Message165552
Thanks a lot for the feedback. The thinking was to use a stand-alone (even testable) construct with dependencies made explicit. For example, it wasn't obvious that the current iterator depended on forever, or whether the args_tuple parameter was a necessary part of it. But I agree it should be simplified.
Here is a simpler patch, keeping iterator semantics. My preference is to avoid adding lock semantics and additional complexity directly to main(), even though doing so would permit a smaller patch. |
|
Date |
User |
Action |
Args |
2012-07-15 23:02:29 | chris.jerdonek | set | recipients:
+ chris.jerdonek, rhettinger, amaury.forgeotdarc, pitrou, flox |
2012-07-15 23:02:28 | chris.jerdonek | set | messageid: <1342393348.96.0.326392259048.issue15320@psf.upfronthosting.co.za> |
2012-07-15 23:02:28 | chris.jerdonek | link | issue15320 messages |
2012-07-15 23:02:28 | chris.jerdonek | create | |
|