Experimental IRC log happs-2008-08-29

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.

00:09:40<sadache>hello
00:09:58<sadache>i am trying to run a sample example of happs
00:10:23<sadache>but it keeps telling me "could not find module 'and' "
00:10:33<sadache>any help?
00:10:54<stepcut>hrm
00:11:08<stepcut>is this sample online somewhere ?
00:11:22<sadache>yes on tha happs homepage
00:11:32<stepcut>did you perhaps accidently put an 'and' in the source code on an import line
00:12:05<sadache>no, i downloaded it and didnt touch to it exactly as it is written on the website
00:12:56<sadache>1. darcs get http://happs.org/repos/happs.org MyProject
00:13:09<sadache>2. cd MyProject
00:13:24<sadache>3. sp ghc -isrc src/Main.hs --make --run --http-port=5000
00:17:26<sadache>this is really worries me, if a sample cant execute even following exactly three simple steps
00:21:20<stepcut>hrm, what is the exact error message you get ?
00:26:11<sadache>http://www.mibbit.com/pb/WIWryQ
00:30:50<sadache>i would like to know did anyone here try the sample at all?
00:34:24<sadache>i guess i found it
00:37:25<stepcut>hrm
00:38:22<stepcut>I have all of happs built and installed via cabal. I did, darcs get http://happs.org/repos/happs.org MyProject && cd MyProject && runhaskell -isrc src/Main.hs
00:38:24<stepcut>and that worked for me
00:38:35<sadache>i guess the problem has something to do with my "Documents and Setting" directory
00:38:40<stepcut>ah
00:38:46<stepcut>yes, spaces in filenames
00:38:58<stepcut>people suck and never think about that
00:39:32<sadache>but I have nothing there, i should find why is it trying to access them
00:40:00<stepcut>might be looking at $HOME or something
00:40:06<sadache>ok, it is the cache directory
00:40:29<sadache>now i have to know where is it configured as a cache directory
00:40:30<stepcut>many people opt to use cabal and forgo sp
00:41:02<stepcut>sp is nice if you can get it to work, but not required by any means
00:41:08<sadache>should i use cabal rather than sp?
00:41:45<stepcut>if none of your files are in 'Documents and Settings' then I would try cabal instead of sp
00:42:12<stepcut>and if that works, complain that sp does not handle spaces in filenames properly
00:42:36<sadache>ok, i am really newbie to this because I always use visual studio for programming in C# and it cant be easier to configure
00:44:28<sadache>but with cabal i should get manually the dependencies right?
00:44:59<stepcut>sadache: there is a program, cabal-install will manually download and install all the dependencies. People say it works with HAppS
00:45:40<sadache>ok
00:45:49<sadache>thank you
00:46:17<Saizan_>s/manually/automatically/
00:46:50<sadache>ok
00:47:45<stepcut>oops ;)
00:47:57<Saizan_>uhm, wait, the hackage packages won't build on linux
00:48:05<Saizan_>s/linux/windows/
00:49:55<Saizan_>sadache: are you familiar with haskell though?
00:51:18<sadache>haskell yes
00:51:40<sadache>i mean the concepts of haskell, i am pretty familiar with
00:51:51<sadache>but i am really bad at configuration
00:51:55<sadache>s
01:07:27<sadache>i am on windows, what am i supposed to do to install cabal and cabal-install?
01:32:04<MarcWeber>He, today I had the longest compilation ever: 43min.. resulting in an error..
01:32:51<MarcWeber>I've fixed some errrors again and I feel I really could type check xhtml if it didn't take that long.. :-(
01:42:18<stepcut>is there a, newtype IdentityM m a = IdentityM { runIdentityM :: m a }, in the the standard libs somewhere ? (or anywhere?)
01:42:32<stepcut>oops, I mean IdentityT
01:45:19<stepcut>also, not quite the right channel (though it is about HAppS+HSP)
01:45:51<Lemmih>There's a HSP channel?
01:46:12<MarcWeber>stepcut: In which way does it differ from IdentityT?
01:46:41<stepcut>MarcWeber: none, I am just wondering if IdentityT is in the standard libs somewhere
01:47:21<stepcut>Lemmih: no, I meant to ask in #haskell, because it has more traffic. But I need this for HAppS+HSP+Formlets -- so it is relevant to this channel.
01:48:01<MarcWeber>stepcut: sorry I mean Identity
01:49:04<stepcut>MarcWeber: mostly the same, but it also implements MonadTrans/lift
01:50:27<stepcut>well, actually, I think it is a fair bit different
01:51:18<stepcut>this is a proposed implementation http://www.haskell.org/pipermail/libraries/2007-June/007560.html
04:18:41<stepcut>hah, I was about to send a patch to make WebT a functor, but did a pull first. First patch: Make WebT a functor
14:21:31<MarcWeber>Saizan_: If you want you can have a try now. After asking Oleg how to implement TypeEq :-) the small XHTML example validates in less than 10 secs
14:23:32<MarcWeber>git fetch; git checkout origin/HListTypeEqGen (for you who don't have the git repo yet use git clone git://mawercer.de/vxml instead of fetch)
14:26:25<Saizan_>oh, so you had a bad performing TypeEq? :)
14:26:54<MarcWeber>Saizan_: Not sure.. I went to bed yesterday because I was tired.. This morning I had a type checked result..
14:27:16<MarcWeber>I estimate compilation time with the TypeToNat approach > 1h
14:30:13<Saizan_>fatal: Entry 'test/test.hs' not uptodate. Cannot merge. <- how do i solve this?:) is there a git revert?
14:30:24<MarcWeber>git revert --hard
14:30:48<MarcWeber>While developping you usually do git stash to save the changes on a stack
