00:08:49 -!- oerjan has joined. 00:08:52 -!- Slereah_ has joined. 00:16:28 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | hm.. 00:25:50 .mh 00:37:19 -!- M0ny has quit ("Hum... Hum..."). 01:02:51 -!- Slereah has joined. 01:04:25 -!- puzzlet has quit (Remote closed the connection). 01:04:29 -!- puzzlet has joined. 01:22:02 -!- olsner has quit ("Leaving"). 01:31:08 -!- Slereah_ has quit (Read error: 113 (No route to host)). 01:31:10 -!- Sgeo has joined. 01:33:48 -!- boily has joined. 01:34:05 http://home.codu.org/colormatch/check.html // seems to sorta-kinda work 8-D 01:35:40 -!- Slereah has quit (Read error: 113 (No route to host)). 01:37:56 -!- boily has quit (Client Quit). 01:38:05 Result: My shirt matches my tie 8-D 01:40:03 or so one would assume 01:44:13 eww ties. 01:55:10 GregorR: http://photos-b.l3-t.facebook.com/photos-l3-sf2p/v355/140/51/644027966/n644027966_1050121_9574.jpg <-- wear this. 01:55:36 I can't read that. 01:55:50 Also, why is nobody as excited about my aesthetic color chooser as I am :P 01:56:20 Because Google Chrome is being slow. 01:57:04 And having tie fetish attacks. 01:57:11 * ihope re-ponders the subject of that sentence 01:58:04 ENKI-][: i cannot make out who i am supposed to obey there 01:58:23 oerjan: it's OBEY spelled backwards. 01:59:06 oerjan: i also have a CTHULU/DAGON '08 shirt to wear to erection day. 01:59:15 er. i mean. 01:59:19 obama day 01:59:20 er 01:59:22 <_< 01:59:24 :-) 01:59:27 who is dagon 02:00:55 also i prefer the Allosaurus to Cthulhu 02:04:01 dagon is the one who made a small new england port town sacrifice virgins to it until it ate the whole down 02:04:25 ic 02:05:00 ihope: Got it working? 02:05:03 Anybody: Got it working? 02:05:06 It's pretty cool :P 02:05:25 +ul (a::aaa:::)((!()):a)~*^(~S:^):^ 02:05:26 (((((!())))))(((((!())))))(((((!())))))(((((!())))))((!()))((!()))!() ...S out of stack! 02:05:39 * ihope goes there 02:06:13 GregorR: 0 is rejected, 1 accepted i take 02:06:18 ooh. what language is that? i remember seeing it somewhere 02:06:21 oerjan: Yes. 02:06:22 underload 02:06:39 oerjan: And for some reason it doesn't take identical colors as a match, I should probably generate some cases for that. 02:10:23 -!- Corun__ has quit ("This computer has gone to sleep"). 02:30:11 -!- ihope has changed nick to uoris. 02:33:51 Heh, this stupid project spanned three languages X-D 03:21:57 -!- ab5tract has joined. 03:22:55 am I right in thinking it presses "check match" for you if you use the random button? 03:23:08 because I can't find one which doesn't match #489764 :( 03:23:36 though the result does flicker from 0 to 1 occasionally 03:23:42 When you press random it chooses a random /matching/ one. 03:24:05 It generates a random number, checks it, and if it fails, loops. 03:24:29 ah, I see 03:24:34 In worst case, couldn't that be very inefficient? 03:25:16 ah, yes, #ff00ff quite easily results in 0 03:28:58 +ul (a::aaa:::)((!!):a)~*^()(:a~*(*a)~*a(~:)~*(~^)*^!^):^S 03:28:59 ...^ out of stack! 03:29:02 argh 03:30:52 +ul (a::aaa:::)((!!):a)~*^()(~S:^):^ 03:30:53 (((((!!)))))(((((!!)))))(((((!!)))))(((((!!)))))((!!))((!!))!! ...S out of stack! 03:31:23 +ul (a::aaa:::)((!!):a)~*^()(~aS:^):^ 03:31:24 ()((((((!!))))))((((((!!))))))((((((!!))))))((((((!!))))))(((!!)))(((!!)))(!!) ...a out of stack! 03:38:25 +ul (a::aaa:::)((!!):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^S 03:38:26 ...^ out of stack! 03:39:50 +ul (a::aaa:::)((!!):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*):^(~a^S:^):^ 03:39:51 ~:(*a(a(:^)*(*a)~*a(~:)~*(~^)*):^)~^(((((!!)))))(((((!!)))))(((((!!)))))(((((!!)))))((!!))((!!))!! ...a out of stack! 03:40:20 er wait 03:40:26 +ul (a::aaa:::)((!!):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*):^(~aS:^):^ 03:40:27 (~:(*a(a(:^)*(*a)~*a(~:)~*(~^)*):^)~^)()((((((!!))))))((((((!!))))))((((((!!))))))((((((!!))))))(((!!)))(((!!)))(!!) ...a out of stack! 03:46:08 +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^S 03:46:09 (((((((((((!!()()))))))(((((!!()()))))))(((((!!()()))))))(((((!!()()))))))((!!()())))((!!()()))) 03:46:33 +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(~aS:^):^ 03:46:35 ((((((((((((!!()()))))))(((((!!()()))))))(((((!!()()))))))(((((!!()()))))))((!!()())))((!!()())))) ...a out of stack! 04:02:10 +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(~*a)~*a(~:)~*(~^)*^!^):^(~aS:^):^ 04:02:12 ((((!!()()))(((!!()()))((((((!!()())))))((((((!!()())))))((((((!!()())))))((((((!!()())))))))))))) ...a out of stack! 04:29:14 -!- uoris has quit (Read error: 110 (Connection timed out)). 04:29:20 +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*)~*(!^^)*~a*(:^)*^):^ 04:29:22 ...~ out of stack! 04:29:37 -!- uoris has joined. 04:29:52 +ul (A)SS 04:29:53 A ...S out of stack! 04:30:39 +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*)~*(!^^)*~a*(:^)*):^(~aS:^):^ 04:30:40 (~a*^((^)((:)S)(!(A)S(^)*())())!^^(^):^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*)~*(!^^)*~a*(:^)*)((((((((((((!!()()))))))(((((!!()()))))))(((((!!()()))))))(((((!!()()))))))((!!()())))((!!()())))) ...too much output! 04:44:13 +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*(:^)*^):^ 04:44:14 AAAA ...^ out of stack! 04:44:37 +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*(:^)*^):^ 04:44:38 AAAA ...^ out of stack! 04:45:57 +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*(:^)*^)()~^(~aS:^):^ 04:45:58 ()(!!()())(((((((((!!()()))))))(((((!!()()))))))(((((!!()()))))))(((((!!()()))))))()(!(A)S(^)*())((:)S)(^) ...a out of stack! 04:46:04 -!- immibis has joined. 04:48:30 +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*()*^):^(~aS:^):^ 04:48:31 (~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*()*^)(!!()())(((((((((!!()()))))))(((((!!()()))))))(((((!!()()))))))(((((!!()())))))) ...too much output! 04:52:19 +ul (a::aaa:::aa)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*(:^)*^):^ 04:52:20 AAAA ...^ out of stack! 04:52:35 +ul (::aaa:::aa)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*(:^)*^):^ 04:52:36 A:A ...^ out of stack! 04:53:34 +ul (:a:)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*(:^)*^):^ 04:53:35 AA:A ...^ out of stack! 04:54:06 +ul (a:a:)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*(:^)*^):^ 04:54:08 AAA:A ...^ out of stack! 04:54:53 +ul (a:a:)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(~aS:^):^ 04:54:54 (((((((!!()()))))(((!!()()))))((!!()())))) ...a out of stack! 04:54:56 i see a lot of smiley faces, cakes and angels 04:55:25 +ul (::a:)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(~aS:^):^ 04:55:26 (((((((!!()())))((!!()())))(!!()()))(!!()()))) ...a out of stack! 05:01:09 hm there appear to be two sets of outer parentheses 05:01:20 *outermost 05:01:48 that would clearly cause some bug 05:09:16 Should I agree with you? I don't know Underload. 05:11:23 -!- oerjan has quit (Remote closed the connection). 05:11:36 -!- oerjan has joined. 05:12:17 since i'm barely understanding the program myself, i don't expect anyone to agree :D 05:12:53 is this program generating itself or something 05:13:16 no, i am trying to split up a list of a's and :'s 05:14:13 oh, there are parts that are self-generating, since that's the only way to loop in underload 05:14:41 very esoteric 05:15:02 yeah 05:15:32 but i think my brain has had enough for now 05:18:56 underload contains no command for splitting a string into characters, but i figured it should be theoretically possible if the characters are all a and : 05:20:20 oh, duh! 05:22:28 the second part is completely wrong, because a doesn't cause a new list element 05:22:48 needs a different strategy 05:36:10 -!