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..
00:37:19 -!- 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 <GregorR> 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 <GregorR> Result: My shirt matches my tie 8-D
01:40:03 <oerjan> or so one would assume
01:55:10 <ENKI-][> GregorR: http://photos-b.l3-t.facebook.com/photos-l3-sf2p/v355/140/51/644027966/n644027966_1050121_9574.jpg <-- wear this.
01:55:50 <GregorR> Also, why is nobody as excited about my aesthetic color chooser as I am :P
01:56:20 <ihope> Because Google Chrome is being slow.
01:57:04 <ihope> And having tie fetish attacks.
01:57:11 * ihope re-ponders the subject of that sentence
01:58:04 <oerjan> ENKI-][: i cannot make out who i am supposed to obey there
01:58:23 <ENKI-][> oerjan: it's OBEY spelled backwards.
01:59:06 <ENKI-][> oerjan: i also have a CTHULU/DAGON '08 shirt to wear to erection day.
02:00:55 <oerjan> also i prefer the Allosaurus to Cthulhu
02:04:01 <ENKI-][> dagon is the one who made a small new england port town sacrifice virgins to it until it ate the whole down
02:05:25 <oerjan> +ul (a::aaa:::)((!()):a)~*^(~S:^):^
02:05:26 <thutubot> (((((!())))))(((((!())))))(((((!())))))(((((!())))))((!()))((!()))!() ...S out of stack!
02:06:13 <oerjan> GregorR: 0 is rejected, 1 accepted i take
02:06:18 <ENKI-][> ooh. what language is that? i remember seeing it somewhere
02:06:39 <GregorR> 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 <GregorR> Heh, this stupid project spanned three languages X-D
03:21:57 -!- ab5tract has joined.
03:22:55 <Asztal> am I right in thinking it presses "check match" for you if you use the random button?
03:23:08 <Asztal> because I can't find one which doesn't match #489764 :(
03:23:36 <Asztal> though the result does flicker from 0 to 1 occasionally
03:23:42 <GregorR> When you press random it chooses a random /matching/ one.
03:24:05 <GregorR> It generates a random number, checks it, and if it fails, loops.
03:25:16 <Asztal> ah, yes, #ff00ff quite easily results in 0
03:28:58 <oerjan> +ul (a::aaa:::)((!!):a)~*^()(:a~*(*a)~*a(~:)~*(~^)*^!^):^S
03:30:52 <oerjan> +ul (a::aaa:::)((!!):a)~*^()(~S:^):^
03:30:53 <thutubot> (((((!!)))))(((((!!)))))(((((!!)))))(((((!!)))))((!!))((!!))!! ...S out of stack!
03:31:23 <oerjan> +ul (a::aaa:::)((!!):a)~*^()(~aS:^):^
03:31:24 <thutubot> ()((((((!!))))))((((((!!))))))((((((!!))))))((((((!!))))))(((!!)))(((!!)))(!!) ...a out of stack!
03:38:25 <oerjan> +ul (a::aaa:::)((!!):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^S
03:39:50 <oerjan> +ul (a::aaa:::)((!!):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*):^(~a^S:^):^
03:39:51 <thutubot> ~:(*a(a(:^)*(*a)~*a(~:)~*(~^)*):^)~^(((((!!)))))(((((!!)))))(((((!!)))))(((((!!)))))((!!))((!!))!! ...a out of stack!
03:40:26 <oerjan> +ul (a::aaa:::)((!!):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*):^(~aS:^):^
03:40:27 <thutubot> (~:(*a(a(:^)*(*a)~*a(~:)~*(~^)*):^)~^)()((((((!!))))))((((((!!))))))((((((!!))))))((((((!!))))))(((!!)))(((!!)))(!!) ...a out of stack!
03:46:08 <oerjan> +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^S
03:46:09 <thutubot> (((((((((((!!()()))))))(((((!!()()))))))(((((!!()()))))))(((((!!()()))))))((!!()())))((!!()())))
03:46:33 <oerjan> +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(~aS:^):^
03:46:35 <thutubot> ((((((((((((!!()()))))))(((((!!()()))))))(((((!!()()))))))(((((!!()()))))))((!!()())))((!!()())))) ...a out of stack!
04:02:10 <oerjan> +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(~*a)~*a(~:)~*(~^)*^!^):^(~aS:^):^
04:02:12 <thutubot> ((((!!()()))(((!!()()))((((((!!()())))))((((((!!()())))))((((((!!()())))))((((((!!()())))))))))))) ...a out of stack!
04:29:14 -!- uoris has quit (Read error: 110 (Connection timed out)).
04:29:20 <oerjan> +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*)~*(!^^)*~a*(:^)*^):^
04:29:37 -!- uoris has joined.
04:30:39 <oerjan> +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*)~*(!^^)*~a*(:^)*):^(~aS:^):^
04:30:40 <thutubot> (~a*^((^)((:)S)(!(A)S(^)*())())!^^(^):^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*)~*(!^^)*~a*(:^)*)((((((((((((!!()()))))))(((((!!()()))))))(((((!!()()))))))(((((!!()()))))))((!!()())))((!!()())))) ...too much output!
04:44:13 <oerjan> +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*(:^)*^):^
04:44:37 <oerjan> +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*(:^)*^):^
04:45:57 <oerjan> +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*(:^)*^)()~^(~aS:^):^
04:45:58 <thutubot> ()(!!()())(((((((((!!()()))))))(((((!!()()))))))(((((!!()()))))))(((((!!()()))))))()(!(A)S(^)*())((:)S)(^) ...a out of stack!
04:46:04 -!- immibis has joined.
04:48:30 <oerjan> +ul (a::aaa:::)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*()*^):^(~aS:^):^
04:48:31 <thutubot> (~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*()*^)(!!()())(((((((((!!()()))))))(((((!!()()))))))(((((!!()()))))))(((((!!()())))))) ...too much output!
04:52:19 <oerjan> +ul (a::aaa:::aa)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*(:^)*^):^
04:52:35 <oerjan> +ul (::aaa:::aa)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*(:^)*^):^
04:53:34 <oerjan> +ul (:a:)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*(:^)*^):^
04:54:06 <oerjan> +ul (a:a:)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(^)(~:a(((:)S)(!(A)S(^)*())())*(^)~a*(~a*^)*~*(!^^)*~a*(:^)*^):^
04:54:53 <oerjan> +ul (a:a:)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(~aS:^):^
04:54:54 <thutubot> (((((((!!()()))))(((!!()()))))((!!()())))) ...a out of stack!
