10:13:53 <HugoDaniel> hi
10:15:34 <bergmark> .o/
21:15:46 <stepkut> bah. TH stage restrictions suck :)
21:35:04 <donri> stepkut: so how's that blog post coming along
21:35:58 <stepcut> http://www.reddit.com/r/haskell/comments/16ede8/comparison_of_4_approaches_to_implementing_url/
21:36:05 <stepcut> worst title ever
21:36:53 <stepcut> also sucks: creating a loop in your Lift instance and causing the compiler to trash your swap
21:36:59 <stepcut> s/trash/thrash/
21:37:49 <donri> oic you posted it, cool!
21:37:58 <stepcut> :)
21:38:19 <stepcut> not sure if it is interesting, but *I* learned something :)
21:38:27 <donri> happstack's blog isn't on planet haskell
21:38:36 <stepcut> it was :(
21:38:54 <stepcut> it is still listed in the right column
21:39:10 <donri> that's what you get for changing the urls!
21:39:20 <donri> wait it is there
21:39:20 <stepcut> and the article is on the top of the page..
21:39:23 <stepcut> :)
21:39:32 <donri> but not in my reader?
21:39:42 <stepcut> well.. it is all based on polling
21:39:59 <stepcut> first planet haskell has to poll my site, then your reader has to poll planet haskell?
21:45:01 <donri> heh yea
21:45:25 <donri> but it's google reader so maybe they poll the planet once
22:15:26 <donri> stepcut: http://hackage.haskell.org/packages/archive/base/
22:16:14 <stepkut> ah, didn't realize that is in base now
22:16:42 <donri> can't you derive functor for the gadt version?
22:17:00 <donri> even with standalone deriving?
22:17:20 <stepkut> I thought not for some reason
22:18:03 <donri> works for me
22:18:13 <stepkut> i see
22:18:16 <donri> didn't even need standalone
22:18:59 <stepkut> for which type?
22:19:55 <stepkut> hmm, works for me as well
22:20:35 <donri> for Segment
22:20:52 <donri> it might depend on if you're really using the gadt stuff?
22:20:59 <donri> not just the syntax
22:23:27 <stepkut> seems to all compile fine
22:23:52 <stepkut> not sure what error I originally got
22:25:56 <stepkut> I updated it to use the Show instance from base and the GADT syntax
22:26:30 <donri> nice :)
22:27:02 <stepkut> I may have run into a case where Free is superior to operational already :)
22:27:48 <donri> you forgot to import the instance i think
22:27:50 <stepkut> in the `operational` approach we have, instr :>>= k, where 'k' is a function.. which makes it hard to inspect using template haskell
22:27:55 <stepkut> oops
22:28:04 <stepkut> I imported it in my code, but forgot to copy the change to live site, thanks
22:29:12 <stepkut> fixed.. I hope
22:29:16 <donri> neither can cope with infinite programs though :(
22:29:27 <donri> evalState (sequence . repeat . state $ \s -> (s,s+1)) 0
22:29:31 <stepkut> yeah
22:29:59 <stepkut> that could screw me in this instance
22:30:01 <donri> not sure if that's a big loss but still
22:30:06 <donri> oh?
22:31:07 <stepkut> I am testing to see if I can implement parser combinators which build a parser value which is inspected via template haskell and used to generate an optimized, verified parser
22:33:17 <donri> cool
22:33:37 <stepkut> for hyperdrive, of course
22:46:26 <donri> stepkut: you forgot to implement routing with the continuation monad!
22:47:42 <stepkut> I forgot nothing!
22:47:48 <stepkut> that post is already too damn long
22:48:14 <donri> muahaha'
22:48:31 <donri> isn't cont and free closely related though
22:49:34 <stepkut> yes
22:50:04 <stepkut> i think free is like Cont but without callCC or something
22:51:36 <donri> or if that was yoneda or codensity or
22:52:00 <stepkut> something
23:11:38 <donri> what about arrow parsers with static parts? http://en.wikibooks.org/wiki/Haskell/Understanding_arrows#Using_arrows
23:12:15 <donri> though it seems it never took of for parsing, so maybe not that useful
23:13:54 <stepkut> Arrows have that 'arr' function which usually means ->, so that is probably not going to work?
23:15:45 <stepkut> hmm. anyChar seems like it could be trouble
23:28:49 <donri> oh i didn't mean for TH
23:29:12 <donri> was speaking generally on the subject of optimizing parsers
23:29:28 <stepkut> I'm more concerned about verified parsers really
23:30:20 <donri> ah yea
23:30:46 <stepkut> seems that maybe monadic parsers are all going to have trouble with the ->
23:30:57 <stepkut> maybe applicative will work
23:34:47 <donri> anyway the post was a good read
23:34:49 <donri> bedtime!
23:35:02 <stepkut> cool
23:35:05 <stepkut> thanks for the corrections