00:00:00 --- log: started haskell/03.05.03 00:20:15 --- join: Radek_MOB (~radek@gprs9211.eurotel.cz) joined #haskell 00:32:33 --- join: irc (sseefried@pill0.orchestra.cse.unsw.EDU.AU) joined #haskell 00:34:12 --- quit: irc (Read error: 54 (Connection reset by peer)) 00:35:40 --- join: seafood2 (sseefried@pill0.orchestra.cse.unsw.EDU.AU) joined #haskell 00:40:20 --- quit: seafood2 (Remote closed the connection) 00:56:21 --- join: Marvin-- (~martin@sjogren.ostkupan.studenthem.gu.se) joined #haskell 00:59:27 --- quit: Radek_MOB (Connection timed out) 02:08:02 I'm trying to get rid of calls to unsafePerformIO. I'm stuck on how to check whether a value of type IO String has zero length. I can write a function that returns IO Bool, but I'm not sure how to code a test (such as == True) for an IO Bool 02:11:04 if you're in IO, you're in IO, you cannot safely get out (that's why it's called *unsafe*PerformIO :) 02:12:02 you could write an ifM :: IO Bool -> IO a -> IO a -> IO a function 02:13:11 Let me be more specific. I have a recursive function, and one of the arguments is set by a call to hGetLine. I want to terminate the recursion when hGetLine returns IO "". 02:13:22 Is there a pattern for that? 02:13:28 Or something I can use in a guard? 02:14:25 uh, what does the call look like? What's the type of the recursive function? 02:14:54 Socket -> IO String -> [IO String] -> [IO String] 02:15:13 and what is it supposed to do? 02:15:32 no IO will "happen" in that function 02:15:37 If the IO String parameter is not "", add it to the list of [IO String] 02:16:00 but IO String is not String, you can't compare IO String to a String 02:16:03 OK, then let's back up. I want to read lines from a file, add them to a list of lines, and stop when I read a blank line 02:16:07 IO is a type, like any other 02:16:25 OK, but IO String is the only thing that the IO functions return. 02:16:25 then you most likely want IO [String] as return type 02:16:40 Oh, OK, but I don't see how that solves this problem. 02:16:45 yes, you have to be in a function of type IO a for some a to be able to inspect the result 02:17:02 I am, but I still don't see where that helps. 02:17:32 what do the various parameters do? What does the [IO String] do? and what is the IO String? 02:17:36 where is the actual reading? 02:18:10 Again, instead of looking at my code, which is probably wrong, let's talk about what I need to do. I have a socket. I read lines from the socket, and I put the lines in a list until I read a blank line. 02:18:21 How do I do that? 02:18:47 I can do it with unsafePerformIO, but as I understand it I should try to avoid that 02:19:01 no you cannot do it with unsafePerformIO, that will break sooner or later :) 02:19:12 ok, then what is the correct way to do it? 02:19:48 I'd do it like this: getLines sock = do { l <- hGetLine sock ; if null l then return [] else do { ls <- getLines sock ; return (l:ls) } } 02:20:02 (feel free to break it up in lines :) 02:20:34 getLines :: Socket -> IO [String] in this case 02:20:42 @type hGetLine 02:20:50 * Marvin-- prods lambdabot 02:20:58 Its IO String 02:21:07 let me try that. 02:22:35 I can't do it that way, because I have to process the hGetLine output before I test it, but maybe I can fit that into this method. 02:23:10 I can just make a function that does what I need and returns IO String, and call that instead of hGetLine 02:24:18 Oh, I see what you did, you put the recursive call inside of a nested combinator 02:24:59 the point is that inside 'do', you use x <- f to be able to inspect the result of f 02:25:35 That's not the point at all. I didn't have any trouble inspecting the string. I had trouble doing something useful after I inspected the string. 