00:00:00 --- log: started haskell/02.03.23 03:43:55 --- quit: shapr (Remote closed the connection) 03:45:27 --- join: ChoJin (~ask@cha213245038031.chello.fr) joined #haskell 03:57:07 --- join: Yurik (~yrashk@gw.telcos.net.ua) joined #haskell 03:58:08 --- join: Pseudonym (~ajb@caffeine.alicorna.net) joined #haskell 04:47:27 --- topic: set to 'Pixar sues Exluna for patent infringement, film at 11' by Pseudonym 04:47:35 Shoops. 04:47:38 Whoops, I mean. 04:47:56 * Pseudonym really needs to get a better IRC client 04:48:06 do you know what the old topic was? 04:48:18 Nope, sorry. 04:48:37 cos it had the URL of the logs and stuff 04:48:53 It'll be in the logs, won't it. 04:49:04 I guess, but where are they? 04:50:05 Someone should add the log location to http://haskell.org/wiki/wiki?HaskellIrcChannel 04:50:21 That way we won't forget. :-) 04:51:01 Found 'em. 04:51:17 --- topic: set to 'Logs are at http://tunes.org/~nef/logs/haskell/' by Pseudonym 04:53:01 --- topic: set to 'Have a library you'd like to donate, or one you'd like to see? Talk to us at http://sf.net/projects/haskell-libs/ | we be loggin' http://tunes.org/~nef/logs/haskell/ | welcome to the Haskell Dojo | julien is up to white belt! jewel and shapr are up to green belt!' by Pseudonym 04:53:13 most of it's kind of out of date anyway 04:53:20 True. 04:53:25 but the logs and the libraries project are useful, I thikn 04:53:30 * Pseudonym nods 04:55:02 How recent is haskell-libs? 04:57:26 --- quit: clausen ("send all money to /dev/clausen") 05:25:58 --- join: shapr (~user@p-c2fbab7e.easy.inet.fi) joined #haskell 05:33:04 /leave #c++ 05:33:08 Whoops. 05:33:21 * Pseudonym isn't doing too well with the IRC client thingy today 05:34:42 I have this bad feeling that Haskell I/O is going to need to be revamped to support sockets properly. 05:35:03 * Pseudonym was just looking through ways that other declarative languages have done it 05:36:43 Systems which added network I/O as an afterthought either needed to be redesigned a bit or are gross hacks. 05:37:56 I think in Haskell, the Handle type would need to be turned into a type class. 05:42:01 Why? 05:42:07 And what would be instances of the class? 05:43:45 Well... Handle isn't extensible. You can't add new kinds of file handle easily. 05:44:08 Ideally, a socket library should not depend on the details of how IO is currently implemented. 05:44:23 I think. 05:44:35 What would this gain the user of the language? 05:44:51 The ability to add new "things" which could be accessed as files. 05:44:59 Like network connections. 05:45:17 Or IPC. 05:45:45 Maybe even file handle filters, like in Java. 05:56:43 It's only important if you would want a file handle for something you could write in Haskell. Admittedly filters may be such a thing 05:57:08 It's also important if you don't want to have to rewrite it for every Haskell implementation. 05:57:11 But you could always use Strings or something for the abstraction 05:58:31 I don't get what you mean about rewriting it 05:58:37 Not that there are that many implementations around at the moment... 05:58:58 Well, at the moment, implementing a socket as a Handle means getting into the implementation of Handle, which is, at the moment, implementation-defined. 05:59:26 If there was a nice way to map between Handles and POSIX fds, that would be sufficient, I think. 05:59:41 But I don't think you can even assume that Haskell is running under POSIX. 06:00:17 So you need a way to do it that doesn't depend on how the Haskell implementation defines the Handle type. 06:00:21 Implementations would still have to write their own sockets stuff, they'd just make the type an instance of a class rather than having a tuple identifying it as a socket rather than normal file handle 06:00:46 That's okay for sockets, but what about the next stream-like abstraction? 06:01:20 Unless you are talking about things you can implement in Haskell code it doesn't make a difference 06:01:32 I can think of a few such things. 06:02:06 I said I agreed in that case up a page or so 06:02:19 Yeah, I know that. :-) 06:02:43 Filters are one such case. Others are devices which don't present to the OS as normal files. 06:03:26 Unix is a bit spoiled in this regard. :-) 06:04:04 It would also, I think, help with different kinds of streams, like those that are seekable and those which aren't. 06:05:30 hGetPosn :: (SeekableHandle h) => h -> IO HandlePosn 06:05:34 ...and so on 06:06:20 *nod* 06:06:50 It might also be good to differentiate between text and binary streams. 