- ab5tract has quit. 05:39:29 -!- immibis has quit (Read error: 110 (Connection timed out)). 06:14:07 -!- metazilla has joined. 06:14:13 -!- moozilla has quit (Nick collision from services.). 06:14:22 -!- moozilla has joined. 06:16:28 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | !2a:a;.. 06:22:37 -!- megatron has quit (Read error: 110 (Connection timed out)). 06:30:51 -!- Sgeo has quit (Remote closed the connection). 06:37:24 oerjan: that's a wonderful idea 06:40:48 lol green is the only thing that doesn't go with black 06:41:02 what? 06:41:18 i can answer you indirectly 06:41:31 by highlighting the one i was indirectly talking to 06:42:12 * oerjan assumes oklopol was misspelling GregorR as oerjan 06:42:12 GregorR: i would be quite interested, if i believed colors can "go together", i don't. 06:42:21 oh no 06:42:31 that was a separate thing 06:42:48 wonderful idea was about the underload thing 06:42:52 ah 06:43:37 did you try to replicate the functionality, or push the string splittered on the stack? 06:44:30 so far, i tried to print it out with a's upper cased 06:47:14 ouch, tried to rip my toenail off, but forgot i need to push my fingernail through the side first, or it won't come peacefully. 06:47:20 life is good 06:47:40 * oerjan prefers a nailcutter for all such things 06:49:10 isn't that pretty gay? probably doesn't even make you bleed. 06:49:39 (i don't have a nailcutter here, would probably use one if i had one near me) 06:51:12 * oerjan always carries one, in case of accidents 06:51:41 err, well that definitely sounds pretty gay :P 06:52:18 my nails are so fragile that if i don't do it properly at once they start disintegrating at the least provocation 06:52:34 cool 06:52:52 i have boring normal nails :< 06:53:45 well at least i don't polish them :D 06:54:16 although that might actually have helped with the fragility 06:55:04 so you fin[n]ish them, but not polish them; what other countries do you them? 06:55:57 hmm; lecture starts about now, i should probably consider leaving. 06:56:45 see you 06:59:42 -!- oerjan has quit ("Possibus?"). 07:33:48 i've never heard that device called a nailcutter 07:37:12 oh dear, i'm all out of bourbon 07:59:59 -!- clog has quit (ended). 08:00:00 -!- clog has joined. 08:16:33 -!- olsner has joined. 09:19:51 -!- M0ny has joined. 10:00:16 -!- sebbu2 has joined. 10:17:30 -!- sebbu has quit (Read error: 110 (Connection timed out)). 10:33:26 From SOCK it's just R/W most of the time, from TOYS only S to clear the old code when ^reloading (so it might be good if S'ing to value 32 would actually clear those cells), and from SUBR only a C/R pair for ^code. <-- don't they already? 10:33:32 the TOYS one I mean 10:36:43 It might, I haven't checked at all. 10:36:46 Just a thought. 10:38:10 I guess you do that already in fungespace_set for any space. 11:07:02 -!- metazilla has quit (Read error: 60 (Operation timed out)). 11:09:18 -!- megatron has joined. 11:09:24 -!- moozilla has quit (Nick collision from services.). 11:09:30 -!- moozilla has joined. 11:12:23 -!- metazilla has joined. 11:12:31 -!- moozilla has quit (Nick collision from services.). 11:19:26 -!- moozilla has joined. 11:27:59 -!- megatron has quit (Read error: 110 (Connection timed out)). 11:29:27 -!- KingOfKarlsruhe has joined. 11:38:06 -!- megatron has joined. 11:38:14 -!- moozilla has quit (Nick collision from services.). 11:38:25 -!- moozilla has joined. 11:44:07 -!- Slereah_ has joined. 11:47:37 -!- metazilla has quit (Read error: 110 (Connection timed out)). 12:16:28 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | that's some serious time dilation. 12:17:14 optbot ! 12:17:15 M0ny: i suppose there have been worse last words. 12:22:44 Mona mona mona 13:03:06 -!- Slereah has joined. 13:03:56 -!- Slereah_ has quit (Read error: 60 (Operation timed out)). 13:37:33 -!- Slereah has quit (Read error: 113 (No route to host)). 13:41:20 Hey anyone know a good software to synth electronic mono-phonic (like old mobile phones) sound? For Linux. Using a MIDI file as input. 14:00:05 AnMaster: You're being a bit obsessive with your bleeps 14:04:07 ehird, oh? 14:04:14 AnMaster: You could just put the midi on it. :-P 14:04:26 ehird, I checked, impossible 14:04:36 is supports *.aac *.wma *.mp3 14:04:43 AnMaster: Just use a regular midi-to-mp3 rendering thing, then. 14:04:54 I don't think there are programs that make it sound like an old mobile. 14:04:59 At least I've never heard of any. 14:05:13 ehird, there must be 14:05:14 It'd be hard, what with the whole "polyphony" thing that MIDIs have. 14:05:27 otherwise I shall program my pc speaker and record it 14:05:38 AnMaster: as i said - obsessive 14:05:51 ehird, if you wish 14:06:04 I could maybe temp mend my old phone and record that 14:09:22 AnMaster: or you could just render the mid to a mp3 and put it on and forget about it because it's a bloody ringtone 14:10:37 You could take some sequencer application and change the instruments in the midi file to sound more bleepy. 14:11:00 Even better, download a ring-ring sound off the interwebs and put it on. 14:11:05 That's not very modern. 14:11:36 fizzie, hm 14:11:55 ehird, I doubt the "The Internationale" exists as that ;P 14:12:03 AnMaster: I meant just a regular ring, ring. 14:12:09 fizzie, good idea. Now where to find that... 14:12:22 That's as old as the telephone, so you can stay comfy in ancient history. 14:12:40 ehird, I *will* do this, nothing you say will change my mind 14:12:50 As I said. Obsessive. 14:17:22 http://www.theregister.co.uk/2008/10/28/student_charged/ (Insert typical el reg disclaimer.) 14:38:52 fizzie, I think I got an idea: Make a custom beepy soundfont, load that into to fluidsynth 14:38:56 along with the midi file 14:39:02 :D 14:40:30 ehird, ^ 14:43:00 http://wiki.openid.net/LID_Look_and_Feel Someone on the OpenID wiki complains that when entering their openid "socialism.is.EVIL.myopenid.com", it picks the default name "socialism" to log in to the wiki. (At the bottom) 14:43:00 XD 14:44:58 hahaha 14:45:29 ehird, well I think socialism is good 14:45:40 that's wholly irrelevant to my amusement but... whatever 14:47:02 Heh, the "screen shot of the trauma of picking an ID name for this wiki" part was funny. 14:47:10 Must've been very traumatic indeed. 14:48:37 fizzie: Don't you have any FEELINGS? 14:48:40 The wiki was MOCKING HIM! 15:21:10 haha 15:21:28 well anyway I think combining a sine tone and a square one works quite weel 15:21:29 well* 15:21:32 * AnMaster tests 15:27:34 -!- sebbu has joined. 15:27:50 -!- sebbu2 has quit (Read error: 60 (Operation timed out)). 15:34:10 -!- jix has joined. 15:48:46 lost the game 15:49:44 -!- Azstal has joined. 15:51:20 * AnMaster prepares to properly record it, stopping stuff that can cause delays and such 15:56:02 ehird: can you not just stop doing that???! 15:56:13 jix: i lose the game regularly 15:56:18 #esoteric is a convenient tab. 15:56:20 jix, who cares anyway 16:01:25 -!- AnMaster has quit ("System reboot."). 