02:25:43 uh, okay 02:26:22 I often get the wrong advise, because everyone assumes that the problem is getting the String from IO String. But that is trivial. 02:26:43 well, I simply must not understand what it is you want 02:26:57 The problem in this case is that I was trying to inspect the string in a separate function, or inspect the string in a guard. 02:27:25 And a separate function returns IO Bool instead of Bool, which doesn't help. And you apparently can't use the tricks for removing the IO within a guard 02:27:53 The reason that what you wrote may work is that you nested a do within another do, and made the recursive call from within the nested do. 02:27:56 of course not, a guard takes a boolean expression, not a monadic expression 02:28:10 well, that's what you always do 02:28:30 Nowhere, in ANY documentation, will you find a single word about that. 02:28:53 And, in a normal (that is, not monadic) setting, you don't have to do that at all. 02:29:07 you can have a pattern like func whatever "" whatever = do something 02:29:16 which you can't do in the monadic setting 02:29:40 but all the documentation shows examples like that, and ONLY examples like that. 02:29:40 huh? I'm not following 02:30:15 I'm just trying to explain why the documentation is so confusing. 02:30:37 because, in most cases, you can't do in the monadic settings any of the techniques shown in all the books and papers. 02:31:16 And you won't find, in Thompson's book, or Hudak's book, a single example with nested do's. I know, I've read both twice. 02:31:49 if you think about it, it's pretty natural though, the if expression is a function :: Bool -> a -> a, and in the monadic code, you've got (>>=) :: Monad m => m a -> (a -> m b) -> m b, so both branches of the if must also have monadic values, how do you usually create monadic values? Using do 02:32:28 You are missing the point. The problem is not that I don't know how to create monadic values. 02:32:50 That's explained in the second paragraph in Thompson's book when talking about monads 02:32:59 and at least Thompson's book doesn't have a whole lot about monads at all 02:33:06 That's quite true. 02:33:41 But look at any of the docs, in wiki, in the doc pages for the various monads. You won't see examples of how to do anything typical. 02:33:55 My point is this. 02:34:15 I'm looking at page 388 in Thompson here.. putNtimes n str = if n <= 1 then putStrLn str else do { putStrLn str ; putNtimes (n-1) str } 02:34:25 well, all right, still only one 'do' 02:34:29 exactly 02:34:35 and not the same situation at all. 02:34:57 The key here is where you make the recursive call, not monadic values. 02:35:01 okay 02:35:15 Now that you showed me, I see it, but it isn't obvious. 02:35:36 on page 392 he creates a while function... while test action = do { res <- test ; if res then do { action ; while test action } else return () } 02:35:44 I think I'll write a "monads for the functionally challanged" 02:35:49 heh 02:36:05 but that won't work, because in any real situation test action would have to be monadic, and we just said you can't do that. 02:36:24 if I could have written "while test xxx" where xxx is that my string is empty, it would have been no problem. 02:36:32 although it is a nested do which I missed. 02:36:38 but test *is* monadic, it's Monad m => m Bool 02:36:44 but the first test action is outside the do 02:36:59 no, that's the lhs of the function definition 02:37:07 you can't say if XXXXX then, if XXXXX is IO Bool 02:37:29 exactly, which is why the code says do { res <- test ; if res then ... 02:37:59 no, you don't get there unless the while test action is true, so that is the wrong part of the statement to look at. 02:38:11 it's not the same situation at all. 02:38:16 I'm not following 02:38:36 I know, I'm not explaining it correctly. 