00:00:10 lol 00:02:59 Play more Gorf! 00:03:01 -!- cpressey has left (?). 00:05:39 -!- jpc has joined. 00:06:06 -!- jpc has quit (Read error: 104 (Connection reset by peer)). 00:06:21 pikhq: until recently I was unaware as well 00:13:12 The only problem with my C lambda stuff is that it is *amazingly* hard to debug. 00:14:50 SPEAKING OF, 00:14:57 ? 00:14:58 pikhq: Did you see my improvement to the Plof parser's debugging? 00:15:06 -!- jpc has joined. 00:15:19 Gregor: I only saw that there was such a commit; I've not actually looked at the details. 00:17:11 oh, btw, is Lisp/LC in C++ templates a common thing that almost everyone has done, or is it just me? 00:17:35 (wondering if I should bother putting my stuff somewhere) 00:24:39 pikhq: what lang is that code in? 00:24:43 the $2 = ... 00:27:01 cheateur: That's GDB output. 00:27:40 hmm 00:28:20 oh ok. 00:34:21 -!- MissPiggy has quit. 00:36:23 -!- coppro has quit (Remote closed the connection). 00:36:49 -!- coppro has joined. 00:50:46 -!- coppro has quit (Remote closed the connection). 01:17:46 "Perl: the only language that looks the same before and after RSA encryption" 01:17:50 http://qntm.org/?msn 01:27:20 -!- MizardX has quit (Read error: 110 (Connection timed out)). 01:39:22 -!- coppro has joined. 02:13:47 -!- oerjan has quit ("Dizzy..."). 02:31:07 -!- Pthing has joined. 02:46:49 http://codu.org/music/op10/GRegor-op10.pdf wooh 02:47:15 C and large church numerals do not mix. 02:47:21 (go go gadget stack overflow!) 02:50:55 What are you using to typeset that? 02:56:19 Gregor: ? 02:56:31 Lilypond 02:56:37 Rosegarden to generate the Lilypond 02:56:45 (Although I then modify it by hand) 02:58:25 (Lilypond, if you don't know, is TeX for musical notation) 02:59:07 pikhq: ^^ 02:59:54 Huh. It appears to have a few issues... For example, measure 88; see the (I dunno what it's called, flag?) on the eighth and half notes in the treble clef? 03:00:41 Are you referring to the stems? 03:00:51 Argh. Yes, the stems. 03:01:00 It displayed two stems because I have a chord of dotted quarter notes overlayed with a dotted half note. 03:01:04 Don't know how to fix that ... 03:01:06 Darn tip-of-tongue phenonmenon. 03:01:50 It's also kinda odd-looking in measure 77 (half note and dotted half note) 03:02:03 -!- oerjan has joined. 03:02:04 I could've sworn Lilypond was smarter than that. 03:03:00 Heh, that's just borken :) 03:03:41 I'll fix that now. 03:03:44 That's an overlay problem again. 03:03:57 Maybe you could get the Lilypond developers to use Opus 10 for the stuff to fix in Lilypond? :P 03:08:38 Nah, they'd just tell me what the "right way" is :P 03:09:14 Which would also be useful. :P 03:12:23 OK, 77 is fixed. 03:12:30 Still have the multiple-stems problem. 03:14:11 -!- bsmntbombdood__ has joined. 03:16:55 Also, a rather large number of eighth notes appear to be detached from their stems... That's just plain odd. 03:17:24 -!- bsmntbombdood_ has quit (Read error: 101 (Network is unreachable)). 03:19:11 -!- jpc1 has joined. 03:21:02 even so 03:22:36 pikhq: ? 03:23:03 I don't know where the detached eighth notes you're referring to are. 03:23:45 Gregor: Only slightly so. 03:23:49 See measure 1. 03:23:52 Note 1. 03:23:56 -!- jpc has quit (Read error: 110 (Connection timed out)). 03:24:14 Even when I zoom in as far as I can, they're not detached for me. 03:24:33 Might be my PDF viewer being odd, then. 03:24:54 It's apparent even when not zoomed in here... 03:25:05 -!- bsmntbombdood_ has joined. 03:25:16 Hrm. The lines also have bizarre width... 03:25:21 I'm going with "PDF viewer bug". 03:28:46 -!- bsmntbombdood__ has quit (Read error: 54 (Connection reset by peer)). 03:34:43 -!- jpc1 has quit ("I will do anything (almost) for a new router."). 03:37:59 -!- jpc has joined. 03:47:55 in my dream, there was this death metal song, crucial to some sort of revolution, that said "space stations / given as nations"; where the latter verse actually means "taken as a given", but i only realized it was wrong after i woke up. oh and there was some spinning too. 03:48:28 (i always spin in my dreams) 04:05:16 -!- bsmntbombdood__ has joined. 04:06:49 -!- bsmntbombdood_ has quit (Read error: 113 (No route to host)). 04:25:51 spinning revolutions 04:34:06 :) 04:34:08 morninn 04:35:58 -!- bsmntbombdood__ has quit (Read error: 101 (Network is unreachable)). 04:39:37 morn' 05:32:11 "using events, by definition, even on a single threaded process, is multithreading" 05:35:01 "redefining terms, by definition, even when obnoxiously done, is genocide" 05:35:05 -!- jpc has quit (Read error: 110 (Connection timed out)). 05:37:50 -!- bsmntbombdood has joined. 05:39:57 Wooh! 05:40:04 I taught someone something new today! 05:40:14 (Had him try an experiment that proved that he was wrong) 05:44:45 -!- oerjan has quit ("Good something"). 05:45:46 -!- jpc has joined. 05:47:59 good for you 05:48:13 owned that bastard 05:49:22 -!- Ayeraw has joined. 06:09:18 -!- jpc1 has joined. 06:10:43 -!- jpc has quit (Read error: 104 (Connection reset by peer)). 06:18:21 -!- augur_ has joined. 06:18:32 -!- augur has quit (Read error: 104 (Connection reset by peer)). 06:31:24 -!- jpc1 has quit ("I will do anything (almost) for a new router."). 06:32:01 -!- jpc has joined. 06:34:55 -!- jpc has quit (Client Quit). 06:35:32 -!- jpc has joined. 06:43:58 -!- augur_ has quit (Read error: 60 (Operation timed out)). 06:55:49 -!- FireFly has joined. 07:20:39 I love using an OS with a scheduler 07:21:39 I take it you're not on linux then :) 07:28:01 I am 07:28:05 :P 07:50:12 -!- augur has joined. 07:59:59 -!- clog has quit (ended). 08:00:00 -!- clog has joined. 08:14:51 -!- FireFly has quit ("Leaving"). 08:36:53 -!- jpc has quit (Read error: 104 (Connection reset by peer)). 08:37:34 -!- jpc has joined. 08:56:12 -!- jpc has quit ("I will do anything (almost) for a new router."). 09:07:33 -!- scarf has joined. 09:22:14 -!- Pthing has quit (Remote closed the connection). 