Experimental IRC log happs-2008-03-24

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.

08:38:17<MarcWeber>Lemmih: Hi. I'd like to add haskell scripting support to vim. Now I'm struggling how to expose the vim api to the ghci session. Do you think some event based communication (the way the HAppS state system is using it) is the way to go?
12:15:19<Lemmih>MarcWeber: I don't know.
12:58:42<MarcWeber>Lemmih: Only using integers for versions can still break data (eg when switching branches).. So what do you think about not using integers but hashes generated from type structure automatically?
12:59:53<Lemmih>You usually don't wanna change the type when upgrading.
13:02:19<Lemmih>And I don't think branches are very common.
13:02:51<MarcWeber>Yes. this is very unlikely to happen without crash I agree.
13:04:15<Lemmih>How would branches be used? You can use a single version of the data at a time.
13:04:54<Lemmih>(There's no reason it should crash, though)
13:05:20<MarcWeber>Lemmih: Think of having data type X version 1. Now me and my friend both extend it in another way to version 2. Then the version number will match, but not the data.
13:06:02<Lemmih>Well, then you'd get a code conflict. You wouldn't be able to compile the program.
13:06:37<MarcWeber>Lemmih: Yeah. But what happens if I sent my state to my friend ?
13:06:48<Lemmih>Without the code to read it?
13:07:34<MarcWeber>This can actually happen.. :)
13:07:50<MarcWeber>It's very unlikely though.
13:08:05<Lemmih>The problem isn't with HAppS then.
13:08:44<MarcWeber>No it's not a problem with HAppS
13:10:00<Lemmih>Either the state you sent him will parse correctly or it won't. I don't see the problem.
13:11:04<Lemmih>You want a guarantee that it doesn't parse?
13:11:18<MarcWeber>I'm not sure wether it's worth implementing it.
13:11:25<MarcWeber>But that's the basic idea yes.
13:12:06<Lemmih>I doubt it very much. Using the type name won't do.
13:12:39<MarcWeber>No. I agree I don't know yet how the hash should be created. Simyly renaming shouold not break reading state.
13:13:48<MarcWeber>Or maybe it shoud ? data X = A | B -> data X = B | A will propably still parse correctly. It's just a thought. Nothing more by now.
13:15:14<Lemmih>We'll look into it if it becomes a problem.
14:41:12<syntaks>hello
14:43:23<syntaks>what would be the correct way of setting the response content type in HAppS? For instanceif I have ... anyRequest $ ok "<foo>bar</foo>", how could I make it text/xml or application/text+xml?
14:47:24<MarcWeber>syntaks: Have a look at all Response implementations. There is alerady one for HTML
14:51:15<MarcWeber>AppS/Server/SimpleHTTP.hs line 275 @ syntaks
14:51:53<Saizan>you can also just use modResponse and set it there
14:52:17<syntaks>thanks guys
14:53:29<syntaks>I wasn't sure whether modifyResponse was the preferred way to do this
17:06:52<Lemmih>syntaks: The preferred way is: ok $ tag "foo" $ stringToHtml "bar"
17:59:20<syntaks>Lemmih: thanks, I wrapped the type for my responses up in an instance of ToMessage that sets text/xml
18:00:43<syntaks>Lemmih: my XML document comes from HXT... for some reason, the HXT serialization of documents to strings is in IO, so I think I can't put the actual HXT document into ToMessage
18:02:10<MarcWeber>syntaks: No, but you can run liftIO and then put the result into ToMessage
18:02:33<syntaks>MarcWeber: yep, that's what I do now
18:04:55<syntaks>MarcWeber: I don't understand enough about HXT to really tell, but it seems unnecessary that their serialization to string is in IO... I guess they just did it so that the signature is the same as for the serialization to file
18:06:22<syntaks>. o O (though they might pull in external entities or something, which would require IO)
23:06:33<Ben_>hello, this link not found darcs get --partial http://happs.org/HAppS
23:06:49<Lemmih>http://happs.org/repos/HAppS
23:08:32<Ben_>not found
23:08:45<Ben_>Invalid repository: http://happs.org/repos/HAppS
23:08:46<Ben_>darcs failed: Failed to download URL http://happs.org/repos/HAppS/_darcs/inventory
23:08:48<Ben_>libcurl: HTTP error (404?)
23:10:29<MarcWeber>Ben_: Try /HAppS-HTTP
23:10:51<Lemmih>Ben_: Sorry, HAppS-Server.
23:12:25<Ben_>ok, thanks
23:15:08<MarcWeber>Lemmih: is HAppS-HTTP depreceated then?
23:15:25<Lemmih>MarcWeber: It's non-existing, I think.
23:16:06<MarcWeber>Lemmih: Seems to be a mirror. Try it
23:16:30<MarcWeber>I've installed HAppS server using this repo .. :)
23:17:38<Lemmih>Hm, ok.
23:24:09<Ben_>mmm...., this explains why it did not work "import HAppS"
23:24:19<Ben_>was using the HAppS-HTTP :P
23:25:00<Ben_>:D
23:29:53<MarcWeber>Ben_: I couldn't find any differences. But maybe I've done a mistake ? I've tried diff -r *-HAPPS/src *-Server/src
23:30:38<Ben_>ok
23:51:38<Ben_>MarcWeber: mmm..., I have compiled HApps-Server, but not exist "import HAppS"
23:52:12<Lemmih>Ben_: import HAppS.Server
23:53:20<Lemmih>Ben_: You might wanna get 0.9.2.1 from Hackage. The unstable branch can be, well, unstable.
23:56:48<Ben_>mmm..., I am compiling the hpaste
23:58:57<Ben_>I needed the "import HAppS"
23:59:27<Lemmih>Ben_: hpaste doesn't work with the latest HAppS.

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