00:01:36 <elliott> Gregor: OKAY, I'm working on lance again, and making it use your far-saner store-info-in-instructions method. :P
00:06:06 -!- GreaseMonkey has joined.
00:10:43 <elliott> Err, how can you die again? Out of bounds pointer, flag zeroing... anything else?
00:10:46 -!- augur has quit (Read error: Connection reset by peer).
00:10:46 -!- augur_ has joined.
00:20:54 <Sgeo> http://codu.org/eso/bfjoust/egojsout/?l=a21e97b13671f40d14ad1fbfd92d9c8ae1d265d6&r=da39a3ee5e6b4b0d3255bfef95601890afd80709&t=30
00:20:58 <Sgeo> What's that dot on the right
00:21:07 <Sgeo> After the second and beyond time I play it
00:21:42 <elliott> switch (p->progs[w][p->ips[w]]->op) {
00:21:44 <elliott> DO I HAVE ENOUGH INDIRECTION YET
00:22:01 <Sgeo> My dumb turtle won against stuff?
00:22:46 <Sgeo> Oh, tripwire avoider, dies against non-decoy
00:23:22 <Sgeo> ais523 won at lengths 29 and 30 though
00:24:23 <Sgeo> http://codu.org/eso/bfjoust/egojsout/?l=044f70240a470b57c4dcbf5765f6ef0e1aebaf78&r=efe98166a86b1fa2c293110c5321b33e8bfc20ea&t=30 I don't get it
00:25:14 <Sgeo> Oh, n/m, I get it
00:25:59 <Gregor> Heh, I built me an ais-alike defense program :P
00:26:16 -!- cheater- has joined.
00:27:31 <cheater-> i just got a captcha with antique greek in it.
00:28:25 <elliott> Gregor: 11.02.12:22:49:35 <Gregor> Damn it, I don't want to be a person who writes BFJ programs as long as defend13 X-P
00:28:36 <Gregor> I didn't write it, I GENERATED it.
00:28:51 <elliott> Gregor: ais523 GENERATED defend9 :D
00:28:58 <elliott> which is near the top of the hill
00:29:03 -!- cheater00 has quit (Ping timeout: 240 seconds).
00:29:07 <elliott> Gregor: so have you implemented quintopia's almost-fixed-point scoring mechanism
00:29:19 <elliott> (fixed-point except in case of cycles, which I've fixed)
00:29:41 <elliott> Gregor: score(p) := sum score(q) for all q that p beats, taking initially score(p) = points(p).
00:29:47 <elliott> Gregor: It is, apparently, literally exactly that.
00:30:03 <elliott> It reaches a fixed-point, astonishingly, unless you have a cycle, and both me and quintopia had an idea for handling cycles.
00:30:21 <Gregor> OK, so my system exactly, but you just iterate :P
00:30:34 <elliott> Gregor: You have all this concept of "worth" of programs and whatnot.
00:30:37 <elliott> There's not a single division in this.
00:30:43 <elliott> You literally sum the score of all the programs it beats to get its score.
00:30:44 <zzo38> Tonight, I watch Double Jeopardy.
00:30:55 <Gregor> elliott: All that does is normalize values, it's all the same.
00:31:01 <elliott> Gregor: And it gave reasonable results like defend9.75 being at the top relative to the other top programs on the hill right now, which is correct: defend9.75 *beats* everything above it.
00:31:05 <Gregor> The only distinction is I don't penalize.
00:31:22 <elliott> Gregor: score(p) := sum score(q) for q in that p beats
00:31:35 <elliott> And you start with score(p) = points(p), and make special arrangements for cycles.
00:31:36 <Gregor> So yeah, it's exactly my system plus iteration, but cycles are what's interesting.
00:31:43 <elliott> Gregor: Well, we both had an idea for that.
00:31:52 <elliott> Mine was: If you ever get recursively asked for a program you've already considered, return 0.
00:32:06 <elliott> His was: If the same scoreboard ever appears twice, average all the scoreboards between them and halt.
00:32:38 <elliott> I'm sceptical of his -- I think that even a quick cycle between the top four programs will cause knock-on effects to the score of the programs below that might last a very long time before cycling, but he disagrees.
00:32:44 <elliott> I was hoping he was coding it up :P
00:33:06 <elliott> Gregor: Obviously this isn't exaaaactly a fixed-point because of that special case, but it gives you all the beneficial properties.
00:33:11 <zzo38> cheater-: I once got a CAPTCHA with Greek text in it. I went to the character map and selected all the Greek letters as much as I could tell, and it accepted that.
00:33:19 <elliott> And is wonderfully simple and therefore immune to disagreement :P
00:33:19 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/eYEL
00:33:35 <elliott> Gregor: Worst polyglot eve.r
00:33:44 <elliott> (Jesus christ at that program.)
00:33:56 <Gregor> elliott: Doing the proper polyglot A) is broken on egojoust and B) is a pain since we have no real comments.
00:33:58 <elliott> Gregor: For bonus points you should have written the base 64 decoder in another language :P
00:34:13 <Gregor> Any other language would have needed parens somewhere ;)
00:34:24 <Gregor> BTW, FFLDG is going to take FOR EV ER
00:34:24 <cheater-> zzo38: i just created an ad-hoc transliteration. it was obvious that the greek part is the "unknown" part of the captcha.
00:35:22 <Gregor> It's also useless against turtles and other defenders.
00:35:24 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 34.7
00:35:32 <Deewiant> !bfjoust pendolino ->+++(>)*6->+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(<)*4(+)*23<(-)*23(<(-)*30<(+)*30)*2(>)*9(>[-([+{[(+)*8[-].[.+]]}])%3])*19
00:35:40 <zzo38> cheater-: Yes, I realized that it is unknown part, too. But I did not use transliteration, I tried to enter the actual Greek letters (although it is possible I might have made a mistake since I am not Greek)
00:35:45 <EgoBot> Score for Deewiant_pendolino: 56.0
00:35:46 <cheater-> gregor: what's furry_furry_leather_discipline_girls?
00:36:03 <Gregor> cheater-: Pretty much a straight-up ripoff of ais-style defenders :P
00:36:04 <cheater-> zzo38: the greek alphabet is super-easy.
00:36:16 <cheater-> GreaseMonkey: ais-style defenders? what's that?
00:36:29 <cheater-> zzo38: greek is just like cyrillic. that makes it trivial.
00:36:45 <zzo38> cheater-: Well, yes, but there might have been some accent marks I got wrong due to distortion.
00:37:00 <elliott> <Gregor> Any other language would have needed parens somewhere ;)
00:37:18 <Gregor> OK, not Perl's OO bastard child *shrugs*
00:38:21 <elliott> Gregor: *not Perl's sort-of-Lisp's Smalltalk's bastard child
00:38:57 <Sgeo> $d =~ tr/_:/\x2B\//; what's that?
00:39:02 <cheater-> how can you say ruby is related to lisp, smalltalk, or perl?
00:39:13 <cheater-> it's like saying macdonald's is related to steaks.
00:39:29 <Gregor> Hah, me and defend9 both just wait for each other to do something forever :P
00:44:25 <Deewiant> !bfjoust pendolino ->+++(>)*5->->+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(<)*4(+)*23<(-)*23(<(-)*30<(+)*30)*2(>)*9(>[-([+{[(+)*8[-].[.+]]}])%3])*19
00:44:33 <EgoBot> Score for Deewiant_pendolino: 57.7
00:46:47 -!- elliott_ has joined.
00:50:38 -!- elliott has quit (Ping timeout: 264 seconds).
00:51:05 -!- Mathnerd314 has joined.
00:51:38 <elliott_> he's omninooinreoktmoinertneroitnioenrtojiojoijojiojo always present
00:52:10 <quintopia> to make sure my closed-form fixed-point scoring system works and does everything the current one does
00:52:48 <quintopia> oh shit i missed jeopardy while i was in the shower
00:52:56 <elliott_> OH NOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
00:53:36 <quintopia> word is that watson destroyed them
00:57:05 <zzo38> quintopia: Maybe later tonight I can tell you the score?
00:58:13 <quintopia> zzo38: 'salright. my friends are telling me all about it
00:59:14 <elliott_> "So many idiots on every side. The downside of public access and pseudo-anonimity" --reddit commenter on Wikipedia's inclusionism/deletionism debacle, apparently not realising that it's the people who aren't anonymous that handle all the egotism...
00:59:22 <zzo38> In the Single Jeopardy round, there was some strange order in which the clues are selected, which is not how they are usually done in Jeopardy (but it is still allowed)
01:01:22 -!- augur_ has quit (Remote host closed the connection).
01:01:31 <elliott_> It's almost as if I'm running... the unstable version.
01:03:50 <Sgeo> Why does Jeopardy have I'm a little teapot as its theme?
01:04:27 <zzo38> Sgeo: It doesn't, it is just a similar theme music.
01:04:47 <elliott_> Deewiant: grats on your victory btw
01:05:26 <Sgeo> zzo38, lies! I want to believe that that's a lie1
01:05:30 <Sgeo> i WANT TO BELIEVE!
01:05:57 <zzo38> Sgeo: OK believe what you want to believe
01:06:14 <quintopia> zzo38: i think watson always picks the least valuable question on the board
01:06:45 <quintopia> Sgeo: that's an insult to merv griffin! he wrote the theme himself!
01:06:47 <zzo38> The $800 were all picked last
01:07:46 <elliott_> yes. we are all chavs here in america. sorry britain.
01:08:14 <quintopia> i wasn't referring to any countries
01:11:02 <quintopia> Deewiant: that is the most ridiculously complicated rusher to date :P
01:11:14 <Sgeo> How soon until we can all have Watson in our phones?
01:11:14 <quintopia> yes, it's even more complicated than space_elevator
01:12:46 <elliott_> I like how a rush is on the top, it's so noob :P
01:12:58 <elliott_> (Not pendolino itself, but the idea of a rusher being on top.)
01:15:09 <quintopia> i had actually considered writing a rusher that did that
01:17:09 <quintopia> it rushes a bit, goes back and builds more decoys, then rushes some more
01:17:27 <quintopia> so basically, its a hybrid fast/slow rusher
01:17:40 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/SNPB
01:18:06 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 35.7
01:18:29 -!- copumpkin has joined.
01:19:19 <quintopia> A database query syntax error has occurred. This may indicate a bug in the software. The last attempted database query was:
01:19:25 <quintopia> from within function "MediaWikiBagOStuff::_doquery". MySQL returned error "1194: Table 'mw_objectcache' is marked as crashed and should be repaired (localhost)".
01:20:37 <Gregor> Oh wow, that was totally broken ...
01:23:07 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/dXVQ
01:23:30 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 28.6
01:24:17 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/hchJ
01:24:30 <elliott_> although that error is worrying
01:24:46 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 42.6
01:25:13 <Gregor> Better, but not better enough.
01:26:32 -!- Behold has quit (Remote host closed the connection).
01:26:58 <Sgeo> zzo38, what is leg?
01:29:44 -!- iconmaster has joined.
01:30:54 <iconmaster> Hey, the Esolang wiki isn't working for me. Has it actually crashed?
01:31:18 <zzo38> Sgeo: Leg? What leg?
01:31:28 <zzo38> Are you talking about Jeopardy?
01:33:41 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/OXDW
01:34:16 <Gregor> iconmaster: Seems that way.
01:34:20 <elliott_> iconmaster: Give it a day or two.
01:34:27 <elliott_> If it doesn't work by then we'll contact Graue.
01:34:34 <elliott_> (We're really terrified of Graue.)
01:34:41 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 40.4
01:34:54 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/hchJ
01:35:30 <zzo38> iconmaster: I think it probably not good to eat?
01:35:33 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 42.6
01:37:06 <iconmaster> I also like how the MediaWiki acsessor object is called 'MediaWikiBagOStuff'.
01:38:17 -!- pikhq_ has joined.
01:38:50 -!- pikhq has quit (Ping timeout: 260 seconds).
01:39:10 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/TebO
01:40:34 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 47.3
01:41:03 -!- iconmaster has quit (Quit: I hope wiki gets back up soon.).
01:46:29 <elliott_> Do you get the joke I made there
01:47:50 <Gregor> Bleh, my strategy really can't win against maglev >_>
01:48:22 <elliott_> Gregor: Just special-case maglev, it's what all the Qool Qids do.
01:57:00 <quintopia> fuck you. ping me if he SAYS SOMETHING HERE
02:05:25 <elliott_> That quintopia fellow wants me to ping him if fizzie says something in here.
02:09:15 <Gregor> lol, what do you do when you /know/ they're already on your flag, other than cry a lot X-D
02:10:47 <Gregor> Naw, speedy3 never jumps off, best I can do is a tie...
02:10:50 <elliott_> Gregor: You may want to use a tripwire to figure out your opponent's cycle count and polarity.
02:11:07 <elliott_> Gregor: Then you can lock it in place.
02:11:23 <Gregor> I use a tripwire, the problem is I need some setup first due to my other strategy, and so it can beat my to the flag before I can even set the tripwire if it's fast.
02:11:26 <elliott_> Gregor: You could do what the Advanced(TM) programs do: go destroy your opponent's flag while returning to your own at regular intervals to boost it back up (knowing the opponent's polarity) and lock it for a bit.
02:11:38 <elliott_> Also, well, if that happens, shit sucks.
02:11:42 <Gregor> Ohhey, I can totally do that 8-D
02:11:46 <elliott_> Set up faster or just resign yourself to losing to those programs on shorter tape lengths.