14:31:03<MarcWeber>You can also git checkout branch -m (to force git doing a merge..)
14:31:53<Saizan_>i don't have "hard" here.. i guess i can just clone a fresh repo
14:32:05<MarcWeber>Sure.
14:32:30<MarcWeber>sorry. It's git reset --hard
14:32:50<MarcWeber>I'm no longer using the commands. I've aliased it to gt.. :-(
14:34:57<MarcWeber>Saizan_: Do you have dist/build/test/test now?
14:38:49<Saizan_>MarcWeber: no, it won't build
14:39:12<MarcWeber>What kind of problem does occur?
14:41:17<Saizan_>http://hpaste.org/10007
14:42:19<MarcWeber>You need a patched HaXmL..
14:42:28<MarcWeber>I've forgotten about that.
14:43:03<MarcWeber>HaXmL can't parse the dtd due to a wrong used hard failure in its parsing description
14:44:44<Saizan_>hah
14:45:37<MarcWeber>Currently checking wether I've already uploaded the fixed version to my server
14:46:41<Saizan_>i've to focus on something else, but i will fetch it and retry later :)
14:47:32<MarcWeber>http://mawercer.de/~nix/repos/haxml.tar.gz
14:47:39<MarcWeber>That's fine
18:00:28<MarcWeber>chr1s: I'm currently reading some formlets and applicative-extras code..
18:01:26<MarcWeber>I'm really wondering why you've introduced Failing ? (which is a bad name IMHO, because it's kind of result.. It needn't be a failure)
18:02:11<stepcut>ACTION is currently hacking on formlets+hsp some more
18:02:23<MarcWeber>I would have used Either a b where a is instance of monoid .. so you'll get the [ErrMsg] ++ [ErrMsg] accumulation for free
18:03:18<stepcut>MarcWeber: so ErrMsg does not have to be String
18:03:19<stepcut>?
18:04:38<MarcWeber>Yes.
18:05:08<MarcWeber>That's another thing I don't understand.. It's a general applicative-extras lib but assumes ErrMsg must be a String
18:06:29<MarcWeber>And there is a class Plus exactly looking like Monoid..
18:07:10<MarcWeber>chr1s: Is there a repo somewhere? May I do such minor refactorings?
18:11:01<MarcWeber>And what about ensurM? Isn't it superfluous? I mean you can use liftM or `ap` or such easily? At least it should call ensure than
18:19:03<MarcWeber>and freshName doesn't give you a fresh one, it is rather returnCurrentAndGenerateFreshForNextUse
18:21:24<MarcWeber>orT looks preety much like Alternative ?
18:36:34<MarcWeber>input' :: Monad m => (String -> String -> xml) -> Maybe String -> Form xml m String
18:37:08<MarcWeber>maybe it would be convinient to have a class here choosing this function based on the xml type? But I don't understand all the code yet
18:39:48<guenni>did understand correctly that every now and then Happs is serializing to disk?
18:40:03<guenni>and keeps all the state in memory
18:55:56<Saizan_>toughly speaking yes
18:56:02<Saizan_>"roughly"
18:58:53<chr1s>MarcWeber: you're right on most of those things!
18:59:56<chr1s>most of those points should be refactored
19:00:05<chr1s>freshName does give you a freshName
19:01:47<chr1s>I also think ensure should be defined in terms of ensureM
19:02:22<chr1s>so I'm currently just adapting formlets for my specific needs, and that's also exactly why I haven't announced it on the haskell mailing list.
19:03:18<chr1s>my plan is to first wait a bit more until the API stabilizes, refactor everything a bit and then announce it.
19:05:17<chr1s>I currently don't have a repos online, but you can always just send patches. I think it would be a good idea to put it in some git repository once haskell.org has git hosting. or maybe on github?
19:05:26<chr1s>anyway, all help is more than welcome!
19:07:34<Saizan_>ACTION advertizes darcs
19:17:55<chr1s>well, I've tried darcs, I really want to use it (just because it's written in Haskell)
19:18:07<chr1s>but it's way WAY to slow
19:18:25<chr1s>once you're used to git's speed...
19:20:18<Saizan_>even for small projects?
19:21:36<Saizan_>it works quite well for me with the Cabal repo, but i've never really tried git, i must confess
19:34:44<chr1s>Saizan_: yeah, even for small projects.
19:37:58<MarcWeber>chr1s: I can host it until you have found another host
19:40:22<MarcWeber>chr1s: Try pushing to git://mawercer.de/formlets
19:41:29<MarcWeber>Saizan_: If you don't mind I can give you a small git introduction (at least all I know)..
19:42:50<Saizan_>MarcWeber: it might be useful :)
19:46:20<MarcWeber>Saizan_: You should tell me when you have time and want to start
19:46:48<MarcWeber>chr1s: ping
19:54:59<MarcWeber>vxml can typecheck this page within 35secs http://rafb.net/p/8AaH7l89.html.
19:55:11<MarcWeber>I feel it's a little bit to slow to be usabe
19:55:14<MarcWeber>usable
19:56:15<Saizan_>considering that people like to use xml files of gigabytes, maybe yes
19:58:30<geezusfreeek>MarcWeber: just coming on the conversation. does vxml check xml documents or xml-generating haskell code or both? i'm just a bit confused
19:59:16<MarcWeber>the last one.
19:59:25<geezusfreeek>aha nice
19:59:43<MarcWeber>But I've implemented a test reading an xml file using HaXml generating template haskell code to write that file again..
20:00:07<MarcWeber>So it could be used (after many enhancements) to really typecheck documentes.. But xmllint is a better tool then
20:02:51<geezusfreeek>well, you can count me interested in it. i will follow the git repository and maybe toy with it later
20:04:07<MarcWeber>geezusfreeek: git://mawercer.de/vxml
20:04:28<geezusfreeek>thx

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