[be] [bug #25785] Better error handling for git subprocesses

[Dan Dennison]

Follow-up Comment #3, bug #25785 (project bibledit):

The remote repository should probably then auto-offline itself after so many
failures. Not having looked at that code much, not sure if it does. In any
event, it sounds like a different task.

As for "failures that were such that they caused so many retries that the
system started to feel slow", I think you're referring to the non-zero return
codes from git. It seems that git returns '1' in many situations where the
command actually succeeds. We need to distinguish git's many success modes
from its outright failures. Outright failure needs to be requeued.

    _______________________________________________________

Reply to this item at:

  <http://savannah.nongnu.org/bugs/?25785>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.nongnu.org/