02:11:55 <elliott_> Gregor: I think that's what definder2 does, so have fun :P
02:12:41 <elliott_> Gregor: But, er, without your tripwire, I dunno how you'll know your opponent's polarity.
02:12:51 <elliott_> So locking it in place and boosting it back up will be non-trivial.
02:12:59 <elliott_> Actually you shouldn't need to lock.
02:13:05 <elliott_> Just boost back up, go ahead, return quickly, etc.
02:14:13 <elliott_> Gregor: I eagerly await furry_furry_married_sex_between_two_christians_in_the_missionary_position_girls.
02:14:16 <elliott_> THE MOST ADVANCED OF ALL THE FF*G PROGRAMS
02:14:39 -!- Slereah_ has quit (Ping timeout: 272 seconds).
02:16:18 <Gregor> LOVE when I can make Chrome crash the living fuck out of my system.
02:17:46 <elliott_> Gregor: But, but, it's PER-PAGE PROCESS SECURITY.
02:17:51 <elliott_> (For a definition of PAGE equal to MORE THAN A PAGE.)
02:18:48 -!- Slereah has joined.
02:21:31 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/MjAB
02:21:54 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 35.1
02:22:08 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/TebO
02:22:30 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 47.3
02:28:46 <elliott_> Man, chevrolet_movie_theater will kick all your asses.
02:29:01 <elliott_> Also: Tomorrow: lance's completion. Guaranteed.
02:30:22 -!- elliott_ has quit (Quit: Leaving).
02:33:10 -!- azaq23 has quit (Quit: Leaving.).
02:36:55 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/cgbB
02:37:23 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 49.5
02:40:03 <quintopia> so what are the odds he'll actually finish tomorrow?
02:40:32 <pikhq_> What are the odds he continues his Funge-98 interpreter?
02:40:38 <quintopia> hey gregor, since fizzie ain't about could you generate this table for me?
02:41:09 <quintopia> i have high hopes that chainlance will actually be completed tho. it seems p cool.
02:41:51 <quintopia> i want the matrix from the scoreboard but with (i,j)=number of configurations in which i beats j
02:42:18 <quintopia> it would take ages for me to get all that data from egojsout
02:43:58 <zzo38> All sorts of strange effects can be created by changing the color spaces and then using another effect and changing back.
02:48:40 * Sgeo is again reminded of Mozilla's Ghostbusters worship
02:52:22 -!- pikhq_ has changed nick to pikhq.
02:52:26 <Sgeo> "Memory Use (No Extensions)Winner: Firefox 3.6!"
02:52:30 <pikhq> Sgeo: Say, did you happen to transfer yet?
02:53:34 <pikhq> Education is wasted on you.
02:54:23 <Sgeo> pikhq, want to convince my dad that Farmingdale is horrible?
02:54:57 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/UTXJ
02:55:08 <pikhq> Sgeo: Why don't you? It is absolutely vital that you take an interest in your *own damned life*.
02:55:19 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 48.0
02:55:29 <Sgeo> pikhq, because I tried and failed?
02:55:51 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/TebO
02:56:11 <pikhq> You mean you didn't pull up a bunch of associates' programs in CS that were pretty much all superior to your course of study?
02:56:15 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 47.3
02:56:26 <pikhq> Maybe you should just stay at Farmingdale.
02:58:32 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/QGaX
02:58:45 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 49.6
03:02:14 <quintopia> not as good as deewiant, who now has the most points by like 3
03:02:29 <Sgeo> Javascript sounds like a fun language
03:03:31 <Sgeo> Of course, I have to google Impromptu language... oh it is a language
03:08:46 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/aFNY
03:09:03 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 49.2
03:09:32 <Gregor> In spite of this being worse, I'm keeping it since my other one was broken :P
03:09:41 <Gregor> Ohwait ... THIS is broken too X_X
03:10:39 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/XIPI
03:11:11 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 50.2
03:11:23 <Gregor> Finally beat 50 again.
03:14:16 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/PQEW
03:14:29 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 45.8
03:14:38 <Gregor> Didn't think that'd help :P
03:14:44 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/XIPI
03:15:29 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 50.2
03:17:31 <quintopia> stop making such good warriors. that sort of behavior is bound to make me start playing again instead of studying
03:18:19 <Gregor> space_elevator is still ahead of it.
03:22:58 <quintopia> eh, space_elevator is obsolete. it's not the WAY OF THE FUTURE
03:26:31 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/ZSAP
03:26:51 <Sgeo> There should be a Javascript MARS interpreter
03:26:56 <Sgeo> Maybe I should make one
03:26:59 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 49.2
03:27:03 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/XIPI
03:27:22 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 50.2
03:36:16 <Sgeo> Is Objective-J evil, by some measure?
03:36:21 <Sgeo> I think I'd like to learn it
03:45:54 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/EWJe
03:46:08 <Gregor> Might be worse, probably not much better.
03:46:19 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 45.5
03:46:27 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/XIPI
03:46:45 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 50.2
03:48:18 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/CDOd
03:48:21 <Gregor> Almost certainly worse.
03:48:47 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 44.6
03:48:52 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/XIPI
03:49:07 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 50.2
03:58:25 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/TiSJ
03:58:37 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 36.7
03:58:59 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/XIPI
03:59:23 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 50.2
04:10:45 <zzo38> Sgeo: Learn it if you want to learn it, then
04:10:57 -!- augur has joined.
04:12:17 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/BNJL
04:12:35 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 52.4
04:34:55 <Gregor> Might be an improvement ... or not:
04:35:07 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/gOBA
04:35:36 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 52.5
04:35:59 <Gregor> ... technically an improvement! X-P
04:37:56 <Gregor> MUST ... GET ... FIRST
04:38:03 <Gregor> LONG way to go though X_X
04:39:58 <Sgeo> Gregor, rig EgoJoust
04:40:05 <Sgeo> ^^don't actually do that
04:47:54 <Gregor> lawl, my philip core is wrong.
04:48:31 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/GTKZ
04:48:51 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 55.6
04:49:12 <Gregor> Still, lawl, my philip core was busted :P
04:49:34 <Sgeo> And I still have no idea what a Philip is, except a kid when I was a kid who I thought needed more salt
04:53:33 <Gregor> Pendolino is the current #1.
04:53:45 <Gregor> And if I don't get #1 tonight, Deewiant will improve it beyond my grasp, losing me a chance.
04:54:14 -!- Mathnerd314 has quit (Remote host closed the connection).
04:54:22 <pikhq> Unless you shoot him.
04:56:39 <Gregor> 2 | + + + + - - + + + + - 0 + + + + + + + + - - + + + + + + + - + + + + - + + + + + + + + + + + | 57.2 | 22.1 | 2 | Deewiant_pendolino.bfjoust
04:56:39 <Gregor> 6 | - + + + + + + + + + - + - + + - - + + + + + + + + - - + + + - - - + + + + + + + + + + + + + | 55.6 | 20.5 | 6 | Gregor_furry_furry_leather_discipline_girls.bfjoust
04:56:51 <pikhq> 23 minutes of awesome.
05:03:02 <Sgeo> I suppose it would be no good to attack Pendolino specifically without regard to general survivalness?
05:08:04 <Gregor> I beat the CRAP out of pendolino.
05:08:31 <Gregor> It only beats me in one configuration.
05:09:09 <Gregor> !bfjoust aspartame_philip <
05:09:16 <Gregor> !bfjoust mapping_turtle <
05:09:20 <EgoBot> Score for Gregor_aspartame_philip: 0.0
05:09:53 <EgoBot> Score for Gregor_mapping_turtle: 0.0
05:10:27 <pikhq> !bfjoust give_it_a_go +[>+]
05:10:43 <EgoBot> Score for pikhq_give_it_a_go: 0.2
05:11:50 <pikhq> At the suck, I presume.
05:13:13 <Sgeo> There exists a Kool-Aid Man video game.
05:15:00 <Sgeo> It occurs to me that there are probably lots of crappy ad games like that even today, except they're just flash games now
05:15:05 <Sgeo> And not notable at all
05:15:15 <Sgeo> Does the Dilberito game count as crappy?
05:17:39 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/GFYT
05:17:48 <Gregor> Sgeo: Not a clue to what you refer.
05:18:12 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 28.5
05:18:31 <pikhq> Is pikhq_give_it_a_go beating it now? :P
05:18:34 <Sgeo> http://www.freegamearchive.com/game.php?id=6507
05:18:38 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/GTKZ
05:18:44 <Gregor> As I recall, I just said "probably worse" >_<
05:19:01 <Sgeo> http://joking.narod.ru/64.htm
05:19:20 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 55.0
05:20:45 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/UbXj
05:20:48 <Gregor> Almost assuredly worse.
05:21:02 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 55.7
05:21:18 <Gregor> Mmmm, or a little bit better.
05:22:48 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/QYjD
05:22:53 <Gregor> I think this is actually better this time.
05:22:59 <Gregor> Mmmm, maybe not though.
05:23:13 -!- pikhq_ has joined.
05:23:30 -!- pikhq has quit (Ping timeout: 264 seconds).
05:23:33 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 54.6
05:23:40 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/UbXj
05:24:04 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 55.7
05:24:36 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/KbQG
05:24:43 <Gregor> If this is better, my whole strategy didn't work like I thought :P
05:25:06 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 46.7
05:25:13 <Gregor> Well, that's a "relief"
05:25:15 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/UbXj
05:25:29 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 55.7
05:26:08 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/OYbY
05:27:06 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 55.7
05:27:35 <Gregor> This new 55.7 is a slightly-higher 55.7 :P
05:30:39 -!- asiekierka has joined.
05:32:07 -!- monodemono has joined.
05:32:22 -!- monodemono has left (?).
05:32:31 <zzo38> Gregor: Make it display more decimal places then? Otherwise how can I know?
05:32:40 <Gregor> By looking at the report.
05:37:30 <Sgeo> My dad is yelling at my step-mom
05:37:36 <Sgeo> Apparently, she was driving badly
05:40:46 -!- sftp has quit (Remote host closed the connection).
05:43:00 <quintopia> Gregor: can you make that table thing? pleeeeeease?
05:49:21 -!- asiekierka has quit (Ping timeout: 276 seconds).
05:55:22 <Gregor> quintopia: I honestly have no idea what you're talking about.
05:56:07 <Gregor> !bfjoust furry_furry_nipple_clamp_girls http://sprunge.us/cSEQ
05:56:19 <Gregor> THIS IS THE FACE OF AN ADDICT
05:56:34 <quintopia> Gregor: a table like the one in the report, but instead of (i,j)=+,0,- have (i,j)=number of configurations in which i beats j
05:57:06 <EgoBot> Score for Gregor_furry_furry_nipple_clamp_girls: 52.9
05:57:22 <Gregor> I really thought that might do better.
05:58:00 <Gregor> quintopia: I'll do that tomorrow if you remind me, it's just a bit more of a pain than I want to get to at 1:30AM when I REALLY ought to be asleep but am instead obsessive :P
05:58:31 <quintopia> since i will be cramming all night
06:00:58 <Gregor> !bfjoust furry_furry_nipple_clamp_girls <
06:00:58 <Gregor> I'm shocked that strategy wasn't better ...
06:01:20 <EgoBot> Score for Gregor_furry_furry_nipple_clamp_girls: 0.0
06:02:38 -!- poiuy_qwert has quit (Ping timeout: 240 seconds).
06:20:43 -!- poiuy_qwert has joined.
06:26:01 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/Dejf
06:26:30 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 58.9
06:27:57 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/ILBB
06:28:20 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 58.9
06:29:23 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/bPJP
06:29:36 <Gregor> Apparently my conclusion that trails were a dead strategy was ... somewhat incorrect :P
06:29:47 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 59.0
06:30:37 <Gregor> Where else did I not add trails ...
06:31:35 <Gregor> !bfjoust furry_furry_leather_discipline_girls http://sprunge.us/FYSi
06:31:56 <EgoBot> Score for Gregor_furry_furry_leather_discipline_girls: 59.0
06:32:08 <Gregor> Aaaaaaaaaaaaaaaaaaaaaaaaaaaaaah
06:32:17 -!- pikhq_ has changed nick to pikhq.
06:36:25 <Vorpal> Gregor, why is it a perl script?
06:36:46 <Gregor> Vorpal: It includes the source for the program I used to generate it.
06:36:57 <Vorpal> Gregor, so how does this manage so well?
06:37:00 <EgoBot> Use: !bfjoust <program name> <program> . Scoreboard, programs, and a description of score calculation are at http://codu.org/eso/bfjoust/
06:37:19 <Gregor> Vorpal: It's a hybrid of virtually every strategy there is :P
06:37:44 <Vorpal> Gregor, that isn't too helpful. How does it select which strategy to use?
06:37:55 <Vorpal> all at once would be too slow
06:38:27 <Gregor> It's 2AM, I'm going to sleep, I'll describe it on the wiki tomorrow :P
06:38:40 -!- zzo38 has quit (Remote host closed the connection).
06:38:41 <Vorpal> Gregor, also.... interesting names here
06:39:08 <pikhq> Gregor: Even the +[>+] "strategy"?
06:39:21 <Gregor> pikhq: That's not a strategy.
06:39:36 <pikhq> Yes it is. It's a piss-poor one.
06:39:45 <pikhq> Though it seems < is worse.
06:40:02 <Vorpal> pikhq, < is *not* a strategy :P
06:40:34 <pikhq> Vorpal: Suicide is a strategy.
06:40:46 <pikhq> Just the most efficiently bad one.
