07:53:05 <chersanya> hi) What may be the cause of error like this: "Server error: lookInput mode failed because the request body has not been decoded yet. Try using 'decodeBody' to decode the body. Or the 'queryString' filter to ignore the body."?
07:53:27 <chersanya> handler has only 'method POST       mode <- lookText "mode" '
08:02:01 <chersanya> anyone here?)
08:04:21 <stepkut> here!
08:05:36 <stepkut> did you try using 'decodeBody'?
08:10:02 <chersanya> i can't get where to use it?
08:10:34 <stepkut> http://www.happstack.com/docs/crashcourse/RqData.html#rqdatapost
08:14:04 <chersanya> thanks, that really helped! probably i need to read the tutorial))
08:15:07 <stepkut> :)
08:15:27 <stepkut> maybe the error message should have a link to the crash course
08:24:20 <stepkut> kxra: things are shaping up! I now have all the plugins ported to the new architecture
08:24:47 <stepkut> kxra: but, not tested. However, I can release what I have before that is all tested
09:48:11 <mercifulhop> Hi sorry to be an annoying n00b
09:48:19 <mercifulhop> but after upgrading to the April Haskell Platform
09:48:27 <mercifulhop> I can't install happstack-hsp
09:48:38 <mercifulhop> http://stackoverflow.com/questions/13540096/cant-install-happstack-hsp
09:48:41 <mercifulhop> any ideas?
13:10:01 <baa_hoombaa> whats happ en in?
13:10:02 <baa_hoombaa> :)
14:57:18 <KBme> hello
14:58:24 <KBme> i'm trying to make render functions in my modules to use as simpleHTTP's argument
14:59:07 <KBme> I'm using the same structure as described in the crash course, and am trying to use the render functions in the "msum"
14:59:49 <KBme> so I have simpleHTTP nullConf msum [myMod1.filterAndRender, …]
15:00:11 <KBme> thing is I'm using a reader monad in one of my types, and I can't figure out how to make it work
15:02:07 <KBme> what is the type I should use for the response, or what do I have to implement in my reader to make this work?
15:02:57 <KBme> because simpleHTTP expects ServerPart Response, but I have MyMonad (serverpartt IO) response
17:12:36 <KBme> thing is I can't find one happstack 7 app
17:36:08 <KBme> oh, yippee I found one
18:00:05 <kxra> stepkut: sweet!
18:00:07 <kxra> stepkut: do you mean you have tests to run, or you need to manually test them?
18:07:45 <stepcut> KBme: you will need to call, runMyMonad / runReaderT at some point
18:08:35 <stepcut> KBme: if all your parts used MyMonad then you could do something like, simpleHTTP nullConf $ runMyMonad (msum $ [… ]) myMonadData
18:09:15 <stepcut> if only one part does then you just do, simpleHTTP nullConf $ msum [ myMod1.fliterAndRender, runMyMonad myMod2.fliterAndRender myMonadData ]
18:09:48 <stepcut> kxra: no tests yet. Though that is a requirement for 1.0
18:09:54 <KBme> oh
18:10:17 <KBme> yeah, that's what I'm doing
18:10:19 <KBme> but
18:10:22 <stepcut> kxra: by 'test' I mean, copy the live data to my local machine, run the binary, and figure out where the paths got screwed up, etc
18:11:11 <KBme> somehow even runmymonad returns a mymonad ((serverpartt io) Response)
18:11:21 <stepcut> :-/
18:11:29 <KBme> that's my problem, eh? ☺
18:11:29 <KBme> ok
18:12:19 <stepcut> you will need a function that goes a bit like, foo :: MyMonad m a -> someOtherArgument -> m a
18:12:35 <stepcut> similar to runReaderT, runStateT, etc
19:27:31 <stepkut> kxra: you around?
19:46:24 <kxra> stepkut: i am!