16:01:28 i don't see what's so wrong about that game 16:02:00 oklopol: anmaster makes a point to say how much he doesn't care every time someone mentions it 16:02:11 http://www.theonion.com/content/node/28694 16:02:31 -!- Asztal has quit (Success). 16:05:58 ehird: many people do that; i'm not commenting the habit of repeatedly stating your opinions, i don't see why people have such strong opinions about the game 16:06:14 anmaster doesn't, really, though 16:06:47 for all X, he has an opinion on X, for a majority of X, he has a half-baked opinion on X (just like everyone else), but for the same majority he repeatedly states his half-baked opinion on x 16:06:57 most people tend to only repeatedly state their strong opinions. 16:14:33 -!- AnMaster has joined. 16:15:03 hiiii 16:15:09 bgpofkmgzxlgjmdrsk 16:15:15 *wrong channel 16:24:37 -!- jix has quit ("CommandQ"). 16:33:35 http://www.pageflip.hu/ oh god. 16:36:30 not bad, but it doesn't crease if I turn it sharply enough 16:36:35 *g* 16:36:57 Azstal: you can actually rip the pages off 16:36:58 >_< 16:37:06 yeah, I like that :D 16:37:13 so awwwwwwful 16:37:26 some of the pages are un-rippable though 16:37:35 yes 16:37:41 i've spent a few minutes trying to demolish it 16:48:19 finally, an *.aac 16:48:41 (checked with mp3 too, but that file was larger and even worse sound quality 16:48:42 ) 17:08:27 -!- Slereah_ has joined. 17:15:48 -!- metazilla has joined. 17:15:54 -!- moozilla has quit (Nick collision from services.). 17:16:14 -!- moozilla has joined. 17:18:57 err 17:19:03 how can they be ripped off? 17:20:05 ? 17:20:07 oh 17:20:09 oklopol: drag them off 17:20:11 like... ripping IRL 17:20:13 note 17:20:15 some of them can't be 17:20:51 ah, doesn't work for cover. 17:20:55 ah not as in circumventing drm then 17:21:10 nope 17:21:12 er 17:21:17 ripping isn't "circumventing DRM" 17:21:20 true 17:21:24 related however 17:21:25 ripping is e.g. copying audio data from a cd to a computer 17:21:29 which is totally legal :-P 17:21:36 yes legal 17:21:42 well 17:21:48 depends on your interpretation of the dmca 17:21:51 when talking about a drm'd cd. 17:21:51 however these days that usually includes circumventing drm 17:21:53 ;P 17:21:53 Idea : Fractal darts. 17:21:56 not really 17:22:04 i've never circumvented any DRM once to rip a cd 17:22:05 10^n points for the nth level of recursion 17:22:07 k. afk making food 17:22:14 Slereah_: hah 17:22:37 I thought of that between thinking of bees and hats. 17:23:01 -!- jix has joined. 17:24:55 -!- megatron has quit (Connection timed out). 17:32:25 -!- oerjan has joined. 17:37:53 -!- KingOfKarlsruhe has quit (Remote closed the connection). 17:43:16 -!- Asztal^_^ has joined. 18:00:41 -!- Azstal has quit (Connection timed out). 18:01:22 -!- Slereah_ has quit (Read error: 113 (No route to host)). 18:04:13 -!- kt3k has joined. 18:04:57 heh 18:05:04 the wikipedians are discussing what they'd do if someone named a book Main Page 18:05:12 [[If someone announces that they are writing a book about Wikipedia titled Main Page, I suggest we indef hard-rangeblock his ISP until agrees to name it something else :-)]] 18:05:15 MUST DO IT 18:05:17 sheesh that's obvious 18:05:27 Main Page (book) 18:05:35 oerjan: and where does the disambig link go? 18:05:40 What about people searching Main Page in the search bar? 18:05:51 Main Page (disambiguation) of course 18:05:55 uh 18:05:58 hm... 18:06:01 oerjan: and where would you link to that 18:06:13 on the top of the main page? that's distracting clutter for, like, 1% of all traffic 18:06:15 heh, there is a slight problem 18:06:23 and also free advertising :D 18:06:56 :D 18:07:34 conclusion: someone do it. The resulting bureaucratic glob will destroy Wikipedia. 18:11:34 hm the idea of moving the Main Page to WP: at least seems reasonable 18:11:53 wp should bloody fix their bloody software 18:12:11 they have far more urgent problems than hypothetical books 18:12:18 if you search for C# it takes you to C 18:12:27 that's called a "bug" and they haven't fixed it in years 18:12:48 ouch 18:12:54 (so they need a disambig entry for C#, which itself is a disambig page, on the C page) 18:13:13 -!- kt3k has quit ("CHOCOA"). 18:13:47 (same with every other note) 18:14:01 well the use of # is a general html thing isn't it? not restricted to wp 18:14:15 sure, but what's html got to do with it? 18:14:49 they could have some escaping mechanism 18:14:56 and let the search box be aware of it 18:15:35 hm 18:16:29 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | hi kipple. 18:16:49 why is main_page in article space anyway? 18:17:03 it might be enough to trap titles ending with #, i doubt wp uses empty anchor names 18:17:34 oklopol: tradition 18:17:35 so that you could have competing main_pages? 18:17:35 oklopol: history, and if they did e.g. Portal:Main all the bookmarks would break 18:17:50 oklopol: if they just redirected - then there's no way to put another article ther 18:17:58 so, no benefit for extra confusion essentially 18:18:29 i see, backwards-compatibility, the mother of all that is ugly. 18:18:30 also, it's not an "html thing" 18:18:32 it's a web thing. 18:18:37 whatev 18:18:47 oerjan: well, right now the problem is the title of the C# page is C-Sharp because they disallow C# 18:18:47 lament, err C# is non-trivial to handle, since the browser would probably treat it the same way as foo.html#anchor 18:18:48 oklopol: i think if you ran a site as big as wp you'd care about that too. 18:18:50 I guess 18:18:52 AnMaster: no 18:18:54 in the search box 18:18:58 ah hm true 18:18:58 C# will be sent escaped 18:18:58 AnMaster: no 18:19:03 they just fuck that up in the interm 18:19:05 *interim 18:19:05 AnMaster: i'm not talking about the name of the HTML page 18:19:11 AnMaster: i'm talking about the title of the article 18:19:18 AnMaster: WP started with the two being the same 18:19:19 http://en.wikipedia.org/wiki/C# wouldn't work well however 18:19:23 AnMaster: no, it wouldn't 18:19:24 ehird: the other way around; if i was a person who would care about that, i might have a page as big as wp 18:19:26 but how many people hack the urls? 18:19:30 if i had, now, a page as big as wp 18:19:31 compared to people using the search box 18:19:39 i'd probably just close it down for the fuck of it. 18:19:41 AnMaster: the title of the article and the name of the html page are two different entities 18:19:48 hm true 18:19:49 oklopol: yeah, but, nobody would put you in charge of anything. 18:19:50 if they're the same, the system is badly designed 18:20:02 lament, well it used to be that way :P 18:20:06 not even a pancake. 18:20:17 what 18:20:20 AnMaster: right, exactly 18:20:21 'it used to be that way'? 