[p4] Problem with integrates

[Stephen Vance]

Matt --

At first glance this looks strange. I'd be interested in knowing how the
workspace in which you are doing the integration is mapped. I would
suspect that this might have to do with that, but the fact that you can
get it to work sometimes and not others would argue against that.

Also try

p4 integ -b branchspec -s @12345, at 12345

The -s flag specifically says that the revision is the source instead of
the "path from the target, revision from the source" interpretation
without it.

Also, is there a possibility that your "changing the names to protect
the innocent" obscures the communication of the problem? Or is your
translation of the details trivial?

All that failing, I would bounce this one off of support.

Steve

Matt Roberts wrote:
> We have 2 branches setup as follows:
>
> //branches/old/...
> //branches/new/...
>
> And a branchspec defining the mapping from //branches/old/... to
> //branches/new/...
>
> There is a change in //branches/old/... that affects only one file.  Now
> we want to merge that change over to //branches/new/...
>
> Normally, we would do the following:
> p4 integrate -b branchspec @12345, at 12345
>
> However, in this particular situation, we get the error that there are
> no such files:
> @12345, at 12345 - no such file(s)
>
> But, I know the file in this changelist exists in both branches and is
> mapped by the branchspec.  So I try this command:
> p4 integrate -b branchspec //branches/old/path/to/file.txt at 12345, at 12345
>
> That works!  So, I start pulling back the file spec in that command
> until I find that this will not work:
> p4 integrate -b branchspec //branches/old/path/... at 12345, at 12345
>
> but this will:
> p4 integrate -b branchspec //branches/old/path/t*/... at 12345, at 12345
>
> Now, the branchspec is rather large since there was a lot of code re-org
> in the move to the new branch. There are also exceptions in the branch
> spec relating to some stuff in old that we didn't want in new.  I have
> double and triple checked that none of the exceptions would affect the
> file in question.  In fact, I would think that the integrate would not
> work when specifying the full path if there was an exception that
> affected this file.
>
> Has anyone seen this behavior before?  What am I missing?
>
> Thanks,
> Matt
>
> _______________________________________________
> perforce-user mailing list  -  perforce-user at perforce.com
> http://maillist.perforce.com/mailman/listinfo/perforce-user
>
>   

-- 
Stephen Vance
www.vance.com