02:39:02 let me finish coding this function and see if the compiler accepts it. 02:39:41 --- join: tic (~vision@h47n2c1o1003.bredband.skanova.com) joined #haskell 02:40:31 no, it almost works, but says: 02:40:42 Couldn't match `[a]' against `IO String' 02:40:43 Expected type: [a] 02:40:43 Inferred type: IO String 02:40:43 In the first argument of `null', namely `li' 02:40:43 In the predicate expression: null li 02:40:57 can you put the code online somewhere? 02:41:13 OK, hold on. 02:41:14 or we could go to #flood 02:41:24 let's do #flood, it isn't very big. 02:44:47 --- join: Radek_MOB (~radek@gprs9211.eurotel.cz) joined #haskell 02:46:14 --- quit: Radek_MOB (Client Quit) 02:48:02 --- join: BlitzNL (DonCo66@node-c-c143.a2000.nl) joined #haskell 02:51:57 --- join: Radek_MOB (~radek@gprs9211.eurotel.cz) joined #haskell 02:56:55 I have this large program almost completed implemented in Haskell (and Ruby and ocaml), and I was stuck there. 02:57:20 ocaml didn't force me to really do functional programming 02:58:21 that's the bad thing about ocaml :) 02:58:58 when I did the first pass from ruby, I implemented a lot of things as classes. Then I went back to see where I _really_ needed a class. Answer: nowhere. 02:59:39 there wasn't a single case where using a class was better. which surprised me a bit. 03:04:23 classes in single instances are irrelevent. 03:04:33 its reuse and interaction that make them useful. 03:05:12 you get exactly the same reuse from a properly constructed module. 03:05:24 The class is only better if dynamic dispatching is necessary 03:06:38 *shrug* 03:16:39 --- quit: ChoJin ("bye !!!") 03:31:41 I am working on my haskell skills lately and it seems that I run into the situation where I want to apply a function 'n' times and eventually apply it to the initial value. To accomplish this I wrote a function exec, but there must be a better way, right ;) 03:31:42 exec f i ini 03:31:42 | i > 0 = f (exec f (i-1) ini) 03:31:42 | otherwise = f ini 03:37:02 --- join: systems (systems@62.12.124.107) joined #haskell 03:37:12 --- part: systems left #haskell 03:46:44 --- part: Jerub left #haskell 03:47:39 --- join: hdaume (~hdaume@c-24-126-229-242.we.client2.attbi.com) joined #haskell 03:51:01 --- quit: hdaume (Client Quit) 03:57:19 --- nick: kani^^off -> kaninchen 04:07:45 BlitzNL: there isn't anything especially wrong with that. 04:07:52 You could do it this way: 04:08:07 no, forget it. 04:08:24 I was going to use a pattern, but that won't work with i > 0, only with i == 0 04:08:40 but wait 04:09:18 is f a function of one variable? 04:10:43 hmm.. what do you guys use to typeset haskell in latex? 04:11:51 opet: I'm not one of the guys who knows that. :) 04:12:33 seth_: f is a function 04:12:51 BlitzNL: right, with how many arguments? 04:12:59 BlitzNL: One? 04:13:17 no two 04:13:37 then I don't think the | otherwise = f ini is what you want. Did you intend a partial application there? 04:14:37 why not it works but it just looks like something that is in the prelude or something. I can paste the rest wait 04:14:54 BlitzNL: no reason why not, if that is what you meant to do. 04:14:55 BlitzNL: take a look at iterate 04:15:05 unbwt enc = exec (shiftenc enc) (length enc - 3) initcolumns 04:15:05 where 04:15:05 initcolumns = sort (shiftenc enc ((map (\c -> [c])) (sort enc))) 04:15:05 shiftenc enc rest = sort (zipWith (\a b -> a : b) enc rest) 04:15:13 dennis: mm ok i will 04:15:37 http://www.haskell.org/onlinereport/standard-prelude.html#$viterate 04:24:42 --- join: tWip (~nobody@ttarvain-dsl.oulu.fi) joined #haskell 04:30:56 dennis: using iterate works fine 'last (take (length n) (iterate (shiftenc enc) initcolums))', thnx 04:31:36 BlitzNL: it's not much easier then to