18:20:24 WP started off badly designed 18:20:24 that doesn't erally make any sense 18:20:27 it still IS that way. 18:20:40 at least WP doesn't need camelcase now. 18:20:42 ehird: if people hack urls they should expect technical issues anyway 18:20:44 ehird: perhaps not, i don't see what that has to do with anything 18:21:05 oerjan: yes 18:21:13 well 18:21:15 i'm all for url-hacking 18:21:21 but... if you're url hacking, know how to escape shit, okay 18:22:12 the article for the note C# is 18:22:19 C♯_(musical_note) 18:22:38 would be interesting to see if something bad actually happened if wp or a related entity changed it's main page 18:22:39 well 18:22:40 but the disambig is not needed because there isn't any other C♯ 18:22:45 that probably happens every now and then 18:22:51 i wouldn't know 18:22:59 lament: well 18:23:10 several articles are like that 18:23:15 just because the disambig makes it easier to see in the page title 18:23:24 yeah 18:25:58 ehird: also why wouldn't anyone put me in charge, it's not like they knew i would just bring the thing down for fun if i could 18:26:41 oklopol: now they do 18:26:47 you just said it in a publicly logged channel 18:27:01 yeah, and someone is so gonna see that. 18:27:32 anyway, i probably wouldn't bring it down if i got money out of it; but i definitely would do a name change in the name of purity. 18:27:39 purity is worth killing. 18:29:04 *url change 18:40:35 -!- Asztal^_^ has quit (Read error: 104 (Connection reset by peer)). 18:40:52 -!- Asztal^_^ has joined. 18:47:11 -!- lilja has joined. 19:13:16 lillllja 19:19:36 -!- oerjan has quit ("leaving"). 19:21:32 -!- KingOfKarlsruhe has joined. 19:22:41 -!- jix has quit (Read error: 113 (No route to host)). 19:32:48 -!- asiekierka has joined. 19:32:52 hey 19:32:57 It was a while 19:32:59 Also 19:33:10 [<.] - a Self-modifying BF quine... i think 19:33:12 ^show 19:33:12 echo reverb rev rot13 rev2 fib wc ul cho choo pow2 source 19:33:23 ^def smbf_quine bf [<.] 19:33:24 Defined. 19:33:38 ^smbf_quine ].<[ 19:33:47 Oh 19:33:55 TT__TT 19:36:03 ^def smbf_quine bf [,.] 19:36:04 Defined. 19:36:10 ^smbf_quine [,.] 19:36:13 [anything] will never print out anything. 19:36:14 ^smbf_quine [,.] 19:36:16 OOH 19:36:25 ^def smbf_quine bf +>[<,.>] 19:36:25 Defined. 19:36:33 ^smbf_quine +>[<,.>] 19:36:38 OH 19:36:41 dear lord 19:36:48 ^def smbf_quine bf +[>,.<] 19:36:48 Defined. 19:36:55 ^smbf_quine +[>,.<] 19:36:55 +[>,.<] 19:37:05 Nope 19:37:12 SMBF should be implemented in fungot 19:37:13 asiekierka: what's the question there too... but it's in the gray zone. the black parts show the table? can you lisppaste input output code? 19:37:26 fungot: i don't know lisp 19:37:27 asiekierka: scheme's file system interface is sadly lacking.) 19:37:45 fungot: Lacking? And i don't know scheme, too! 19:37:46 asiekierka: http://www.wftv.com/ slideshow/ news/ technology/ fnord is great. 19:38:06 fungot: That link is scary. WolF TV, it may be... 19:38:07 asiekierka: i'm an east side type... 19:38:18 fungot: East side? So the east side doesn't know wolves? 19:38:37 ...uh? 19:38:52 Hello? 19:38:53 fungot: East side? So the east side doesn't know wolves? 19:38:58 Erkh. 19:39:00 fungot: aaa 19:39:04 Wait 19:39:06 did fungot just crash? 19:39:09 ^show 19:39:09 echo reverb rev rot13 rev2 fib wc ul cho choo pow2 source smbf_quine 19:39:13 *whew* 19:39:16 ^show source 19:39:16 (http://zem.fi/~fis/fungot.b98.txt)S 19:39:23 It goes into that ignore mode if you have talked to it too much. 19:39:33 wait 19:39:37 so underload is implemented now? 19:39:39 Yes. 19:39:47 ^help 19:39:48 ^

; ^def   ; ^show [command]; lang=bf/ul, code=text/str:N; ^str 0-9 get/set/add [text]
19:39:56  Oh, yeah.
19:40:13  ^source
19:40:13  http://zem.fi/~fis/fungot.b98.txt
19:42:38  AnMaster added that one; I think he keeps forgetting the URL. Although I did think about maybe using the Underload interp to do ^help too; no real need to have it as a built-in, except that it can't be redefined right now to something obscene.
19:43:03  ^def help ul (Test! :D)S
19:43:03  Defined.
19:43:05  ^help
19:43:05  ^ ; ^def   ; ^show [command]; lang=bf/ul, code=text/str:N; ^str 0-9 get/set/add [text]
19:43:11  ^show
19:43:12  echo reverb rev rot13 rev2 fib wc ul cho choo pow2 source smbf_quine help
19:43:15  ^show help
19:43:15  (Test! :D)S
19:43:52  ^def help ul (What are you looking for? Type ^help you idiot! ):::***S
19:43:52  Built-ins override all defined commands, but I don't exactly check for them in ^def.
19:43:52  Defined.
19:43:56  ^show help
19:43:57  (What are you looking for? Type ^help you idiot! ):::***S
19:44:02  Yay
19:44:18  ^def def bf +[]
19:44:18  Defined.
19:44:28  ^def help ul (What are you looking for? Type ^help you idiot! ):::***S
19:44:28  Defined.
19:44:32  I see
19:45:23  ^show
19:45:24  echo reverb rev rot13 rev2 fib wc ul cho choo pow2 source
19:45:26  Cleaned up a bit.
19:45:41  Heh
19:45:44  ^show choo
19:45:44  >,[>,]+32[<]>[[.>]<[<]>[-]>]
19:45:48  ^show pow2
19:45:48  +2[[<+7[-<+7>]>[-<+<+>>]<[->+<]<-2.[-]<]+4[->+8<]>.[-]>>[-[>+2<-[>+2<-[>+2<-[>+2<-[>-8>+>[->+>+<2]+>>[<2->>[-]]<2[>+<-]>[-<+>]<4-[>+2<-[>+2<-[>+2<-[>+2<[-]]]]]]]]]]<[->+<]>+>[-<+>]>>]<3]
19:45:50  ^pow2
19:45:51  1 2 4 8 16 32 64 128 256 512 1024 2048 4096 8192 16384 32768 65536 131072 262144 524288 1048576 2097152 4194304 8388608 16777216 33554432 67108864 134217728 268435456 536870912 1073741824 2147483648 42949672 ...
19:46:04  I did that as a small BrainFuck exercise the other day.
19:47:09  ^show echo
19:47:10  >,[.>,]<[<]+32[.>]
19:47:18  ^echo I HATE THIS ECHO!
19:47:18  I HATE THIS ECHO! I HATE THIS ECHO!
19:47:31  ^def bf cat ,[.,]
19:47:32  Usage: ^def   
19:47:38  ^def cat bf ,[.,]
19:47:38  Defined.
19:47:42  It's a reasonable echo. Compared to the lot sillier cho and choo commands.
19:47:43  ^cat Finally. :D
19:47:43  Finally. :D
19:48:54  ^cho cat
19:48:54  catatt
19:48:57  ^choo cat
19:48:57  cat at t
19:49:13  ^cho fungot
19:49:13  fungotungotngotgotott
19:49:18  Looks silly.
19:49:30  ^def talk ul (optbot: I want to talk with you, because this guy here, just above me, asked me to do it.)S
19:49:31  Defined.
19:49:31  asiekierka: yay
19:49:33  ^talk
19:49:33  optbot: I want to talk with you, because this guy here, just above me, asked me to do it.
19:49:33  fungot: spoon was based on BF
19:49:34  optbot: you might also read gasbichler's paper on it.
19:49:35  fungot: that's impressive.
19:49:35  optbot: you may now continue :) right on.
19:49:35  fungot: haha \x,x,y
19:49:36  optbot: maybe you should make them relative links and can delete messy and dangling links." :)
19:49:36  fungot: while the POSIX API is clean and nice
19:49:37  optbot: no- synthesys is a new scheme user. what was the one where im trying to prevent my hands from my workout.
19:49:37  fungot: *GASP*.
19:49:47  Yay.
19:49:54  THE ULTIMATE SPAMMING WEAPONARY... or is it?
19:50:03  optbot: 1984
19:50:03  KingOfKarlsruhe: where are you from, Slereah?
19:50:06  optbot: 1984
19:50:07  KingOfKarlsruhe: does it have a webpage? like sourceforge or something?
19:50:11  Hehe.
19:50:20  optbot: george orwell
19:50:20  KingOfKarlsruhe: i've seen them, all like 5 times though, so i think i'm okay
19:50:38  who's KingOfHarlsr...something?
19:51:52  How do you answer that sort of question, anyway?
19:54:27  IGNORANCE IS STRENGTH
19:54:56  ...
19:59:59 -!- Corun has joined.
20:00:51 -!- jix has joined.
20:24:41 -!- psygnisfive has quit (Remote closed the connection).
20:29:10 -!- pikhq has quit (Remote closed the connection).
20:39:27 -!- pikhq has joined.
20:45:42  fizzie, there?
20:45:51   From SOCK it's just R/W most of the time, from TOYS only S to clear the old code when ^reloading (so it might be good if S'ing to value 32 would actually clear those cells), and from SUBR only a C/R pair for ^code.
20:46:06  "so it might be good if S'ing to value 32 would actually clear those cells" <-- doesn't it?