09:32:43 hey guys 09:32:59 hi 09:33:05 someone set a ban on the ident that i'm using on #haskell 09:33:11 i'm scared 09:34:17 get a ghost 09:34:27 or whatever they call it 09:34:35 cloak, that's it 09:34:47 does the cloak change the ident tho? 09:34:53 I have a cloak 09:35:01 scarf is n=scarf@unaffiliated/ais523 09:42:55 but.. is your ident normally scarf? 09:43:24 cheater: oh, my IRC client lets me change it 09:43:39 the ident isn't checked in any way, you can put whatever you like there 09:44:15 well, there might be a ban on it; that's unlikely though 09:44:32 there's a ban on it in #haskell :S 09:44:56 What's the banmask? 09:45:54 -!- pikhq has quit (Read error: 104 (Connection reset by peer)). 09:47:57 The ident *is* checked, by doing an ident protocol request (see RFC1413); of course, if you control the machine you're connecting from, or if it doesn't run an identd, or if the identd allows for user-specific replies anyhow, then you can put anything there. 09:49:10 actually, Freenode seems to try to check it 09:49:19 but an identd request wouldn't get past the firewall here 09:49:48 Okay, so that's one more way to avoid the check. 09:50:27 If identd doesn't work or is blocked, one normally gets 'user' as user part. 09:50:59 One normally gets whatever one's IRC client specifies in the initial USER message at the beginning of the connection. 09:51:15 Though here with a "n=" prefix. And in many other places, with a ~ prefix. 09:51:48 I think Freenode's new server, the one they're going to switch to soonishly, also does the ~ prefix instead of the current n=foo/i=foo thing. 09:54:10 To be even more exact, the traditional IRCnet ircd has six different cases; no prefix for I-line (normal connection) with ident, ^ for ident with type "OTHER", ~ for no ident; and correspondingly +, =, - for i-lines (restricted connections). 09:54:49 Don't know if Freenode's seven will do all those. 09:58:03 Apparently it does just the ~ prefix. And even that's a configurable option. 09:59:37 Freenode's policy document or some-such used to justify the n=/i= prefixes so that you can set a ban on "?=bar", instead of having to use two bans ("bar" and "~bar") or "*bar" (which'd match "foobar" too). 10:07:24 -!- pikhq has joined. 10:08:36 -!- Pthing has joined. 10:23:11 -!- coppro has quit ("I am leaving. You are about to explode."). 10:41:12 -!- BeholdMyGlory has joined. 10:41:15 Ilari: it's *!gast@*.* can you do something about it? 10:44:27 cheater: Block identd? 10:46:16 Based on the fact that his current username has the "n=" prefix (as opposed to "i="), it's already blocked; it's just a matter of setting the client to send something else. 10:46:35 In general ban evasion is a good way of obtaining even more bans, though. 10:47:37 yeah 10:47:51 stupid xchat needs me to actually install ident -_- 10:48:01 * cheater is trawling through the manual. 10:48:29 fizzie: i talked to one of the ops and he said 'it's an old ban, i don't know why it is set, i am not going to touch it' 10:48:43 so, it's not a ban against .me. 10:49:19 I thought blocking/disabling identd causes the lhs to be 'n=user@' (if you aren't identified). 10:49:44 that's the thing, i am identified 10:49:49 nope, it uses whatever you put in the USER line 10:49:55 identification confirms the nick 10:49:58 it has nothing to do with the ident 10:51:01 ah. any idea how to change the user in xchat? 10:51:22 "User name" field in the network list/connect screen/whatever. 10:51:34 "network list" seems to be the dialog title. 10:51:58 there we go 10:52:05 -!- cheater has quit ("Verlassend"). 10:52:16 -!- cheater has joined. 10:53:25 thank you fizzie 10:55:13 Ah, right, that *!*Gast*@* ban might be related to what's being talked at in http://irclogs.ubuntu.com/2008/05/15/%23ubuntu-ops.txt -- just search for "gast". 10:56:04 Banning "*gast*" is a bit overkill; I would think it matches a number of perfectly reasonable words, for example. 10:56:31 Freenode's +d "realname bans" are funky. 10:57:51 but it is not a realname ban, only an ident ban 11:00:13 Yes, that was just an aside; the log I linked to considered a "Java user" realname ban. 11:01:18 There seems to be some sort of a general rule that channel #x has a huge list of bans for all x where x is the name of a Linux distribution. 11:04:46 ah ok 11:20:03 -!- MigoMipo has joined. 11:21:08 -!- MizardX has joined. 11:29:00 -!- scarf has changed nick to scarf|away. 11:36:30 -!- scarf|away has changed nick to scarf. 11:39:19 -!- Pthing has quit (Remote closed the connection). 12:17:40 -!- Pthing has joined. 12:36:20 anyone know why in haskell [0.1, 0.3 .. 1] turns out to contain 1.099999...? 12:58:41 -!- bsmntbombdood has quit (Read error: 113 (No route to host)). 13:00:10 -!- bsmntbombdood has joined. 13:05:10 -!- scarf has changed nick to scarf|away. 13:09:26 cheater: http://docs.sun.com/source/806-3568/ncg_goldberg.html 13:12:19 Deewiant: non-helpful 13:12:37 Or did you mean that it's greater than 1? enumFromThenTo works like that, it goes up to the last thing that's >= the end point 13:13:09 Or was this just an artifact of floating points being Enum, I forget. 13:13:19 according to the haskell98 report, it should end when the next item would be higher than a3 13:15:44 Actually, it does a takeWhile (<= a3 + (a2-a1)/2) 13:16:24 http://www.haskell.org/onlinereport/standard-prelude.html 13:18:21 so this is an implementation bug 13:18:22 right? 13:18:44 No 13:19:26 takeWhile (<= a3 + (a2-a1)/2) i.e. takeWhile (<= 1 + (0.3-0.1)/2) i.e. takeWhile (<= 1.1) 13:19:26 it is, because the implementation does not do what haskell 98 defines 13:19:40 The above is what Haskell 98 defines 13:19:47 Copied from the link I pasted above 13:20:01 (With different parameter names since you already used a3) 13:20:01 can you find the exact wording of the haskell 98 definition? 