04:54:56 <immibis> i see a lot of smiley faces, cakes and angels
04:55:25 <oerjan> +ul (::a:)((!!()()):a)~*^()(a(:^)*(*a)~*a(~:)~*(~^)*^!^):^(~aS:^):^
04:55:26 <thutubot> (((((((!!()())))((!!()())))(!!()()))(!!()()))) ...a out of stack!
05:01:09 <oerjan> hm there appear to be two sets of outer parentheses
05:01:48 <oerjan> that would clearly cause some bug
05:09:16 <immibis> 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 <oerjan> since i'm barely understanding the program myself, i don't expect anyone to agree :D
05:12:53 <immibis> is this program generating itself or something
05:13:16 <oerjan> no, i am trying to split up a list of a's and :'s
05:14:13 <oerjan> oh, there are parts that are self-generating, since that's the only way to loop in underload
05:15:32 <oerjan> but i think my brain has had enough for now
05:18:56 <oerjan> 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 :
05:22:28 <oerjan> the second part is completely wrong, because a doesn't cause a new list element
05:22:48 <oerjan> 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 <oklopol> oerjan: that's a wonderful idea
06:40:48 <oklopol> lol green is the only thing that doesn't go with black
06:41:18 <oklopol> i can answer you indirectly
06:41:31 <oklopol> by highlighting the one i was indirectly talking to
06:42:12 * oerjan assumes oklopol was misspelling GregorR as oerjan
06:42:12 <oklopol> GregorR: i would be quite interested, if i believed colors can "go together", i don't.
06:42:48 <oklopol> wonderful idea was about the underload thing
06:43:37 <oklopol> did you try to replicate the functionality, or push the string splittered on the stack?
06:44:30 <oerjan> so far, i tried to print it out with a's upper cased
06:47:14 <oklopol> 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.
06:47:40 * oerjan prefers a nailcutter for all such things
06:49:10 <oklopol> isn't that pretty gay? probably doesn't even make you bleed.
06:49:39 <oklopol> (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 <oklopol> err, well that definitely sounds pretty gay :P
06:52:18 <oerjan> my nails are so fragile that if i don't do it properly at once they start disintegrating at the least provocation
06:52:52 <oklopol> i have boring normal nails :<
06:53:45 <oerjan> well at least i don't polish them :D
06:54:16 <oerjan> although that might actually have helped with the fragility
06:55:04 <oklopol> so you fin[n]ish them, but not polish them; what other countries do you them?
06:55:57 <oklopol> hmm; lecture starts about now, i should probably consider leaving.
06:59:42 -!- oerjan has quit ("Possibus?").
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 <AnMaster> <fizzie> 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?
10:36:43 <fizzie> It might, I haven't checked at all.
10:38:10 <fizzie> 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.
12:17:15 <optbot> M0ny: i suppose there have been worse last words.
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 <AnMaster> 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 <ehird> AnMaster: You're being a bit obsessive with your bleeps
14:04:14 <ehird> AnMaster: You could just put the midi on it. :-P
14:04:43 <ehird> AnMaster: Just use a regular midi-to-mp3 rendering thing, then.
14:04:54 <ehird> I don't think there are programs that make it sound like an old mobile.
14:04:59 <ehird> At least I've never heard of any.
14:05:14 <ehird> It'd be hard, what with the whole "polyphony" thing that MIDIs have.
14:05:27 <AnMaster> otherwise I shall program my pc speaker and record it
14:05:38 <ehird> AnMaster: as i said - obsessive
14:06:04 <AnMaster> I could maybe temp mend my old phone and record that
14:09:22 <ehird> 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 <fizzie> You could take some sequencer application and change the instruments in the midi file to sound more bleepy.
14:11:00 <ehird> Even better, download a ring-ring sound off the interwebs and put it on.
14:11:05 <ehird> That's not very modern.
14:11:55 <AnMaster> ehird, I doubt the "The Internationale" exists as that ;P
14:12:03 <ehird> AnMaster: I meant just a regular ring, ring.
14:12:09 <AnMaster> fizzie, good idea. Now where to find that...
14:12:22 <ehird> That's as old as the telephone, so you can stay comfy in ancient history.
14:12:40 <AnMaster> ehird, I *will* do this, nothing you say will change my mind
14:12:50 <ehird> As I said. Obsessive.
14:17:22 <ehird> http://www.theregister.co.uk/2008/10/28/student_charged/ (Insert typical el reg disclaimer.)
14:38:52 <AnMaster> fizzie, I think I got an idea: Make a custom beepy soundfont, load that into to fluidsynth
14:43:00 <ehird> 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)
14:45:29 <AnMaster> ehird, well I think socialism is good
14:45:40 <ehird> that's wholly irrelevant to my amusement but... whatever
14:47:02 <fizzie> Heh, the "screen shot of the trauma of picking an ID name for this wiki" part was funny.
14:47:10 <fizzie> Must've been very traumatic indeed.
14:48:37 <ehird> fizzie: Don't you have any FEELINGS?
14:48:40 <ehird> The wiki was MOCKING HIM!
15:21:28 <AnMaster> well anyway I think combining a sine tone and a square one works quite weel
15:27:34 -!- sebbu has joined.
15:27:50 -!- sebbu2 has quit (Read error: 60 (Operation timed out)).
15:34:10 -!- jix has joined.
15:49:44 -!- Azstal has joined.
15:51:20 * AnMaster prepares to properly record it, stopping stuff that can cause delays and such
15:56:02 <jix> ehird: can you not just stop doing that???!
15:56:13 <ehird> jix: i lose the game regularly
15:56:18 <ehird> #esoteric is a convenient tab.
16:01:25 -!- AnMaster has quit ("System reboot.").
16:01:28 <oklopol> i don't see what's so wrong about that game
16:02:00 <ehird> oklopol: anmaster makes a point to say how much he doesn't care every time someone mentions it
16:02:11 <ehird> http://www.theonion.com/content/node/28694
16:02:31 -!- Asztal has quit (Success).
16:05:58 <oklopol> 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 <ehird> anmaster doesn't, really, though
16:06:47 <ehird> 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 <ehird> most people tend to only repeatedly state their strong opinions.
