<<16-11-2012>>

03:45:12*fowl quit (Ping timeout: 268 seconds)
03:47:47*Amrykid3 joined #nimrod
03:50:21*Amrykid2 quit (Ping timeout: 276 seconds)
04:18:33*fowl joined #nimrod
05:07:12*XAMPP_ joined #nimrod
05:09:49*XAMPP quit (Ping timeout: 246 seconds)
05:43:38*Amrykid3 quit (Quit: Leaving)
07:53:43*Boscop quit (Ping timeout: 260 seconds)
08:14:05*Araq_ joined #nimrod
08:17:58*Araq_ quit (Client Quit)
09:29:50*rking quit (*.net *.split)
09:32:50*fowl quit (Quit: Leaving)
09:40:56*rking joined #nimrod
09:59:32*rking quit (*.net *.split)
10:10:05*rking joined #nimrod
10:35:28*rking quit (*.net *.split)
10:56:03*Arcanum_za joined #nimrod
10:56:11*Arcanum_za is now known as Trixar_za
10:59:41*Trixar_za quit (Client Quit)
11:23:14*Trix[a]r_za quit (Ping timeout: 252 seconds)
12:06:57*Araq_ joined #nimrod
12:12:31*Araq_ quit (Quit: ChatZilla 0.9.89 [Firefox 16.0.2/20121024073032])
12:28:55*rking joined #nimrod
12:49:03*Trixar_za joined #nimrod
13:48:21*q66 joined #nimrod
15:21:22*Trixar_za quit (Quit: Leaving)
16:24:04dom96hello
16:52:01dom96This looks interesting: http://aisandbox.com
17:35:38Araqhi dom96
17:38:23dom96hey Araq
17:39:05Araqthe first class iterators can now be used for a light task system :P
17:39:27Araqdunno if it helps async IO though ...
17:57:05Araqso ... we can have:
17:58:07Araqvar html = download("http://nimrod-code.org")
17:58:10Araqyield
17:58:29Araq# scheduler gurarantes the IO completed when control comes back here
17:58:41Araqecho html.get()
17:59:11Araqit can be wrapped in a template
17:59:28Araqbut the 3 statements need to occur in an iterator
18:00:25Araqbasically the whole pseudo-linear flow needs to be within an iterator
18:01:03Araqhowever, with the right async library design it should be really sweet
18:05:54Araqnow aren't you happy you already switched to threading for nimbuild? :P
18:34:39dom96most of nimbuild is still async.
18:41:27dom96I'm not entirely sure how these new iterators work, so I'm not sure how to adjust the asyncio module to work with them.
18:41:45AraqI'm looking at the current api again ...
18:42:31dom96In fact I think a new module might be necessary
18:44:40AraqI don't think so
18:49:47Araqyour TDelegate stuff is pretty weird
18:50:06AraqI suppose you left it that way to not break any code?
18:53:13dom96what's weird about it?
18:53:36Araqwell TAsyncSocket supports closures but TDelegate does not
18:53:55Araqand both have handleRead etc. proc vars
18:54:19dom96mmm, perhaps it should be a closure
18:55:31dom96I simply forgot to change it I think.
18:56:07dom96but then again I think it would break some code.
18:56:39Araqyeah
18:57:10dom96You don't really need a closure there though.
18:57:20dom96You're already given a custom object.
18:57:23Araqtrue.
18:57:38Araqit's fine the way it is but my point was:
18:57:56AraqTDelegate and TAsyncSocket look like they should be merged somehow
18:58:29dom96no.
18:58:46Araqwhy not?
18:58:57dom96TDelegate is the generic object which works of file descriptors
18:59:01dom96*works on
18:59:54dom96If you merge it, asyncio will turn into asyncsock
19:00:16Araqasyncio already contains the TAsyncSocket
19:00:37Araqin fact these should all be streams instead
19:01:13AraqI'm not saying that you should change all this code btw
19:01:16AraqI'm merely thinking
19:03:57Araqftpclient.listDirs returns @[] when in async mode
19:04:18Araqand existsFile always blocks
19:04:47AraqI think 'yield' should be a nop in a 'proc'
19:05:02Araqthen the same template could be used in a task and in a proc
19:07:23Araqbrb
19:29:12Araqback
20:55:11*fowl joined #nimrod
20:57:56*ekselkiu joined #nimrod
23:14:19*ekselkiu quit (Ping timeout: 260 seconds)
23:22:23*fowl quit (Remote host closed the connection)
23:41:10*fowl joined #nimrod