[01:15] <wgrant> cjwatson: If you're still around, could you please look at https://code.launchpad.net/~wgrant/launchpad/bug-1500576/+merge/272676?
[08:32] <cjwatson> wgrant: Right, I'd wondered if that was going to require ILaunchBag in the model, but when it was reported it was a bit late for me to start in on it.  Thanks.  r=me
[08:34] <wgrant> cjwatson: Thanks.
[08:34] <wgrant> Yeah, no choice here, unfortunately.
[08:36] <StevenK> wgrant: You mean you haven't abolished the bag? :-(
[08:36] <wgrant> StevenK: I removed two callsites today.
[08:36] <wgrant> But added one...
[08:36] <StevenK> Net -1!
[19:41] <blr> morning
[20:37] <cjwatson> blr: morning :)
[20:38] <blr> cjwatson: morning!
[20:39] <blr> just found a bug in my token invalidation that did have a test, but the test was too naive..
[20:40] <blr> tests can certainly engender a false sense of security heh
[20:50] <cjwatson> Earlier today I had a bug where I'd left out "self" in a parameter list - almost all the tests passed anyway due to luck, but I was saved by one that now had a *very* confusing traceback ...
[21:44] <blr> cjwatson: wgrant: in the token purge script called by cron, is there any harm in reporting the number of tokens purged to stdout? (useful for a manual test)
[21:50] <wgrant> blr: You could also log it at INFO or something and have that not shown by default.
[21:51] <cjwatson> Yep, I'd do that
[21:51] <cjwatson> Lots of our cron scripts have that kind of status output
[21:52] <blr> right, thanks