lp:~cmiller/desktopcouch/getport-at-call-time

Branch merges

Propose for merging

No branches dependent on this one.

Merged into lp:desktopcouch

Elliot Murphy (community): Approve on 2009-08-13

Eric Casteleijn (community): Approve on 2009-08-13

Diff: None lines api

Related source package recipes

No recipes using this branch. (?)

Create packaging recipe

Related snap packages

Related bugs

Bug #405612: Fix failing tests in desktopcouch

High

Fix Released

Link a bug report

Related blueprints

44. By Chad Miller on 2009-08-13

Replace twisted TestCase with testtiils TestCase.

43. By Chad Miller on 2009-08-13

Trade one buggy process spawner for another. os.spawn* suffers, and
so does subprocess.Popen , but subprocess is newer and has a better
chance of being fixed one day soon.

Chad filed a patch for it in Python's bug tracker.
  http://bugs.python.org/issue1068268#msg91502

42. By Chad Miller on 2009-08-13

Handle case where pid-file of couchdb doesn't have a valid integer in it.

41. By Chad Miller on 2009-08-12

Make a way for find_pid to only test, and never start a couchdb
instance.

Also, abstract the names of the heavily Linux-kernel-specific
functions so it's easier to add other OS support.

Make find_port return a None, instead of the string 'None' on
errors.

40. By Chad Miller on 2009-08-12

To test the automatic starting of couchdb, stop the db one quarter of
the time a the end of a test.

39. By Chad Miller on 2009-08-12

Add method to stop a running local couchdb.

38. By Chad Miller on 2009-08-12

Don't wait unnecessarily for couchdb to start. Also plug a hole that
allowed more than one couch to start.

37. By Chad Miller on 2009-08-11

Use os.spawn* instead of subproces.call, because subprocess is broken.

Simplify and make pid-finding function not wait unnecessarily.

36. By Chad Miller on 2009-08-11

Fixed a NameError that thisfred noticed.

35. By Chad Miller on 2009-08-11

Clean up func that tests for running couchdb, given a kind of pid value.

Remove IOError race and LBYL.