Experimental IRC log happs-2008-10-01

Available formats: content-negotiated html turtle (see SIOC for the vocabulary)

Back to channel and daily index: content-negotiated html turtle

These logs are provided as an experiment in indexing discussions using IRCHub.py, Irc2RDF.hs, and SIOC.

02:09:07<pflynn>The cabal install worked.
06:54:56<insitu>hello, I am trying to build happs using 'getting started' info but run into a problem: sp cannot find System.Log.Logger
06:57:12<insitu>ok, found the answer on google group archive. will have to install hslogger myself
14:28:09<fxr>hi
14:28:46<fxr>http://hpaste.org/10810 do you see any problem with this component? I'm getting "method not general enough".
16:53:24<Saizan>fxr: i'm not sure, but maybe you've to use Proxy (Session a) like in numSessions
16:57:01<fxr>Saizan: hmm lemme try
17:14:41<fxr>nope same err
17:19:09<fxr>I think this happens when methodTyVars are LT than methodKeys but I don't know what they are.
17:24:39<Saizan>yeah
17:24:54<Saizan>methodTyVars are all the tyvars in the type of the method
17:25:07<Saizan>while methodKeys are the tyvars in the state type
17:25:29<Saizan>and indeed they are not equal if isSession :: SessionKey -> Query (Session a) Bool
17:25:57<Saizan>but isSession :: Proxy (Session a) -> SessionKey -> Query (Session a) Bool should work
17:32:10<Saizan>to be precise methodTyVars are all the tyvars except those appearing in the state
20:17:49<fxr>Saizan: but there is, getSession :: SessionKey -> Query (Session a) (Maybe a) which breaks the same rule but it comples.
20:27:12<fxr>hmm I moved the mkMethods line before instance declaration and it compiled
20:44:31<Saizan>getSession doesn't break that rule
20:46:15<fxr>Saizan: that doesn't because of (Maybe a)?
20:46:29<Saizan>yeah
20:49:54<fxr>thank you
20:50:34<fxr>basically I saw a comment in AllIn.hs "really you should put components in their own modules."
20:50:50<fxr>and looked at the examples on web
20:51:16<fxr>they all implemented state as a big component
20:51:58<Saizan>heh
20:53:39<fxr>my first experiment was same, now I separated them and defined a state component which depends others.
20:53:42<Saizan>the rationale for this rule is that all the tyvars in the state must appear in the type of the generated events, so that you can have both a (Session Foo) and a (Session Bar) component and still know which one a particular event belongs to
20:57:10<fxr>but I really don't know the practical usage of a (Session a) type
20:58:00<Saizan>what do you mean?
20:58:48<fxr>is that because some people thinks that alphanumeric session ids will give better security? I mean I cannot see any difference between (Session Int) and (Session String) in practice.
20:59:41<Saizan>the session id is SessionKey
20:59:55<Saizan>the 'a' is the kind of data you want to keep in your session
21:00:07<fxr>oh sorry you're right
21:01:09<fxr>so this means that in practice, integers stored as Integer type which occupies less mem.
21:02:35<Saizan>for the session ids, yeah
21:04:34<fxr>Saizan: is this M.Map thing has O(log n) complexity?
21:05:27<Saizan>fxr: yes, for insert and lookup
21:05:47<Saizan>http://haskell.org/ghc/docs/latest/html/libraries/containers/Data-Map.html
21:06:26<fxr>thanks
21:07:32<Saizan>http://haskell.org/ghc/docs/latest/html/libraries/index.html is a very handy page :)
21:08:54<fxr>Great

Back to channel and daily index: content-negotiated html turtle