16:14:33 -!- AnMaster has joined.
16:24:37 -!- jix has quit ("CommandQ").
16:33:35 <ehird> http://www.pageflip.hu/ oh god.
16:36:30 <Azstal> not bad, but it doesn't crease if I turn it sharply enough
16:36:57 <ehird> Azstal: you can actually rip the pages off
16:37:26 <Azstal> some of the pages are un-rippable though
16:37:41 <ehird> i've spent a few minutes trying to demolish it
16:48:41 <AnMaster> (checked with mp3 too, but that file was larger and even worse sound quality
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.
17:19:03 <oklopol> how can they be ripped off?
17:20:09 <ehird> oklopol: drag them off
17:20:11 <ehird> like... ripping IRL
17:20:15 <ehird> some of them can't be
17:20:51 <oklopol> ah, doesn't work for cover.
17:20:55 <AnMaster> ah not as in circumventing drm then
17:21:17 <ehird> ripping isn't "circumventing DRM"
17:21:25 <ehird> ripping is e.g. copying audio data from a cd to a computer
17:21:29 <ehird> which is totally legal :-P
17:21:48 <ehird> depends on your interpretation of the dmca
17:21:51 <ehird> when talking about a drm'd cd.
17:21:51 <AnMaster> however these days that usually includes circumventing drm
17:22:04 <ehird> i've never circumvented any DRM once to rip a cd
17:22:05 <Slereah_> 10^n points for the nth level of recursion
17:22:37 <Slereah_> 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.
18:05:04 <ehird> the wikipedians are discussing what they'd do if someone named a book Main Page
18:05:12 <ehird> [[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 :-)]]
18:05:35 <ehird> oerjan: and where does the disambig link go?
18:05:40 <ehird> What about people searching Main Page in the search bar?
18:05:51 <oerjan> Main Page (disambiguation) of course
18:06:01 <ehird> oerjan: and where would you link to that
18:06:13 <ehird> on the top of the main page? that's distracting clutter for, like, 1% of all traffic
18:06:15 <oerjan> heh, there is a slight problem
18:06:23 <oerjan> and also free advertising :D
18:07:34 <ehird> conclusion: someone do it. The resulting bureaucratic glob will destroy Wikipedia.
18:11:34 <oerjan> hm the idea of moving the Main Page to WP: at least seems reasonable
18:11:53 <lament> wp should bloody fix their bloody software
18:12:11 <lament> they have far more urgent problems than hypothetical books
18:12:18 <lament> if you search for C# it takes you to C
18:12:27 <lament> that's called a "bug" and they haven't fixed it in years
18:12:54 <lament> (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 <lament> (same with every other note)
18:14:01 <oerjan> well the use of # is a general html thing isn't it? not restricted to wp
18:14:15 <lament> sure, but what's html got to do with it?
18:14:49 <lament> they could have some escaping mechanism
18:14:56 <lament> and let the search box be aware of it
18:16:29 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | hi kipple.
18:16:49 <oklopol> why is main_page in article space anyway?
18:17:03 <oerjan> it might be enough to trap titles ending with #, i doubt wp uses empty anchor names
18:17:35 <oklopol> so that you could have competing main_pages?
18:17:35 <ehird> oklopol: history, and if they did e.g. Portal:Main all the bookmarks would break
18:17:50 <ehird> oklopol: if they just redirected - then there's no way to put another article ther
18:17:58 <ehird> so, no benefit for extra confusion essentially
18:18:29 <oklopol> i see, backwards-compatibility, the mother of all that is ugly.
18:18:30 <ehird> also, it's not an "html thing"
18:18:47 <lament> oerjan: well, right now the problem is the title of the C# page is C-Sharp because they disallow C#
18:18:47 <AnMaster> 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 <ehird> oklopol: i think if you ran a site as big as wp you'd care about that too.
18:18:58 <ehird> C# will be sent escaped
18:19:03 <ehird> they just fuck that up in the interm
18:19:05 <lament> AnMaster: i'm not talking about the name of the HTML page
18:19:11 <lament> AnMaster: i'm talking about the title of the article
18:19:18 <lament> AnMaster: WP started with the two being the same
18:19:19 <AnMaster> http://en.wikipedia.org/wiki/C# wouldn't work well however
18:19:23 <ehird> AnMaster: no, it wouldn't
18:19:24 <oklopol> 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 <ehird> but how many people hack the urls?
18:19:30 <oklopol> if i had, now, a page as big as wp
18:19:31 <ehird> compared to people using the search box
18:19:39 <oklopol> i'd probably just close it down for the fuck of it.
18:19:41 <lament> AnMaster: the title of the article and the name of the html page are two different entities
18:19:49 <ehird> oklopol: yeah, but, nobody would put you in charge of anything.
18:19:50 <lament> if they're the same, the system is badly designed
18:20:02 <AnMaster> lament, well it used to be that way :P
18:20:06 <ehird> not even a pancake.
18:20:20 <lament> AnMaster: right, exactly
18:20:21 <ehird> 'it used to be that way'?
18:20:24 <lament> WP started off badly designed
18:20:24 <ehird> that doesn't erally make any sense
18:20:27 <ehird> it still IS that way.
18:20:40 <lament> at least WP doesn't need camelcase now.
18:20:42 <oerjan> ehird: if people hack urls they should expect technical issues anyway
18:20:44 <oklopol> ehird: perhaps not, i don't see what that has to do with anything
18:21:15 <ehird> i'm all for url-hacking
18:21:21 <ehird> but... if you're url hacking, know how to escape shit, okay
18:22:12 <lament> the article for the note C# is
18:22:38 <oklopol> would be interesting to see if something bad actually happened if wp or a related entity changed it's main page
18:22:40 <lament> but the disambig is not needed because there isn't any other C♯
18:22:45 <oklopol> that probably happens every now and then
18:23:10 <ehird> several articles are like that
18:23:15 <ehird> just because the disambig makes it easier to see in the page title
18:25:58 <oklopol> 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 <ehird> oklopol: now they do
18:26:47 <ehird> you just said it in a publicly logged channel
18:27:01 <oklopol> yeah, and someone is so gonna see that.
18:27:32 <oklopol> 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.
18:40:35 -!- Asztal^_^ has quit (Read error: 104 (Connection reset by peer)).
