Experimental IRC log happs-2008-03-09

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.

12:08:08<fxr>knock knock
12:08:49<fxr>I'm trying to use RJson with HAppS but got some errors: http://hpaste.org/6232, any ideas?
12:17:05<MarcWeber>fxr: Don't know what is happening but after removing import RJson (don't hav it installed ye) it compiles
12:17:10<MarcWeber>http://rafb.net/p/Aie13k55.html
12:17:32<MarcWeber>But I get the same warning.
21:28:45<fxr>according to RJson author the problem is FilePath type synonym. I changed the type from FilePath to String and it worked.
21:29:21<fxr>"Hi, the problem here is with FilePath. It seems that trying to
21:29:41<fxr>serialize type synonyms can sometimes cause problems"
21:36:30<mae>hello mortals
21:46:06<akamaus>mae: who is guilty? unsafePerformIO in show instance of SockAddr or happs for not wrapping call to show with catch?
21:48:40<mae>lol
21:48:49<mae>the network guys i blame ; )
21:49:11<mae>oh wait
21:49:15<mae>yeah happs fault
21:49:18<mae>if they did both of those ; )
21:49:41<mae>heh
21:50:05<mae>such a simple foregoing of the haskellish strictness brought the whole titanic down i suppose ; )
21:50:07<mae>one icebergf
21:51:29<akamaus>yeah
21:52:25<mae>but hey
21:52:29<akamaus>not all uses of unsafePerformIO are safe )
21:52:30<mae>look on the bright side
21:53:02<mae>at least it threw its' hands up in frustration rather than continuing in an exceptional situation ; P
21:59:23<akamaus>now i'm puzzled how to fix this. From one side, show can be looked as a partial function here. From another - hostnames assigned to ip addresses are supposed to change from time to time.
22:01:02<akamaus>by the way, is there a way to locate all calls to particular instance of overloaded function?
22:18:32<mae>could someone explain the difference between ServerPartT and WebT
22:19:13<mae>hmm
22:19:20<mae>i don't know whater your asking exactly
22:20:01<mae>but i would simply not make the call at all, and attach the ip resolution to the logging facility, as the reverse hostname is not needed for core operation
22:20:59<mae>as for hostname lookups for proxies, the hostname lookup should be deferred to the last possible moment before a connection is opened imo
22:21:05<mae>and can this not be converted into an action?:
22:21:10<mae>since its in the network code
22:23:06<Saizan>ServerPartT = WebT + a Request
22:23:25<mae>so
22:23:39<mae>when i pass a ServerPartT to simpleHTTP
22:23:47<mae>does it evaluate left associatively or right associatively
22:23:55<mae>is it drilling down or bubbling up
22:24:13<Saizan>what do you mean?
22:25:13<mae>because i can chain xslt in there for instance
22:25:20<mae>or nested dir functors
22:25:39<Saizan>yeah
22:26:11<Saizan>well, when you provide a list the elements that comes first have precedence
22:27:09<mae>hmm ok
22:27:11<Saizan>(since the combinators are written with multi)
22:27:21<mae>so when i do multi [parts] for instance
22:27:31<mae>this does the operation of selecting the first match
22:28:07<mae>yeah my problem is that i was rewriting xslt because when i hand it data it is returning the wrong Content-Length
22:28:20<mae>and it made me consider, why does this combinator calculate content length at all
22:28:33<mae>why wouldn't that be done in the last step automatically
22:28:35<Saizan>do you understand the Maybe monad?
22:28:45<mae>yeah'/
22:29:56<akamaus>mae: big thanks for your help
22:33:18<mae>akamaus: heh, how did i help?
23:11:07<Lemmih>fxr: FilePath/String should already be an instance of Serialize.
23:11:36<Lemmih>mae: It searches for a match from head to tail.

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