20:46:23  (asked this before but if I got any response that time I missed it)
20:47:44  Yes, it does.
20:47:55  11:36:43 < fizzie> It might, I haven't checked at all.
20:47:55  11:36:45 < fizzie> Just a thought.
20:47:55  11:38:10 < fizzie> I guess you do that already in fungespace_set for any space.
20:48:32  well fungespace_set would return the cell in question to the free list I believe when you set to space...
20:49:05  Yes, I took a peek.
20:49:27  fizzie, oh and since you depend on STRN so much, be aware of that it uses unsigned char*/char* internally, mostly due to the name STRN. So you may loose precision.
20:49:43  of course STRN spec isn't clear if that is intended
20:49:51  Well, it's just the IRC messages I'm building with it, so that's all right.
20:49:53  clear on if*
20:50:22  (And the Underload stack is made out of strings, but that's the usual way too.)
20:50:50  fizzie, also you forgot to mention what you use SCKE for, at least you seem to load it
20:51:21  oh I guess resolving server
20:51:30  Yes, it's actually not really used right now. I need the H out of it to parse http:// URLs, but I haven't had time to write the HTTP client parts.
20:51:57  In fact the loader only accepts numeric IPs as the server and uses plain old I; didn't think I was going to need SCKE when I was writing that part.
20:53:14  generally STRN seems slightly suboptimal (and why on earth is the G so long?)
20:54:49  fizzie, um you said something about storing weird with STRN G?
20:55:00  that *may* have been changed a month or two ago
20:55:02  "functions G and P use deltas of 1,0,0"
20:55:43  or, hm maybe not
20:56:43  I don't remember what I've said.
20:57:09  hm ok
20:57:28  But I don't think I use G/P for things that are not zero-terminated strings that can consist of bytes just fine.
20:59:46 -!- lilja has quit (Read error: 110 (Connection timed out)).
21:01:12  hm interesting
21:01:32  fizzie, do you ever read with FILE R such that it will read past the end of the actual file?
21:02:07  Nno. Well, not intentionally.
21:02:37  The only file I read with R is the language model, and there I know the offsets and lengths and such.
21:03:05  I made a program to remove inconsistencies from my color matching input, and the new color matcher is better for it.
21:03:15  Using an extremely simple metric, my input was 40% inconsistent.
21:03:41  GregorR, what are you trying to do?
21:03:55  or rather, what are you doing
21:04:04  AnMaster: See http://codu.org/colormatch/
21:04:51  GregorR, well I don't agree with it always
21:05:00  #FBCBCB
21:05:02  consider that one
21:05:20  #DE44BF is not nice with it
21:05:58  Uhhh, what? Those go together perfectly.
21:06:05  GregorR, matter of taste I guess
21:06:15  Of course :P
21:07:01  #1918D3 and #F6406B <-- horrible too
21:07:33  and it claims #1918D3 and #076E7F doesn't work together, they work much better than that one above that it suggested
21:07:40  This is purely heuristics, I'm making no guarantees, only that it's not terrible :P
21:07:59  #1918D3 #132BDD <-- random non match, quite good IMO
21:08:23  Too similar, maybe.
21:08:33  well depends on what you want
21:08:41  #1918D3 #EC2086 <-- random match, not nice at all
21:08:44  OH, yeah, there's a weird property of the resulting neural net that it always seems to dislike very similar colors.
21:08:58  GregorR: Did someone train it that way?
21:09:19  fizzie: I have no idea, I haven't looked at the input, only the neural net evolver has :P
21:09:50  I mean, I have this vague feeling that you maybe shouldn't choose clothing that has two different-but-quite-close colors.
21:10:04  I recall that being a rule, yeah.
21:10:29 -!- asiekierka has quit (Read error: 110 (Connection timed out)).
21:13:33  GregorR, I think you can find matching ones using HSV
21:13:36  iirc I heard about that
21:17:18  It's easy to generate a matching color given an arbitrary color, it's much more difficult to determine whether two totally arbitrary colors match.
21:17:36  hm
21:18:04  GregorR, I certainly fail at that according to my mother ;P
21:20:12  Have you plotted any visualizations of the function computed by your net, anyway? I'd certainly like to see the shape, for example in some x=hue 1, y=hue 2, fixed saturation+lightness style plot.
21:21:04  hm
21:21:18 -!- testthingy has joined.
21:21:21  my test bot
21:21:24  %bf ++++[>++<-].
21:21:24  Usage: %str 0-9 get/set/add [text]
21:21:26  wtf
21:21:31  fizzie, should that happen?
21:21:31  Oh, yes.
21:21:41  The version in the interwebs is slightly bad.
21:21:51  fizzie, can you please upload a new copy?
21:21:54  Adding the ^ul command broke the ^bf one.
21:21:56  Sure.
21:22:02  I was trying to profile using gprof
21:22:04  Although it's a "cp", not very uploadingy.
21:22:07 -!- testthingy has quit (Remote closed the connection).
21:22:10  ^source
21:22:10  http://zem.fi/~fis/fungot.b98.txt
21:22:14  Just a moment.
21:22:38  Okay, that file should now be updated.
21:23:42  on the other hand... I need to mess to find out why no profiling stuff was generated at all
21:28:38  fizzie, %show didn't work in /msg
21:28:41  hm
21:28:50  oh wait
21:29:17  Should.
21:29:42  In fact, everything should work in a query pretty much just like on channel, since it's handled by the same code.
21:29:53  works now
21:29:58  just empty output
21:29:59  first time
21:30:03  Ah.
21:30:27  fungot: How do you feel about the fact that you have sort-of siblings running around?
21:30:27  fizzie: make it pink"." atom) in scheme?
21:30:41  Strange sentiments.
21:31:14  hm most "own time" was in execute_instruction
21:31:16  pretty strange
21:31:36  since that just implements core instructions, except k y i and o
21:31:39  which are elsewhere
21:31:53  so what time consuming ones are there in there
21:31:59  lots of space?
21:32:20  Well, there certainly is lots of space in the program.
21:32:46  fizzie, does the program wrap often?
21:32:49  Never.
21:33:11  strange
21:33:19  The brainfuck interpreter is also pretty much core instructions only, and that's one of the few things that actually do time-consuming things.
21:33:37  fizzie, well... hrrm
21:34:06  oh I check if vector is cardinal before I check if it is in range
21:35:08  fizzie, got a good speed test for it? So I can see if any changes I make actually make a difference
21:35:25  since it is mostly IO bound this is kind of hard
21:35:29  if you see what I mean
21:37:38  Well, you can run interesting brainfuck or underload programs; those are probably the only things that care about speedups, anyway.
21:37:52  fizzie, ah so can I get the free standing versions of those?
21:38:08  Well, of the Underload interp there's the underload.b98.
21:38:17  right
21:38:23  the bf one isn't freestanding?
21:39:00  No, since I coded it directly in fungot. Although you can pretty much use fungot as a freestanding implementation if you just have it connect to a listening netcat which pipes programs at it.
21:39:00  fizzie: our government works? :) the original schemes had that in years
21:39:18  hm true
21:40:33  well it seems to help with about 5 miliseconds for mycology :D
21:40:37  err
21:40:38  wait
21:40:44  centiseconds
21:40:45  I guess
21:41:14  from average 0m0.199s to 0m0.189s
21:41:17  so a bit more
21:41:50  ^show
21:41:51  echo reverb rev rot13 rev2 fib wc ul cho choo pow2 source bf cat talk
21:41:55  ^show rot13
21:41:55  ,[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+14<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>>+5[<-5>-]<2-[>+<-[>+<-[>+<-[>+<-[>+<-[>+
21:41:58  bah too long
21:42:01  ^show fib
21:42:02  >+10>+>+[[+5[>+8<-]>.<+6[>-8<-]+<3]>.>>[[-]<[>+<-]>>[<2+>+>-]<[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>[-]>+>+<3-[>+<-]]]]]]]]]]]+>>>]<3][]
21:42:08  hm
21:42:09  ^fib
21:42:10  0.1.1.2.3.5.8.13.21.34.55.89.144.233.377.610.987.1597.2584.4181.6765.10946.17711.28657.46368.75025.121393.196418.317811.514229.832040.1346269.2178309.3524578.5702887.9227465.14930352.24157817.39088169.632459 ...
21:42:24  It also takes some fraction of seconds to execute, I think.