18:40:52 -!- Asztal^_^ has joined.
18:47:11 -!- lilja has joined.
19:19:36 -!- 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.
19:33:10 <asiekierka> [<.] - a Self-modifying BF quine... i think
19:33:12 <fungot> echo reverb rev rot13 rev2 fib wc ul cho choo pow2 source
19:36:13 <fizzie> [anything] will never print out anything.
19:37:13 <fungot> 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?
19:37:27 <fungot> asiekierka: scheme's file system interface is sadly lacking.)
19:37:45 <asiekierka> fungot: Lacking? And i don't know scheme, too!
19:37:46 <fungot> asiekierka: http://www.wftv.com/ slideshow/ news/ technology/ fnord is great.
19:38:06 <asiekierka> fungot: That link is scary. WolF TV, it may be...
19:38:07 <fungot> asiekierka: i'm an east side type...
19:38:18 <asiekierka> fungot: East side? So the east side doesn't know wolves?
19:38:53 <asiekierka> fungot: East side? So the east side doesn't know wolves?
19:39:09 <fungot> echo reverb rev rot13 rev2 fib wc ul cho choo pow2 source smbf_quine
19:39:16 <fungot> (http://zem.fi/~fis/fungot.b98.txt)S
19:39:23 <fizzie> It goes into that ignore mode if you have talked to it too much.
19:39:48 <fungot> ^<lang> <code>; ^def <command> <lang> <code>; ^show [command]; lang=bf/ul, code=text/str:N; ^str 0-9 get/set/add [text]
19:40:13 <fungot> http://zem.fi/~fis/fungot.b98.txt
19:42:38 <fizzie> 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.
19:43:05 <fungot> ^<lang> <code>; ^def <command> <lang> <code>; ^show [command]; lang=bf/ul, code=text/str:N; ^str 0-9 get/set/add [text]
19:43:12 <fungot> echo reverb rev rot13 rev2 fib wc ul cho choo pow2 source smbf_quine help
19:43:52 <asiekierka> ^def help ul (What are you looking for? Type ^help you idiot! ):::***S
19:43:52 <fizzie> Built-ins override all defined commands, but I don't exactly check for them in ^def.
19:43:57 <fungot> (What are you looking for? Type ^help you idiot! ):::***S
19:44:28 <asiekierka> ^def help ul (What are you looking for? Type ^help you idiot! ):::***S
19:45:24 <fungot> echo reverb rev rot13 rev2 fib wc ul cho choo pow2 source
19:45:44 <fungot> >,[>,]+32[<]>[[.>]<[<]>[-]>]
19:45:48 <fungot> +2[[<+7[-<+7>]>[-<+<+>>]<[->+<]<-2.[-]<]+4[->+8<]>.[-]>>[-[>+2<-[>+2<-[>+2<-[>+2<-[>-8>+>[->+>+<2]+>>[<2->>[-]]<2[>+<-]>[-<+>]<4-[>+2<-[>+2<-[>+2<-[>+2<[-]]]]]]]]]]<[->+<]>+>[-<+>]>>]<3]
19:45:51 <fungot> 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 <fizzie> I did that as a small BrainFuck exercise the other day.
19:47:18 <fungot> I HATE THIS ECHO! I HATE THIS ECHO!
19:47:32 <fungot> Usage: ^def <command> <lang> <code>
19:47:42 <fizzie> It's a reasonable echo. Compared to the lot sillier cho and choo commands.
19:49:30 <asiekierka> ^def talk ul (optbot: I want to talk with you, because this guy here, just above me, asked me to do it.)S
19:49:33 <fungot> optbot: I want to talk with you, because this guy here, just above me, asked me to do it.
19:49:33 <optbot> fungot: spoon was based on BF
19:49:34 <fungot> optbot: you might also read gasbichler's paper on it.
19:49:35 <optbot> fungot: that's impressive.
19:49:35 <fungot> optbot: you may now continue :) right on.
19:49:36 <fungot> optbot: maybe you should make them relative links and can delete messy and dangling links." :)
19:49:36 <optbot> fungot: while the POSIX API is clean and nice
19:49:37 <fungot> optbot: no- synthesys is a new scheme user. what was the one where im trying to prevent my hands from my workout.
19:49:54 <asiekierka> THE ULTIMATE SPAMMING WEAPONARY... or is it?
19:50:03 <optbot> KingOfKarlsruhe: where are you from, Slereah?
19:50:07 <optbot> KingOfKarlsruhe: does it have a webpage? like sourceforge or something?
19:50:20 <optbot> KingOfKarlsruhe: i've seen them, all like 5 times though, so i think i'm okay
19:51:52 <fizzie> How do you answer that sort of question, anyway?
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.
20:45:51 <AnMaster> <fizzie> 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 <AnMaster> "so it might be good if S'ing to value 32 would actually clear those cells" <-- doesn't it?
20:46:23 <AnMaster> (asked this before but if I got any response that time I missed it)
20:47:55 <fizzie> 11:36:43 < fizzie> It might, I haven't checked at all.
20:47:55 <fizzie> 11:36:45 < fizzie> Just a thought.
20:47:55 <fizzie> 11:38:10 < fizzie> I guess you do that already in fungespace_set for any space.
20:48:32 <AnMaster> well fungespace_set would return the cell in question to the free list I believe when you set to space...
20:49:27 <AnMaster> 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 <AnMaster> of course STRN spec isn't clear if that is intended
20:49:51 <fizzie> Well, it's just the IRC messages I'm building with it, so that's all right.
20:50:22 <fizzie> (And the Underload stack is made out of strings, but that's the usual way too.)
20:50:50 <AnMaster> fizzie, also you forgot to mention what you use SCKE for, at least you seem to load it
20:51:30 <fizzie> 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 <fizzie> 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 <AnMaster> generally STRN seems slightly suboptimal (and why on earth is the G so long?)
20:54:49 <AnMaster> fizzie, um you said something about storing weird with STRN G?
20:55:00 <AnMaster> that *may* have been changed a month or two ago
20:55:02 <AnMaster> "functions G and P use deltas of 1,0,0"
20:56:43 <fizzie> I don't remember what I've said.
20:57:28 <fizzie> 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)).
21:01:32 <AnMaster> fizzie, do you ever read with FILE R such that it will read past the end of the actual file?
21:02:07 <fizzie> Nno. Well, not intentionally.
