[petsc-dev] v3.5 release schedule

Barry Smith bsmith at mcs.anl.gov
Sun Oct 6 17:55:30 CDT 2013


  What does *completely* merged here in this context? Would a branch that was merged, but then modified and that modified version not merged be "incompletely" merged?  How do we make sure those modifications get into master if the person who "owns" the branch is not aware that the modification hasn't gotten into master (or next).

   In other words, what is needed in the bitbucket (or some other) gui to track this stuff?


   Barry

On Oct 6, 2013, at 5:25 PM, Jed Brown <jedbrown at mcs.anl.gov> wrote:

> Matthew Knepley <knepley at gmail.com> writes:
>> Something is wrong with this determination. I am pretty sure all those
>> branches from me have
>> been merged to master, and you are missing my important ones that need to
>> go in.
> 
> Branches must be *completely* merged to 'next' in order to be candidates
> and must be completely merged to 'master' in order to be considered merged.
> 
> $ git log --oneline  master..knepley/fix-dm-clone
> 0663011 DMPlex: clones are already setup
> $ git log master..knepley/fix-plex-subdm 
> 2a4c499 DMPlex: DMGetSubDM_Plex() had a memory leak - It was also doing extra work
> $ git log --oneline  master..bb/knepley/submesh-declarations 
> 89a07f4 DMInterpolation: Fixed declarations
> 120d0d3 DMPlex: Added declaration




More information about the petsc-dev mailing list