[Mailman-Developers] failure upgrading to current CVS

Barry A. Warsaw bwarsaw@python.org
Mon, 5 Jun 2000 10:53:17 -0400 (EDT)


>>>>> "CVR" == Chuq Von Rospach <chuqui@plaidworks.com> writes:

    CVR> I tried to move to the current CVS tonight, and ran into a
    CVR> glitch.

Patch appended.

    CVR> Looks like b2 left something around that caused the ugprade
    CVR> to lurch.  It might make sense ot have some kind of cron job
    CVR> that deletes locks older than N days, just to be safe?

I don't think so.  The old locks shouldn't do more than take up a
little disk space, and you can just rm them easily enough.

-Barry

Index: LockFile.py
===================================================================
RCS file: /cvsroot/mailman/mailman/Mailman/LockFile.py,v
retrieving revision 1.18
retrieving revision 1.19
diff -c -r1.18 -r1.19
*** LockFile.py	2000/06/01 22:08:10	1.18
--- LockFile.py	2000/06/05 14:49:39	1.19
***************
*** 404,411 ****
          # unlink their temp file -- this doesn't wreck the locking algorithm,
          # but will leave temp file turds laying around, a minor inconvenience.
          winner = self.__read()
!         if winner:
!             os.unlink(winner)
          # Now remove the global lockfile, which actually breaks the lock.
          try:
              os.unlink(self.__lockfile)
--- 404,414 ----
          # unlink their temp file -- this doesn't wreck the locking algorithm,
          # but will leave temp file turds laying around, a minor inconvenience.
          winner = self.__read()
!         try:
!             if winner:
!                 os.unlink(winner)
!         except OSError, e:
!             if e.errno <> errno.ENOENT: raise
          # Now remove the global lockfile, which actually breaks the lock.
          try:
              os.unlink(self.__lockfile)