write your own function, but at least it's a prelude function 04:32:10 dennis: yeah my goal is indeed to use as much prelude as possible 04:33:30 dennis: and it also turns out to be slightly more efficient ((3600 reductions, 7856 cells) against (3531 reductions, 7763 cells)) ;-) 04:34:51 BlitzNL: you could make your more efficent also I think. Also, this is in the interpreter, a compiler might take away the difference 04:35:28 BlitzNL: you don't have to pass around ini in each recursive call, maybe it would make it faster 04:35:45 dennis: don't spoil my moment ;) 04:36:06 BlitzNL: ehhh... last (take (length n) ...) sounds silly, why not head (drop (length n - 1) ...) ? 04:37:11 or use !! if one never use it with negative values 04:38:29 Marvin--, dennis: mmm 04:44:49 Marvin--: you're right, it much better thnx 05:07:21 * shapr boings happily 05:07:25 tmoertel: you called? 05:12:02 --- quit: jemfinch (Read error: 60 (Operation timed out)) 05:12:07 --- quit: lambdabot (Remote closed the connection) 05:14:21 --- join: lambdabot (~lambdabot@h224n1fls32o1017.telia.com) joined #haskell 05:15:51 shapr! ;) 05:16:02 y0 sjj 05:16:03 wassup? 05:16:15 doing a bit of work, hacking selectmodule.c .. you? 05:16:23 --- join: esap (~esap@212-246-110-78.adsl.tpo.fi) joined #haskell 05:16:27 doing a bit of paying work 05:16:35 *nod* 05:16:48 and playing with lambdabot in the bits of spare time 05:17:01 I know this is the wrong channel, but do you know much about hotshot? 05:19:10 never heard of it 05:19:40 what is it? 05:19:42 profiler module 05:20:35 oh, for python 05:28:24 * shapr wishes for a Haskell nntp module 05:29:33 no genie about? 05:30:38 I have a nice little art deco lamp here I could rub 05:30:49 though it might electrocute me if I'm not careful 05:34:31 I can do something like 'data Bla = ... deriving Show', but can I influence the way a datatype is pprinted ? 05:35:40 do the instance declaration yourself 05:37:02 opet: ok I'll look into it thnx 05:46:58 --- nick: kaninchen -> kani^^off 06:07:14 --- join: nerdlor (~hjass@Waterloo.fw.opentext.com) joined #haskell 06:08:34 --- join: elmex (~elmex@pandora.labs.dn-systems.de) joined #haskell 06:10:15 --- join: Vincenz (Vincenz@D5770839.kabel.telenet.be) joined #haskell 06:10:48 hey, i don't want to post this to the ghc-users-mailing list, maybe it's a stupid question, but when looking at the examples in the directory ghc/misc/examples/ (in the current ghc5.05.20030430 tarball) and trying to compile them, it comes out that it is missing readSocket 06:11:00 (compiled with ghc -o tet Main.hs -package net) 06:19:02 * shapr looks 06:19:24 hi elmex 06:19:25 it's using SocketPrim, which seems to be moved to Network.Socket 06:19:31 shapr: hi ;-)) 06:20:57 the name of your gateway box is cute 06:21:11 pandora.labs.dn-systems.de ? 06:21:22 it's just temporary... from my company 06:21:30 it should be elmex.x-paste.de ;-) 06:21:34 I call a computer a "box" so that would be "pandora's box" 06:21:42 ;-)) 06:22:12 btw. how experimental is the Concurrent stuff? 06:22:23 has it at least a bit matured? ;-) 06:22:25 works for me 06:22:29 okay 06:22:31 lambdabot is using concurrent 06:23:01 lambdabot: @hello 06:23:05 well *g* i just thinking about some irc-client too... but argh... it just sucks writing an irc-clinet in every new language i learn 06:23:22 * shapr pokes lambdabot 06:23:23 lambdabot: y0 06:23:35 oh, it's lagging 06:23:41 Hello world. 