13:20:16 numericEnumFromThenTo :: (Fractional a, Ord a) => a -> a -> a -> [a] 13:20:16 numericEnumFrom = iterate (+1) 13:20:16 numericEnumFromThen n m = iterate (+(m-n)) n 13:20:16 numericEnumFromTo n m = takeWhile (<= m+1/2) (numericEnumFrom n) 13:20:16 numericEnumFromThenTo n n' m = takeWhile p (numericEnumFromThen n n') 13:20:18 where 13:20:20 p | n' >= n = (<= m + (n'-n)/2) 13:20:23 | otherwise = (>= m + (n'-n)/2) 13:20:46 And relating to the expression itself: "Arithmetic sequences satisfy these identities: .. [ e1,e2..e3 ] = enumFromThenTo e1 e2 e3" 13:21:05 (From http://www.haskell.org/onlinereport/exps.html#sect3.10 ) 13:21:28 according to http://www.haskell.org/onlinereport/derived.html enumFromThenTo x y z = map toEnum [fromEnum x, fromEnum y .. fromEnum z] 13:22:03 cheater: For derived instances of Enum 13:22:14 Not all instances 13:22:26 i am not sure what derived instances are and why they are different 13:22:55 Have you been reading the report in order? :-P 13:23:02 of course not 13:23:06 Top of that page: "A derived instance is an instance declaration that is generated automatically in conjunction with a data or newtype declaration." 13:23:26 I can do "data MyBoolean = MyTrue | MyFalse deriving (Enum)" 13:23:49 And enumFromThenTo would then work as you described. 13:24:17 ok 13:25:52 thanks Deewiant 13:25:56 that was quite helpful 13:26:35 As to the deeper "why" of why it's defined like that for Float and Double, I can't say. 13:26:55 i think it's fairly dumb 13:27:38 Oh, right 13:27:42 -!- MigoMipo has quit ("Page closed"). 13:27:43 toEnum truncates to Int 13:28:05 not sure what that gives us 13:28:13 So you need to take more elements than necessary because you've lost info at that point 13:28:38 "Technology overview: Cognitive radio". Wow, that's so buzzword-compliant; I have no idea what it actually means. 13:28:43 Except that that doesn't make any sense because there's no need to use toEnum anywhere 13:28:43 isn't that quite shit, Deewiant 13:28:52 cheater: The "shit" is that Float and Double are instances of Enum 13:29:04 That, in itself, is shit 13:29:16 TBH I don't care about the details of how their implementation is shit :-P 13:29:20 i don't know haskell enough to fully follow that through but ok 13:35:10 "The demo presents the use of micro-array LEDs for displays embedded in contact lenses." It's things like these that at least partially alleviate my disappointment of it being 2010 already. 13:35:22 Things still suck and all, but a display in a contact lens. 13:36:44 cognitive radio is the stupidest name, yes 13:39:16 fizzie: Incidentally: is something other than Java acceptable for the AI tournament? 13:40:00 Deewiant: If you can compile it to JVM bytecode and make it work in the tournament system, yes. 13:40:06 -!- BeholdMyGlory has quit (Remote closed the connection). 13:40:13 Alright, cool. 13:40:42 I don't have my workstation online, but I think someone participated with Scala last year. 13:41:28 Presumably it's a handicap as anything non-Java tends to be slower on the JVM 13:42:08 That's possible, but I don't think I could find a fair way of mitigating that. 13:42:36 No, certainly not. I wouldn't expect you to. 13:42:58 Hm, last year's bot file sizes are all pretty tiny; perhaps it was the year before. 13:43:32 Can you tell what language it's in from the file size? :-) 13:44:17 Not in the general case; but the single non-Java participant I remember seeing was a megabyte or two, since they had to bundle the language's standard-library-equivalent in there. 13:44:40 Ah, yes, that can happen. 13:45:31 I may have had some sort of file size limit in place, mostly due to the Computing Centre quota. 13:45:59 You can request a quota-expansion, you know. You'd probably get it for this. 13:46:46 Probably, I just haven't bothered. It's not like there are more than one or two freaks, uh, I mean... differingly language-opinionated folks, in each year's list of participants. 13:46:56 :-) 13:48:11 Deewiant: Since you hang around nearby, I guess I could copy on the link: http://research.nokia.com/Otaniemi_demo_house 13:48:15 I'll translate CCBI to Clojure, write the thing in Befunge and bundle that along with the Clojure interp 13:48:41 Our speech-to-text demo is going to be there, though it's nothing too fancy. 13:49:36 The others are a bit more... flashy. Thought controlled games and so on. 13:50:11 My wednesdays are rather booked by the school as it is, I probably won't be there 13:50:31 Isn't your software freely available? ;-) 13:51:26 Not really, though I guess that's mostly a matter of not having enough resources to construct a distributable thing out of it. We have some components downloadable, like that VariKN language-modelling toolkit, and the Morfessor thing. 13:52:11 And of course it's just a trivial matter of implementation to reconstruct our system from the published articles about it. 13:52:17 Of course. 13:52:25 It's just code, after all. 13:54:00 I dislike the fact that the bot-submission CGI script for the AI competition says "my $deadline = ...", since it's not really *my* deadline. 13:54:16 So write "their $deadline = ..." instead. 13:55:17 I guess http://www.cis.hut.fi/projects/speech/demo/ is the closest equivalent to "freely available". 13:56:04 Yes, and that one requires an account too. Though you get that via email. 13:56:25 A more flashier online-demo -- perhaps even something running client-side -- has been talked about, but it's obviously not very high on the priority list. 13:57:13 Presumably you have something like http://www.cis.hut.fi/projects/speech/srdemo.jpg - just tarball it and torrent the presumably-multigigabyte language model :-P 13:58:40 Ooh, that's an old image; it looks better nowadays. 13:59:55 Anyway, yes, we have that thing; it's just not very user-friendly. I don't really know, software-distributionary policy decisions aren't really my department. 