19:46:55 <stepkut> http://hub.darcs.net/stepcut/web-plugins, http://hub.darcs.net/stepcut/clckwrks
19:46:59 <stepkut> that is the work in progress
19:47:11 <stepkut> here is the new Main.hs, http://hub.darcs.net/stepcut/clckwrks/browse/example-dot-org
19:47:35 <stepkut> I am pushing a bunch of changes at the moment, in a few minutes things should be a bit more stable
19:56:58 <stepkut> urgh. A few hours. Having guests over.
20:03:12 <donri> stepkut: hm i didn't know about cabal manual flags before. we maybe could use flags instead of impl(ghc) for directory, then
20:03:59 <donri> i assumed before that all flags are manual:True and that manual:False didn't even exist
20:05:16 <stepkut> kxra: actually, i think it may be working now
20:05:51 <stepkut> donri: I am aware of manual flags. Not sure what problem we are trying to solve at the moment though.
20:06:03 <kxra> awesome!
20:06:11 <kxra> should i try to test this?
20:06:27 <donri> stepkut: yea, i don't remember if this is relevant or not but just making a note
20:06:48 <kxra> stepkut: what remaining changes are going into the release?
20:06:53 <donri> if we do make an old-time-compat package, it could be relevant to that
20:06:54 <stepkut> donri: regarding the validation stuff in H8. I suspect we can do away with it in the sense that it can be generalized into whatever we use to support things like e-tags
20:07:03 <donri> ok
20:07:40 <stepkut> kxra: I have not actually tested that media/bugs/ircbot plugins are working correctly, use the same paths as before, and probably upgrade old data
20:08:26 <stepkut> kxra: also, right now the page that shows all the recent blog posts is broken
20:08:33 <kxra> ah, cool
20:08:43 <stepkut> kxra: probably requires a minor change to Theme type
20:09:03 <stepkut> kxra: I don't expect any more major changes, just a bunch of minor cleanup
20:29:04 <kxra> sounds good
20:29:07 <kxra> how many contributors does clckwrk have now?
20:34:59 <stepkut> mostly me, with a few other patches here and there
20:35:12 <stepkut> so.. it's easy to be the #2 contributor!
20:36:24 <stepkut> hoping to attract more interest with this next release, since the code is a lot less scary now ;)
20:37:10 <stepkut> would be nice if someone redid the clckwrks theme using bootstrap (or similar)
20:40:33 <stepkut> the main focus now is getting the core blogging functionality really nice
20:40:41 <stepkut> so that people want to actually use it for their haskell blogs
20:41:45 <stepkut> which will lead to more contributors
21:50:00 <donri> stepkut: so things like trhsx work by parsing with all extensions on, right? but couldn't that mess up the code? i.e. $ having different meanings with TH on or off
21:50:11 <donri> or does HSE take LANGUAGE pragmas into consideration?
21:50:59 <stepkut> donri: by default trhsx enables them all and ignores the pragma i think
21:53:04 <stepkut> HSE itself appears to allow you to turn individual extensions on and off, so we could possible have trhsx pay attention to the PRAGMAs
21:54:18 <stepkut> that reminds me.. I started an experimental port of hsx to the exact parser/printer
21:54:28 <stepkut> no idea if I am doing it right ;)
21:54:33 <stepkut> probably 2 hours left on that to see
21:54:33 <donri> i'm curious if we *need* to; i.e. if it could actually affect the generated code
21:55:08 <donri> i imagine it might be a lesser issue with the exact printer, though, but maybe still could possibly be tricked to fail
21:55:33 <stepkut> probably, though I am not sure how many extensions create syntax ambiguities
21:55:39 <stepkut> the $ is the only thing I know of offhand
21:55:55 <donri> any extension that steals a keyword...
