(19:33:30) bantu: I think I've found another one

(19:33:37) nn-: with \n ?

(19:33:44) nn-: its use in sed is unlikely to work

(19:34:00) nn-: but i don't feel like fixing that right now

(19:34:35) bantu: you kind of broke it

(19:34:41) bantu: you're using exit here https://github.com/p/phpbb3/blob/develop-olympus/git-tools/hooks/commit-msg#L251

(19:35:15) bantu: but there is a git config value https://github.com/p/phpbb3/blob/develop-olympus/git-tools/hooks/commit-msg#L15 that is stupposed to make it only show warnings instead of actually abort

(19:35:23) bantu: *supposed

(19:35:52) nn-: i see

(19:36:05) ***nn- slaps whoever did _not_ document that :p

(19:36:43) nn-: except your blame is misplaced :p

(19:37:05) nn-: i did not author that line

(19:37:55) bantu: right

(19:38:02) bantu: that's chris' code

(19:38:24) bantu: Guess I mean https://github.com/p/phpbb3/commit/e55228712831a2422cc1c171a6a1160c20d25f1e

(19:38:48) nn-: hm

(19:39:09) bantu: not sure right now

(19:39:14) bantu: can check that tomorrow again

(19:39:24) nn-: are you saying it is an entirely useless change if the exit was conditional like it should have been?

(19:40:13) nn-: i will probably vastly prefer non-fatal behavior

(19:40:14) bantu: nn-: no, it's not useless

(19:40:29) nn-: since you can always amend the commit and keep seeing the warnings until you get it right

(19:40:41) bantu: I guess it just has to use quit instead of exit

(19:40:45) bantu: yes

(19:40:50) bantu: that's exactly what I thought

(19:41:07) bantu: we can just make it non-fatal all the time

(19:41:08) nn-: i remember now

(19:41:18) nn-: didn't igorw specifically requested this change? :P

(19:41:23) bantu: no idea

(19:41:24) nn-: request*

(19:41:52) nn-: http://tracker.phpbb.com/browse/PHPBB3-9824

(19:41:58) nn-: "It would be nice if the message rejection were a bit less strict and a bit more helpful in general. "

(19:42:02) nn-: that's how i interpreted that

(19:42:26) bantu: I guess that makes sense