21:02:37 <fizzie> The only file I read with R is the language model, and there I know the offsets and lengths and such.
21:03:05 <GregorR> I made a program to remove inconsistencies from my color matching input, and the new color matcher is better for it.
21:03:15 <GregorR> Using an extremely simple metric, my input was 40% inconsistent.
21:03:41 <AnMaster> GregorR, what are you trying to do?
21:04:04 <GregorR> AnMaster: See http://codu.org/colormatch/
21:04:51 <AnMaster> GregorR, well I don't agree with it always
21:05:58 <GregorR> Uhhh, what? Those go together perfectly.
21:06:05 <AnMaster> GregorR, matter of taste I guess
21:07:01 <AnMaster> #1918D3 and #F6406B <-- horrible too
21:07:33 <AnMaster> and it claims #1918D3 and #076E7F doesn't work together, they work much better than that one above that it suggested
21:07:40 <GregorR> This is purely heuristics, I'm making no guarantees, only that it's not terrible :P
21:07:59 <AnMaster> #1918D3 #132BDD <-- random non match, quite good IMO
21:08:41 <AnMaster> #1918D3 #EC2086 <-- random match, not nice at all
21:08:44 <GregorR> OH, yeah, there's a weird property of the resulting neural net that it always seems to dislike very similar colors.
21:08:58 <fizzie> GregorR: Did someone train it that way?
21:09:19 <GregorR> fizzie: I have no idea, I haven't looked at the input, only the neural net evolver has :P
21:09:50 <fizzie> 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 <GregorR> I recall that being a rule, yeah.
21:10:29 -!- asiekierka has quit (Read error: 110 (Connection timed out)).
21:13:33 <AnMaster> GregorR, I think you can find matching ones using HSV
21:17:18 <GregorR> 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.
21:18:04 <AnMaster> GregorR, I certainly fail at that according to my mother ;P
21:20:12 <fizzie> 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.
21:21:18 -!- testthingy has joined.
21:21:41 <fizzie> The version in the interwebs is slightly bad.
21:21:51 <AnMaster> fizzie, can you please upload a new copy?
21:21:54 <fizzie> Adding the ^ul command broke the ^bf one.
21:22:02 <AnMaster> I was trying to profile using gprof
21:22:04 <fizzie> Although it's a "cp", not very uploadingy.
21:22:07 -!- testthingy has quit (Remote closed the connection).
21:22:10 <fungot> http://zem.fi/~fis/fungot.b98.txt
21:22:38 <fizzie> Okay, that file should now be updated.
21:23:42 <AnMaster> on the other hand... I need to mess to find out why no profiling stuff was generated at all
21:28:38 <AnMaster> fizzie, %show didn't work in /msg
21:29:42 <fizzie> In fact, everything should work in a query pretty much just like on channel, since it's handled by the same code.
21:30:27 <fizzie> fungot: How do you feel about the fact that you have sort-of siblings running around?
21:30:27 <fungot> fizzie: make it pink"." atom) in scheme?
21:31:14 <AnMaster> hm most "own time" was in execute_instruction
21:31:36 <AnMaster> since that just implements core instructions, except k y i and o
21:31:53 <AnMaster> so what time consuming ones are there in there
21:32:20 <fizzie> Well, there certainly is lots of space in the program.
21:32:46 <AnMaster> fizzie, does the program wrap often?
21:33:19 <fizzie> The brainfuck interpreter is also pretty much core instructions only, and that's one of the few things that actually do time-consuming things.
21:34:06 <AnMaster> oh I check if vector is cardinal before I check if it is in range
21:35:08 <AnMaster> fizzie, got a good speed test for it? So I can see if any changes I make actually make a difference
21:35:25 <AnMaster> since it is mostly IO bound this is kind of hard
21:37:38 <fizzie> Well, you can run interesting brainfuck or underload programs; those are probably the only things that care about speedups, anyway.
21:37:52 <AnMaster> fizzie, ah so can I get the free standing versions of those?
21:38:08 <fizzie> Well, of the Underload interp there's the underload.b98.
21:38:23 <AnMaster> the bf one isn't freestanding?
21:39:00 <fizzie> 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 <fungot> fizzie: our government works? :) the original schemes had that in years
21:40:33 <AnMaster> well it seems to help with about 5 miliseconds for mycology :D
21:41:14 <AnMaster> from average 0m0.199s to 0m0.189s
21:41:51 <fungot> echo reverb rev rot13 rev2 fib wc ul cho choo pow2 source bf cat talk
21:41:55 <fungot> ,[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+14<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>>+5[<-5>-]<2-[>+<-[>+<-[>+<-[>+<-[>+<-[>+
21:42:02 <fungot> >+10>+>+[[+5[>+8<-]>.<+6[>-8<-]+<3]>.>>[[-]<[>+<-]>>[<2+>+>-]<[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>+<-[>[-]>+>+<3-[>+<-]]]]]]]]]]]+>>>]<3][]
21:42:10 <fungot> 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 <fizzie> It also takes some fraction of seconds to execute, I think.
21:45:45 <fungot> 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 <fizzie> Well, it's not exactly very slow. But still.
21:46:18 <AnMaster> fizzie, got a slow but non-infinite underload program around?
21:46:33 <fungot> echo reverb rev rot13 rev2 fib wc ul cho choo pow2 source bf cat talk
21:46:36 <fungot> (optbot: I want to talk with you, because this guy here, just above me, asked me to do it.)S
21:46:37 <optbot> 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 <fungot> optbot: are there any levels to download for me,
21:46:38 <fungot> optbot: what is maclisp? i realized it actually is
21:46:39 <fungot> 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 <optbot> fungot: There are a bunch of BF compilers that compile BF into C.
21:46:40 <fungot> optbot: the particular problem is that it?
21:46:40 <optbot> fungot: its so good :O
21:46:49 <fizzie> As far as BrainFuck programs go, one of the shorter rot13s was pretty slow.
21:47:03 <fizzie> I don't know very many Underload programs, and they rarely seem to terminate.
21:47:15 <fizzie> Actually the 99 bottles of beer program is pretty slow.
21:47:25 <fizzie> 'bf' is an empty program, it's been there for a while.
21:48:26 <fizzie> ^def bf ul Sorry, ^bf is just a builtin.
