Mark Phippard wrote:
> On Wed, Mar 4, 2009 at 5:46 PM, Stefan Sperling <stsp_at_elego.de> wrote:
>
>> On Wed, Mar 04, 2009 at 05:27:15PM -0500, Ruslan Sivak wrote:
>>
>>
>>> Now I would like to fully catch up to the trunk. I follow the
>>> instructions here: http://blog.red-bean.com/sussman/?p=92
>>>
>> These instructions only merge changes one way before reintegrating.
>>
>> The problem you are seeing is not a bug.
>> It's a problem enforced by Subversion's design, and is explained here:
>> http://blogs.open.collab.net/svn/2008/07/subversion-merg.html
>>
>
> It is not so black and white to just say it is not supported. It is
> actually merge, without --reintegrate, that you cannot really use once
> you get into bidirectional merges. I think there might be enough of a
> "recipe" here to see if we can recreate it in a test and possibly
> enhance the detection process.
>
> --reintegrate is just trying to prevent problems. It detects a
> difference that it does not think should be there. In this case, that
> detection was a false positive. Maybe we an enhance it. The example
> given shows how much better is has already gotten in that it gave a
> very intelligible error that allowed Ruslan to figure out why it was
> complaining and pinpoint the reason.
>
> I'd use the --record-only option to record the revision that it thinks
> is missing.
>
>
Mark,
Thank you for your response.
I tried using --record-only (after deleting the mergeinfo property on
this folder). It added /trunk/somefolder:433-559 to the mergeinfo.
When trying to reintegrate after this, it still gave me an error about
revision 432.
I edited the property and changed it to /trunk/somefolder:432-559, which
seems to have made reintegrate happy.
Russ
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=1065&dsMessageId=1269242
To unsubscribe from this discussion, e-mail: [users-unsubscribe_at_subversion.tigris.org].
Received on 2009-03-05 00:56:50 CET