06:48:53 -!- pikhq_ has joined.
06:49:46 -!- pikhq has quit (Ping timeout: 276 seconds).
06:57:48 <quintopia> !bfjoust wireless_frownie (>)*7+<-<(+)*18<(-)*19<(-)*19<(+)*19<(-)*17<(-)*22(>)*9([[[[-[++[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[(-)*26.[.-].[.++-------]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]>)*21
06:58:08 <EgoBot> Score for quintopia_wireless_frownie: 42.6
06:59:28 <quintopia> let's see if you're right about trails!
06:59:35 <quintopia> !bfjoust wireless_frownie (>)*7+<-<(+)*18<(-)*19<(-)*19<(+)*19<(-)*17<(-)*22(>)*9([[[[-[++[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[+[(-)*26.[.-].[.++-------]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]+>)*21
06:59:45 <EgoBot> Score for quintopia_wireless_frownie: 46.5
07:00:54 <quintopia> !bfjoust wireless (>)*8+<-<(+)*18<(-)*19(<(-)*33<(+)*33)*2<(-)*17(>)*9([[[[-[++[(+)*9.[.-].[.++-------]]]]]]]->[[[[-[++[(+)*9.[-].[.++-------]]]]]]]+>)*10[[[[-[++[(+)*9.[.-].[.++-------]]]]]]]
07:01:19 <EgoBot> Score for quintopia_wireless: 42.4
07:01:46 <quintopia> !bfjoust wireless (>)*8+<-<(+)*18<(-)*19(<(-)*33<(+)*33)*2<(-)*17(>)*9([[[[-[++[(+)*9.[.-].[.++-------]]]]]]]>[[[[-[++[(+)*9.[-].[.++-------]]]]]]]+>)*10[[[[-[++[(+)*9.[.-].[.++-------]]]]]]]
07:01:55 <EgoBot> Score for quintopia_wireless: 43.9
07:02:11 <quintopia> !bfjoust wireless (>)*8+<-<(+)*18<(-)*19(<(-)*33<(+)*33)*2<(-)*17(>)*9([[[[-[++[(+)*9.[.-].[.++-------]]]]]]]>[[[[-[++[(+)*9.[-].[.++-------]]]]]]]>)*10[[[[-[++[(+)*9.[.-].[.++-------]]]]]]]
07:02:21 <EgoBot> Score for quintopia_wireless: 45.3
07:02:42 <quintopia> conclusion: the efficacy of trails is extremely strategy dependent
07:03:19 <quintopia> !bfjoust wireless (>)*8+<-<(+)*18<(-)*19(<(-)*23<(+)*23)*2<(-)*17(>)*9([[[[-[++[(+)*9.[.-].[.++-------]]]]]]]>[[[[-[++[(+)*9.[-].[.++-------]]]]]]]>)*10[[[[-[++[(+)*9.[.-].[.++-------]]]]]]]
07:04:10 <EgoBot> Score for quintopia_wireless: 46.3
07:04:24 <quintopia> smaller decoys on the other hand...ALWAYS FAVORABLE
07:04:41 <quintopia> !bfjoust wireless (>)*8+<-<(+)*18<(-)*19(<(-)*23<(+)*23)*2<(-)*29(>)*9([[[[-[++[(+)*9.[.-].[.++-------]]]]]]]>[[[[-[++[(+)*9.[-].[.++-------]]]]]]]>)*10[[[[-[++[(+)*9.[.-].[.++-------]]]]]]]
07:04:59 <EgoBot> Score for quintopia_wireless: 47.5
07:05:08 <quintopia> also favorable: bumping your own flag
07:14:04 -!- FireFly has joined.
07:23:56 -!- deadPrincess has joined.
07:25:20 -!- deadPrincess has quit (Client Quit).
07:27:20 -!- pingveno has quit (Ping timeout: 250 seconds).
07:29:09 -!- pingveno has joined.
07:41:47 -!- FireFly has quit (Remote host closed the connection).
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:03:13 -!- pikhq_ has quit (Ping timeout: 240 seconds).
08:03:18 -!- pikhq has joined.
08:25:13 -!- aloril has quit (Ping timeout: 240 seconds).
08:37:32 -!- Wamanuz2 has joined.
08:37:39 -!- aloril has joined.
08:37:53 -!- GreaseMonkey has quit (Quit: The Other Game).
08:40:38 -!- Wamanuz has quit (Ping timeout: 246 seconds).
08:58:12 -!- Lymia has joined.
09:03:30 -!- pikhq_ has joined.
09:04:23 -!- pikhq has quit (Ping timeout: 276 seconds).
09:52:19 -!- cheater- has quit (Ping timeout: 272 seconds).
09:53:20 -!- cheater- has joined.
09:55:20 -!- mtve has joined.
10:02:33 -!- Sgeo has quit (Read error: Connection reset by peer).
10:02:59 -!- Sgeo has joined.
10:18:53 -!- pikhq_ has quit (Ping timeout: 255 seconds).
10:18:53 -!- pikhq has joined.
11:39:30 -!- Tritonio has joined.
11:46:06 -!- Tritonio has quit (Quit: Leaving).
11:59:19 -!- Lymia has quit (Read error: Connection reset by peer).
11:59:46 -!- Lymia has joined.
12:19:53 -!- cheater- has quit (Ping timeout: 240 seconds).
12:20:48 -!- cheater- has joined.
12:23:59 -!- ais523 has joined.
12:32:59 -!- ais523 has quit (Read error: Connection reset by peer).
12:35:15 -!- copumpkin has quit (Quit: Computer has gone to sleep.).
12:35:40 -!- ais523 has joined.
12:43:29 -!- augur has quit (Remote host closed the connection).
12:48:18 -!- pikhq has quit (Ping timeout: 246 seconds).
12:48:32 -!- pikhq has joined.
13:03:42 -!- Sgeo has quit (Ping timeout: 246 seconds).
13:06:57 <Gregor> You bumped me to second by modifying wireless_frownie >: (
13:11:01 <fizzie> !bfjoust evo-1 http://p.zem.fi/evo-1
13:11:03 <fizzie> This is probably just a direct ripoff of your furry_furry_leather_discipline with some random changes, and it doesn't even perform so well (at least with crank), but for the record... (feel free to wipe these things out if you don't want them contaminating the scoreboard)
13:12:00 -!- ais523_ has joined.
13:12:22 -!- ais523 has quit (Disconnected by services).
13:12:27 -!- ais523_ has changed nick to ais523.
13:12:38 <Deewiant> fizzie: It's not parsing, apparently.
13:12:44 <fizzie> Deewiant: So it seems.
13:12:48 <Gregor> Very random changes :P
13:12:55 <EgoBot> Score for fizzie_evo-1: 0.0
13:13:28 <fizzie> They're random-but-structural, but I guess there could be a buggery.
13:15:31 <Gregor> (...(..{..)...}...) is my bet :P
13:15:40 <fizzie> That shouldn't happen.
13:15:44 <fizzie> It also works in egojsout.
13:15:59 <Gregor> Then (...(...{...{...}...}...)...) is my bet
13:16:25 <fizzie> That shouldn't happen either, but I guess it's always possible.
13:17:26 <fizzie> Sure would be nice to get a more verbose error message than just "failed to parse".
13:20:00 <fizzie> I'll try a marginally different version just for giggles.
13:20:06 <fizzie> !bfjoust evo_2 http://p.zem.fi/evo-2
13:20:16 <EgoBot> Score for fizzie_evo_2: 0.0
13:20:18 <fizzie> Underscores seem to be the more common delimiter hereby.
13:20:29 <fizzie> Well, that's not surprising.
13:24:42 <quintopia> can you make me a table for the current hill where (i,j)=number of configurations in which i beat j?
13:25:27 <fizzie> Isn't that what the dpoints.png table is? Or do you mean a text version?
13:28:19 -!- ais523_ has joined.
13:28:20 -!- ais523_ has quit (Changing host).
13:28:20 -!- ais523_ has joined.
13:28:36 -!- ais523 has quit (Disconnected by services).
13:28:37 -!- ais523_ has changed nick to ais523.
13:31:28 <fizzie> http://users.ics.tkk.fi/htkallas/dpoints.txt http://users.ics.tkk.fi/htkallas/proglist.txt
13:31:42 <fizzie> It's pretty horribly formatted from numpy.savetxt, but after loading to matlab it looks just fine.
13:31:51 <fizzie> Diagonal values are NaNs.
13:32:02 <fizzie> And proglist gives the row/column number → program mappings.
13:33:26 -!- ais523_ has joined.
13:33:30 -!- ais523 has quit (Disconnected by services).
13:33:34 -!- ais523_ has changed nick to ais523.
13:46:46 <quintopia> so how do i go from this to (i,j)=number of configurations in which i beat j. do i just add 21 to everything?
13:49:27 -!- cheater- has quit (Ping timeout: 240 seconds).
13:49:57 -!- cheater- has joined.
13:50:56 <fizzie> Oh, right; that was the sum of duel points, which takes losses as -1s.
13:51:49 <fizzie> I'm not sure you can get exactly that value back from it, since win+win+loss+loss = tie+tie+tie+tie in the sum.
13:52:16 <fizzie> I can do another table with the "sum of positive points only" values.
13:54:22 -!- ais523 has quit (Remote host closed the connection).
13:54:45 -!- poiuy_qwert has quit (Ping timeout: 240 seconds).
13:55:42 -!- ais523 has joined.
13:56:54 <fizzie> http://users.ics.tkk.fi/htkallas/dppoints.txt (with a double-p) should be that, unless I miswrote something.
13:57:48 -!- ais523 has quit (Remote host closed the connection).
13:58:08 -!- sebbu has quit (Read error: Connection reset by peer).
13:58:49 -!- sebbu has joined.
13:58:50 -!- sebbu has quit (Changing host).
13:58:50 -!- sebbu has joined.
14:00:32 -!- sebbu2 has joined.
14:00:32 -!- sebbu2 has quit (Changing host).
14:00:32 -!- sebbu2 has joined.
14:01:09 <fizzie> Let's try a generated Deewiant_pendolino mutation for a chance now; it's shorter, so maybe parse errors haven't crept in.
14:01:17 <fizzie> !bfjoust evo_3 http://p.zem.fi/evo-3
14:04:01 <fizzie> According to crank and a local hill-copy, it's pretty close to identical except it beats some elliott_awkwardly_long_name.
14:04:46 <EgoBot> Score for fizzie_evo_3: 57.2
14:05:03 -!- sebbu has quit (Ping timeout: 276 seconds).
14:05:04 -!- sebbu2 has changed nick to sebbu.
14:05:51 <Deewiant> fizzie: After that first *19 everything is dead code
14:06:36 <quintopia> Deewiant: you're just upset that it beat pendolino
14:06:40 <Deewiant> !bfjoust pendolino ->+++(>)*5->->+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23(<(-)*30<(+)*30)*2(>)*9++++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
14:06:50 <fizzie> Mhm, well, you can't expect my thingie to know about that.
14:06:52 <EgoBot> Score for Deewiant_pendolino: 56.0
14:07:09 <quintopia> fizzie: what sort of GA are you using?
14:07:20 <Deewiant> !bfjoust pendolino ->++(>)*5->->+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23(<(-)*30<(+)*30)*2(>)*9++++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
14:07:31 <fizzie> quintopia: It's mostly just a hack.
14:07:36 <EgoBot> Score for Deewiant_pendolino: 55.5
14:08:03 <fizzie> There's some crossover though only on the top-level; and some random mutation.
14:08:37 <Deewiant> !bfjoust pendolino ->++(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23(<(-)*30<(+)*30)*2(>)*9++++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
14:09:03 <fizzie> quintopia: So that loops and ()-reps count as a single indivisible element for the purposes of that.
14:09:03 <EgoBot> Score for Deewiant_pendolino: 55.7
14:09:48 <Deewiant> !bfjoust pendolino ->++(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23(<(-)*30<(+)*30)*2-(>)*9++++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
14:10:01 <EgoBot> Score for Deewiant_pendolino: 56.0
14:10:07 -!- copumpkin has joined.
14:11:02 <Deewiant> !bfjoust pendolino ->+++(>)*5->->+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
14:11:13 <EgoBot> Score for Deewiant_pendolino: 55.6
14:11:41 <Deewiant> !bfjoust pendolino ->++(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
14:11:55 <EgoBot> Score for Deewiant_pendolino: 58.7
14:13:01 <fizzie> Even though the scoring function supposedly optimizes the same score, the latest-generation variants dropped down to the end of the list according to my dpoints.png-generating code. Weirdness.
14:13:33 <Deewiant> !bfjoust pendolino ->++(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*2(-<)*2(<)*5(+)*29>(-)*30>(+)*30>(-)*30>(-)*23>(+)*23(>)*9++++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
14:13:40 <EgoBot> Score for Deewiant_pendolino: 30.4
14:14:27 <Deewiant> !bfjoust pendolino ->++(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*28(>)*9(->[-([+{[(+)*8[-].[.+]]}])%3])*19
14:14:47 <Deewiant> I thought it probably wouldn't be but wanted to be sure
14:14:54 <EgoBot> Score for Deewiant_pendolino: 51.4
14:15:04 <fizzie> !bfjoust evo_3 http://p.zem.fi/evo-3b
14:15:05 <Deewiant> !bfjoust pendolino ->++(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*28(>)*9(>[-([+{[(+)*8[-].[.+]]}])%3])*19
14:15:11 <fizzie> That's what it spit out now, which is strange.