21:56:15 <donri> hm that's easy to check, let's see if we can use 'proc' in a hsp file that's not using Arrow
21:56:32 <stepkut> yeah
21:56:50 <donri> hsp.hs:3:6:  Parse error: =
21:56:50 <donri> phase `Haskell pre-processor' failed (exitcode = 1)
21:56:54 <donri> so yea, this is a problem
21:57:42 <stepkut> right
21:57:59 <stepkut> i can push my changes to the exact printer/parser somewhere if you want to hack on it ;)
21:58:05 <donri> if it fails, it ain't that bad, but i imagine in theory it could be tricked into working but producing the wrong code
21:58:35 <donri> although unlikely to pass type checking, i guess
21:59:21 <stepkut> another reasons why it will be nice when the QQ is released
21:59:28 <donri> (but you'll be baffled by the errors, though)
21:59:35 <stepkut> though, in practice, i am not sure if I have ever run int othe issue in the wild
22:00:14 <donri> hm, QQ might do worse, though. well, the parse will be localized, but it can't sensibly know about LANGUAGE pragmas (actually you can, by looking at the containing source file which TH allows)
22:00:43 <stepkut> true
22:01:14 <donri> i mean it might still be confusing that [hsx| let proc = "hi" in <p><% proc %> |] fails when you don't have LANGUAGE Arrow in the module
22:01:27 <stepkut> right
22:01:30 <donri> but at least you'll be able to use proc outside the QQ
22:01:41 <donri> (but then you can't splice that proc variable into the QQ!)
22:02:19 <donri> yea i haven't had any issues either, but it always confused me how trhsx could "get away" with just enabling every extension
22:03:17 <stepkut> i think it is because there are very few extensions that create ambiguities
22:03:57 <stepkut> a lot of the extensions are things like multiparameter type classes, overlapping instances, etc
22:04:45 <donri> http://www.haskell.org/ghc/docs/latest/html/users_guide/syntax-extns.html#syntax-stolen
22:05:47 <stepkut> nice
22:06:12 <donri> i could imagine trying to name something 'rec' and hsp getting in the way
22:06:26 <stepkut> fortunately, when it does get in the way, it is easy to fix
22:06:40 <donri> magic hash also affects how # as an operator parses, similar to $ with TH
22:06:41 <stepkut> either add a space after $ or use a different variable name
22:06:47 <donri> yea
22:07:09 <stepkut> so maybe i ran into it and didn't even think about it long enough to remember it
22:07:44 <donri> what bugs me is that it *might* in theory generate incorrect code that type checks :D but probably unlikely in practice
22:08:17 <stepkut> that would be annoying
22:08:43 <donri> i.e. we should fix this, however unlikely and however uncommon in practice, but maybe it's not a priority task
22:09:06 <donri> does HSE parse LANGUAGE pragmas?
22:09:13 <stepkut> no idea
22:09:35 <donri> if so we could first do a parse with all extensions, just to get the pragmas, then re-parse with selected extensions
22:09:56 <donri> (yay double parsing)
22:11:21 <donri> i guess LANGUAGE pragmas must appear before any code, so we could add a specialized LANGUAGE parser to HSE
22:13:13 <stepkut> right
22:15:24 <donri> downside is it won't work with extensions only in .cabal, but that's already something to avoid for sake of e.g. ghci ...
22:15:37 <donri> worth it for correctness IMO
22:18:26 <donri> i wish nibro would enable issue tracking on hub :)
22:18:54 <donri> looks like it does parse LANGUAGE http://hackage.haskell.org/packages/archive/haskell-src-exts/1.13.5/doc/html/Language-Haskell-Exts-Syntax.html#g:14
22:28:47 <donri> heh we're not alone with issues on 7.6 due to directory http://gentoohaskell.wordpress.com/2012/11/24/eapi5-ghc-7-6-and-other-goodies/
22:29:17 <donri> oops (dyre) "Now that package has random signature type depending on which directory version it decided to build against."
22:30:23 <stepkut> :-|
22:30:47 <donri> "TODO: find a nice and simple ‘:: ClockTime -> IO UTCTime’ compatibility function to end that keep creeping mess. (I wish the directory package to provide that)."
22:30:56 <donri> old-time-compat! who will do it? :)
22:31:30 <donri> i actually couldn't figure out how to do that at a glance
22:35:55 <stepkut> :)
22:57:02 <stepkut> donri: HSE does has CorePragme, SCCPragma, and GenPragma
22:57:26 <donri> i don't know what that means
22:58:09 <stepkut> meaning, it does at least have the code for parsing the Pragmas to some degree
23:01:07 <donri> yea