14:00:30 Since when is any research project of the university user-friendly anyway 14:03:49 News headlines talked about in the lunch table: "U.S. group sends solar-powered Bibles to Haiti -- Not any Bible. These are solar-powered audible Bibles that can broadcast the holy scriptures in Haitian Creole to 300 people at a time." 14:04:13 "Called the 'Proclaimer,' the audio Bible delivers 'digital quality' and is designed for 'poor and illiterate people,' the Faith Comes By Hearing group said." 14:04:20 That's an awesome name. 14:04:34 money well spent 14:10:00 -!- scarf|away has changed nick to scarf. 14:54:25 -!- Sgeo_ has joined. 14:54:32 -!- Sgeo_ has quit (Read error: 104 (Connection reset by peer)). 15:19:51 Surely it's better to let people die and go to heaven than let them live and have them end up in hell! 15:23:18 they're catholics, reading the bible at them isn't how you save catholics 15:23:21 ignorant americans 15:24:46 -!- Pthing has quit (Remote closed the connection). 15:33:53 -!- cpressey has joined. 15:55:52 My dad apparently believes that barackobama.com is a scam site 15:56:10 Because it doesn't end with .gov 15:57:10 -!- FireFly has joined. 16:17:54 -!- BeholdMyGlory has joined. 16:41:51 -!- Leonidas has quit (lindbohm.freenode.net irc.freenode.net). 16:42:01 -!- Leonidas has joined. 16:44:57 -!- MissPiggy has joined. 16:45:54 -!- augur has quit (Read error: 110 (Connection timed out)). 17:09:12 -!- Pthing has joined. 17:14:11 -!- augur has joined. 17:24:34 -!- augur has quit (Read error: 60 (Operation timed out)). 17:36:44 -!- augur has joined. 17:55:45 fizzie, do you (or anyone else in this channel reading this) know of any program that is able to fix the "bent" image from scanning a book 17:55:56 near the middle of the book you know 17:58:27 * Sgeo keeps campaigning for coakley 17:58:37 for who? 17:59:35 Martha Coakley for U.S. Senator from MA 17:59:49 As opposed to Scott Brown, who would kill Healthcare Reform 18:01:04 oh you are in US? hm 18:01:45 ah yes it was Slereah who was in France. Someone with a nick starting with S anyway 18:20:31 -!- cheater2 has joined. 18:21:51 -!- kar8nga has joined. 18:24:57 -!- Asztal has joined. 18:30:46 Again I must ask ... why is it that everyone writes left-recursive grammars, then writes left-to-right parsers, when right-to-left parsers make left-recursive grammars SO much easier to parse. 18:33:17 -!- zeotrope has quit (Read error: 60 (Operation timed out)). 18:36:03 -!- cheateur has quit (Connection timed out). 18:37:57 Because Yacc does left-to-right parsers? :P 18:42:22 The "healthcare reform" is giant piece of CF anyway. Doesn't fix the real problems. 18:44:30 -!- augur_ has joined. 18:45:07 -!- augur has quit (Read error: 104 (Connection reset by peer)). 18:46:01 I blame it on the ubiquity of left-to-right lexers. 18:46:12 (the problems with the US health system, I mean) 18:46:37 -!- scarf has quit (Read error: 113 (No route to host)). 18:47:26 Actually, I advocate outlawing all medicine. (Seriously. A black market would probably work better than the current system.) 18:51:40 -!- oerjan has joined. 18:52:01 -!- MigoMipo has joined. 18:55:16 Gregor, write a right recursive parser then 18:55:33 err 18:55:37 right recursive grammar 18:55:39 I meant 18:57:58 Ilari: it's *!gast@*.* can you do something about it? 18:58:37 since that means "guest" in german, i take it they had a german spammer problem once, and didn't realize (or maybe care) that it's a very generic word 18:59:31 AnMaster: right-recursive grammars suck. 18:59:43 1-2-3 != 2 19:01:33 cpressey: Actually, that made me think ... probably the only reason why everybody does everything left-to-right is that early compilers were all one-pass, and you can't do right-to-left in one pass with most filesystems (they'll only give you the file left-to-right). So the lexers are l->r because the file I/O is l->r, and the parsers are l->r because the lexers are l->r. 19:02:22 And nobody will indulge themselves in the sweet awesomeness of an R->L parser because they're stuck believing that they can't buffer the lexical tokens and read them the other way. 19:02:35 -!- cpressey1 has joined. 19:02:54 I just sent a message to cpressey that he apparently didn't receive! :P 19:03:11 temporary network outage. 19:03:22 cpressey: Actually, that made me think ... probably the only reason why everybody does everything left-to-right is that early compilers were all one-pass, and you can't do right-to-left in one pass with most filesystems (they'll only give you the file left-to-right). So the lexers are l->r because the file I/O is l->r, and the parsers are l->r because the lexers are l->r. 19:03:23 And nobody will indulge themselves in the sweet awesomeness of an R->L parser because they're stuck believing that they can't buffer the lexical tokens and read them the other way. 19:05:15 Gregor: yes, I was going to blame the left-to-right-ness of streams first. But then I considered that files are often memory mapped by the OS these days, and the order in which the tokens come is really the lexer's responsibility... 19:05:25 ouroboros parsing, anyone? :D 19:05:52 also, only L->R makes sense for interactive use... 19:06:34 oerjan: So what you're saying is, file I/O is l->r because time goes forward, not backward! 19:06:44 and also, if you are parsing simultaneously with downloading from the net... 19:07:02 yeah 19:07:24 Not that a file has a "left" anyway 19:07:28 time travelers' computers must be _so_ messed up 19:07:34 Bloody human conventions. 19:07:51 One way one might try to fix the US healthcare would be to restrict how low deductibles insurance can have. Except that would be unconsitutional (unless done by states, and there are 50 of them). 19:09:58 Not that feds care much what is constitutional or not... 19:10:06 * oerjan thinks that "prior condition" thing is messed up too, and seems impossible to solve unless everyone has mandatory insurance from birth... 19:11:22 oerjan: Many people here seem to think that any form of mandatory health care is evil and communist. 