06:23:42 Sorry, I'm not a very smart bot yet, try "lambdabot: @listcommands" 06:23:47 does it understand ping? 06:24:01 oh, no, he doesn't know ctcp ;-) 06:24:04 I don't think so 06:24:08 you could add that ;-) 06:24:09 @fact source 06:24:09 http://sf.net/projects/haskell-libs/, http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/haskell-libs/libs/lambdabot/ 06:24:28 oh, I have to update the chess plugin 06:24:30 lambdabot: @quit 06:24:30 --- quit: lambdabot (Remote closed the connection) 06:26:45 hm, i guess, i will look for some other crap to create some code for, nooone ever needs 06:27:08 I need an nntp module 06:27:21 elmex: look at http://sf.net/projects/haskell-libs/ 06:27:23 uh, hm, i don't read much usenetnews 06:27:33 bunch of people getting together to write libraries for Haskell 06:28:31 i'm not much of a haskell/functional programmer (yet) 06:28:57 that sourceforge projects is sort of a cvs-wiki for this channel 06:29:05 and for haskell developers in general 06:29:18 cvs-wiki? everyone throws code in? 06:29:30 something like that, yah 06:29:49 right now we have a web-server, an irc 'bot, chess playing code, some crypto code, some email code 06:29:58 exact real number code 06:30:02 lots of interesting bits 06:30:24 you need a ircd in haskell ;-))) 06:30:42 but i hate to write the routing-code for it. so i won't write any ;))) 06:31:25 lambdabot uses two threads, one for reading one for writing 06:31:35 I bet it would fun to extend it to bridging between networks 06:31:57 you mean, using a client as gateway? 06:32:03 yah 06:32:06 between two networks 06:32:16 uh, hm, why? 06:32:24 I wonder if there's a #haskell on oftc, EFNet, etc 06:32:34 just for the fun of writing code :-) 06:32:44 ;-)))) well, yes 06:32:45 it's virtually empty on efnet 06:32:47 --- join: lambdabot (~lambdabot@h224n1fls32o1017.telia.com) joined #haskell 06:32:55 @listmodules 06:32:55 I have the following modules installed: ["chess","dict","eval","fact","fortune","hello","karma","more","searchml","state","system","topic","type","yow"] 06:33:05 and only has ibid in on oftc 06:33:20 i wrote a quick and dirty client with my knowledge of haskell yet, here is the main routine: 06:33:29 do args <- getArgs 06:33:30 case args of 06:33:30 [] -> putStr "Usage: client \n" 06:33:30 x -> do putStr $ "Connecting " ++ (head args) ++ " ...\n" 06:33:30 conn $ head args 06:33:37 wait 06:33:53 the cool thing, i love are those forkIO 06:33:54 forkIO $ readLines sock 06:33:54 forkIO $ sendLines sock 0 06:33:54 readConsole sock 06:34:14 elmex: lambdabot will stretch your brain 06:34:27 at some points this is MUCH easier in haskell than in Perl 06:34:46 @board 06:34:46 r n b q k b n r 06:34:46 p p p p p p p p 06:34:46 - - - - - - - - 06:34:46 - - - - - - - - 06:34:46 - - - - - - - - 06:34:47 - - - - - - - - 06:34:49 P P P P P P P P 06:34:51 R N B Q K B N R 06:34:57 stretch my brain? 06:35:17 ohmygodachessinircchannelpleasenot 06:35:21 yah, lambdabot sources range in difficulty from the simplicity of writing plugins to the insanity of layered monad transformers 06:35:55 monad transformers? 06:36:19 you know about monads? 06:36:44 * tmoertel makes steel-cut oats for breakfast 06:36:45 well, yes, i know how to do some stuff with it and that >> >>= stuff. and do and all that. 06:37:02 right, so you know that "IO a" is the IO monad, yah? 06:37:11 good morning tmoertel! 06:37:21 'morning, shapr! 06:37:38 tmoertel: I saw the new dict commands, spiffy! 06:37:45 shapr: well.. uh.. hm. not really 06:37:55 i'm still a bit confused about that type-stuff 06:37:59 @prelude monad 06:38:01 *** "monad" prelude "Haskell Standard Prelude": text follows 06:38:01 Monad 06:38:01 class Monad m where 06:38:01 return :: a -> m a 06:38:02 (>>=) :: m a -> (a -> m b) -> m b 06:38:04 (>>) :: m a -> m b -> m b 06:38:06 fail :: String -> m a 06:38:09 -- Minimal complete definition: (>>=), return 06:38:14 @more 06:38:15 tmoertel: whoa, where'd that come from? 06:38:15 p >> q = p >>= \ _ -> q 06:38:15 fail s = error s 06:38:18 [ DONE ] 06:38:21 @dict-help 06:38:23 I perform dictionary lookups