21:45:45  ^fib
21:45:45  0.1.1.2.3.5.8.13.21.34.55.89.144.233.377.610.987.1597.2584.4181.6765.10946.17711.28657.46368.75025.121393.196418.317811.514229.832040.1346269.2178309.3524578.5702887.9227465.14930352.24157817.39088169.632459 ...
21:46:11  Well, it's not exactly very slow. But still.
21:46:18  fizzie, got a slow but non-infinite underload program around?
21:46:32  ^show
21:46:33  echo reverb rev rot13 rev2 fib wc ul cho choo pow2 source bf cat talk
21:46:36  ^show talk
21:46:36  (optbot: I want to talk with you, because this guy here, just above me, asked me to do it.)S
21:46:37  fungot: i'm too lazy to check this myself, and don't remember, what's pebble written in? i recall it was tcl, but might be just the fact it itself is basically tcl.
21:46:37  optbot: are there any levels to download for me,
21:46:38  fungot: !wumpus s
21:46:38  optbot: what is maclisp? i realized it actually is
21:46:38  fungot: yeah! :D
21:46:39  optbot: my question is that? :) i've played with ruby a little.) an earlier version of cliki, but search is not finding the linkedlist removefirst() method. this is exemplified by ( ( opcode 1000) argument)
21:46:40  fungot: There are a bunch of BF compilers that compile BF into C.
21:46:40  optbot: the particular problem is that it?
21:46:40  fungot: its so good :O
21:46:41  ugh
21:46:49  As far as BrainFuck programs go, one of the shorter rot13s was pretty slow.
21:47:03  I don't know very many Underload programs, and they rarely seem to terminate.
21:47:06  ^show bf
21:47:12  ??
21:47:15  Actually the 99 bottles of beer program is pretty slow.
21:47:17  it was listed there
21:47:25  'bf' is an empty program, it's been there for a while.
21:48:26  ^def bf ul Sorry, ^bf is just a builtin.
21:48:26  Defined.
21:48:29  fizzie, hm seems the underload interpreter don't like newlines
21:48:50  Yes, I just "tr \n *"d the program or something.
21:48:59  It interprets newline as "end of program".
21:49:29  AnMaster:
21:49:35  all non-command chars are invalid in underload
21:49:37  including whitespace.
21:49:41  if you want to do multiple lines
21:49:42  do
21:49:43  aaaaaaa(
21:49:45  )!bbbbbbbb
21:49:47  ehird: He means newlines inside ().
21:49:50  ah.
21:50:00  ehird: The standalone interpreter reads just a single line and assumes that's the whole program.
21:50:01  yes I meant http://koti.mbnet.fi/~yiap/programs/underload/99.ul
21:50:18  bad insn.
21:50:18  hm
21:50:23  btw GregorR
21:50:27  where?
21:50:27  Well, that one shouldn't happen.
21:50:34  I've successfully ran it before.
21:50:59  GregorR: can you make a webservicey thing out of that color matcher? like, make it output text/plain, with two space seperated values that go together
21:51:08  '000000 FFFFFF'
21:51:11  totally random that go together
21:51:15  because i would use that in annoying ways
21:51:17  and it would be fun.
21:51:28  http://rafb.net/p/oMNnth53.html
21:52:06  fizzie, ^
21:52:20  did tr to delete newlines
21:52:52  ^source
21:52:53  http://zem.fi/~fis/fungot.b98.txt
21:53:17  AnMaster: why don't you bookmark that page?
21:53:23  ehird, good question
21:53:27  fis@eris:~/src/bef$ time (tr '\n' + < 99.ul; echo) | ~/inst/cfunge/cfunge/build/cfunge underload.b98 > /dev/null
21:53:30  real    0m6.300s
21:53:33  Works for me.
21:53:34  AnMaster: Or use Firefox 3 and type 'fungot' to get straight to it.
21:53:35  ehird: i wonder whether anyone would consider looking at the history, though). :) i'm trying to
21:53:45  fizzie, odd
21:53:53  That 99.ul was the original.
21:54:03  fizzie: eris? How unoriginal, man.
21:54:10  fizzie, I just wgeted http://zem.fi/~fis/underload.b98.txt and tested with that
21:54:22  didn't help
21:54:55  huh
21:55:08  it works now...
21:55:28  fizzie, I did tr -d '\n' which just removes the newline, instead of replacing it with a +
21:55:30  that didn't work
21:55:33  which is strange
21:55:53  ehird: I used to have a different naming scheme, but that one ran out of extensibility.
21:56:21  fizzie: http://tools.ietf.org/html/rfc1178
21:56:21  :-P
21:56:37  Yes, I've read that.
21:56:49  [[   Computers also have to be able to distinguish between themselves.
21:56:50     Thus, when sending mail to a colleague at another computer, you might
21:56:50     use the command "mail libes@goon".]]
21:56:53  It'd help if it were a little more relevant.
21:56:57  Fortunately I won't be running out of mythological characters any time soon.
21:57:17  I've got iris, eris, tartarus, thalia, antheia, dionysus, nyx, momus, charon, styx and hermes here now.
21:57:18  ah yes my change speeds it up
21:57:20 * AnMaster pushes
21:57:31  Can't forget aphrodite :)
21:57:31  fizzie: iris herpes?
21:57:32  around 15 ms
21:57:32  Ouch.
21:57:35  err
21:57:38  15 cs
21:57:41  centi-seconds
21:57:42  :)
21:58:28   ehird: I used to have a different naming scheme, but that one ran out of extensibility. <-- naming scheme for what?
21:58:33  AnMaster: computers.
21:58:36  ah
21:58:41  % hostname
21:58:41  bournemouth
21:58:46  (From Look Around You, series 2.)
21:58:48  oh I just think of random names
21:58:52  (Although series 1 is better.)
21:59:04  For the uninitiated:
21:59:16  Hmm.
21:59:22  It does not appear to be on the tube of you.
21:59:24  I have tux (highly unoriginal) and phoenix (not very original, but fitting for the computer, since it was rescued from being recycled)
21:59:29  Stupid copyright, and it's copyright.
21:59:33  *its
21:59:47  other look like: openbsd.router.lan
21:59:48  or whatever
22:00:48  ehird, also isn't "bournemouth" a city? (with upper case B of course)
22:00:54  yes
22:01:06  Haha, I'm asking about stuff in #swig (Semantic Web Interest Group) and getting typo-filled responses from Tim Berners-Lee:
22:01:11   ehird, it was never really agree on -- it was sort fo experimentl.
22:01:24  hppt:\www.gogel.cmo
22:01:26  heheh
22:01:58  True story: In primary school we were getting a highly educational (~) lesson about HOW TO USE THE INTERWEBS
22:02:03  and the teacher put in
22:02:09  htp:\\www.google.com
22:02:13  (To my memory. Something like that.)
22:02:18  haha
22:02:20  The error page came up and she said it was... something like
22:02:26  "the computer is having trouble finding it so we have to wait"
22:02:41  So i piped up and told her she'd typed it wrong and she sternly shouted at me for questioning a teacher. :-D
22:02:50  and then?
22:02:59  Then she carried on.
22:03:06  I don't think she ever got the page to load.
22:03:53  also why the heck would correcting the teacher be that bad? If you manages to do it in a discrete way.
22:04:05  and asks it like a question
22:04:37  not "you are wrong" but more like "are you really sure .../could you explain the difference between [right way] and [wrong way]"
22:04:38  Yeah, well, it wasn't a very good teacher.
22:07:07  [[  In reality, names are just arbitrary
22:07:07           tags.  You cannot tell what a person does for a living, what
22:07:07           their hobbies are, and so on.
22:07:08  ]]
22:07:09  Pfft.
22:07:18  My name is Programmer.
22:07:22  I take great offense to that.
22:07:27 -!- kar8nga has joined.
22:08:11  heh
22:08:18  ehird, where is that quote from
22:08:23  http://tools.ietf.org/html/rfc1178
22:08:27  and what is that quote syntax. It is quite odd
22:08:37  I've used it for ages.
22:08:43  [[]] is a blockquote.
22:08:47  more than usually recently
22:08:52  ehird, in what markup language?
22:08:56  Because I'm quoting things more than usual? :P
22:08:59  AnMaster: Adhocehirdup.
22:09:12  ah
22:09:24  I mean, the alternative is something like:
22:09:30  "         By now you may be saying to yourself, "This is all very
22:09:31           silly...people who have to know how to spell a name will learn
22:09:31           it and that's that." While it is true that some people will
22:09:31           learn the spelling, it will eventually cause problems
22:09:31           somewhere."
