01:54:43 <alpounet> stepcut, around?
01:56:32 <stepcut> alpounet: yup
01:56:36 <stepcut> a bit
01:57:03 <alpounet> stepcut, i think i did the necessary to send a whois request to the irc server
01:57:10 <stepcut> nice!
01:57:11 <alpounet> how can i get the result from that ?
01:57:55 <stepcut> ah..
01:57:57 <stepcut> one moment
01:58:14 <alpounet> i mean, the server will answer right after
01:58:14 <stepcut> this is one of the things I would like to formalize more
01:58:24 <alpounet> how can i get this?
01:58:43 <stepcut> one moment
01:59:37 <stepcut> we have a similar problem when changing nicks
01:59:38 <stepcut> http://patch-tag.com/r/stepcut/ircbot/snapshot/current/content/pretty/Network/IRC/Bot/Part/NickUser.hs
01:59:54 <stepcut> you send a request to the server, and then asynchronously you receieve a response back later
02:00:05 <stepcut> but it does not directly refer to the request you sent
02:00:27 <stepcut> in that example, we send the 'changeNick' request in the changeNick function
02:00:41 <stepcut> and then there is a part that looks at all incoming messages to see if there was a nick changing error
02:01:16 <stepcut> an improvement would be if there was a part where  you could register a callback for a certain responses
02:01:59 <stepcut> but for now, NickUser is the best working example that I have
02:02:51 <stepcut> in irc, in general, there are a bunch of different states you can be in, and state transitations that can happen
02:03:10 <stepcut> it would be nice if there was some pure model of that which we could tie into instead of relying on an IO driven model
02:03:53 <stepcut> a starting point would be to create some sort of abstraction to deal with issuing a 'request' to the server and waiting for ' response'
02:32:14 <alpounet> stepcut, yeah ok i see
02:32:35 <alpounet> i'll try to implement a first version of my whois thing
02:32:39 <alpounet> tomorrow
02:32:53 <alpounet> with the same method you used for NickUser
02:33:14 <alpounet> and then once it works we can try to have a proper abstraction
02:33:23 <alpounet> anyway, off to bed, good evening/night
03:08:28 <stepcut> sweet!
07:55:08 <donri> stepcut: you're right it works wonderfully to just put everything in the same sitemap https://github.com/dag/kibr/commit/91db88788cbed8565f1b62e32aed03b1b37f5b9c
07:55:31 <donri> i was just stuck in a line of thinking
10:20:53 <donri> stepcut: i'm such a noob, do you understand what roconnor is saying about ixLens or how to fix it?
10:22:13 <donri> in your reddit post
12:58:49 <alpounet> ACTION is stuck in the whois nightmare
13:03:13 <donri> :(
13:03:43 <donri> lpeterse: ohai, higgsset no compile for me on any ghc
13:04:44 <lpeterse> donri: are you trying to compile the cabal or the github version?
13:04:53 <alpounet> it seems i have to talk to nickserv
13:04:55 <donri> i've tried both
13:05:36 <lpeterse> donri: what does the compiler say?
13:06:35 <donri> crap, now th-expand-syns is failing too and blocking me from seeing the error for higgs
13:07:44 <donri> let's try clean env
13:08:45 <donri> lpeterse: http://hpaste.org/57374
13:11:03 <lpeterse> donri: what version is your 'containers'?
13:11:20 <donri> containers-
13:12:54 <donri> looks like nfdata is new in
13:13:07 <lpeterse> try updating it to at least the new version contains the missing instances
13:15:52 <lpeterse> I'll add a constraint for this dependency.
13:16:43 <donri> ah, that's what's bothering th-expand-syns
13:17:27 <donri> http://hpaste.org/57375
13:18:56 <donri> guess i'll have to talk to its maintainer
13:21:43 <lpeterse> i can confirm this. just found out that my own containers version was I'm wondering where the NFData instances were coming from
13:22:01 <donri> hehe
13:22:28 <donri> i emailed the author of th-expand-syns
13:23:07 <donri> when that's resolved you should set a lower bound constraint for the containers dependency
13:24:19 <donri> possibly adding th-expand-syns as an explicit dependency constrained to the version he'll release next?
13:24:43 <donri> even if you don't use it directly, to ensure it builds
13:26:24 <lpeterse> I'm wondering why the new version of set defines it's own versions of foldl and foldr although having an instance of Foldable
20:17:42 <donri> stepkut: HiggSet from git should build now, in case you wanted to try it
20:18:32 <stepkut> donri: sweet!
20:18:45 <stepkut> donri: i do.. but not for a few days
20:19:57 <donri> stepkut: do you know if kdtree will be hackaged soon?
20:21:22 <stepkut> donri: dunno.. the author said he was going to be using it for something.. but I am not sure if he has actually had time or not
20:21:32 <stepkut> you could write him and ask
21:43:26 <stepkut> sometimes I think that ircbot needs to use FRP
21:44:10 <stepkut> but then I think, "I should work on happstack instead"
21:44:23 <dsfox> indeed
21:46:49 <stepkut> if you accidentally delete a line of imperative code.. there is a reasonable chance your code will still run, but do the wrong thing.
21:47:09 <stepkut> seems like that is less likely to happen in Haskell code.. because almost every line binds a new variable that you use later?
21:49:01 <stepkut> irc is mostly discrete events.. though timers/timeouts could be continuous signals
22:26:00 <donri> argh higgsset derives *no* instances and the constructor/fields are not exported -.-
22:27:12 <stepkut> -.-
22:27:16 <stepkut> better send a patch
22:27:24 <donri> i wonder how he's even using it himself
22:27:39 <stepkut> dunno
22:28:57 <donri> it also makes you write some weird code
22:29:30 <donri> https://github.com/lpeterse/HiggsSet/blob/master/src/Data/HiggsSet.hs#L117
22:29:55 <donri> lots of "undefined" hackery
22:36:32 <stepkut> you can always write, Foo {}, instead :p
22:44:47 <donri> for patterns yea, but for values that gives a warning
22:44:54 <donri> at least with -Wall
22:45:16 <donri> see the instances for Bounded and Enum further down
22:56:55 <mulletron> I started writing some automatic crud stuff for happstack+acid-state based on Data.Data/Data.Typeable - if I get enough time to get it working properly, do you think this would be of use to people?
22:59:14 <stepcut> yes ?
22:59:29 <stepcut> it sounds useful..
22:59:35 <stepcut> people like CRUD and automatic things :)
23:03:55 <mulletron> cool