10:34:56 <killing-joke> as predicted in the tutorial, i am stalled while installing with some kind of Quick-Check deadlock
10:36:09 <killing-joke> when i try to install manually with 'runhaskell', my Cabal packages are not found, even though the repo is in the PATH
10:53:32 <stevenmarky> try moving to the directory that contains the extracted tar.gz and doing runhaskell there
11:02:01 <killing-joke> ok, thanks
11:03:43 <killing-joke> stevenmarky, no go   http://pastebin.org/153574
11:08:00 <stevenmarky> was the QuickCheck deadlock coming from the happstack-helpers?
11:16:39 <killing-joke> looking at that ...
11:17:45 <killing-joke> it said "Resolving dependencies...    cabal: cannot configure Crypto-4.2.1. It requires QuickCheck >=2"
11:19:04 <killing-joke> funny -- i installed Crypto and QuickCheck-2.1.0.3 before trying to install happstack-helpers
11:19:25 <killing-joke> ghc-pkg list shows both
11:20:34 <killing-joke> so 'cabal install happstack-helpers' produces the puzzling message about not being able to configure Crypto, which was already installed
11:27:47 <killing-joke> stevenmarky, i am using GHC 6.12.1 and Cabal 1.8.0.4
11:33:26 <stevenmarky> killing-joke: Sorry I don't know the solution to this problem
11:33:39 <killing-joke> that's ok  :)
11:33:44 <killing-joke> thatnks for your insight
11:34:06 <killing-joke> i'll read the source. it will help me learn Haskell
12:11:31 <McManiaC> hmmm
12:14:26 <McManiaC>     No instance for (Data.Generics.SYB.WithClass.Basics.Data
12:14:27 <McManiaC>                        DefaultD UserId)
12:14:32 <McManiaC> wtf is that?
12:15:01 <McManiaC> UserId is just a new type with a Word64 (which should have a Default instance)
12:21:47 <burp> a newtype?
12:21:58 <burp> or a new type? ;)
12:22:40 <McManiaC> newtype
12:35:56 <McManiaC> wtf
12:36:12 <McManiaC> if I remove that Default instance everywhere I get a "no Migrate instance" error
12:43:49 <McManiaC> eeeeh
12:44:36 <McManiaC> lets say I have some datatype of version 5
12:44:48 <McManiaC> with migrate instances from 0 to 5 for each version
12:45:19 <McManiaC> if another lib is using "migrate" on a version 3 to migrate to version 5 - do I need a "instance Migrate Version0 Version5 where..." instance? :O
12:45:25 <McManiaC> *Version3
12:47:14 <McManiaC> lol
12:47:17 <McManiaC> I really do :O
12:48:17 <McManiaC> whats the point of having version x -> version x+1 migration instances when I have to define them by hand again?
13:52:53 <McManiaC> can I use Control.Exception with ServerPart?
13:53:00 <McManiaC> or how can I catch exceptions?
13:56:36 <st3pcut> McManiaC: dunno, I should add that to my tutorial sometime
13:57:26 <stepcut> McManiaC: there are some instances related to supporting that though I believe
17:10:08 <peti`> Hi. I've noticed that happstack-server (among others) is restricted to Parsec < 3. This means that projects that need Parsec 3 cannot use happstack. :-( So I was wondering what the rationale behind that choice is? Would it be possible to update the code so that it works with recent versions of parsec?
17:22:18 <stepcut> peti`: network uses parsec <2, so.. there isn't much choice
17:22:49 <stepcut> peti`: parsec 2 is what is in the current version of the haskell platform
17:23:45 <stepcut> peti`: dcoutts has proposed that cabal needs a concept of 'private dependencies' so we can indicated that happstack-server does not actually export any of the parsec 2 types, and therefore it is ok to use happstack in an app that uses parsec 3
17:24:57 <stepcut> I believe for parsec specifically, there was a discussion of given parsec 3 a completely new name, so that you could use both parsec 2 and parsec-alt-name 3 with out conflicts
17:32:45 <stepcut> peti`: as a 'workaround' you can use, runhaskell Setup configure, etc, instead of cabal configure. But then people won't be able to cabal install from hackage
19:12:11 <shapr> peti`: Are you the original peti?
19:12:17 <shapr> Aha! You are!
19:12:23 <shapr> peti`: Hiya! Long time no see!
19:53:50 <McManiaC> stepcut: npaste is becoming grown up :D
19:53:52 <McManiaC> www.n-sch.de/mypastes.png www.n-sch.de/edit.png
19:54:15 <McManiaC> testing system running on test.npaste.de
19:54:24 <McManiaC> going online as soon as I cant find any more bugs =)
19:57:03 <aavogt> quick, before hpaste comes back!
19:58:35 <McManiaC> :D
20:02:32 <aavogt> perhaps npaste shouldn't encourage useless uses of cat?
20:02:39 <aavogt> you can write:
20:02:54 <aavogt> curl -F "content=<-" npaste.de < file
20:02:55 <McManiaC> I love cats :>
20:05:16 <McManiaC> better? :P
20:06:48 <aavogt> :)
20:44:03 <McManiaC> aaand online
20:44:07 <McManiaC> http://npaste.de/
20:53:37 <siracusa> McManiaC: The heading looks like `npaste.de \n :: \n IO String' on Firefox 2
20:54:38 <McManiaC> firefox 2? :O
20:55:05 <siracusa> Yeah, don't ask what IE version I'm using ...
20:55:14 <McManiaC> lol
20:56:26 <McManiaC> it should work for all current webbrowsers
20:59:00 <burp> that's why it says IO ;)
20:59:46 <McManiaC> you'll never know what'll happen =)
21:58:45 <gwern> 'useless use of cat' considered harmful
21:59:16 <gwern> clarity beats microseconds of performance any day
22:01:21 <aavogt> clarity of what?
22:01:48 <gwern> cat foo | bar | baz | quux
22:02:10 <aavogt> vs. something involving parentheses
22:03:17 <aavogt> actually:
22:03:26 <aavogt> < foo | bar | bas | quux
22:03:50 <aavogt> or    bar < foo | ...
22:04:27 <gwern> my eyes!
22:04:27 <aavogt> this is just like writing:     getLine >>= return . words
22:04:36 <aavogt> I mean your version
22:05:09 <aavogt> vs.   fmap words getLine
22:05:13 <aavogt> imho
22:05:50 <aavogt> but cat instead of < isn't that bad
22:09:05 <McManiaC> 22:02    McManiaC I love cats :>
22:11:30 <aavogt> > cat
22:11:31 <lambdabot>   No instance for (Test.SmallCheck.Serial
22:11:31 <lambdabot>                     Text.PrettyPrin...
22:11:40 <aavogt> @type cat
22:11:41 <lambdabot> [Doc] -> Doc
22:12:02 <McManiaC> < cat