via the following 13 commands: 06:38:23 @all-dicts .. Query all databases on dict.org 06:38:24 @devils ..... The Devil's Dictionary 06:38:26 @easton ..... Easton's 1897 Bible Dictionary 06:38:28 @elements ... Elements database 06:38:30 @foldoc ..... The Free On-line Dictionary of Computing 06:38:32 @gazetteer .. U.S. Gazetteer (1990) 06:38:34 @hitchcock .. Hitchcock's Bible Names Dictionary (late 1800's) 06:38:35 oy 06:38:36 @jargon ..... Jargon File 06:38:41 @more 06:38:41 @prelude .... Haskell Standard Prelude 06:38:41 @vera ....... V.E.R.A.: Virtual Entity of Relevant Acronyms 06:38:42 @web1913 .... Webster's Revised Unabridged Dictionary (1913) 06:38:44 aha 06:38:44 @wn ......... WordNet (r) 1.7 06:38:45 there it is! 06:38:46 @world02 .... CIA World Factbook 2002 06:38:59 wow, that's handy 06:39:21 that was kind of the point of the dict module 06:39:27 :-) 06:39:29 the other dictionaries are icing on the cake, so to say 06:39:57 I'd like to write a haddock module, but I rely on ghci to play with the code I've written, and FastMutInt.hs kills ghci 06:40:41 I had to tease the definitions out of the Prelude code in Hugs 06:41:01 I hacked -- and I *do* mean hacked -- together a bit of Perl to do it 06:41:04 have you seen fit.c2.com ? 06:41:10 * tmoertel looks 06:41:31 lets you use a wiki to input unit tests 06:42:11 interesting 06:43:20 I have this idea for some combination of haddock and hws-wp to allow you to input unit tests and see the results immediately on the haddock docs 06:44:55 any thoughts? 06:44:59 So you specify a category for test cases in your source code? 06:45:14 And haddock/hws-wp/shapr-thing then grabs all the cases in that 06:45:26 category, runs them, and places the results into the commentary for 06:45:31 your code in the haddock docs? 06:46:17 yup 06:47:16 shapr: the type IO a means i have to use this action stuff >> or do? 06:47:20 of course, haddocks could include both quickcheck and hunit tests 06:47:32 elmex: yah, sort of 06:47:58 shapr: hmm, ok... well 06:48:13 elmex: thing is, there's some invisible state being passed along with the >> or >>= calls 06:48:40 tmoertel: one advantage would be the ability to write demo code that is both unit test and demo 06:48:49 tmoertel: what do you think? would it be useful/interesting? 06:49:07 shapr: ah, ok 06:49:27 elmex: http://www.dcs.gla.ac.uk/~nww/Monad.html 06:49:31 it would be useful in that the results of test cases would be visible to users of the docs 06:50:25 might make it easier to get unit tests for ghc stuff also 06:50:36 shapr: ah, okay. thanks. will read that... once i got out of this flirt-chat here ;-)) 06:50:42 (not #haskell) 06:50:45 * shapr laughs 06:50:46 aww 06:51:54 there is no other way for a geek to find contact to girls 06:57:11 'He, 18, searchs a girl for sex and exchanging haskell sources.' 06:57:37 well, my fiancee just fixed her mandrake installation today 06:57:43 and I wasn't able to do that yesterday ;-) 07:04:15 mandrake... uh 07:04:32 she got tired of having to ask me to fix her debian installation. 07:04:43 she wanted one she could install and admin by herself 07:04:44 whats bad about debia? 07:04:45 an 07:04:50 uh, ok 07:09:43 --- part: nerdlor left #haskell 07:18:16 --- quit: BlitzNL (Read error: 60 (Operation timed out)) 07:18:47 haskell r0xx0rs! 07:19:53 hm 07:20:01 ? 07:20:02 h42k311 07:20:09 whoa 07:20:10 ah, no 07:20:12 h45k311 07:20:17 that too 1337 for me 07:21:52 hm, my post-code is totally eleet 07:22:08 it's 31137 07:22:24 post code? 07:22:29 power on self test? 07:22:44 hm, no i mean for the mail-delivery (real mail ;) 07:22:50 it's germany 07:22:59 i don't know how to translate it 07:23:31 oh 07:23:48 for the US it's zip code, we use post code in the UK 07:23:49 well, post is the correct british way to say it 07:24:25 yes, postcode, or zipcode 07:32:18 --- nick: kani^^off -> kaninchen 07:34:21 --- quit: o3 (sterling.freenode.net irc.freenode.net) 07:34:40 * shapr bounces happily 07:37:00 anyone here knows whether MS Access supports "select * from bar where bar.foo like '%blah%';" ? 