[petsc-dev] Pushing non-working code

[Jed Brown]

It seems like every time I pull PETSc, I get more build noise from unused
or uninitialized variables. I have a suggestion: if it's not clean yet,
just don't push it. If you want a backup or to communicate the messy code
with someone else, push it to bitbucket.org/YOURNAME/petsc-dev. Pushing to
petsc-dev should mean that it's either ready for users or that it's at
least ready for other people to start working with.

Pushing obviously broken code to petsc-dev discourages people from reading
the commit logs because it's exhausting and it's hard to argue with "yeah,
I know none of that works" (except to fix the process). Lack of review
lowers the quality of code in PETSc.


It's Matt's code today, but we've all been guilty on occasion.

/home/jed/petsc/src/dm/impls/plex/plex.c:10062:19: warning: unused variable
'numSubPoints' [-Wunused-variable]
  PetscInt       *numSubPoints, *coneNew;
                  ^
/home/jed/petsc/src/dm/impls/plex/plex.c[ 68%] :10209:7: warning: variable
'coneNew' is uninitialized when used here [-Wuninitialized]
      coneNew[c] = firstSubVertex + subvertex;
      ^~~~~~~
/home/jed/petsc/src/dm/impls/plex/plex.c:10062:41: note: initialize the
variable 'coneNew' to silence this warning
  PetscInt       *numSubPoints, *coneNew;
                                        ^
                                         = NULL
/home/jed/petsc/src/dm/impls/plex/plex.c:10280:38: warning: variable
'subpointMap' is uninitialized when used here [-Wuninitialized]
  ierr = DMPlexSetSubpointMap(subdm, subpointMap);CHKERRQ(ierr);
                                     ^~~~~~~~~~~
/home/jed/petsc/src/dm/impls/plex/plex.c:10060:76: note: initialize the
variable 'subpointMap' to silence this warning
  IS              subvertexIS, subedgeIS, subfaceIS, subcellIS, subpointMap;
                                                                           ^

= NULL
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20130202/897e0029/attachment.html>