19:11:28 Yeah, it is messed up. When I explained to my father what insurance companies do with "prior condition", he wouldn't believe me at first... 19:11:51 "I HAVE THE RIGHT TO NOT HAVE HEALTH CARE", they assert. 19:15:24 Then there is "preventative healthcare". Some of it (like most vaccines in standard vaccination programs) is spectacularly cost-effective. Some it is are complete BS. And stuff in between. 19:16:26 People seem to be ignorant that we already pay through the nose for healthcare, universally... 19:16:39 Emergency rooms are required to accept anyone. 19:17:04 Most of the charges from uninsured people going to an ER are footed by the government. 19:17:24 So, we're already paying for health care, we're just doing it in the most spectacularly inefficient way we can manage. 19:17:34 -!- cpressey has quit (Read error: 110 (Connection timed out)). 19:17:46 -!- cpressey1 has changed nick to cpressey. 19:17:55 And "death panels". The _already_ exist in US. 19:18:17 Yup. We call them "the insurance company". 19:20:38 I don't think the requirement that ER accepts anyone is that bad. There is probably much much worse stuff that really screws things up. 19:20:51 anyone know why in haskell [0.1, 0.3 .. 1] turns out to contain 1.099999...? 19:21:56 cheater2: i think it is because if they put the cutoff exactly at 1, they could lose 1 itself because of rounding errors. so they do it at the halfway point instead, sine that seems the next most obvious choice... 19:22:22 yeah 19:22:25 Ilari: It's not that requirement that's bad. 19:22:29 use rationals instead of floats 19:22:33 -!- zeotrope has joined. 19:22:44 Ilari: It's the fact that that is the only universal healthcare available, and therefore what many people *soley* use, that's bad. 19:23:01 -!- zeotrope has changed nick to Guest83037. 19:23:06 Ilari: It's what makes the US government spend more per capita on healthcare than any other nation. 19:30:29 * oerjan notes that norway is no. 3 on the http://www.nationmaster.com/graph/hea_spe_per_per-health-spending-per-person list, and it's _not_ because of quality 19:30:39 I don't think that is the cause why healthcare spending is out of control. 19:30:51 yeah our spending is out of control too 19:31:55 oerjan: How many people are financially ruined because of your spending? 19:32:15 ok we may not have _that_ problem (much, anyway) 19:32:51 One thing is for sure. The US healthcare system is giant Charlie Foxtrot. 19:33:30 Also, that graph is off... 19:33:44 The per capita spending here is more like $7,000. 19:34:09 huh. 19:35:24 maybe the rest is insurance company profits ;D 19:36:32 There is some 30% administrative overhead. 19:37:29 And 15% of the population has literally no coverage... 19:40:00 And what percentage of citizens that would want coverage don't have one? 19:41:31 i think the reverse of that percentage is more meaningful 19:42:39 (i.e. the percentage of citizens that don't have coverage who would want one) 19:44:02 Yeah that could be meaningful too. But the "citizens" numbers are more meaningful than "all people" numbers. 19:44:04 -!- kar8nga has quit (Remote closed the connection). 19:44:47 -!- coppro has joined. 19:47:38 The difference between "citizens" and "all people" is that latter also includes legal aliens (shouldn't be much problem) and illegal aliens (these pretty much can't have coverage anyway). 19:48:58 Again I must ask ... why is it that everyone writes left-recursive grammars, then writes left-to-right parsers, when right-to-left parsers make left-recursive grammars SO much easier to parse. 19:49:45 * pikhq notes that "immigration" is an outmoded concept. 19:49:55 Why can't we all just sign onto Schengen? 19:49:58 it's not necessary LR(n) from the right just because it's from the left, i think, so you might get hideous ambiguity 19:50:25 oerjan: Yes, that's true. 19:50:58 oerjan: I'm talking more about hand-written parsers though, so lookahead isn't a huge issue necessarily I think maybe. 19:51:02 in fact i think most languages are designed to have little ambiguity from the left, while no one cares about the other way 19:51:25 I'm going to finish writing my R-L JavaScript parser and see where my hangups are :P 19:51:53 R-L? 19:51:54 (of course you still wouldn't get _global_ ambiguity) 19:51:56 mmapping won't solve all the problems with lexing R->L. Pipes and that sort of stuff is just the most obivious, even mmaps are mostly designed to be read forwards or randomly, not backwards. 19:51:59 oh, parser 19:52:02 nvm 19:52:34 Ilari: We're living in the future, it's not that big a deal. 19:53:03 * oerjan wonders if it would have been the other way around if the japanese had invented the main computer languages 19:54:03 (japanese being a natural language which branches the other way iiuc) 19:54:15 One test one could try: How fast can computer read large file forwards and how fast it can read it backwards (if file is too large to fit in RAM). 19:54:18 Odds are, we'd be using RPN languages. 19:54:30 -!- Pthing has quit (Remote closed the connection). 19:55:10 although human cognitive load considerations or something would probably imply left-to-right parseability regardless 19:55:14 Ilari: Unless it's a pipe or in FAT, "just as fast". 19:55:46 oerjan: I've wondered that too, but I suspect that Forth would have a large cult following in Japan if that effect were in play, and I've not seen evidence of that... 19:56:16 Instead... Ruby :) 19:56:20 heh 19:57:58 Also, I feel I must point out that if there's one language where there's an actual use case for parsing while it's coming to you over a pipe and not a disk file, it's Javascript. 19:58:51 But that's just me being pedantic. 19:58:54 Play more Gorf! 19:58:58 pikhq: Well, try it (the defintion of processing could be function that takes chunk of data and does nothing)? 19:59:01 well html itself too, it's a language for parsing purposes 19:59:14 it's an SGML profile 19:59:31 * oerjan knows that 20:00:24 `swedish Play more Gorf. 20:00:26 Pley mure-a Gurff. \ Bork Bork Bork! 20:01:33 mmmm, leverpastej med gurka! 