14:15:18 <EgoBot> Score for Deewiant_pendolino: 53.6
14:15:20 <EgoBot> Score for fizzie_evo_3: 0.0
14:15:41 <Deewiant> !bfjoust pendolino ->++(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
14:15:43 <fizzie> I think my crank isn't exactly egojoust-compatible.
14:15:57 <EgoBot> Score for Deewiant_pendolino: 58.1
14:16:18 <Deewiant> !bfjoust pendolino ->++(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9(>[-([+{[(+)*8[-].[.+]]}])%3])*19
14:16:27 <EgoBot> Score for Deewiant_pendolino: 58.0
14:16:41 <Deewiant> !bfjoust pendolino ->++(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
14:16:50 <EgoBot> Score for Deewiant_pendolino: 58.2
14:17:18 <Deewiant> !bfjoust pendolino ->++(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*2(-<)*2(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
14:17:32 -!- copumpkin has quit (Ping timeout: 276 seconds).
14:17:33 <EgoBot> Score for Deewiant_pendolino: 58.0
14:17:35 <Gregor> I love how Deewiant's strategy is just "poke integers until victory"
14:17:38 <Gregor> Admittedly my strategy INVOLVES that.
14:17:39 <Gregor> But most of my improvements from last night were due to bugfixes :P
14:17:41 <Gregor> (e.g. my philip strategy lacking its steamroller core, or not having a tail)
14:17:47 <Gregor> Hmmmm, fizzie: Do you distinguish * and %?
14:17:54 <fizzie> Yes, for the generated output.
14:18:01 <Deewiant> My biggest improvement yesterday was a bugfix as well
14:18:12 <Deewiant> Not running off the tape intentionally helps
14:18:13 <fizzie> From the evolver thingie, I mean. Crank doesn't when it's running those.
14:18:58 <Deewiant> !bfjoust pendolino ->++(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23<<<(-)*30>(+)*30>(-)*30>(+)*29(>)*6++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
14:19:13 <EgoBot> Score for Deewiant_pendolino: 47.1
14:19:16 * Gregor waits for Deewiant to finish adjusting integers until victory.
14:19:31 <Deewiant> Most of the time here I've not only been adjusting integers :-P
14:19:58 <Deewiant> !bfjoust pendolino ->++(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*30(>)*9++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
14:20:04 <Deewiant> But I'll do it once just for you
14:20:09 <EgoBot> Score for Deewiant_pendolino: 57.7
14:20:16 <Deewiant> !bfjoust pendolino ->++(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
14:20:41 <Gregor> Mainly I can't counter until pendolino is stable enough that I can actually guess how I'll do against it :P
14:20:41 <EgoBot> Score for Deewiant_pendolino: 58.2
14:21:14 <Gregor> Well ... I also can't counter until I'm not at work and don't have other things to do X-P
14:31:35 -!- katy has joined.
14:32:48 <katy> what is this room about?
14:33:33 <Gregor> Esoteric programming languages and esoteric topic in computing.
14:33:56 <Gregor> At present, mostly a particular competitive programming language (BF Joust)
14:34:38 <Gregor> Yeah, we're a strange bunch :P
14:35:01 <katy> lol, takes all kinds to make this stuff happen
14:36:04 <katy> whoosh over my head... thanks tho. bye
14:36:44 -!- katy has left (?).
14:40:05 -!- augur has joined.
14:41:41 -!- BeholdMyGlory has joined.
14:47:49 -!- Behold has joined.
14:48:01 -!- Sgeo has joined.
14:51:58 -!- BeholdMyGlory has quit (Ping timeout: 276 seconds).
15:05:40 <cheater-> should've told her it's about tantric sex therapy
15:05:59 <Deewiant> You really have a way with the ladies, don't you
15:06:35 <Deewiant> I suggest you don't tell them that
15:06:53 <cheater-> every time i do it they become even more mine.
15:07:32 <Gregor> cheater- just has poor grammar, he means they're "mines", and he's talking metaphorically.
15:07:42 <Gregor> Suffice it to say his love life blows up in his face a lot.
15:12:56 -!- pikhq has quit (Ping timeout: 250 seconds).
15:13:02 -!- pikhq_ has joined.
15:17:08 -!- azaq23 has joined.
15:27:33 <Ilari> APNIC down by about 0.2 today.
15:27:42 <Sgeo> WTF shit did I almost get myself into?
15:27:48 <Sgeo> And then failed to ever look at again?
15:28:06 <Sgeo> "As you can see from the project page, I am the only active developer on this
15:28:07 <Sgeo> project yet, so you would be number two. That's a high-risk investment of
15:28:07 <Sgeo> your time. If the project takes off, you'll be in from the start. But the
15:28:07 <Sgeo> risk that it doesn't take off is quite high."
15:29:35 <Sgeo> Just from reading the email, it looks like the node.js idea
15:29:41 <Sgeo> Except this wasn't node.js
15:29:58 <Deewiant> !bfjoust pendolino ->---(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
15:30:09 <EgoBot> Score for Deewiant_pendolino: 57.8
15:30:35 <Deewiant> !bfjoust pendolino ->+++(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
15:30:40 <Ilari> 3x/12 for China (and a /21 for Japan). On IPv6 front, a /32 to Japan.
15:30:45 <Sgeo> http://jsext.sourceforge.net/
15:30:48 <EgoBot> Score for Deewiant_pendolino: 57.8
15:31:04 <Deewiant> !bfjoust pendolino ->--(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
15:31:13 <EgoBot> Score for Deewiant_pendolino: 57.9
15:31:45 <Deewiant> !bfjoust pendolino ->++(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
15:31:58 <EgoBot> Score for Deewiant_pendolino: 58.1
15:32:43 <Deewiant> !bfjoust pendolino ->--(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
15:32:43 -!- copumpkin has joined.
15:32:50 <EgoBot> Score for Deewiant_pendolino: 57.9
15:33:04 <Deewiant> !bfjoust pendolino ->---(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
15:33:15 <EgoBot> Score for Deewiant_pendolino: 57.7
15:33:30 <Deewiant> !bfjoust pendolino ->++(>)*5->+>+(>[-([+{[(+)*8[-].[.+]]}])%3])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[-([+{[(+)*8[-].[.+]]}])%3])*19
15:33:37 <EgoBot> Score for Deewiant_pendolino: 58.2
15:36:25 -!- Sgeo has quit (Ping timeout: 255 seconds).
15:45:39 -!- Lymia_ has joined.
15:49:47 -!- Lymia has quit (Ping timeout: 276 seconds).
15:58:25 <Gregor> IF THE UNIVERSE EXPLODED
15:58:43 <quintopia> if every warrior handed every other warrior a fraction of its score based on the number of times that warrior beat it
15:58:54 <quintopia> and if, in doing so, all the scores stayed the same
15:59:00 <quintopia> what the scoreboard would look like?
15:59:22 <Gregor> Does this come 'round to you reminding me to add the more specific point breakdown to the report? :P
15:59:22 <quintopia> well, here's how, based on a limited scoreboard from data fizzie gave me: http://sprunge.us/SGda
15:59:56 <quintopia> no, it's my way of telling you i worked out the fixed point scoring algorithm :P
16:01:07 <Gregor> And then convince me, ais and elliott.
16:02:25 <quintopia> well that above explains the process the algorithm produces the limit scoreboard of
16:02:34 <quintopia> the closed form algorithm is this:
16:02:58 <quintopia> construct the matrix P(i,j)=number of configurations in which i beats j
16:03:04 <Gregor> "that above"? What above?
16:03:21 <quintopia> the thing about each warrior handing a fraction of its points to the warriors that beat it
16:03:34 <quintopia> based on the number of times they beat it
16:03:49 <Gregor> But that's how it works already, and that's not fixed-point :P
16:04:11 <Gregor> Well, OK, just gimme the algo
16:04:32 <quintopia> let n be the number of warriors, so P is nxn, yes?
16:05:19 <quintopia> also, initially set P(i,i)=0 for all i
16:05:57 <quintopia> now, for i=1 to n, let P(i,i)=1-(sum of column i)
16:06:51 <quintopia> finally take the eigenvector of P corresponding to eigenvalue 1 (it should be the first/highest eigenvalue), make sure its entries are positive, and multiply by 100 to put it on a 100 scale
16:07:10 <quintopia> you'll find then, that if you multiply P by this vector, you get the same vector back
16:08:53 <quintopia> alternately, you could repeatedly multiply the P so constructed by the vector of all 1's until it settles down to the same score
16:09:03 <quintopia> if you don't want to solve the closed form problem
16:14:45 -!- pumpkin has joined.
16:15:10 <Gregor> Hmmmmmmmmmmmmmmmmmmmmmmmm
16:15:23 <quintopia> does it make sense what P actually is?
16:15:30 -!- copumpki_ has joined.
16:15:49 <Gregor> I'm just wondering if this is getting to the point that it's too complicated for $RANDOM_PERSON_X to get ... maybe that's fine?
16:16:12 <quintopia> I WAS GOING FOR A FIXED POINT ALGORITHM THAT DID EVERYTHING THE CURRENT ONE DOES, NOT SOMETHING UNDERSTANDABLE
16:16:12 -!- copumpk__ has joined.
16:16:16 <Gregor> I guess even the current system isn't understood by most people :P
16:16:39 -!- copumpkin has quit (Ping timeout: 240 seconds).
16:16:41 <quintopia> but anyone who has studied markov chains will understand it
16:18:50 <quintopia> the major difference from the current scoring system is that it emphasizes more how MUCH you beat your opponents more and how many opponents you beat less (and the quality of the opponents you beat by about the same amount)
16:19:09 -!- pumpkin has quit (Ping timeout: 246 seconds).
16:19:48 -!- copumpki_ has quit (Ping timeout: 260 seconds).
16:20:46 -!- copumpk__ has quit (Ping timeout: 264 seconds).
16:21:46 -!- cheater- has quit (Ping timeout: 240 seconds).
16:24:03 -!- cheater- has joined.
16:28:42 <quintopia> Gregor: why do you spend 66 cycles on your full tape clear lock? why not the full 256?
16:28:49 <quintopia> does it do something special at 66?
16:29:31 <Gregor> I'm spending 66 cycles at their end of the tape while they're doing whatever they want to to my decoy.
16:29:36 <Gregor> I spend 255 cycles on the decoy.
16:30:25 <Gregor> Because 33 is floor(100/3) :P
16:30:31 <Gregor> And it's a two-cycle clear :P
16:31:04 <Gregor> There's no real logic behind it except that it plus the time getting back and forth was sufficiently less than 128 that I didn't fear that they'd drop the decoy in the interim.
16:32:04 <quintopia> oh, i thought you were doing something special where 2 and 3 cycle clear in each polarity would both get locked for sufficiently long
16:32:36 <quintopia> did you try doing a 256 cycle full tape clear?
16:32:37 <Gregor> Nope, it was just bean counting.
16:32:55 <Gregor> Ohohohohowait, I think I see the confusion.
16:33:12 <Gregor> I spend 66 cycles at the other end of the tape each time I go over there between futzing with the decoy.
16:33:25 <Gregor> I actually go all the way 'round the 256 values, 33 at a time.
16:33:49 <quintopia> did you try doing it where you spend 256 cycles going to the other end, clearing and coming back?
16:34:25 -!- augur has quit (Remote host closed the connection).
16:34:53 <Gregor> I'm not sure if I understand what you mean ...
16:35:05 <Gregor> Optimize it for a two-cycle clear?
16:35:32 <quintopia> because it looks like you don't have it optimized for anything :P
16:36:04 -!- sftp has joined.
16:36:05 <Gregor> I don't actually detect the direction they're clearing or the speed, I Just set it to a prime number and futz with it often enough that they're unlikely to hit zero in any strategy.
16:36:43 <quintopia> you're very likely to cross zero though, which may let them escape
16:36:57 <quintopia> aka, start doing a different clear
16:37:06 <quintopia> that does bring it to zero while you're gone
16:37:08 <Gregor> I do, and those that don't have any antishudder protection walk right through me, yes.
16:37:11 -!- augur has joined.
16:37:17 <Gregor> Luckily most have antishudder protection, which I'm taking advantage of >: )
16:37:47 <quintopia> a good anti shudder protection will get it out of phase so much that it still eventually breaks the lock
16:38:24 <Gregor> And yet, I was still beating most of them :P
16:38:33 <quintopia> for instance: [.-].[.++-------[...-]]
16:38:35 <Gregor> I admit freely that my locking is very simple, but it worked for me *shrugs*
16:39:20 <quintopia> how well does it work against careless clears?
16:39:40 <quintopia> not the lock i mean, the whole strategy
16:40:32 <Gregor> They usually cause me to switch to a philip, which does well against careless clears since it's just careless + small decoy avoidance + multi-decoy avoidance.
16:41:27 -!- azaq23 has quit (Quit: Leaving.).
16:43:27 -!- augur_ has joined.
16:43:42 -!- Lymia__ has joined.
16:43:44 -!- augur has quit (Ping timeout: 246 seconds).
16:44:22 <quintopia> so did you like my rewrite of the BF Joust page? huh?
16:46:04 -!- augur_ has quit (Read error: Connection reset by peer).
16:46:31 -!- augur has joined.
16:47:36 -!- Lymia_ has quit (Ping timeout: 276 seconds).
16:47:56 <Gregor> Having the current rules well described at the beginning is much better.