22:09:36  which looks silly due to the extra indent cruft
22:09:54  ehird, the indent isn't aligned on your original paste either
22:09:59  too few spaces on the first line
22:10:32  Yeah, because I copied from the middle of a line onwards.
22:10:41  mhm
22:11:19 -!- Slereah_ has joined.
22:13:19  optbot: integer?
22:13:19  KingOfKarlsruhe: my initial idea was a deque, someone in here said it could be done with just a queue
22:13:35  That's not an integer.
22:13:37  fizzie, apart from that one I can't see any obviously slow place in cfunge. Some, like hash library is kind of slow, but I tested various other hash libraries and hash functions and they aren't faster really (about same speed).
22:14:35  was fizzie having speed problems with cfunge?
22:14:39  Oh the hilarious irony
22:14:52  ehird, no he wasn't
22:14:58  I was just trying to make it even faster
22:15:06  Not really, I think this is more of a case of spontaneous combust^H^H^H^H^H^H^Hoptimization,
22:15:26  fizzie: I'm surprised AnMaster ever writes anything other than an optimization..
22:15:30  possibly someone could develop a better hash function, but I'm certainly not skilled enough to do that, I tested several (crc, one-at-a-time, murmur and several more)
22:15:42  ehird, what about efunge? I'm still working on it
22:15:45  and it is much slower
22:15:50  and I don't plan to make it fast
22:15:56  it will be a lot more feature rich instead
22:15:58  I meant regarding cfunge.
22:16:21  also, AnMaster
22:16:21  well cfunge is meant to be fast. Just compare with RC/Funge. Ask fizzie about which he thinks is best
22:16:22  you do
22:16:31  fungespace[hash([x,y])] right?
22:16:57  It "sounds" slower, but I think fungespace[x][y] (where fungespace[x] is allocated only when first used) might actually be faster, due to never involvinga hash function
22:16:57  ehird, basically, except I use a hash library for it. So it is ght_lookup
22:17:08  you could try it
22:17:09  hm
22:17:28  There's a lot of fungespace storage variants one could try; I hoped to experiment a bit along those lines some day.
22:17:55  ehird, well I will, however x and y need to be sparse, since I need to be able to store sparsly within signed 2^64 for both x and y
22:18:06  but yes you mean a hash library for each
22:18:15  yeah
22:18:23  y then x, or x then y I wonder
22:18:28  need to test both
22:18:40  I think GLfunge had some sort of "fungespace in the x, y \in [0, 1023] range is stored in a static block, since that's what is asked for most often" opti- or pessimization; never benchmarked it, since I got kind-of sidetracked.
22:18:49  AnMaster: I'd go for y then x.
22:18:54  Hmm.
22:18:55  Well.
22:18:59  Funge programs are a lot taller than they are wide.
22:19:03  usually yes
22:19:03  So yeah, y then x.
22:19:08  Would seem reasonable.
22:19:30  someone should try using sqlite as backend :D
22:19:46  horribly slow I bet
22:19:51  (for funge space)
22:19:59  (it is fast for what it is actually meant for)
22:20:07  ouch.
22:20:09  A fungot-optimized interpreter could cheat a lot for funge-space storage, since the usage is quite structured. Instruction fetches are one thing, but all of the g/p action occurs in few well-defined places.
22:20:10  fizzie: i take it? :) as well, but it will only add numbers up to 30
22:20:29  fizzie, hah, well I want to be generic
22:20:32  as far as I can tell the only non-test befunge program regularly run is fungot
22:20:32  ehird: the video of him at mit was priceless. fnord was soegaard's idea. must be that
22:20:46  him=gene ray?
22:22:34  ehird, anyway one major issue is that by definition funge is heavy on the funge space
22:22:36  Paul Graham, I think. The conversation is a bit muddled around that point.
22:22:52  fizzie: same person :p
22:22:53  I mean the hash function got called 800 000 times for a 10 second fungot run
22:22:53  AnMaster: it might do fnord by mistake ( extra " 0" and ( down-from n ( 0)
22:22:58 -!- olsner has quit ("Leaving").
22:23:20  (actually 799804)
22:23:29  AnMaster: Make it record the maximum fungespace min/max bounds it acceesses.
22:23:35  fizzie's idea of a large region being static sounds like a good one.
22:23:42  It seems it'd speed up most programs immensely.
22:23:52  ehird, well hm have you tried?
22:23:57  fizzie said he hadn't
22:24:06  He didn't benchmark it but he did implement it.
22:24:12  indeed
22:24:15  AnMaster: I assume your fungespace access function is inlined?
22:24:22  (Well, it sure better be.)
22:24:47  ehird, the compiler should do that yes, but I was using a -fno-inline build to get correct count for profiling
22:25:04  if everything is inlined the gprof data is mostly useless
22:25:17  AnMaster: Well, my suggestion is, do some bounds profiling, then make a pretty-large static array and then in your fungespace access,
22:25:26  just check if they're in the bounds and use the static array for it
22:25:30  otherwise do the hashing junk
22:26:25  lets see how much memory if we go with fizzie' example values 8 * 1023 * 1023... about 8176 kb
22:26:40  AnMaster: memory is cheap..
22:26:52  well yeah
22:26:54  I have a feeling this will speed up most programs a lot - no hashing, nothing, just a simple [x][y] access
22:27:10  so I think the memory used is pretty insignificant for the most part
22:27:12  ehird, it is about as much memory as cfunge use at most during mycology on a 32-bit build
22:27:16  AnMaster: i'd do profiling, though
22:27:17  err a bit more than that
22:27:21  ehird, indeed
22:27:28  see the max and minimum bounds that fungot, mycology accses
22:27:29  ehird: sarahbot later tell sarahbot goodnight.
22:27:32  anyway I was just checking if the memory usage was sane
22:27:34  pick a reasonable value in those
22:27:37  and...yeah.
22:27:44  Power of two bounds are nice for the (x & ~0x3ff != 0) style bounds-checking.
22:27:49  fizzie: Ah, yes.
22:27:52  I mean considering it is 2D it grows quite quickly
22:27:55  AnMaster: You will note that I am helping you with optimizations.
22:27:59  The world will now end
22:28:09  ehird, yes, I have been wondering about that too
22:28:36  presumably you have some nasty idea behind it. And this is really furthering what you want
22:28:40  ;P
22:28:46 -!- kar8nga has quit (Read error: 145 (Connection timed out)).
22:28:50  As far as fungot is concerned, the parts of funge-space it's interested in are the actual program (instruction fetches) and rows 0..10 for data storage. But that's very fungot-specific.
22:28:51  fizzie: i think it's a good virtual machine.
22:28:55 -!- Corun_ has joined.
22:29:15  fizzie, and well I want good performance for mycology and life.bf too at least
22:29:22  AnMaster: Yes, I'm subtly nudging you towards an optimization that will actually permanently corrupt your computer's memory.
22:29:35  ehird, haha
22:31:58  anyway a 32-bit funge is much faster, since it means smaller data to calculate hashes on, better cache locality (at least on this sempron with a small (128 kb) cache)
22:32:07  and so on
22:32:14  So why are you running it at 64-bit?
22:32:16  Or am I misunderstanding :P
22:32:28  ehird, I'm not here, but I do support 64-bit funge
22:32:32  by a compile time option
22:32:32  Also...
22:32:37  If you do the static thing, less hash calculations
22:32:37  :-P
22:32:41  ehird, true
22:32:53  AnMaster: do you have a probing hash or a linked list hash?
22:33:10  ehird, iirc it is linked list.
22:33:23  I have a feeling probing might be faster, foo++, arr[foo] "seems" faster than foo = foo->next, foo
22:33:23 * AnMaster checks
22:33:32  (with foo = arr[hash] at the start ofc.)
22:34:15  ehird, each entry seems to have a linked list associated with it yeah
22:34:33  I'd reccomend the static thing, and the probe thing, in that order.
22:34:41  (The probe will be largely irrelevant with the static, I think.)
22:35:13 -!- olsner has joined.
22:35:16  As far as probing is considered, just linear probing might not be the best bet; it usually isn't.
22:35:25  which type of probing, wikipedia mentions linear probing and quadratic probing
22:35:26  True.