21:48:29 <AnMaster> fizzie, hm seems the underload interpreter don't like newlines
21:48:50 <fizzie> Yes, I just "tr \n *"d the program or something.
21:48:59 <fizzie> It interprets newline as "end of program".
21:49:35 <ehird> all non-command chars are invalid in underload
21:49:37 <ehird> including whitespace.
21:49:41 <ehird> if you want to do multiple lines
21:49:47 <fizzie> ehird: He means newlines inside ().
21:50:00 <fizzie> ehird: The standalone interpreter reads just a single line and assumes that's the whole program.
21:50:01 <AnMaster> yes I meant http://koti.mbnet.fi/~yiap/programs/underload/99.ul
21:50:27 <fizzie> Well, that one shouldn't happen.
21:50:34 <fizzie> I've successfully ran it before.
21:50:59 <ehird> 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
21:51:11 <ehird> totally random that go together
21:51:15 <ehird> because i would use that in annoying ways
21:51:17 <ehird> and it would be fun.
21:51:28 <AnMaster> http://rafb.net/p/oMNnth53.html
21:52:53 <fungot> http://zem.fi/~fis/fungot.b98.txt
21:53:17 <ehird> AnMaster: why don't you bookmark that page?
21:53:27 <fizzie> fis@eris:~/src/bef$ time (tr '\n' + < 99.ul; echo) | ~/inst/cfunge/cfunge/build/cfunge underload.b98 > /dev/null
21:53:34 <ehird> AnMaster: Or use Firefox 3 and type 'fungot' to get straight to it.
21:53:35 <fungot> ehird: i wonder whether anyone would consider looking at the history, though). :) i'm trying to
21:53:53 <fizzie> That 99.ul was the original.
21:54:03 <ehird> fizzie: eris? How unoriginal, man.
21:54:10 <AnMaster> fizzie, I just wgeted http://zem.fi/~fis/underload.b98.txt and tested with that
21:55:28 <AnMaster> fizzie, I did tr -d '\n' which just removes the newline, instead of replacing it with a +
21:55:53 <fizzie> ehird: I used to have a different naming scheme, but that one ran out of extensibility.
21:56:21 <ehird> fizzie: http://tools.ietf.org/html/rfc1178
21:56:49 <ehird> [[ Computers also have to be able to distinguish between themselves.
21:56:50 <ehird> Thus, when sending mail to a colleague at another computer, you might
21:56:50 <ehird> use the command "mail libes@goon".]]
21:56:53 <ehird> It'd help if it were a little more relevant.
21:56:57 <fizzie> Fortunately I won't be running out of mythological characters any time soon.
21:57:17 <fizzie> I've got iris, eris, tartarus, thalia, antheia, dionysus, nyx, momus, charon, styx and hermes here now.
21:57:31 <ehird> fizzie: iris herpes?
21:58:28 <AnMaster> <fizzie> ehird: I used to have a different naming scheme, but that one ran out of extensibility. <-- naming scheme for what?
21:58:33 <ehird> AnMaster: computers.
21:58:46 <ehird> (From Look Around You, series 2.)
21:58:48 <AnMaster> oh I just think of random names
21:58:52 <ehird> (Although series 1 is better.)
21:59:04 <ehird> For the uninitiated:
21:59:22 <ehird> It does not appear to be on the tube of you.
21:59:24 <AnMaster> 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 <ehird> Stupid copyright, and it's copyright.
21:59:47 <AnMaster> other look like: openbsd.router.lan
22:00:48 <AnMaster> ehird, also isn't "bournemouth" a city? (with upper case B of course)
22:01:06 <ehird> 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> <timbl> ehird, it was never really agree on -- it was sort fo experimentl.
22:01:24 <ehird> hppt:\www.gogel.cmo
22:01:58 <ehird> True story: In primary school we were getting a highly educational (~) lesson about HOW TO USE THE INTERWEBS
22:02:03 <ehird> and the teacher put in
22:02:09 <ehird> htp:\\www.google.com
22:02:13 <ehird> (To my memory. Something like that.)
22:02:20 <ehird> The error page came up and she said it was... something like
22:02:26 <ehird> "the computer is having trouble finding it so we have to wait"
22:02:41 <ehird> So i piped up and told her she'd typed it wrong and she sternly shouted at me for questioning a teacher. :-D
22:02:59 <ehird> Then she carried on.
22:03:06 <ehird> I don't think she ever got the page to load.
22:03:53 <AnMaster> also why the heck would correcting the teacher be that bad? If you manages to do it in a discrete way.
22:04:37 <AnMaster> 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 <ehird> Yeah, well, it wasn't a very good teacher.
22:07:07 <ehird> [[ In reality, names are just arbitrary
22:07:07 <ehird> tags. You cannot tell what a person does for a living, what
22:07:07 <ehird> their hobbies are, and so on.
22:07:18 <ehird> My name is Programmer.
22:07:22 <ehird> I take great offense to that.
22:07:27 -!- kar8nga has joined.
22:08:18 <AnMaster> ehird, where is that quote from
22:08:23 <ehird> http://tools.ietf.org/html/rfc1178
22:08:27 <AnMaster> and what is that quote syntax. It is quite odd
22:08:37 <ehird> I've used it for ages.
22:08:43 <ehird> [[]] is a blockquote.
22:08:52 <AnMaster> ehird, in what markup language?
22:08:56 <ehird> Because I'm quoting things more than usual? :P
22:08:59 <ehird> AnMaster: Adhocehirdup.
22:09:24 <ehird> I mean, the alternative is something like:
22:09:30 <ehird> " By now you may be saying to yourself, "This is all very
22:09:31 <ehird> silly...people who have to know how to spell a name will learn
22:09:31 <ehird> it and that's that." While it is true that some people will
22:09:31 <ehird> learn the spelling, it will eventually cause problems
22:09:36 <ehird> which looks silly due to the extra indent cruft
22:09:54 <AnMaster> ehird, the indent isn't aligned on your original paste either
22:09:59 <AnMaster> too few spaces on the first line
22:10:32 <ehird> Yeah, because I copied from the middle of a line onwards.
22:11:19 -!- Slereah_ has joined.
22:13:19 <optbot> KingOfKarlsruhe: my initial idea was a deque, someone in here said it could be done with just a queue
22:13:35 <fizzie> That's not an integer.