16:48:02 <Gregor> Having to read the history to know what the rules are was silly :P
16:55:43 <quintopia> alright, nap time. lemme know if anyone else likes the fixed point scoring alg :P
17:03:47 <Vorpal> Gregor, seems like Deewiant beat you on bfjoust!
17:06:36 -!- Lymia_ has joined.
17:10:20 -!- Lymia__ has quit (Ping timeout: 276 seconds).
17:46:42 -!- augur_ has joined.
17:47:45 -!- augur has quit (Ping timeout: 246 seconds).
17:48:01 -!- FireFly has joined.
17:48:53 <Gregor> !bfjoust furry_furry_nipple_clamp_girls http://sprunge.us/EhPH
17:49:12 <EgoBot> Score for Gregor_furry_furry_nipple_clamp_girls: 50.3
17:49:29 <Gregor> !bfjoust furry_furry_nipple_clamp_girls <
17:49:36 <EgoBot> Score for Gregor_furry_furry_nipple_clamp_girls: 0.0
17:50:11 <Gregor> Wait, forgot to get the breakdown X-P
17:50:13 <Gregor> !bfjoust furry_furry_nipple_clamp_girls http://sprunge.us/EhPH
17:51:00 <EgoBot> Score for Gregor_furry_furry_nipple_clamp_girls: 50.3
17:51:26 <Gregor> !bfjoust furry_furry_nipple_clamp_girls <
17:51:44 <EgoBot> Score for Gregor_furry_furry_nipple_clamp_girls: 0.0
17:56:37 <Gregor> Heh, wireless_frownie tricked my tripwire :P
18:00:05 <Gregor> !bfjoust furry_furry_nipple_clamp_girls http://sprunge.us/gThU
18:00:51 <EgoBot> Score for Gregor_furry_furry_nipple_clamp_girls: 45.5
18:07:00 <Gregor> !bfjoust furry_furry_nipple_clamp_girls http://sprunge.us/NdPg
18:07:46 <EgoBot> Score for Gregor_furry_furry_nipple_clamp_girls: 48.9
18:07:54 -!- asiekierka has joined.
18:12:22 <Gregor> !bfjoust furry_furry_nipple_clamp_girls http://sprunge.us/ZYOU
18:13:07 <EgoBot> Score for Gregor_furry_furry_nipple_clamp_girls: 51.2
18:14:15 <Gregor> !bfjoust furry_furry_nipple_clamp_girls http://sprunge.us/SYGP
18:14:58 <EgoBot> Score for Gregor_furry_furry_nipple_clamp_girls: 52.6
18:15:10 <Deewiant> !bfjoust allegro ->++(>)*5->+>+(>[(+)*9[-].[.+]])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[(+)*9[-].[.+]])*19
18:15:26 <EgoBot> Score for Deewiant_allegro: 53.5
18:15:39 <Gregor> !bfjoust furry_furry_bondage_girls <
18:15:48 <EgoBot> Score for Gregor_furry_furry_bondage_girls: 0.0
18:17:39 <Gregor> !bfjoust furry_furry_nipple_clamp_girls http://sprunge.us/SYGP
18:18:15 <EgoBot> Score for Gregor_furry_furry_nipple_clamp_girls: 50.9
18:19:10 -!- augur_ has quit (Ping timeout: 250 seconds).
18:22:42 <Gregor> !bfjoust furry_furry_nipple_clamp_girls http://sprunge.us/LCcd
18:22:53 -!- ais523_ has joined.
18:23:48 <EgoBot> Score for Gregor_furry_furry_nipple_clamp_girls: 47.6
18:23:50 <Gregor> ais523_: Comments on quintopia's fixed-point scoring algo?
18:23:55 <Deewiant> !bfjoust allegro ->++(>)*5->+>+(>[(+)*9[-].[.+]])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[(+)*9[-].[.+][-[+]]])*19
18:23:56 -!- elliott has joined.
18:24:10 <ais523_> Gregor: I rather like the existing one, and haven't seen the new one
18:24:13 <EgoBot> Score for Deewiant_allegro: 55.7
18:24:24 <ais523_> [18:48] <ais523_> grr, elliott isn't here
18:24:24 <Gregor> ais523_: It's basically the same, but fixed-point.
18:24:37 <elliott> ais523_: I support the new score system
18:24:47 <EgoBot> Use: !bfjoust <program name> <program> . Scoreboard, programs, and a description of score calculation are at http://codu.org/eso/bfjoust/
18:24:56 <elliott> umm, anyone care to clarify
18:25:24 <Gregor> !bfjoust furry_furry_nipple_clamp_girls http://sprunge.us/CXRJ
18:25:27 <ais523_> Gregor: is it in use on the hill yet?
18:25:28 <elliott> ais523_: what did you want me for then
18:25:40 <Gregor> ais523_: No, I wanted to make sure top contenders agree before I go effing things up.
18:25:40 <ais523_> I was attending a series of lectures today
18:25:46 <ais523_> and someone brought up a point I disagreed with
18:25:48 <elliott> Gregor: Is it quintopia's?
18:26:09 <elliott> Gregor: What cycle-breaker?
18:26:21 <EgoBot> Score for Gregor_furry_furry_nipple_clamp_girls: 49.5
18:26:25 <ais523_> which can most easily be answered by asking someone who's both rather younger than me, and who is decent at programming/computer science/that sort of thing
18:26:25 <Gregor> elliott: Just logread durf :P
18:26:32 <elliott> Gregor: I'd rather you tell me
18:26:41 <Deewiant> !bfjoust allegro ->++(>)*5->+>+(>[(+)*9[-]])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[(+)*9[-][-[+]]])*19
18:26:52 <ais523_> elliott: they claimed that nowadays, there was no expectation under young people that any data was unavailable
18:26:55 <EgoBot> Score for Deewiant_allegro: 52.6
18:27:05 <elliott> ais523_: hmm, that's a little vague
18:27:11 <ais523_> e.g. talking about young children confused as to why they couldn't see themselves on Google Street View
18:27:18 <Gregor> finally take the eigenvector of P corresponding to eigenvalue 1 (it should be the first/highest eigenvalue), make sure its entries are positive, and multiply by 100 to put it on a 100 scale // the "cycle-breaker" is MATH
18:27:38 <elliott> Gregor: Uhh, wow. Lemme ping quintopia.
18:27:39 <Deewiant> !bfjoust allegro ->++(>)*5->+>+(>[(+)*9[-].[.+]])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[(+)*9[-].[.+][-[+]]])*19
18:27:49 <elliott> Perhaps it's the eigenvectorial equivalent to a cyclebreaker he mentioned.
18:27:49 <EgoBot> Score for Deewiant_allegro: 55.8
18:28:01 <ais523_> it seems that this would most likely translate, in people slightly older, into a) a belief that copyright shouldn't exist, and b) a belief in a complete lack of privacy
18:28:06 <ais523_> and I'm wondering if those are actually the case
18:28:06 -!- pikhq_ has quit (Ping timeout: 240 seconds).
18:28:08 <elliott> ais523_: well, privacy is rather dead among the DARN YOUNGSTERS nowadays
18:28:16 <elliott> ais523_: just look at facebook, nobody cares about their personal information
18:28:17 * Gregor stops posting new FFNCG until Deewiant calms down :P
18:28:23 -!- pikhq has joined.
18:28:27 <Deewiant> Gregor: That beats FFLDG in egojsout but not in egojoust
18:28:47 <ais523_> btw, I was thinking about fixed-point scoring last night, and came to the conclusion that the correct algorithm was "calculate wins+draws/2 for each pair of players, place in a matrix (i.e. all elements positive), take principle eigenvector"
18:29:02 <ais523_> which seems to be the same algo quintopia came up with
18:29:05 <elliott> ais523_: as for copyright, I'm not sure; I don't get the feeling that any DARN KID would bother to obey it, after all piracy is rather common, but I don't know if they'll have an actual intellectual objection to it
18:29:09 <elliott> rather than just simply not caring
18:29:26 <ais523_> I mean, more in the sense of not understanding why it should exist
18:29:35 <ais523_> or why anyone would come up with it in the first place
18:29:44 <elliott> ais523_: the theory seems at least plausible, though of course generalising to "all youngsters" is incredibly impossible
18:29:55 <Gregor> Deewiant: FFLDG beats it in EgoJSout.
18:30:02 <ais523_> what about your opinions personally? I know you disagree with me at least slightly on this
18:30:10 <elliott> ais523_: I couldn't really /imagine/ a WHIPPERSNAPPER coming up with copyright, but then I can't really imagine them coming up with any laws at all
18:30:27 <Gregor> Deewiant: You probably have an old FFLDG in your buffer.
18:30:28 <elliott> ais523_: I'm not a big fan of copyright, but I think privacy is an important right
18:30:31 <elliott> although I'm not paranoid about my own data
18:30:40 <ais523_> the thing is, I think I'm already an old codger/traditionalist in this sense
18:31:00 <ais523_> elliott: hmm, OK; that's what I would have guessed, I think
18:31:25 <ais523_> and your response to the alleged "problem" of people keeping algos, etc., secret and never disclosing them, is that it isn't actually a problem?
18:31:39 <Deewiant> Gregor: Evidently... that's weird, did you update it recently?
18:31:44 <elliott> ais523_: I'm not sure how that's related to copyright
18:31:46 <ais523_> I think that's the response I'd give to people who asked me, too
18:31:53 <elliott> ais523_: People already reverse-engineer algorithms in closed-source software.
18:32:01 <Deewiant> Gregor: Or ha, I know the reason
18:32:03 <ais523_> elliott: oh, the "justification" for copyright is that people would never produce anything copyable if they didn't have copyright to protect them
18:32:16 <Deewiant> !bfjoust allegro ->++(>)*5->+>+(>[(+)*9[-].[.+]])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[(+)*9[-].[.+]][-[+]])*19
18:32:16 <ais523_> I'm not sure I agree with that justification at all
18:32:24 <Gregor> Deewiant: Blame the software on your own failure :P
18:32:25 <EgoBot> Score for Deewiant_allegro: 58.9
18:32:57 <elliott> ais523_: Right; I do disagree with that. In fact I think a business model based around information is even possible without copyright, just not in the existing form; I'm not sure what model would be viable, however
18:32:58 <ais523_> likewise, the idea behind patents is that they persuade people to disclose hidden details of an invention, in return for a temporary monopoly, and anyone can use them freely afterwards
18:33:04 <Deewiant> !bfjoust allegro ->++(>)*5->+>+(>[(+)*9[-]])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[(+)*9[-].[.+]][-[+]])*19
18:33:15 <EgoBot> Score for Deewiant_allegro: 58.8
18:33:20 <ais523_> which to me makes no sense, as the world blatantly doesn't work like that at the moment
18:33:28 <elliott> ais523_: wrt algorithms, though, mathematicians have been publishing algorithms for public use since time began and I don't see much worry about people not disclosing their findings there
18:33:34 <ais523_> and I agree with you on that, I think; most likely it'd be a variant on the bounty method
18:33:39 <Deewiant> !bfjoust allegro ->++(>)*5->+>+(>[(+)*9[-][.+]])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[(+)*9[-].[.+]][-[+]])*19
18:33:49 <ais523_> which some people already use, and which isn't perfect but does at least seem to be vaguely viable
18:34:04 <EgoBot> Score for Deewiant_allegro: 58.7
18:34:11 <elliott> ais523_: indeed -- or even as simple as (for a film) "give away the film for free, encourage derivative works (= free marketing), sell DVDs and merchandise"
18:34:15 <elliott> ais523_: this worked for Sita Sings the Blues, at least
18:34:22 <elliott> which has earned its maker a handy profit
18:34:34 <ais523_> well, what about other people selling merchandise too?
18:34:36 <ais523_> or would you enforce rules against that?
18:34:37 <elliott> (it is possible that the novelty of releasing a film this way was part of the reason that it generated a lot of money, but I'm not so sure)
18:34:39 <Deewiant> !bfjoust allegro ->++>>-->>>->+>+(>[(+)*9[-].[.+]])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[(+)*9[-].[.+]][-[+]])*19
18:34:41 <ais523_> (I don't think it's necessary)
18:34:56 <elliott> ais523_: I don't think it's necessary either, brand names are a valuable thing
18:34:56 <EgoBot> Score for Deewiant_allegro: 59.8
18:35:01 <elliott> (I think trademark law is perfectly reasonable)
18:35:04 -!- Sgeo has joined.
18:35:08 <ais523_> hmm, I suppose I should finish with this: what's your opinion of trademark law? I think it's a lot more justifiable than copyright/patents
18:35:16 <Deewiant> !bfjoust allegro ->++>>->+>>->+>+(>[(+)*9[-].[.+]])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[(+)*9[-].[.+]][-[+]])*19
18:35:27 <EgoBot> Score for Deewiant_allegro: 60.0
18:35:29 <ais523_> I think it isn't perfectly reasonable atm, but only due to problems with some of the details
18:35:33 <ais523_> rather than with the concept
18:35:52 <ais523_> hmm, what does it lose to?