06:06:58 I think I'd rather see efforts focussed on efficient implementation of IO personally 06:07:03 Well true. 06:07:27 Haskell at the moment is highly optimised for the efficient implementation of Haskell compilers. 06:22:41 in what way? 06:24:02 Let me put it this way: Do you know of an I/O-bound Haskell application? 06:26:36 I have an I/O-bound shell script that I rewrote in Haskell :-) 06:26:44 :-) 06:27:34 I'm not saying it can't be done, but I certainly don't know of any "real" (and I do use that word advisedly) Haskell applications which mostly do I/O. 06:27:59 All the ones I know of mostly do complex structure manipulation. 06:33:36 You know a language has taken off when it has a Comprehensive Archive Network. 06:35:56 Hmmm, I only know of 2 OTTOMH, and neither of them are C 06:37:08 I think it's impossible for any single C archive to win the title of "comprehensive", somehow. 06:38:01 I didn't mean ubiquitous languages, obviously. 06:40:10 * Pseudonym thinks of it as a sufficient rather than a necessary condition 06:41:53 --- quit: Pseudonym ("using sirc version 2.211+ssfe") 07:02:23 --- quit: ChoJin (Remote closed the connection) 07:20:41 --- quit: shapr (Remote closed the connection) 07:29:02 --- join: ChoJin (~ask@cha213245038031.chello.fr) joined #haskell 07:55:36 --- quit: ChoJin (Remote closed the connection) 07:59:00 --- join: ChoJin (~ask@cha213245038031.chello.fr) joined #haskell 07:59:53 --- join: jadrian (~none@195-23-147-15.nr.ip.pt) joined #haskell 08:00:01 --- part: jadrian left #haskell 08:22:24 * Yurik is away: just for few minutes, will be back shortly 08:27:39 * Yurik is away: I'm busy 08:47:26 * Yurik is away: home, beer, sleep 08:48:24 --- quit: ChoJin ("bye !!!") 09:31:57 --- join: zorb (~nobody@66-90-218-146.grandecom.net) joined #haskell 09:32:55 --- topic: set to '' by zorb 09:33:01 doh 09:33:08 what is it with today? 09:33:22 ? 09:33:35 I haven't been here in a while. 09:33:37 you're the second person to mess up the topic today 09:33:46 after it survived for months completely unchanged 09:33:58 did anybody have op status? 09:34:04 channel isn't +t 09:34:22 yeah. 09:34:34 (was it?) 09:34:34 --- topic: set to 'Have a library you'd like to donate, or one you'd like to see? Talk to us at http://sf.net/projects/haskell-libs/ | we be loggin' http://tunes.org/~nef/logs/haskell/ | welcome to the Haskell Dojo | julien is up to white belt! jewel and shapr are up to green belt!' by Heffalump 09:34:39 dunno 09:34:42 hi, anyway :-) 09:34:56 heh, hi 09:35:24 * zorb checks the recent channel logs 09:35:50 ooh, nice. 10:00:54 oye. 10:01:09 I need to get back to working on Strictness FAQ 10:02:19 anybody here want to comment on things that once confused them? 10:02:24 (if anybody is available. 10:04:58 ah, for future reference, if anybody wants to know where the logs are, just /whois clog 10:27:38 --- quit: xbill ("Leaving") 10:39:09 --- join: xbill (wli@DOMINIA.MIT.EDU) joined #haskell 11:02:00 brb 11:02:02 --- quit: zorb () 11:28:52 --- join: tmoertel (~chatzilla@pa-mtlebanon2a-68.pit.adelphia.net) joined #haskell 14:02:29 --- join: clausen (~andrew@c17997.eburwd3.vic.optusnet.com.au) joined #haskell 14:26:28 --- join: gene9 (dna@194.158.216.190) joined #haskell 14:52:44 --- quit: gene9 ("b") 15:25:58 --- join: ChoJin (~ask@cha213245038031.chello.fr) joined #haskell 16:21:38 --- quit: ChoJin ("bye !!!") 16:34:44 --- join: gene9 (dna@194.158.216.190) joined #haskell 16:43:01 --- quit: gene9 ("b") 18:19:04 --- join: zorb (~nobody@66-90-218-103.grandecom.net) joined #haskell 20:01:40 --- quit: clausen ("send all money to /dev/clausen") 20:37:28 --- join: Pseudonym (~ajb@caffeine.alicorna.net) joined #haskell 22:04:04 --- quit: zorb ("Changing server...") 22:04:16 --- join: nrut (~nobody@66-90-218-92.grandecom.net) joined #haskell 22:05:53 --- nick: nrut -> zorb 22:14:03 --- quit: Pseudonym ("using sirc version 2.211+ssfe") 22:31:13 --- quit: smkl (Remote closed the connection) 22:41:22 --- join: smkl (~sami@glubimox.yok.utu.fi) joined #haskell 23:15:09 ick! must sleep 23:15:16 --- quit: zorb ("KVirc 2.1.0 'Dark Star'") 23:36:52 --- quit: smkl (Remote closed the connection) 23:59:59 --- log: ended haskell/02.03.23