopenworld.com> wrote:
> On Fri, 2008-11-21 at 07:10 -0800, pburba_at_tigris.org wrote:
>> Author: pburba
>
>> Follow-up to r34091. Make reintegrate even more tolerant of subtree
>> mergeinfo on the reintegrate source.
>
>> This change also improves the error message that reintegrate produces
>> when the source hasn't had the target uniformly merged to it. [...] it
>> skips right to showing what needs to be merged for the reintegrate to work.
>
> Paul, both of those improvements sound great! I look forward to seeing
> it in action.
Julian,
It's worth noting that since WC-to-WC copy/moves no longer create
explicit mergeinfo on the destination (r34184) and the cleanup of
subtree mergeinfo in r34275, that the fixes in r34036 are not likely
to have any noticeable effect (beneficial or otherwise) in the short
term, as least as concerns our own repository.
But as merge_tests.py 126 demonstrates, it is still easy to end up in
a situation where prior to r34184 we should be able to reintegrate,
but couldn't.
Also, there is still one case we don't handle:
If a merge adds a subtree to a branch with explicit mergeinfo already
on it. Currently that subtree doesn't get additional explicit
mergeinfo set on it describing the merge that was responsible for its
addition. This causes any subsequent reintegration attempt to fail
because it looks like the branch isn't completely synched up with the
trunk. IMO this isn't a problem with reintegrate, but rather with how
merge sets mergeinfo on added paths with mergeinfo. I saw this
problem a lot when testing r34036 with our own pre-r34275 repository.
The problem is closely related to the one fixed in r32975 (explicit
mergeinfo added to existing paths by merge) and the fix is probably
similar.
Paul
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe_at_subversion.tigris.org
For additional commands, e-mail: dev-help_at_subversion.tigris.org
Received on 2008-11-21 16:58:25 CET