18:36:01 <elliott> ais523_: In practice I don't support an outright reveal of copyright because any sort of abrupt dismantling of social structures like that would be chaotic
18:36:17 <Deewiant> defend9_*, FFNCG, quapping_turtle, and 3pass
18:36:23 <ais523_> elliott: yep, doing it suddenly would make no sense
18:36:30 <elliott> ais523_: The Pirate Party's position seems perfectly viable, though: keeping copyright around, but loosening the laws on non-profit derivative works, and making its duration much shorter
18:36:33 <ais523_> you'd do something like reduce term rates at the rate of 2 years per year
18:36:58 <ais523_> elliott: oh, I think their official position's entirely sane, yes
18:37:00 <Deewiant> !bfjoust allegro ->++>->->+>+>->+>+(>[(+)*9[-].[.+]])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[(+)*9[-].[.+]][-[+]])*19
18:37:02 <elliott> I think the Pirate Party's main plan is five-year commercial copyright, where commercial copyright = copyright except e.g. remixing not for profit is allowed
18:37:05 <ais523_> I think some of their members are rather more extreme than that
18:37:21 <EgoBot> Score for Deewiant_allegro: 59.1
18:37:22 <elliott> ais523_: that's Swedes for you >:D
18:37:37 <Deewiant> !bfjoust allegro ->++>>-->>>->+>+(>[(+)*9[-].[.+]])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[(+)*9[-].[.+]][-[+]])*19
18:37:47 <ais523_> hmm, a sort of enforced CC-by-nc on everything is an interesting idea
18:38:11 <ais523_> although it'd disrupt most present commercial software models (also music and film, but not really most other things)
18:38:15 <EgoBot> Score for Deewiant_allegro: 59.8
18:38:16 <Deewiant> !bfjoust allegro ->++>>-->+>>->+>+(>[(+)*9[-].[.+]])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[(+)*9[-].[.+]][-[+]])*19
18:38:21 <elliott> ais523_: that's practically how everything works already, take a look at YouTube
18:38:21 <ais523_> I'm far from convinced that'd be a bad thing, though
18:38:32 <EgoBot> Score for Deewiant_allegro: 60.0
18:38:34 <ais523_> elliott: yep, indeed, that's why I don't think it's a bad thing
18:38:54 <elliott> I'm not sure how a non-profit derivative work of closed-source software would work -- modifying the machine code? :)
18:39:06 <ais523_> what about patching out DRM?
18:39:13 <Deewiant> !bfjoust allegro ->++>-->>+>>->+>+(>[(+)*9[-].[.+]])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[(+)*9[-].[.+]][-[+]])*19
18:39:16 <ais523_> people do that routinely and illegally nowadays
18:39:21 <ais523_> the only real change would be to make it not illegal
18:39:27 <EgoBot> Score for Deewiant_allegro: 60.4
18:39:33 <elliott> well, exactly; an unenforcable law that nobody obeys is a rather pointless thing
18:39:44 <elliott> ais523_: the last I recalled you were basically in favour of copyright, has that changed or do I misremember?
18:40:02 <ais523_> elliott: I'm still in favour of it as a concept, I think
18:40:09 <ais523_> but current copyright laws are insane and make no sense
18:40:30 -!- asiekierka has quit (Ping timeout: 264 seconds).
18:40:41 <ais523_> I'm beginning to believe that abolishing it altogether would be saner than the current situation, in the abstract; but I think somewhere in between is likely better still
18:41:19 <ais523_> hmm... my parents seem to believe that I'm entitled to make a profit from good ideas I have
18:41:39 -!- asiekierka has joined.
18:41:40 <ais523_> and whether that's correct or not, spending years believing that because it's the only thing you're told leaves quite an impression on someone
18:41:55 <Deewiant> !bfjoust allegro ->++>-->>+>>->+>+(>[(+)*9[-].[.+]])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[(+)*16[-].[.+]][-[+]])*19
18:42:32 <EgoBot> Score for Deewiant_allegro: 61.4
18:42:47 <ais523_> Deewiant: are you just optimising for winning on greater tape length ranges now?
18:42:52 <ais523_> or trying to overfit to current programs?
18:43:16 <Gregor> ais523_: Nobody is /entitled/ to make a profit in any circumstance, and no system is going to be so perfect that luck doesn't overwhelm creativity w.r.t. actually economic or social uplift.
18:43:46 <ais523_> Gregor: I think you're highlighting the law in current copyright and patent laws
18:44:02 <Deewiant> More of the latter, I guess, although I try to justify every change I make more generally
18:44:07 <Gregor> No, I'm highlighting human nature.
18:44:10 <Gregor> Whatever system you place to try to encourage creativity, luck will always outstrip it.
18:44:22 <Gregor> Luck, and breast size.
18:44:28 <Deewiant> !bfjoust allegro ->++>-->>+>>->+>+(>[(+)*16[-].[.+]])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[(+)*16[-].[.+]][-[+]])*19
18:44:47 <EgoBot> Score for Deewiant_allegro: 60.4
18:44:48 <elliott> `addquote <Gregor> Whatever system you place to try to encourage creativity, luck will always outstrip it. <Gregor> Luck, and breast size.
18:45:08 <Deewiant> !bfjoust allegro ->++>-->>+>>->+>+(>[(+)*6[-].[.+]])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[(+)*16[-].[.+]][-[+]])*19
18:45:24 <EgoBot> Score for Deewiant_allegro: 61.1
18:45:26 <ais523_> I don't think breast size is relevant at the very highest levels, because discrimination against women is even stronger there
18:45:39 <Deewiant> !bfjoust allegro ->++>-->>+>>->+>+(>[(+)*9[-].[.+]])*2(+<)*4(+)*23<(-)*23<(-)*30<(+)*30<(-)*30<(+)*29(>)*9++(>[(+)*16[-].[.+]][-[+]])*19
18:45:46 <EgoBot> Score for Deewiant_allegro: 61.4
18:45:48 <Gregor> ais523_: I was actually covering popular music with that one :P
18:46:08 <Gregor> Please hold while the gremlin who runs HackEgo cranks the gear that makes the mouse move cheese to calculate things.
18:46:29 <HackEgo> 307) <Gregor> Whatever system you place to try to encourage creativity, luck will always outstrip it. <Gregor> Luck, and breast size.
18:48:00 <Gregor> Also, *whoosh!* CAPTAIN DEPRESSO TO THE RESCUE
18:50:52 <ais523_> Gregor: I'm disappointed in your latest naming series, anyway
18:51:03 <Gregor> Whaaaaaaaaaaaaaaaaaaaaaa? :(
18:51:11 <ais523_> furry_furry_bondage_girls was an excellent name, but the various variations of it don't live up to its promise
18:51:13 <ais523_> because they don't scan well
18:51:32 <ais523_> is it targeting defend9.75, btw? or beating it by chance?
18:51:50 <Gregor> I added a little bit of targeting for it, then sort of gave up, but ended up getting better anyway.
18:52:24 <ais523_> (every time I type "targeting" I'm reminded of a bitter online flamewar about what the present participle of the verb "target" was in British English; quite a lot of people thought it was "targetting", but nobody could find any evidence)
18:52:27 <Gregor> And how does furry_furry_leather_discipline_girls not scan well X-P
18:53:00 <Gregor> How about if I cut off one "furry"? furry_leather_discipline_girls
18:53:21 <Gregor> (But then it's not furry_furry :( )
18:53:54 <ais523_> it'd probably be better to change the naming scheme altogether
18:54:34 <Gregor> (Or rather, not until I substantially change the strategy)
18:55:08 -!- Lymia_ has quit (Read error: Connection reset by peer).
18:55:12 <ais523_> btw, I'm wondering if your latest BF Joust naming scheme outranks porn sites yet on searches for the names for the programs
18:55:19 <ais523_> (probably only if you use the underscores)
18:55:33 -!- Lymia_ has joined.
18:56:15 <ais523_> you could try it all three ways
18:56:52 <Gregor> The only result for "furry furry bondage girls" (w/ quotes) is a channel log.
18:57:13 <ais523_> I'm not entirely sure what that says about rule 34
18:57:46 <elliott> ais523_: it means that we're rapidly headed towards NSFW BF Joust-derived material
18:58:18 * Gregor arranges his generator to put the code in the shape of ASCII furry hentai porn
18:58:22 <ais523_> surely the name itself is potentially non-worksafe?
18:58:33 <ais523_> depending on what your workplace is like?
18:58:45 <elliott> ais523_: It's safe for work if you're a furry furry bondage girl.
18:59:22 <Gregor> "If you work at a sex shop ... is it still NSFW?"
18:59:27 <elliott> Gregor: How much ram does EgoBot have? http://www.mooseyscountrygarden.com/cats-dogs/ram-merino.jpg
18:59:51 <Gregor> elliott: EgoBot has plenty, egojoust is run with -v 100240
18:59:52 -!- asiekierka has quit (Ping timeout: 272 seconds).
18:59:57 <Gregor> ulimit -v 100240 that is
19:00:13 <elliott> What's that measured in, kio?
19:00:37 <elliott> (HAHA IT'S NOT THIEVERY DEEWIANT, I'M USING "ki")
19:00:53 <Gregor> Err, sorry, 102400 rather :P
19:00:57 <ais523_> "kibioctet" is a great word
19:01:14 <Gregor> Only retards say "octet"
19:01:18 <elliott> Units are always separated from their value by a space.
19:01:22 <elliott> Even if you are anti-retard.
19:01:56 <elliott> Gregor: 64-bit system right?
19:02:30 <Deewiant> elliott: It wouldn't be thievery if you also used "O"
19:02:55 <elliott> So my ins_t structure is... 4 + 8 + 8 + 8 octets = 28 octets. 1 Mio = 1048576 o. So...
19:03:02 <elliott> A 1 Mio program consisting entirely of isntructions takes up 28 Mio in RAM.
19:03:15 <elliott> I'll just allocate two 32 Mio ins_t buffers, then :P
19:05:00 <ais523_> elliott: are you sure it's a 0x20 space?
19:05:09 <ais523_> I thought it was one of the other spaces used for that purpose
19:05:14 <elliott> ais523_: It's not a non-space, that's for sure
19:05:20 <ais523_> yep, there is meant to be something there
19:06:34 -!- Sgeo_ has joined.
19:08:46 <elliott> Hmm, storing things that will always be under like 100K in a size_t is pretty WASTEFUL
19:09:10 -!- Sgeo has quit (Ping timeout: 255 seconds).
19:09:27 <elliott> BEST IODJEA CANT EVEN TYPOE
19:09:57 <elliott> int tp; /* signed to allow for < at position 0 in step */
19:10:04 <elliott> Maybe I should just set ->overflowed :P
19:10:05 <ais523_> WP:MOSNUM strongly implies (but doesn't say outright) that it should be a non-breaking space
19:10:08 <Gregor> BIT SHIFT BIT SHIFT BIT SHIFT BITSHIFT BITSHIFT BITSH IF BITCH IF BITCH IF YOU DON'T SHIFT THOSE BITS NOW something something
19:10:23 <elliott> ais523_: Wikipedia, the authority on style!
19:10:48 -!- copumpkin has joined.
19:10:53 <ais523_> elliott: not /directly/, but 3/4 of the world's population who really care about these sort of things and have endless arguments about it end up taking it to Wikipedia
19:10:57 <elliott> Deewiant: Man! Using unsigneds is WEIRD!
19:11:00 <ais523_> thus it nearly always ends up with the actual standard
19:11:05 <elliott> I have an enum that I suspect is represented as int, not unsigned :P
19:11:32 <Deewiant> Unless you explicitly stick a negative number in there it doesn't make a difference
19:11:51 <elliott> IT MAKES A MENTAL DIFFERENT
19:12:07 <ais523_> ah, it does say it outright, much later on
19:12:23 <elliott> Hmm, maybe I should use "unsigned char" rather than "unsigned" for all these SMALL FIELDS.
19:12:34 <elliott> Except that'd have to be signed.
19:13:13 <ais523_> the page contains explanations like the degree symbol being ° rather than º or ˚
19:13:27 <ais523_> I imagine it's easy to get wrong if you aren't a Unicode expert
19:14:28 <ais523_> also, apparently Ki, Mi, Gi are sufficiently unfamiliar that they shouldn't be used unless most of the sources on the topic use them, in quoting sources that use them, or in articles about the prefixes themselves
19:14:48 <ais523_> I suspect there was an edit war over that, although I'm not sure
19:15:28 <quintopia> every time it comes time to go to this class, i decide it's not worth it
19:15:39 <ais523_> bleh, Wikipedia suggests B for byte, and kB for kilobyte
19:15:45 <ais523_> and says that o for octet should only be used in French
19:15:50 <quintopia> but the class really is useless, when i could get all that information from the assigned readings
19:16:20 <elliott> <ais523_> also, apparently Ki, Mi, Gi are sufficiently unfamiliar that they shouldn't be used unless most of the sources on the topic use them, in quoting sources that use them, or in articles about the prefixes themselves
19:16:32 <elliott> to hell with the IEC, that's not SI-consistent!
19:16:49 <elliott> <ais523_> bleh, Wikipedia suggests B for byte, and kB for kilobyte
19:16:56 <elliott> B /is/ correct for byte if you have to use something for byte
19:17:09 <ais523_> here's another fun one from the Wikipedia advice: liter/litre = L, milliliter/millilitre = ml
19:17:19 <ais523_> on the basis that l on its own looks too much like 1
19:17:34 <Gregor> That's not a rational basis, but that is almost always how I see L and ml.
19:17:34 <ais523_> and as Wooble keeps pointing out, in SI B = bel
19:17:45 <elliott> ais523_: the bit/byte confusion is the worst ever... ask anyone you know (who isn't a geek) how fast their internet connection is
19:17:53 <elliott> "but I only get 1 megabyte tops..."
19:17:56 <quintopia> EVEN THOUGH EVERYONE ALWAYS USES dB!
19:18:10 <ais523_> all the standard school maths textbooks in the UK use a different font for l, when using it as an abbreviation for litres
19:18:25 <elliott> ais523_: hmm, isn't dB the base unit?