22:13:37 <AnMaster> 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 <ehird> was fizzie having speed problems with cfunge?
22:14:39 <ehird> Oh the hilarious irony
22:14:58 <AnMaster> I was just trying to make it even faster
22:15:06 <fizzie> Not really, I think this is more of a case of spontaneous combust^H^H^H^H^H^H^Hoptimization,
22:15:26 <ehird> fizzie: I'm surprised AnMaster ever writes anything other than an optimization..
22:15:30 <AnMaster> 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 <AnMaster> ehird, what about efunge? I'm still working on it
22:15:50 <AnMaster> and I don't plan to make it fast
22:15:56 <AnMaster> it will be a lot more feature rich instead
22:15:58 <ehird> I meant regarding cfunge.
22:16:21 <AnMaster> well cfunge is meant to be fast. Just compare with RC/Funge. Ask fizzie about which he thinks is best
22:16:31 <ehird> fungespace[hash([x,y])] right?
22:16:57 <ehird> 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 <AnMaster> ehird, basically, except I use a hash library for it. So it is ght_lookup
22:17:28 <fizzie> 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 <AnMaster> 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 <AnMaster> but yes you mean a hash library for each
22:18:23 <AnMaster> y then x, or x then y I wonder
22:18:40 <fizzie> 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 <ehird> AnMaster: I'd go for y then x.
22:18:59 <ehird> Funge programs are a lot taller than they are wide.
22:19:03 <ehird> So yeah, y then x.
22:19:08 <ehird> Would seem reasonable.
22:19:30 <AnMaster> someone should try using sqlite as backend :D
22:19:59 <AnMaster> (it is fast for what it is actually meant for)
22:20:09 <fizzie> 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 <fungot> fizzie: i take it? :) as well, but it will only add numbers up to 30
22:20:29 <AnMaster> fizzie, hah, well I want to be generic
22:20:32 <ehird> as far as I can tell the only non-test befunge program regularly run is fungot
22:20:32 <fungot> ehird: the video of him at mit was priceless. fnord was soegaard's idea. must be that
22:22:34 <AnMaster> ehird, anyway one major issue is that by definition funge is heavy on the funge space
22:22:36 <fizzie> Paul Graham, I think. The conversation is a bit muddled around that point.
22:22:52 <ehird> fizzie: same person :p
22:22:53 <AnMaster> I mean the hash function got called 800 000 times for a 10 second fungot run
22:22:53 <fungot> AnMaster: it might do fnord by mistake ( extra " 0" and ( down-from n ( 0)
22:22:58 -!- olsner has quit ("Leaving").
22:23:29 <ehird> AnMaster: Make it record the maximum fungespace min/max bounds it acceesses.
22:23:35 <ehird> fizzie's idea of a large region being static sounds like a good one.
22:23:42 <ehird> It seems it'd speed up most programs immensely.
22:23:52 <AnMaster> ehird, well hm have you tried?
22:24:06 <ehird> He didn't benchmark it but he did implement it.
22:24:15 <ehird> AnMaster: I assume your fungespace access function is inlined?
22:24:22 <ehird> (Well, it sure better be.)
22:24:47 <AnMaster> ehird, the compiler should do that yes, but I was using a -fno-inline build to get correct count for profiling
22:25:04 <AnMaster> if everything is inlined the gprof data is mostly useless
22:25:17 <ehird> 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 <ehird> just check if they're in the bounds and use the static array for it
22:25:30 <ehird> otherwise do the hashing junk
22:26:25 <AnMaster> lets see how much memory if we go with fizzie' example values 8 * 1023 * 1023... about 8176 kb
22:26:40 <ehird> AnMaster: memory is cheap..
22:26:54 <ehird> I have a feeling this will speed up most programs a lot - no hashing, nothing, just a simple [x][y] access
22:27:10 <ehird> so I think the memory used is pretty insignificant for the most part
22:27:12 <AnMaster> ehird, it is about as much memory as cfunge use at most during mycology on a 32-bit build
22:27:16 <ehird> AnMaster: i'd do profiling, though
22:27:28 <ehird> see the max and minimum bounds that fungot, mycology accses
22:27:29 <fungot> ehird: sarahbot later tell sarahbot goodnight.
22:27:32 <AnMaster> anyway I was just checking if the memory usage was sane
22:27:34 <ehird> pick a reasonable value in those
22:27:44 <fizzie> Power of two bounds are nice for the (x & ~0x3ff != 0) style bounds-checking.
22:27:52 <AnMaster> I mean considering it is 2D it grows quite quickly
22:27:55 <ehird> AnMaster: You will note that I am helping you with optimizations.
22:27:59 <ehird> The world will now end
22:28:09 <AnMaster> ehird, yes, I have been wondering about that too
22:28:36 <AnMaster> presumably you have some nasty idea behind it. And this is really furthering what you want
22:28:46 -!- kar8nga has quit (Read error: 145 (Connection timed out)).
22:28:50 <fizzie> 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 <fungot> fizzie: i think it's a good virtual machine.
22:28:55 -!- Corun_ has joined.
22:29:15 <AnMaster> fizzie, and well I want good performance for mycology and life.bf too at least
22:29:22 <ehird> AnMaster: Yes, I'm subtly nudging you towards an optimization that will actually permanently corrupt your computer's memory.
22:31:58 <AnMaster> 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)
22:32:14 <ehird> So why are you running it at 64-bit?
22:32:16 <ehird> Or am I misunderstanding :P
22:32:28 <AnMaster> ehird, I'm not here, but I do support 64-bit funge
22:32:37 <ehird> If you do the static thing, less hash calculations
22:32:53 <ehird> AnMaster: do you have a probing hash or a linked list hash?
22:33:10 <AnMaster> ehird, iirc it is linked list.
22:33:23 <ehird> I have a feeling probing might be faster, foo++, arr[foo] "seems" faster than foo = foo->next, foo
22:33:32 <ehird> (with foo = arr[hash] at the start ofc.)
22:34:15 <AnMaster> ehird, each entry seems to have a linked list associated with it yeah
22:34:33 <ehird> I'd reccomend the static thing, and the probe thing, in that order.
22:34:41 <ehird> (The probe will be largely irrelevant with the static, I think.)
22:35:13 -!- olsner has joined.
22:35:16 <fizzie> As far as probing is considered, just linear probing might not be the best bet; it usually isn't.
