All times shown according to UTC.

Time Nick Message
00:33 melinath z3p: perhaps I'm missing something, but wouldn't workers be overriding the settings object to potentially different values at the same time?
00:50 z3p melinath: each worker is only doing task at a time
00:51 melinath: so while it's doing its task, it's safe
00:51 melinath: at least, that's how celery works as I understand it
00:51 melinath: (for the processes backend)
00:51 melinath: it would be an issue for the eventlet backend, yes
00:52 melinath: but the eventlet backend has other issues for our uses anyways
00:52 melinath: we could conceviably move some stuff (network IO) to separate tasks and run another celery
00:53 melinath: but the tasks aren't broken up like that ATM
01:46 melinath z3p: ok, if the workers don't share global state, then yeah, there's no issue.
01:47 z3p: We can always refactor it more later.
01:47 z3p: If we feel like it.
02:01 Odysimus joined #miro-hackers
06:35 Jarrhead joined #miro-hackers
07:48 maggie_s joined #miro-hackers
08:23 Jarrhead joined #miro-hackers
11:08 arpu joined #miro-hackers
11:51 arpu_ joined #miro-hackers
12:08 arpu joined #miro-hackers
13:06 willkg joined #miro-hackers
13:07 willkg hi!
13:26 janetPCF joined #miro-hackers
13:56 z3p melinath: I had a problem with them sharing state on the eventlet backend, but given that we do a bunch of blocking stuff anyways it breaks there for lots of reasons
13:58 willkg z3p, melinath: you two have conversations that span days.  it's like the ents in LOTR.
13:58 z3p willkg: hah
13:59 willkg: we are on pretty different schedules
13:59 willkg i can see that!  :)
13:59 z3p willkg: luckily, the magic of scrollback means we can /have/ the conversations!
14:00 willkg yay!
14:00 scrollback and whatever you're using for irc that keeps you on all the time.
14:03 z3p willkg: "not turning off my computer" except when gnome3 freezes
14:04 willkg oh, wow.  i thought you did something like what paulproteus does where he's got some other computer and does something and thus his irc sessions last foever.
14:04 z3p willkg: it's kinda of like that
14:04 willkg true.
14:04 janetPCF magicians
14:05 z3p willkg: I'm actually going to experiment with http://yieldthought.com/post/1[…]cbook-for-an-ipad out at the café sometime, just connecting back to this machine
14:08 willkg huh.
14:08 that's kind of funky.
14:21 z3p I can't do it everywhere; I love having a giant monitor
14:21 but it seems easier than lugging my laptop around
14:34 willkg totally.  especially since your laptop is _huge_.
14:51 z3p and the battery life is getting worse
14:52 willkg i bet.
15:06 ajonas joined #miro-hackers
15:54 GitHub139 joined #miro-hackers
15:54 GitHub139 [miro] bendk pushed 2 new commits to master: https://github.com/pculture/mi[…]7f6b6a0...8362464
15:54 [miro/master] bz18422: Index active database objects via (id, table name). - Geoffrey Lee
15:54 [miro/master] Merge pull request #74 from geoffl/bz18422 - bendk
15:54 GitHub139 left #miro-hackers
16:29 arpu joined #miro-hackers
16:30 janetPCF joined #miro-hackers
17:15 janet joined #miro-hackers
18:12 z3p melinath: I checked in the FeedImport code
18:12 melinath: but there's a problem; since we aren't keeping track of the results of of the tasks
18:12 melinath: we can't know when the feed import is really done
18:12 melinath: oh, but maybe we can see if the task completed, without having a result?
18:14 melinath: nope
19:31 melinath z3p: Why do we need to know when the feed import is done?
19:34 DGMurdockIII joined #miro-hackers
19:38 melinath z3p: I guess what I mean is more: Could you just add a status field to the FeedImport model to track whether it's finished?
19:39 z3p: Although if you have an end time field, you could just check whether it has a value stored there already.
19:40 z3p: Which you might already be doing...
20:28 Odysimus joined #miro-hackers
21:40 maggie_s joined #miro-hackers
22:11 maggie_s joined #miro-hackers
23:49 mattjohnson joined #miro-hackers

← Previous day | Index | Server Index | Channel Index | Today | Next day → | Atom Feed | Search | Google Search | Plain-Text | plain, newest first