00:25:27 <Entroacceptor> you failed to be entertaining enough ;)
00:27:20 <Entroacceptor> but there's a remote backend for acid-state now? awesome
00:32:17 <donri> :)
00:32:52 <donri> been around since november '11 it seems
21:37:20 <donri> stepkut: when you have time and will :) if you could look at the end of http://www.happstack.com/I/IrcLog/happs-2012-07-08.txt (you pinged out yesterday)
21:38:38 <stepkut> no idea
21:38:45 <stepkut> someone would need to dig into the code and see what is happening
21:39:39 <donri> in part i'm wondering if it's just me, or if it is even expected/correct behavior and i'm misunderstanding ti
21:39:41 <donri> it
21:40:07 <stepkut> you expectations sound correct to me
21:40:14 <stepkut> the behaviors you are seeing.. not so much
21:40:19 <donri> :)
21:40:33 <donri> don't you use the remote backend yourself?
21:40:56 <stepkut> i do.. in clckwrks to set the admin user
21:41:11 <stepkut> clckwrks-cli connects to the running server via the remote backend (via a UnixSocket)
21:41:11 <donri> and you haven't seen these issues?
21:41:22 <stepkut> no
21:43:41 <donri> also things i'd expect to fail work, like, i can open a state twice
21:44:37 <donri> hm i think i even issued updates on two opened states from the same log
21:45:10 <stepkut> openning a state twice is definitely a bad idea
21:45:27 <donri> yea, but i'd expect acid-state to complain
21:45:37 <donri> it seems even the updates succeed
21:45:38 <stepkut> it really hard to write cross-platform file locking code
21:47:35 <donri> and i can query the state after abusing it like that, just fine; i'd expect the log to be messed up by that?
21:48:16 <stepkut> can your restart successfully though?
21:48:52 <donri> i'm issuing the queries from ghci in a freshly opened state after shutting down the others
21:49:19 <stepkut> dunno, I am sure some data was lost
21:49:36 <stepkut> I don't really care much about the behavior of acid-state when you do that though
21:49:38 <donri> ok, but then i'd expect acid-state to fail to open the state again?
21:49:47 <stepkut> why?
21:50:19 <donri> oh, data loss, i misinterpreted that to mean corrupted state
21:51:08 <donri> and no it's not terribly important :)
21:51:15 <stepkut> well.. it would be 'corrupted' in the sense that you are getting the wrong result.. but not in the sense that it can't read the data
21:51:15 <donri> just trying to understand it
21:59:46 <donri> kibr: too few bytes
21:59:46 <donri> From:	demandInput
21:59:46 <donri> ^ when I ^C the remote client after it hangs
22:00:10 <donri> kibr is the program name
22:03:03 <donri> my code and all the "update" calls exit, so i guess it's some lingering thread
22:06:15 <donri> does acid-state have a bug tracker?
22:39:15 <stepkut> donri: no? You could put it in happstack's bug tracker. that is what I usually do
22:45:07 <donri> ok