20:01:49 Actually, I advocate outlawing all medicine. (Seriously. A black market would probably work better than the current system.) 20:02:03 i hear that didn't quite work for abortion, back in the day 20:02:42 * oerjan just had leverpastej, but no gurka 20:03:03 oerjan: fail 20:03:41 actually red beets is the mandatory condiment for leverpastej here 20:04:00 -!- jpc has joined. 20:04:11 woah! weird 20:04:16 but sadly i didn't have that either 20:04:27 maybe something to try after the can of cucumber runs out 20:04:47 pickled (?) red beets, mind you 20:05:11 ok cucumbers are not exactly unheard of either 20:05:29 although with red beets, _beware_ of spilling and soiling 20:07:15 I can imagine... I have vinegar from the cucumber all over my desk, but luckily it's colorless 20:07:49 yeah 20:08:55 oerjan: Well, that was more of a comment on the abysmalness of the current system than on the effectiveness of any proposed alternative. 20:09:35 i detected _no_ sarcasm in that <_< >_> 20:11:17 http://www.youtube.com/watch?v=MjMYQyhjiYA 20:11:52 COMPLETELY OBVIOUS DISCOVERY: Although there are more cases of left-recursion than right-recursion, the right-recursive cases are REALLY REALLY NASTY (IfStatement : 'if' ... Statement, Statement : ... | IfStatement | ... ) 20:13:02 * oerjan had to turn off that youtube before getting to the actual sarcasm. sitcom allergy, sorry 20:13:43 oerjan: It's not a sitcom, it's a skit by a comedy troupe. 20:13:53 well, same poison 20:14:07 oerjan is against ... media. 20:14:09 Or humor. 20:14:10 Or something. 20:14:42 Monty Python too? 'Cos KitH is just the Canadian version, really. 20:15:04 Except much better, of course. 20:15:07 oh monty python i can stomach, some of it 20:16:48 i just cannot cope with looking at people really embarassing themselves 20:16:55 Ohhh, Kids in the Hall! 20:16:58 You're dissing Kids in the Hall? 20:17:01 Pfffff. 20:17:26 oerjan's rank is now lowered to 2/5ths of a person. 20:17:37 who, me? i don't even know who kids in the hall are. i don't watch tv, remember? 20:17:50 However, speaking of Canadians and actual sitcoms which are funny, Corner Gas is hilarious. 20:19:02 what in a who now? 20:20:02 oerjan: Well, now you do. They're Canada's answer to Monty Python. 20:20:34 i don't think those guys were very good actors 20:20:40 Well, "answer" is maybe the wrong word, but you get the idea. 20:20:49 everyone is everyone's answer to monty python these days, aren't they 20:20:55 <3 Corner Gas 20:21:23 heck norway's answers to monty python started dying _off_ in the last decade :D 20:22:09 These days? I don't think there are comedy troupes anymore. Culture has disintegrated too far. KitH was like 1992. 20:22:10 i hear this monty python thing is pretty good 20:22:12 I prefer Monty Python. They're Britain's answer to Monty Python. 20:22:31 i just realized that after making this comment 20:24:03 cpressey, link to this Canadian answer? I can't locate it. Unless it is that last youtube link? 20:24:16 * coppro listens to ISIHAC 20:24:17 *the previous 20:24:18 (sparse log reading does have some downsides) 20:24:27 oerjan, hm? 20:24:38 oerjan, was it http://www.youtube.com/watch?v=MjMYQyhjiYA ? 20:24:45 yes 20:24:46 AnMaster: the last youtube link is a fairly good introduction. 20:25:14 mhm 20:25:39 cpressey, btw I don't watch tv either 20:27:00 * AnMaster agrees with oerjan 20:27:07 on the unwatchability of it 20:27:29 cpressey, ^ 20:27:39 I think the only modern tv show I've been able to watch has been "Good Eats". 20:28:00 cpressey, star trek tng, I voyager maybe 20:28:04 nothing newer than that 20:28:18 s/I / 20:28:42 ... You enjoyed Voyager? 20:28:46 -!- Guest83037 has quit (Read error: 110 (Connection timed out)). 20:29:00 The show where trans-infinite velocity has meaning? 20:29:22 i should (or maybe it would have been better not to) point out that by "embarassing themselves" i was referring to the characters, not the actors. 20:29:25 pikhq, I said "maybe" 20:29:36 a few of the episodes certainly, some definitely not 20:29:48 pikhq, I didn't watch all episodes 20:30:00 I probably missed that one with "trans-infinite velocity" 20:30:19 in case that has something to do with me being relegated to 2/5th person 20:30:23 AnMaster: Voyager had a range from horrible to mediocre, with a couple actually being good... 20:30:54 pikhq, I guess I mostly hit the mediocre and one of two of the good ones then 20:31:01 Probably. 20:31:12 oerjan: don't worry, to me, you are at least 2 and a half persons 20:31:23 "Break the warp 10 barrier!" "Warp 10... Isn't that like infinity?" "Yes." "So... Very fast?" 20:31:24 i guess that cancels out :D 20:31:26 pikhq, also, if isn't quite as bad if you are watching it to experience the treknobabel 20:31:34 ^ Actual. Conversation. 20:31:47 pikhq, well that's bad indeed 20:32:04 pikhq, "so good it is bad"? 20:32:05 errr 20:32:08 so bad it is good 20:32:09 That sounds funnier than any comedy troupe I can think of. 20:32:09 XD 20:32:14 No, so bad it's painful. 20:32:26 pikhq, starwars holiday special (not that I watched it) 20:32:34 So bad it's horrible. 20:32:50 THEY ACCELERATE TO IT. 20:32:52 Mind you, the Voyager writers have disowned the infamously-retarded "Warp 10" episode. 20:32:57 ACCELERATE TO INFINITY AND BEYOND. 20:33:11 http://tvtropes.org/pmwiki/pmwiki.php/Main/SoBadItsGood http://tvtropes.org/pmwiki/pmwiki.php/Main/SoBadItsHorrible (MWAHAHA) 20:33:12 Gregor: True. Still... 20:33:37 pikhq, well to infinity isn't impossible. Just double your speed the first minute, then double it again after half a minute, and again after a quarter of a minute, and so on 20:33:45 beyond I doubt 20:34:08 Going to warp 10 was only the /beginning/ of the retardedness of that episode. 