[p4] automated resolve issue

Rick Macdonald rickmacd at shaw.ca
Thu Oct 8 09:30:09 PDT 2009


What flags do you have on the integration and resolve steps? You would 
lose Branch A modifications if the second merge had "resolve -at" but 
there are probably other scripting errors that you could make to cause this.

In a case like yours, after the merge of Branch A to the trunk, we would 
first integrate/merge from trunk to Branch C, and compile and retest as 
necessary. Then we would integrate from Branch C back to trunk. This is 
the "Merge Down - Copy Up" that Laura writes about.

My scripts use resolve -as only. Not -am and certainly not -at or -ay. 
Well, I use -at in a script that performs Laura's complicated "Copy Up" 
recipe, but it is a very specific sequence of steps that does the right 
thing.

What does your script do if there are resolve conflicts?

     -- A
   /
=========================== trunk
     \
      ---------- B
           \
            -------- C
Rick

Nicolas Brault (2K Czech) wrote:
> Hi,
>
>  
>
> I would like to know if me and my company are the only ones having
> trouble during a resolve process.
>
>  
>
> I explain:
>
> Imagine a file which is available on 3 branches (Trunk, BranchA,
> BranchC).
>
> (BranchC has been created from BranchB which has been created from
> Trunk. 
>
>  BranchA has been created from Trunk)
>
>  
>
> - The file has been modified on branchA and BranchC.
>
> - The file is merge from branchA to Trunk (automated resolve, no
> conflict). At this point there is no problem.
>
> - the file is from BranchC to trunk (automated resolve, no conflict). 
>
>  
>
> The problem is that after the second merge, we lost some modifications
> that came from the merge BranchA to trunk. And I don't understand why...
> but it's really dangerous because we can lost lot of code
>
>  
>
> Thanks,
>
> Nico
>
>  
>
>  
>
>  
>
>  
>
> _______________________________________________
> perforce-user mailing list  -  perforce-user at perforce.com
> http://maillist.perforce.com/mailman/listinfo/perforce-user
>
>   



More information about the perforce-user mailing list