19:18:26 <ais523_> so it looks very different from 1, and indeed very different from most of the rest of the book
19:18:42 <ais523_> cm is the base unit of a different system, CGS
19:18:47 <elliott> which is the base unit, because SI are stupids
19:18:56 <ais523_> and which doesn't even have a constant value
19:19:13 <elliott> ais523_: apparently it's kg because the other natural units that fall out of those are more useful with kg than g
19:19:21 <ais523_> apparently, they found recently that the official kilogram was measurably lighter than all the backup kilograms
19:19:34 <elliott> ais523_: did the scales report the weight in kg? :D
19:19:43 <elliott> "Let's weigh the kilogram. [...] BEEP! "0.999 kg""
19:19:43 <ais523_> elliott: if they did, they were wrong
19:19:56 <ais523_> I think they were relative scales that checked the difference between the two
19:20:01 <elliott> ais523_: hasn't there been a successful proposal to fix the definition?
19:20:08 <elliott> <ais523_> I think they were relative scales that checked the difference between the two
19:20:20 <elliott> maybe they got one made that used pounds specially
19:20:25 <elliott> just to be technically correct
19:20:29 <Gregor> Pounds are defined by kilograms.
19:20:39 <elliott> Gregor: Only in the evil modern US of A.
19:20:53 <elliott> (MODERN: an adjective never before applied to the USA.)
19:20:58 <ais523_> Gregor: no, they're defined in terms of pence
19:21:01 <quintopia> that would help with specifying kg
19:21:19 <Gregor> http://en.wikipedia.org/wiki/File:CGKilogram.jpg Why is this a computer-generated image.
19:21:44 <elliott> "A kilogram is defined as 1/1,000,000,000,000,000,000,000,000,000,000,000,000 the weight of your mother."
19:21:49 <ais523_> quintopia: how, exactly? kilograms don't have much to do with gravity, and I don't think there are any fundamental constant weights to compare against that would let us use gravitational formulae to convert
19:22:13 <quintopia> so, did someone make "the kilogram" out of uranium on accident? that would be a problem...
19:22:17 <elliott> ais523_: there's been like a NERD COMPETITION (ok SCIENTIST COMPETITION) to come up with a definition of kg
19:22:21 <elliott> to replace the crappy existing non-definition
19:22:22 <Deewiant> Gregor: The description says why.
19:22:30 <elliott> <Gregor> http://en.wikipedia.org/wiki/File:CGKilogram.jpg Why is this a computer-generated image.
19:22:34 <elliott> Gregor: the light aimed at it would BASH IT
19:22:51 <ais523_> apparently, two different suggestions have been made to calculate the kilogram from fundamental constants
19:23:11 <elliott> wasn't there one involving fixing avogadro's constant?
19:23:15 <elliott> that seems reasonable to me
19:23:15 <ais523_> but they actually give different masses, even taking experimental error into account
19:23:26 <ais523_> elliott: yep, the major issue of that is trying to count an exact number of atoms
19:23:28 <elliott> ais523_: who cares, pick one now so we don't have to worry later :)
19:23:37 <Gregor> That's such a poor reason, instead of taking a picture of one of the many duplicates which is less well-protected, they generate an image which is less authentic because it's GENERATED ...
19:23:38 <ais523_> I fear trying to count well over 10^23 of /anything/ is a pretty tricky task
19:23:50 <elliott> Gregor: untrue, it was generated to be exactly one kilogram!
19:23:55 <quintopia> ais523_: define it to be the mass that has a specific weight at a specific point where the measurement is always taken. define that weight absolutely in terms of graviton interaction.
19:24:04 <Gregor> elliott: How meaningful :P
19:24:08 <ais523_> elliott: anyway, someone suggested averaging them
19:24:10 <elliott> quintopia: gravitons are your solution to everything!
19:24:18 <ais523_> which went and annoyed a lot of scientists
19:24:30 <elliott> ais523_: we need to define the mean of two definitions
19:24:31 <ais523_> more so as the suggestion's actually being taken seriously
19:24:48 <ais523_> arithmetic, geometric, or harmonic?
19:25:00 <elliott> ais523_: harmonic, to promote harmony in the scientific community
19:25:12 -!- Lymia__ has joined.
19:26:03 <Gregor> "that is to say, the 25 milligram difference shows that the scientists over 212 years ago managed to make the mass of the Kilogram of the Archives equal that of a cubic decimeter of water at 4 °C to within the mass of a single excess grain of rice."
19:26:10 <ais523_> elliott: fun fact: the harmonic average of 1 and -1 is plus or minus infinity
19:26:15 <Gregor> Is there a reason why that's not the standard instead of the prototype kilogram?
19:26:35 <Deewiant> Because a grain of rice is heavy.
19:26:43 <ais523_> Gregor: basing it on water? it's not just the temperature that matters, but also the pressure
19:26:57 <ais523_> and it turns out that you can't easily define pressure except in units that depend on kilograms
19:27:00 <Gregor> ais523_: Oh yeah, and pressure will ultimately have kg in it :P
19:27:09 <ais523_> so the issue is that it's a circular definition
19:27:25 <Gregor> Surely you can define a pure vacuum without requiring any particular measure ... a bit wonky and VERY hard to verify I suppose.
19:27:30 <elliott> clearly we just need to use natural units
19:27:42 <ais523_> some of the saner people on Slashdot said "just take the triple point", although I fear the issue /there/ is that the volume of water is likely hard to measure while it's simultaneously solid, liquid, and gas
19:28:05 <ais523_> Gregor: it wouldn't be a pure vacuum if you put a cubic decimetre of water in it
19:28:29 <elliott> unsigned line = 1, col = 0;
19:28:34 <elliott> Deewiant: USEFUL ERROR MESSAGES OH YEAH
19:28:53 <Gregor> ais523_: Surely a cubic decimetER of water would remain liquid by its own gravitation?
19:29:16 <ais523_> elliott: Farenheit was originally defined as: 0°F is the freezing point of saturated salt water; 100°F is body temperature
19:29:22 -!- Lymia_ has quit (Ping timeout: 276 seconds).
19:29:29 <elliott> ais523_: Farenheit is stoopid
19:29:36 <ais523_> liquid water can't exist in a vacuum, although both gaseous and solid water can
19:29:48 <ais523_> at sufficiently low pressures, it sublimes
19:29:50 -!- aloril has quit (*.net *.split).
19:30:05 <Gregor> Well, it doesn't have to be water, just something that's non-gaseous in a vacuum at a reasonable temperature.
19:32:10 <elliott> Gregor: EgoBot doesn't support sending multiple lines to the channel, yah?
19:32:49 <elliott> Gregor: But, but, I want SO NICE error messages.
19:33:01 <ais523_> put them in breakdown, rather than the channel
19:33:04 <Gregor> elliott: I don't want EgoBot to spam the channel, so too bad.
19:33:04 <elliott> With LITTLE ARROWS POINTING TO THE OFFENDING CHARACTER
19:33:11 <elliott> ais523_: The breakdown won't even start being generated if your program is invalid :P
19:33:24 <elliott> Deewiant: ais523_: If you ever get unhelpful syntax errors, DOOOOOOOOON'T BLAME ME
19:33:39 <ais523_> elliott: what about putting the arrow inline, like Perl's regex error messages
19:34:14 <ais523_> they go something like "Unexpected quantifier in regex, /this.*is.*a.?*<-- HERE regex/
19:34:17 <Gregor> Use a unicode error :)
19:34:39 <Gregor> <3 Unicode SO HARD YOU GUYS
19:34:39 <Deewiant> Quantifier follows nothing in regex; marked by <-- HERE in m/? <-- HERE / at -e line 1.
19:35:46 <ais523_> I'm wondering if the "marked by" changes if the regex actually contains <-- HERE naturally
19:36:00 <ais523_> it's the sort of thing Perl /would/ do
19:36:11 -!- aloril has joined.
19:36:44 <fizzie> $ perl -e '/? <-- HERE/'
19:36:44 <fizzie> Quantifier follows nothing in regex; marked by <-- HERE in m/? <-- HERE <-- HERE/ at -e line 1.
19:36:47 <fizzie> Not in that case, at least.
19:37:40 <elliott> lance.c:184: warning: ‘printf’ attribute directive ignored
19:37:58 <elliott> <ais523_> I'm wondering if the "marked by" changes if the regex actually contains <-- HERE naturally
19:38:05 <elliott> why else would it point out how it's marked?
19:38:27 <ais523_> yep, I'm surprised and disappointed now
19:38:35 <elliott> It's format(printf, m, n).
19:39:12 <elliott> ais523_: quick, what's the printf format for size_t again? :P
19:39:44 <ais523_> I'm not completely certain that ptrdiff_t would be %zd, although that would be logical
19:40:32 <Deewiant> And they don't have to have the same size
19:40:47 <ais523_> Deewiant: hmm, was I wrong?
19:41:07 <ais523_> it actually has a separate size letter? that's interesting
19:41:20 <ais523_> %zd would be ssize_t, then?
19:41:45 <Deewiant> ssize_t is a POSIX extension so the standard doesn't cover it :-P
19:42:04 <ais523_> I assumed ptrdiff_t was just a standard version
19:42:10 <ais523_> but you implied it was actually more complex than that
19:42:20 <Deewiant> It's not, which is why ssize_t exists, AIUI
19:42:25 <elliott> what's the thing you need a header for the format string for?
19:42:33 <elliott> that defines a macro containing a string literal
19:42:36 <Deewiant> Not that I know of any place where ptrdiff_t != ssize_t (assuming both exist)
19:43:21 <ais523_> except in Microsoft Visual C++ (in C mode), where long long always == int64_t, and the format specifier for both is "%I64d"
19:43:48 <ais523_> nice way to completely ignore C99, Microsoft!
19:44:01 -!- augur has joined.
19:44:02 <elliott> hmm, this thing appears to be under the impression that ADVANCE == RETREAT
19:44:10 <ais523_> (the worst bit is, there's no reason why they couldn't support both the standard %lld and their nonstandard version, as capital letters in printf specifiers is undefined behaviour)
19:45:38 <fizzie> Deewiant: I think there was one Cray that has a 32-bit size_t and 32-bit most pointers, except char *s and void *s are 64 bits. I guess it's possible ptrdiff_t there could be 64 bits, though not necessarily, given that size_t is 32 bits too for some bug-compatibility reasons.
19:47:15 <ais523_> 32-bit char is not uncommon on DSPs
19:47:37 <elliott> wow, looks like I just restarted agora
19:47:46 <elliott> (for some very loose definition of restart)
19:47:55 <elliott> perhaps ADVANCE _is_ RETREAT!
19:48:07 <elliott> ah, no, it's just parsing everything as <
19:48:47 <ais523_> elliott: it's the computer definition, I think; people are trying to fix Agora by turning it off and on again
19:49:11 <elliott> ais523_: I mean my suggestion for a currency, which BobTHJ proto'd and Rouououojoujojo said e liked
19:49:16 <ais523_> which uses that sort of expression for everything
19:49:20 <elliott> which is close enough to restarting considering the state of the game as it is
19:49:40 <elliott> ais523_: (the idea was to use voting power as a currency itself, since that's basically the only thing with true, permanent value in the game)
19:49:53 <ais523_> hadn't checked Agora for a while, it means logging into Yahoo! on this computer
19:50:05 <elliott> "hadn't"? surely "haven't"? or did you just?
19:50:40 <ais523_> in response to your comment
19:50:50 <elliott> hmm, perhaps make_ins is broken
19:50:57 <elliott> static inline ins_t make_ins(op_t op) { return (ins_t) { .op = op, .jump = 0, .repeat = 0, .iter = 0 }; }
19:50:59 <ais523_> and then replied in the pluperfect, to both state the situation honestly and to imply I'd just checked it
19:51:04 <elliott> perhaps you can't inline a function like that for some insane reason
19:51:10 <Gregor> `addquote <fizzie> $ perl -e '/? <-- HERE/' Quantifier follows nothing in regex; marked by <-- HERE in m/? <-- HERE <-- HERE/ at -e line 1.
19:51:11 <HackEgo> 307) <fizzie> $ perl -e '/? <-- HERE/' Quantifier follows nothing in regex; marked by <-- HERE in m/? <-- HERE <-- HERE/ at -e line 1.
19:51:26 <ais523_> elliott: "static inline"? IIRC, there's a gcc C99 bug related to that
19:51:35 <ais523_> which is easy to fix but would break a bunch of existing code
19:51:43 <ais523_> so they're reluctant to fix it
19:51:50 <elliott> ais523_: Agora activity summary: G.'s repeal-half-of-the-game proposal is succeeding, which is good; we want a currency and auctions; and my vote-currency idea seems popular enough.
19:52:05 <ais523_> I'm not sure what it is, though, and it's probably really subtle
19:52:10 <elliott> nope, even non-static, non-inlined has the same issue
19:52:13 <ais523_> and unlikely to be causing your problem
19:52:21 <elliott> < (jump=0, repeat=0, iter=0)
19:52:21 <elliott> < (jump=0, repeat=0, iter=0)
19:52:21 <elliott> < (jump=0, repeat=0, iter=0)
19:52:26 <elliott> can't be uninitialised memory, then
19:52:40 <elliott> OTOH, the program is "<>."; I'll try "><."
19:53:00 <elliott> must be a bug in parse(), then
19:54:08 -!- Sgeo_ has quit (Ping timeout: 255 seconds).
19:54:25 <ais523_> how did you not notice that your program was swapping < and > earlier?