20:34:09 oerjan, I use a custom css so I can't see the links for that site 20:34:26 Ah, well. Is your velocity *countably* infinite or *uncountably* infinite? Which aleph are we talking about, here? 20:34:28 AnMaster: for safety? 20:34:34 oerjan, yeah 20:34:41 darn my plans foiled 20:34:47 Yes. That was the stupidity that *started* the episode... 20:34:59 I refuse to even mention the rest of the stupidity that was that episode. 20:35:15 In a series that was pretty mundane, that was an episode that was astoundingly awful. 20:35:40 The vast majority of the series was rather... Meh. 20:35:43 Which is worse: using SQL Server Compact, or using SQLite but installing Visual Studio? 20:35:43 Not terrible, just meh. 20:35:52 But when it was bad, it was horrible. 20:36:39 I'll just note that there is one, and only one, consistent way to explain everything that occured in it: infinite improbability drive. 20:38:01 Sgeo: They're both pretty awful, but with the latter you have a prayer of portability. 20:50:09 TNG or TOS, which was best? 20:50:49 TNG > TOs 20:50:50 pikhq, what was that episode called btw? 20:50:51 *TOS 20:51:08 coppro, what about DS9? (I never watched any episodes from it) 20:51:22 some of DS9 is really good 20:51:22 AnMaster: Threshold. 20:51:30 pikhq, mhm 20:58:00 AnMaster: I don't know of a specialized program for that scan thing, but certainly you could try adding short line segments that trace the rows of letters in the book marked as horizontal lines in Hugin, then letting it optimize all those lens distortion parameters; it might find something that straightens them, though probably not very well. 20:59:35 The DS9 abbr always reminds me of the DeathStation 9000 first. 20:59:46 fizzie, I did some experiments, no luck 20:59:55 also it should do colour correction IMO 21:00:06 since it is way darker towards the middle 21:00:19 fizzie, that would be DS9K 21:00:40 I know, but that's always my first connotation anyhow. 21:08:28 -!- oerjan has quit ("Later"). 21:09:46 Deewiant: Checked; it was indeed in 2008, and Scala, and a 3212277-byte .jar, and MAIMTRON 9000 in http://www.cs.hut.fi/Opinnot/T-93.4400/2008/results/ 21:10:51 I don't suppose you have stats for how often they timed out? 21:11:28 Deewiant: I'm not sure. There's something about CPU usage in the "details" page. 21:11:57 Ah, so there is. 21:12:00 Didn't notice that page. 21:12:30 319.6 seconds/game on average; 3600 seconds was the limit. Probably didn't run out of time very much. 21:12:35 Oh, and I guess there might be the plot too. 21:13:02 "Per-bot statistics" and the CPU usage graph. 21:13:44 Well, seems like it wasn't slowed down by not being in Java. Or then it was just coded exceptionally well. 21:14:09 Or it didn't do very deep searching in the situation-space. :p 21:14:38 Did relatively well though. 21:15:16 Yes. Though that might be more dependant on how game-relevant a heuristic the author can figure out for evaluating a position, than on how clever the move search part is. 21:15:35 Meh, heuristics. Boo @ them. 21:15:59 There have been some "used practically no time at all" contenders in the top-N for very small N, either last year or the one earlier. 21:18:13 In 2008 there's Joojoojotti, which used 2.5 seconds of thinking per game (and a game is typically over a hundred half-moves) and still was 23rd out of 46. 21:18:32 which game? 21:19:14 coppro: There's this locally-developed board game that is used traditionally for the programming project in our "introduction to AI" course; to keep people interested, there's a bit of competition going on. 21:19:56 coppro: http://www.niksula.cs.hut.fi/~svirpioj/hierarkia/rules_en.html 21:21:31 Or http://www.niksula.cs.hut.fi/~svirpioj/hierarkia/ for a less detailed overview. 21:22:30 Deewiant: Come to think of it, were you asking about the language rules out of plain curiosity? 21:22:54 -!- augur_ has quit (Read error: 110 (Connection timed out)). 21:23:00 Well, I'm sort-of intending to complete the course 21:23:28 It may have been curiosity in the sense that I don't have any particular language wishes at the moment 21:24:05 -!- MigoMipo has quit. 21:24:46 Deewiant: Ah. But still, potentially participating. Now, I asked the question for some particular reason, which I still remembered back when I wrote it, but now I've completely forgotten what it might have been. 21:25:22 So, uh, "never mind", I guess. 21:25:27 :-P 21:25:53 -!- oklopol has quit (Read error: 60 (Operation timed out)). 21:29:58 -!- BeholdMyGlory has quit (Read error: 104 (Connection reset by peer)). 21:31:49 Deewiant: Checked; it was indeed in 2008, and Scala, and a 3212277-byte .jar, and MAIMTRON 9000 in http://www.cs.hut.fi/Opinnot/T-93.4400/2008/results/ <-- checked what? 21:32:07 Read more backlog. 21:32:29 Nooooooo! 21:33:27 Since when is any research project of the university user-friendly anyway <-- what if it *is* about user-friendly user interfaces? 21:34:25 They probably still suck 21:38:32 By "user-friendly" in this context I mostly mean "can manage to get it compiled and figure out what commands to run without calling us". 21:39:27 Admittedly it wouldn't be so hard to provide a simple script for recognizing audio files using existing models. 21:40:13 -!- oklopol has joined. 21:40:37 But the training part is all done with crafty little bits and pieces, and everyone pretty much has their own collection of scripts, since the sort of research that is done differs so much. 21:41:43 Documentation's been partially replaced with aural tradition, passed on from older shamans to younger acolytes by the campfires. Okay, so that last part was fiction. 21:42:43 s/au/o/ I guess too. 21:42:48 oh my I just found out there exists such a thing as "chess boxing" 21:42:51 as, in a sport 21:43:00 I have known of this for several years 21:43:25 it's a ridiculous idea though 21:44:06 fizzie, what is the project in question? 21:44:23 AnMaster: Our speech recognizer in general. 21:44:35 fizzie, ah. Open source? 