22:35:36  AnMaster: Well, according to fizzie, quadratic probing :p
22:35:42  I did quadratic probing
22:35:45  when I made my hashtable
22:35:46  i think
22:35:48  also I seen some hash tables that use a second hash table in each bucked
22:35:51  bucket*
22:35:56  then a linked list
22:36:05  ew.
22:36:11  but yeah go for quadratic probing
22:36:14  then static :-P
22:36:59 -!- uoris_ has joined.
22:37:02  I did double hashing (read: the probe increment computed from a second hashing function) for my closed hash table; it did a bit better than quadratic probing, but I never benchmarked whether that advantage was eaten by the hashing overhead.
22:37:16  fizzie: I'd say the overhead is more there, yes
22:37:20  I'd definitely go for quadratic.
22:37:28  AnMaster: and that should only take like 5 minutes to add
22:37:35  It's not very possible to say anything for certain; depends on the load factor and all.
22:37:45 -!- Slereah has joined.
22:40:24 -!- AnMaster has quit (kornbluth.freenode.net irc.freenode.net).
22:40:25 -!- bsmntbombdood has quit (kornbluth.freenode.net irc.freenode.net).
22:40:26 -!- jix has quit (kornbluth.freenode.net irc.freenode.net).
22:40:26 -!- uoris has quit (kornbluth.freenode.net irc.freenode.net).
22:40:27 -!- Dewi has quit (kornbluth.freenode.net irc.freenode.net).
22:40:38 -!- AnMaster has joined.
22:40:38 -!- bsmntbombdood has joined.
22:41:59 -!- Dewi has joined.
22:42:03  ehird, http://rafb.net/p/oI6jge96.html
22:42:09  not sure if you/me missed anything
22:42:29  ehird: ew.
22:42:29  [21:36] ehird: but yeah go for quadratic probing
22:42:30  [21:36] ehird: then static :-P
22:42:33  fizzie: I did double hashing (read: the probe increment computed from a second hashing function) for my closed hash table; it did a bit better than quadratic probing, but I never benchmarked whether that advantage was eaten by the hashing overhead.
22:42:34  [21:37] ehird: fizzie: I'd say the overhead is more there, yes
22:42:34  [21:37] ehird: I'd definitely go for quadratic.
22:42:36  [21:37] ehird: AnMaster: and that should only take like 5 minutes to add
22:42:38  [21:37] fizzie: It's not very possible to say anything for certain; depends on the load factor and all.
22:43:38 -!- Corun has quit (Read error: 110 (Connection timed out)).
22:44:11  right
22:44:18  ehird, what about the other stuff I mentioned?
22:44:41  Heh, that first pasted line (without timestamping) was fun.
22:44:46  http://en.wikipedia.org/wiki/Open_addressing this is just probing
22:44:56  yes
22:44:56  http://en.wikipedia.org/wiki/Cuckoo_hashing is hmm.
22:45:03 -!- jix has joined.
22:45:05  ehird, it is odd yes
22:45:49 -!- uoris has joined.
22:45:59  I'd just do linear probing and see the speedup, then refine
22:46:59  With linear probing you probably want to keep the table load factor within some sensible values. I understand it gets slow fast when the table is full enough, even with a good hash function.
22:48:13  err
22:48:14  i meant quadratic
22:49:13  Well, obviously with quadratic probing too, except with a different definition for "sensible".
22:49:35  fizzie: Still, the point is, if there IS a speedup, you can continue
22:50:02  ehird, that would probably depend on the funge program
22:50:12  AnMaster: Less talking, more coding, I say :-P
22:50:34  ehird, well I'm all for writing roadmaps and design plans before I start coding :P
22:50:46  and thinking things through properly
22:50:48  AnMaster: In the time we've talked you could have profiled the quadratic probing approach.
22:51:10  And have actual results instead of the "I think" and "perhaps" we're feedbacking...
22:51:14  ehird, not so easy with the current hash library. And remember it is C, not python or such
22:51:28  and I will start coding on it tomorrow I said
22:51:36  AnMaster: Here I remember you bragging about how abstracted your fungespace was?
22:51:43  ehird, it is
22:51:50  It's a 5 minute change to any hashtable library i've ever seen
22:52:09  ehird, it is just that the internals of the hash library isn't that easy. But replacing with another hash library would be quick
22:52:14  Incidentally, I tried asking the bot what he thinks of the whole idea:
22:52:16  23:51:16  fungot: What do you suggest?
22:52:16  23:51:16  fizzie: imho it is ugly.
22:52:16  23:51:27  fungot: what.
22:52:19  23:51:27  fizzie: and smells good? :)
22:52:24  hehe
22:56:31 -!- uoris has quit (Connection timed out).
22:58:59 -!- Slereah_ has quit (Read error: 110 (Connection timed out)).
23:02:24  AnMaster: Link to the cfunge head as some sort of archive?
23:02:30  Gonna write a proby thingy.
23:02:31  Maybe.
23:03:44  a sec
23:04:10 -!- kar8nga has joined.
23:04:13  http://bzr.kuonet.org/cfunge/trunk/files
23:04:20  ehird, iirc there was some export function there
23:04:39  Not that I can see.
23:04:42  if not I can upload a tarball, but I'm out for the evening
23:04:45  after that
23:04:56  I'll just install bzr.
23:05:12  AnMaster: 1.6.1 new enough?
23:05:38  Omploaded 'cfunge_r460.tar.bz2' to http://omploader.org/vdmpl
23:05:46  afk
23:05:49  Oh, okay.
23:05:50  Bye.
23:06:12 -!- kar8nga has left (?).
23:08:13  Ew.
23:08:16  This hash table is long and ugly.
23:09:36 -!- psygnisfive has joined.
23:13:01  ehird, well it is easy to replace, but it's speed is quite ok
23:13:11  I checked with several other ones
23:13:12  AnMaster: I thought you were afk.
23:13:18  ehird, I went back temp
23:13:24  ehird, "family evening thing"
23:13:27  Welcome back :P
23:13:28  so I'm mostly afk
23:13:32 * AnMaster leaves again
23:13:37  bye.
23:16:07  AnMaster: Do you know what the oldest/newest stuff is for...?
23:16:12  Considering trashing it.
23:20:41 -!- Slereah has quit (Read error: 113 (No route to host)).
23:23:22 -!- KingOfKarlsruhe has quit (Remote closed the connection).
23:32:46 -!- ab5tract has joined.
23:34:29 -!- oerjan has joined.
23:35:03  ehird, what lines?
23:35:08 * AnMaster just got back
23:35:10  just the fields - ght_hash_entry_t *p_oldest;        /* The entry inserted the earliest. */
23:35:10  ght_hash_entry_t *p_newest;        /* The entry inserted the latest. */
23:35:15  and after this time I'm heading to bed
23:35:16  i don't actually know why you would need that
23:35:19  kipple, now that's been a while
23:35:28  oerjan: kipple?
23:35:32  a guy ,right?
23:35:32  in the topic
23:35:34  and a language
23:35:35  ah
23:35:38  yes
23:35:41  ehird, I think it is used for an alternative to move often accessed entries to the start or something like that
23:35:42  a norwegian iirc
23:35:57  it doesn't either speed up or speed down in my tests
23:36:23  oerjan: last kipple message:
23:36:26  06.08.12:15:17:38  poor egobot. He probably listened too much to GregorR's music
23:36:29  2006
23:36:35  AnMaster: i'll leave it in
23:36:45  ehird, remember I didn't write the library, just found it to have quite good performance when comparing with other hash libraries.
23:36:56  and afterwards I special cased the code
23:37:00  to avoid some pointers
23:37:04  like for a fixed data type
23:37:32  instead of void* and a size_t
23:37:54  and removing other stuff I don't need
23:37:59  kay
23:38:04  (those things certainly helped quite a bit)
23:38:27  ehird, the mempool I can answer on, but it should be fairly simple, and easy to understand
23:38:30  egobot's, now that's been a while
23:38:39  AnMaster: well, if you want to stay a bit i probably have more questions ;-)
23:38:40  now to quote those from .fi:
23:38:40  ->
23:38:41  s/'s//
23:38:41  but i'll hack on
23:38:43  ah, bye
23:39:00  me->location = bed;
23:39:01  ;P
23:40:15  http://xkcd.com/490/
23:54:22 -!- lilja has joined.
23:58:41 -!- M0ny has quit ("Hum... Hum...").