07:37:32 I've been unable to figure out a valid Jet 4.0 query that uses "like" 07:37:36 --- join: o3 (andrep@c17736.belrs2.nsw.optusnet.com.au) joined #haskell 07:41:33 shapr: is there no grammar for valid queries? 07:41:37 (hi btw) 07:41:40 hi 07:41:42 I don't know 07:41:58 hmm. got no docs? 07:42:01 I don't own a copy of ms access, but the mdac components come with the Jet 4.0 engine 07:42:15 what's jet 4.0? 07:42:25 the actual engine that ms access uses 07:42:32 ah 07:43:50 a client asked me to do some MS Access work, and I don't feel like putting more money into Bill's pocket. 07:45:39 hmm. I only found a 3.0 quickref which doesn't seem to completely describe where clauses. 07:46:26 I found an intro that shows an example with one single percent wildcard 07:46:33 maybe it just doesn't like me using two wildcards 07:47:23 hmm 07:54:14 --- quit: o3 (sterling.freenode.net irc.freenode.net) 07:54:17 * shapr wishes for a working HaskellDB 07:54:56 --- join: o3 (andrep@c17736.belrs2.nsw.optusnet.com.au) joined #haskell 07:55:13 y03 07:55:37 --- quit: kawfee (Read error: 110 (Connection timed out)) 08:03:10 * shapr boings 08:03:16 @yow 08:03:16 My LIBRARY CARD expired... 08:03:22 hi dnm 08:11:40 <-- graduate! 08:11:45 * Igloo bounces 08:12:25 * Smerdyakov has to take 4 final exams before graduating =( 08:12:31 Igloo: you graduated? 08:13:12 Yup 08:13:19 So. Does anyone want to recommend interesting summer jobs that I could still arrange? 08:13:31 I can travel anywhere, especially if you pay ;-) 08:14:07 Igloo: cool, congratulations 08:14:18 Igloo: congraduations! 08:14:51 shapr: *groan* 08:15:09 :-) 08:21:32 --- quit: o3 (sterling.freenode.net irc.freenode.net) 08:22:36 --- join: o3 (andrep@c17736.belrs2.nsw.optusnet.com.au) joined #haskell 08:28:12 --- quit: o3 (sterling.freenode.net irc.freenode.net) 08:28:35 * shapr boings 08:31:02 --- join: o3 (andrep@c17736.belrs2.nsw.optusnet.com.au) joined #haskell 08:37:45 * Segora bum-tchaks. 08:38:21 * shapr shebangs 08:45:12 hrm, how does UNIQUE fit into a SQL query? 08:45:15 I've forgotten :-/ 08:46:00 I don't think it has to do with queries, but rather with table definitions. 08:46:38 I thought I could "select unique foo from bar" 08:46:47 Do you mean DISTINCT? 08:46:53 ohhh maybe so 08:47:34 Smerdyakov: thanks! 08:47:41 that was a braino on my part 08:47:48 Draino in the braino? 08:48:46 yah, something like that 08:50:17 --- join: jak (~jak97@holt.doc.ic.ac.uk) joined #haskell 08:50:22 hi jak 08:50:30 yo shapr 08:50:34 wassup? 08:50:37 lol 08:50:45 trying to build a debian kernel atm 08:50:55 but I am sorely confused by --initrd option to make-kpkg 08:51:54 what you up to? 08:54:21 I've never used the initrd option with make-kpkg 08:54:29 but I do like make-kpkg lots 08:55:15 --- quit: Darius (Read error: 104 (Connection reset by peer)) 08:56:17 --- quit: o3 (sterling.freenode.net irc.freenode.net) 08:57:28 --- join: o3 (andrep@c17736.belrs2.nsw.optusnet.com.au) joined #haskell 08:57:28 --- quit: o3 (sterling.freenode.net irc.freenode.net) 08:59:08 --- join: o3 (andrep@c17736.belrs2.nsw.optusnet.com.au) joined #haskell 09:02:38 the hawiki is pretty cool 09:02:59 --- quit: archie_sihir (Remote closed the connection) 09:04:10 yah it is 09:04:22 btw. is there some mysql binding for haskell? 09:04:35 @fact htoolkit 09:04:35 http://sf.net/projects/htoolkit, postgresql mysql and odbc bindings 09:04:49 the @fact plugin is using the postgresql binding 09:05:22