22:35:25 <AnMaster> which type of probing, wikipedia mentions linear probing and quadratic probing
22:35:36 <ehird> AnMaster: Well, according to fizzie, quadratic probing :p
22:35:42 <ehird> I did quadratic probing
22:35:45 <ehird> when I made my hashtable
22:35:48 <AnMaster> also I seen some hash tables that use a second hash table in each bucked
22:36:11 <ehird> but yeah go for quadratic probing
22:36:59 -!- uoris_ has joined.
22:37:02 <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.
22:37:16 <ehird> fizzie: I'd say the overhead is more there, yes
22:37:20 <ehird> I'd definitely go for quadratic.
22:37:28 <ehird> AnMaster: and that should only take like 5 minutes to add
22:37:35 <fizzie> 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 <AnMaster> ehird, http://rafb.net/p/oI6jge96.html
22:42:09 <AnMaster> not sure if you/me missed anything
22:42:29 <ehird> [21:36] ehird: but yeah go for quadratic probing
22:42:30 <ehird> [21:36] ehird: then static :-P
22:42:33 <ehird> 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 <ehird> [21:37] ehird: fizzie: I'd say the overhead is more there, yes
22:42:34 <ehird> [21:37] ehird: I'd definitely go for quadratic.
22:42:36 <ehird> [21:37] ehird: AnMaster: and that should only take like 5 minutes to add
22:42:38 <ehird> [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)).
22:44:18 <AnMaster> ehird, what about the other stuff I mentioned?
22:44:41 <fizzie> Heh, that first pasted line (without timestamping) was fun.
22:44:46 <ehird> http://en.wikipedia.org/wiki/Open_addressing this is just probing
22:44:56 <ehird> http://en.wikipedia.org/wiki/Cuckoo_hashing is hmm.
22:45:03 -!- jix has joined.
22:45:49 -!- uoris has joined.
22:45:59 <ehird> I'd just do linear probing and see the speedup, then refine
22:46:59 <fizzie> 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.
22:49:13 <fizzie> Well, obviously with quadratic probing too, except with a different definition for "sensible".
22:49:35 <ehird> fizzie: Still, the point is, if there IS a speedup, you can continue
22:50:02 <AnMaster> ehird, that would probably depend on the funge program
22:50:12 <ehird> AnMaster: Less talking, more coding, I say :-P
22:50:34 <AnMaster> ehird, well I'm all for writing roadmaps and design plans before I start coding :P
22:50:46 <AnMaster> and thinking things through properly
22:50:48 <ehird> AnMaster: In the time we've talked you could have profiled the quadratic probing approach.
22:51:10 <ehird> And have actual results instead of the "I think" and "perhaps" we're feedbacking...
22:51:14 <AnMaster> ehird, not so easy with the current hash library. And remember it is C, not python or such
22:51:28 <AnMaster> and I will start coding on it tomorrow I said
22:51:36 <ehird> AnMaster: Here I remember you bragging about how abstracted your fungespace was?
22:51:50 <ehird> It's a 5 minute change to any hashtable library i've ever seen
22:52:09 <AnMaster> 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 <fizzie> Incidentally, I tried asking the bot what he thinks of the whole idea:
22:52:16 <fizzie> 23:51:16 <fizzie> fungot: What do you suggest?
22:52:16 <fizzie> 23:51:16 <fungot> fizzie: imho it is ugly.
22:52:16 <fizzie> 23:51:27 <fizzie> fungot: what.
22:52:19 <fizzie> 23:51:27 <fungot> fizzie: and smells good? :)
22:56:31 -!- uoris has quit (Connection timed out).
22:58:59 -!- Slereah_ has quit (Read error: 110 (Connection timed out)).
23:02:24 <ehird> AnMaster: Link to the cfunge head as some sort of archive?
23:02:30 <ehird> Gonna write a proby thingy.
23:04:10 -!- kar8nga has joined.
23:04:13 <AnMaster> http://bzr.kuonet.org/cfunge/trunk/files
23:04:20 <AnMaster> ehird, iirc there was some export function there
23:04:39 <ehird> Not that I can see.
23:04:42 <AnMaster> if not I can upload a tarball, but I'm out for the evening
23:04:56 <ehird> I'll just install bzr.
23:05:12 <ehird> AnMaster: 1.6.1 new enough?
23:05:38 <AnMaster> Omploaded 'cfunge_r460.tar.bz2' to http://omploader.org/vdmpl
23:06:12 -!- kar8nga has left (?).
23:08:16 <ehird> This hash table is long and ugly.
23:09:36 -!- psygnisfive has joined.
23:13:01 <AnMaster> ehird, well it is easy to replace, but it's speed is quite ok
23:13:11 <AnMaster> I checked with several other ones
23:13:12 <ehird> AnMaster: I thought you were afk.
23:16:07 <ehird> AnMaster: Do you know what the oldest/newest stuff is for...?
23:16:12 <ehird> 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.
23:35:10 <ehird> just the fields - ght_hash_entry_t *p_oldest; /* The entry inserted the earliest. */
23:35:10 <ehird> ght_hash_entry_t *p_newest; /* The entry inserted the latest. */
23:35:15 <AnMaster> and after this time I'm heading to bed
23:35:16 <ehird> i don't actually know why you would need that
23:35:19 <oerjan> kipple, now that's been a while
23:35:41 <AnMaster> ehird, I think it is used for an alternative to move often accessed entries to the start or something like that
23:35:57 <AnMaster> it doesn't either speed up or speed down in my tests
23:36:23 <ehird> oerjan: last kipple message:
23:36:26 <ehird> 06.08.12:15:17:38 <kipple> poor egobot. He probably listened too much to GregorR's music
23:36:35 <ehird> AnMaster: i'll leave it in
23:36:45 <AnMaster> 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 <AnMaster> and afterwards I special cased the code
23:37:54 <AnMaster> and removing other stuff I don't need
23:38:04 <AnMaster> (those things certainly helped quite a bit)
23:38:27 <AnMaster> ehird, the mempool I can answer on, but it should be fairly simple, and easy to understand
23:38:30 <oerjan> egobot's, now that's been a while
23:38:39 <ehird> AnMaster: well, if you want to stay a bit i probably have more questions ;-)
23:54:22 -!- lilja has joined.
23:58:41 -!- M0ny has quit ("Hum... Hum...").