21:45:15 fizzie, why is one entry between 27 and 28 at http://www.cs.hut.fi/Opinnot/T-93.4400/2008/results/ greyed out? 21:45:24 Not really. That, I think, was Deewiant's whole point. Why don't you read the logs?-) 21:45:26 and another one between 45 and 46 21:45:42 AnMaster: Those are unofficial participants. 21:45:59 fizzie, oh I see. How did they work 21:46:32 also why didn't any bot play against itself. It might have been funny 21:46:39 Vince being 2007's winner, and RandomBot choosing a random legal move. 21:47:07 Because it wouldn't have affected the scoring. 21:47:07 ah 21:47:19 fizzie, try, but might have been amusing 21:47:23 also what were the rules for the game 21:47:37 Already linked twice 21:47:47 That was linked to just a moment ago, yes. 21:47:49 ah at the bottom there 21:48:03 funny it crashed firefox 21:48:50 fizzie, oh yeah, china-chess (iirc it is called that in Sweden at least) 21:48:57 If "at the bottom" means the bottom of the results page, then no, that's just detailed statistics. 21:49:05 well, not exactly 21:49:06 almost 21:49:31 The game is reasonably different from most things. 21:49:46 The way to capture pieces is rather peculiar. 21:50:20 I don't dare to say unique since games are not my speciality. But not well-known, anyway. 21:51:14 ah 21:52:21 In 2009's results there are more unofficial participants, since last year's top 5 was also included. http://www.cs.hut.fi/Studies/T-93.4400/2009/results/ 21:52:56 Evidently the 2007 folks sucked then 21:53:09 Or the results are just that random 21:53:30 Deewiant: Well, 2007 was done with Scheme in a different system. 21:53:52 How'd you run that versus the 2008 folks? 21:54:14 I ported personally the Scheme winner over to Java, but did not tune anything so it assumed a lot less computing time was available. 21:54:42 Ah. 21:54:57 It's just that they wanted something that made more sense than RandomBot to compare against. 21:55:09 Sensible. 21:55:39 fizzie, why java? 21:55:42 2008's top three were all better than 2009's winner, though. 21:56:16 Not by much, though. 21:56:43 AnMaster: It pretty much follows from what is used in the introductory programming courses here. 21:56:55 Python next year, then? ;-P 21:56:58 And Java's got that sandbox in place. 21:57:11 fizzie, where are the 2007 results? 21:57:28 fizzie, why was it scheme before then? 21:57:41 Same reason. 21:57:46 hah 21:59:23 Yes. But the majority of 2007 already complained about Scheme, since teaching using it had already stopped. 21:59:38 Not 2006? 22:00:17 I wasn't the assistant back then. 22:00:41 fizzie, still where is the 2007 results. Neither http://www.cs.hut.fi/{Opinnot,Studies}/T-93.4400/2007/results/ worked 22:00:46 neither of* 22:01:10 It's a text file somewhere. 22:01:23 I'll try to look for it in a bit. 22:01:40 Different systems, different results reports. 22:02:27 well, going to sleep. doesn't matter if it isn't easy to reach 22:03:48 It's a lot less fancy, just the scores bsically. 22:04:15 Bit busy right now though; probably no great loss there. 22:11:00 wtf 22:11:21 I independently invented chess boxing as a joke in the early 90's. I had forgotten about it until now :) 22:12:06 Well, that's about the age of the idea 22:12:12 Early 90s, I mean. 22:12:26 Hey, I found it: http://www.cs.hut.fi/Opinnot/T-93.4400/2007/airesults.html 22:12:45 In case AnMaster somehow is still here. 22:12:54 And understands Finnish. 22:12:54 Parallel evolution of ideas, I guess... 22:13:11 And it wasn't just a text document after all. 22:13:40 It's not too hard to come up with the premise "combine something extremely intellectual with something extremely physical" 22:13:47 Oh, wait. 22:14:00 Maybe it wasn't chess boxing I came up with... what was it? 22:14:12 I think it might have actually been cycle boxing. 22:14:21 Deewiant: The game-browser was also cgi-based back then and has rotted away. 22:14:30 Box a round, then once around a track on a bicycle, then box another round... 22:14:52 Two extremely physical but otherwise very dissimilar activities. 22:15:40 Cycle around a track and try to punch the other guy with your fists while at it. 22:15:50 That would be... harder. 22:16:21 Also more accident-prone, and therefore would draw larger audiences. 22:27:42 -!- augur has joined. 22:28:29 At least in western countries popularity of something seems approximately inversely proportional to amount of intellect in it... :-> 22:32:35 Hmm... From one study "Intake of TFA [trans fatty acids] was strongly associated with CHD mortality..." -- not surprising, TFAs (not to confused with CLA) are known to be dangerous for health. 22:36:19 And if you think "trans fat problem" has been solved, check some nutrient database querying: Most trans fats, milk free (milk free is to try to minimize interference from CLA, as most databases classify that as trans fat). 22:41:49 cpressey: oh, i thought like riding bicycles while boxing. 22:42:07 now *that's* harder. 22:45:27 oklopol: A regular boxing match, with the ring and all, but put both participants on unicycles. 22:45:32 oklobox that's a great idea lol bike with one hand and whack the guy with the other 22:45:48 the speed of the bike gives extra power to the bike, just like jousting 22:46:03 oklobox could be my bi-boxer name. 22:46:26 (bi-boxing = bicycle boxing, obviously) 22:46:55 "the speed of the bike gives extra power to the bike" <<< yeah speed is powah 22:47:42 oops, I meant punch 22:47:55 i know 23:10:05 -!- cheater2 has quit (lindbohm.freenode.net irc.freenode.net). 23:10:05 -!- MizardX has quit (lindbohm.freenode.net irc.freenode.net). 23:10:05 -!- pikhq has quit (lindbohm.freenode.net irc.freenode.net). 23:10:06 -!- fungot has quit (lindbohm.freenode.net irc.freenode.net). 23:10:06 -!- yiyus has quit (lindbohm.freenode.net irc.freenode.net). 23:12:21 -!- pikhq has joined. 23:13:37 -!- cheater2 has joined. 23:13:53 -!- yiyus has joined. 23:13:55 -!- MizardX has joined. 23:31:11 -!- oklopol has quit (Read error: 110 (Connection timed out)).