19:54:31 <ais523_> I'd think it would be a little obvious...
19:54:41 <elliott> ais523_: also, this is a rewritten parser
19:54:50 <elliott> also, rewritten interpreter, so actually nothing's left of the original :)
19:55:08 <Gregor> Somebody should write a strategy assistant as a plugin for EgoJSout :P
19:55:17 <elliott> ais523_: the problem is that the first operation is being used for all the subsequenti nstructions
19:55:26 <elliott> case '+': *p++ = make_ins(DINK); break;
19:55:26 <elliott> case '-': *p++ = make_ins(DONK); break;
19:56:16 <elliott> Gregor: "You lost. Try losing less."
19:56:53 <Gregor> elliott: "Your problem at cycle 3240 is that the enemy dropped your flag to zero. Try not letting them do that."
19:57:13 <elliott> Gregor: "You ran off the end of the tape. Try not moving forward once you reach the last cell."
19:57:49 <ais523_> that's pretty generic advice
19:58:01 <ais523_> knowing where the last cell is is the issue...
19:58:32 <Gregor> !bfjoust dumbtarded (>)*9((-)*128.>)*21
19:58:42 <Gregor> I have no idea why I felt like making that :P
19:58:47 <EgoBot> Score for Gregor_dumbtarded: 9.2
19:58:54 <Gregor> Yessssssssssssssssssssssssssssssssss
19:59:37 <EgoBot> Use: !bfjoust <program name> <program> . Scoreboard, programs, and a description of score calculation are at http://codu.org/eso/bfjoust/
19:59:53 <Deewiant> Oh, right, it'll suicide otherwise :-P
19:59:54 <elliott> Deewiant: To avoid going off the end of the tape
19:59:58 <EgoBot> Score for elliott_WINNER: 3.0
20:00:06 <elliott> WOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
20:00:10 <Deewiant> Well, thanks for pushing allegro above 62 :-P
20:00:17 <ais523_> Gregor: simple hint: make it (>)*8, then move the > at the end of the loop to the start
20:00:31 <ais523_> I should mention that in the wiki somewhere
20:00:32 <elliott> ais523_ is our strategy guide
20:00:37 <Gregor> ais523_: Oh yes, that gives me a negligible chance of winning at length 30 instead of none :P
20:00:43 <ais523_> it's still rule of nine, just a neater way to write it that avoids special cases
20:01:22 <Gregor> !bfjoust dumbtarded (>)*8(>[(-)*120[[-]]])*21
20:01:31 <EgoBot> Score for Gregor_dumbtarded: 23.6
20:02:28 <elliott> Gregor: Can you move the hill to the fixed-point scoring system and then easily change it back?
20:03:02 <Gregor> elliott: Right now I can't, but in principle I should be able to, the only issue is that warriors will get removed by the other system for a while.
20:03:52 <elliott> Gregor: Just back up everything :P
20:04:17 <elliott> And tell people not to submit any serious programs :P
20:04:26 <Gregor> 'cuz that so often works.
20:04:45 <Gregor> I also don't know how to implement calculating eigenvectors in C X-D
20:06:16 <ais523_> it's a pretty standard thing to want to do, it's probably in libraries
20:06:18 <elliott> I don't see what's wrong with myyyy cycle-breaking method :P
20:06:32 <Gregor> elliott: I like quintopia's because it's so pure.
20:06:34 <ais523_> elliott: it isn't an eigenvector
20:06:40 <ais523_> eigenvectors get this correct by definition
20:06:41 <elliott> ais523_: Libraries = hurr hurr use our specific data structure
20:06:51 <ais523_> as in, they're defined to be the fixed point of the scoring system
20:07:08 <ais523_> the interesting thing is the algos for calculating them
20:07:28 <elliott> WRITE THE SCORING SYSTEM IN MATHEMATICA
20:08:08 <ais523_> hmm, apparently most existing libraries for solving eigenproblems are written in Fortran
20:08:25 <elliott> ais523_: The heavy numeric libraries are Fortran that people never use from Fortran.
20:08:31 <ais523_> you could possibly steal an implementation from GNU Octave; I'd be surprised if it couldn't do eigenvectors
20:08:39 <ais523_> and it's probably written in C
20:09:05 <Gregor> Frankly I'd like to see lance get in there first since there's talk of effing with report and report is where this would be implemented.
20:09:09 <ais523_> hmm, I wonder if Wolfram Alpha does eigenvectors?
20:09:21 <elliott> Gregor: The effing will be minor.
20:09:22 <elliott> http://bytes.com/topic/c/answers/648589-how-compute-eigenvalues-eigenvectors-c-programming-language ;; USENET WISDOM
20:09:25 <elliott> "surely you could find the accompanying source on the net... I found it
20:09:45 <ais523_> elliott: what, did the answer just provide the binaries?
20:10:10 <ais523_> that snippet doesn't really make sense otherwise
20:10:27 <elliott> source = the book it's from
20:11:06 <elliott> Try this search on sourceforge:
20:11:06 <elliott> (+matrix +complex eigenvalue eigenvector eigenvalues eigenvectors) AND
20:11:20 <elliott> people /search/ sourceforge?
20:13:01 <elliott> "Nowadays, the state of the art in dense-matrix eigensolvers is usually
20:13:01 <elliott> considered to be LAPACK (the successor to LINPACK and EISPACK)."
20:13:14 <elliott> Gregor: http://en.wikipedia.org/wiki/LAPACK ENJOY
20:13:30 <Gregor> I am not using LAPLACK for fucking SCOREBOARD CALCULATION
20:13:34 <elliott> [[A LAPACK subroutine name is in the form pmmaaa, where:
20:13:34 <elliott> p is a one-letter code denoting the type of numerical constants used. S, D stand for real floating point arithmetic respectively in single and double precision, while C and Z stand for complex arithmetic with respectively single and double precision. The newer version LAPACK95 use generic subroutines in order to overcome the need to explicitly specify the data type.
20:13:34 <elliott> mm is a two-letter code denoting the kind of matrix expected by the algorithm. The codes for the different kind of matrices are reported below; the actual data are stored in a different format depending on the specific kind; e.g., when the code DI is given, the subroutine expects a vector of length n containing the elements on the diagonal, while when the code GE is given, the subroutine expects a array containing the entries of the matrix.
20:13:37 <elliott> aaa is a one- to three-letter code describing the actual algorithm implemented in the subroutine, e.g. SV denotes a subroutine to solve linear system, while R denotes a rank-1 update.]]
20:13:49 <elliott> Gregor: Also, you might have little choice, I think what quintopia's suggesting is non-trivial :P
20:14:18 -!- BeholdMyGlory has joined.
20:14:21 <ais523_> elliott: it's actually pretty easy to do, but relatively hard to do efficiently
20:14:42 <elliott> Gregor: You could probably segregate the LAPACK pain into one function :P
20:14:50 <ais523_> I think there's a relatively simple O(n^3) algo, perhaps it's even O(n^2)
20:15:03 <Gregor> elliott: One function and two hours of trying to get the MOFO thing compiled and running.
20:15:10 <elliott> Gregor: sudo aptitude install
20:15:30 <elliott> Or even "clapack for C (especially useful if there is no Fortran compiler available, as it is already preprocessed with f2c)" :P
20:16:19 <ais523_> elliott: one side of the matrix (which has to be square)
20:16:39 <elliott> ais523_: I meant, what does n equal for the scoreboard situation?
20:16:45 -!- Sgeo has joined.
20:17:15 <ais523_> possibly O(n^4), I have difficulty working it out in my head, especially as I haven't had to calculate an eigenvector in about five years now
20:17:18 <elliott> ais523_: then n is constant, 50
20:17:27 <elliott> hmmm, ^3 might be feasible, ^4 not, then
20:17:32 -!- Behold has quit (Ping timeout: 255 seconds).
20:17:33 <ais523_> yep, but that doesn't give a good indication of how long it would take
20:17:34 <elliott> unless it's 6,250,000 times very quick
20:17:45 <elliott> but that still seems impractical
20:17:54 <elliott> after all, verifying a simple predicate on all 32-bit integers takes a few seconds at least
20:25:16 <Sgeo> Is what O(n^4)?
20:25:56 <Deewiant> Getting eigenvectors from a matrix
20:28:47 <elliott> wow, GNU C has constructors and destructors
20:31:02 <ais523_> so do most C impls, as nonstandard pragmas
20:31:18 <ais523_> and so impls can't resist letting people do the opposite
20:31:25 <Gregor> For atexit to be sufficient to support destructors, you still need constructors, but any ELF system has .init
20:31:38 <elliott> Anyone planning to nest their BF Joust program more than 1,048,576 levels deep?
20:31:51 <ais523_> nesting more than timeout makes no sense
20:31:55 <ais523_> as you could never reach the inside of the nest
20:32:00 <Gregor> elliott: In spite of the <1MB file size limit
20:32:18 <elliott> Gregor: Actually it's 1 Mii exactly right now, where i = instruction.
20:32:39 <elliott> I'm only gonna expand that, not shrink it, it's just like that because it's convenient for my parser :P
20:32:44 <elliott> Actual limits should be done at a higher level.
20:33:08 <elliott> ("(", ")", "{" and "}" count as instructions)
20:34:57 <Gregor> "Brain damage in football players may be hidden." ... by the fact that football players are otherwise indistinguishable from brain-damaged individuals.
20:34:57 -!- Behold has joined.
20:36:20 <fizzie> If you don't like LAPACK, you can always use lapack++; you get the best (or was that the worst?) parts of both LAPACK and C++.
20:37:39 <elliott> <Gregor> "Brain damage in football players may be hidden." ... by the fact that football players are otherwise indistinguishable from brain-damaged individuals.
20:38:13 -!- BeholdMyGlory has quit (Ping timeout: 255 seconds).
20:38:20 -!- augur has quit (Remote host closed the connection).
20:39:59 <Gregor> elliott: Remember that these are American-football players.
20:42:14 <elliott> Gregor: I meant, source? :P
20:42:31 <elliott> (Association football players are hardly leading intellectuals either)
20:42:37 <Gregor> elliott: It was in this stupid email all Purdue students get every day X_X
20:42:43 <elliott> (^^ Name courtesy of Wikipedia pedantry)
20:42:52 <elliott> Gregor: No unsubscribe link? :P
20:43:49 <Sgeo> What's the best first language for someone who is not particularly likely to enter the field, or do much recreational programming?
20:44:09 <Sgeo> I'm kind of thinking Pytho.. yeah
20:44:10 <Gregor> elliott: Oh, I guess it's just all Purdue employees, not students.
20:44:20 <ais523_> elliott: you know how in every organisation, there's someone who sends out lots of messages, all of which are useless?
20:44:20 <Gregor> elliott: But no, no unsubscribe link.
20:44:23 <fizzie> The unsubscribe link will just let the university to know the addres is valid, and then they start sending that email thrice a day.
20:44:44 <elliott> Gregor: An unsubscribe link.\1?
20:44:49 <ais523_> here in the CS department, someone went and persuaded the mail server admins to block mail sent internally from it at the server, so they could honestly say they hadn't received it
20:45:29 <Gregor> fizzie: It's to my university address, they know it's valid :P
20:46:56 <ais523_> I know that the emails I get from the guild are HTML-only, filled with something like 6 or 7 different tracking mechanisms
20:47:14 <ais523_> I asked them why they didn't send plaintext, and they said it's because the marketing people said more people read the HTML emails
20:47:15 <Sgeo> You don't mean the web series, do you?
20:47:32 <ais523_> it has a royal charter or something like that, so it's a guild rather than a union
20:48:06 <elliott> ais523_: if G.'s proposal passes, you might want to reregister in Agora
20:48:33 <ais523_> I'm thinking that the reason the marketing people said that were a) you can't track whether someone reads a plaintext email anyway (nor HTML in sensible clients); b) the plaintext version is a link to an online version of the same email
20:48:40 <ais523_> so they can track you that way instead
20:49:06 <fizzie> Our per-study-program student organizations are called "guilds" too. Or at least the CS one. They don't have any special reason, though.
20:49:07 <elliott> ais523_: well, yes, I just meant that you'll want to want to register in Agora when the timeout expires
20:49:11 <elliott> ais523_: especially if this currency thing gets in
20:49:49 <elliott> 06:01:09 <fizzie> Let's try a generated Deewiant_pendolino mutation for a chance now; it's shorter, so maybe parse errors haven't crept in.
20:49:49 <elliott> 06:01:17 <fizzie> !bfjoust evo_3 http://p.zem.fi/evo-3
20:49:49 <elliott> 06:04:01 <fizzie> According to crank and a local hill-copy, it's pretty close to identical except it beats some elliott_awkwardly_long_name.
20:49:51 <elliott> 06:04:46 <EgoBot> Score for fizzie_evo_3: 57.2
20:50:13 <elliott> 06:05:51 <Deewiant> fizzie: After that first *19 everything is dead code
20:50:13 <elliott> 06:06:36 <quintopia> Deewiant: you're just upset that it beat pendolino
20:50:23 <fizzie> It didn't actually beat elliott_long_name with egojoust, sadly.
20:50:47 <Gregor> AND YET WHERE IS THE FIX BAWK BAWK BAWK
20:51:00 <fizzie> Deewiant's own subsequent submissions were more impressive, though.
20:51:21 <elliott> Gregor: BEING TYPED IN RIGHT AS YOU SPEAK
20:51:34 <Gregor> I am very annoyed that my awesomely complex metastrategy is defeated by Deewiant's rush :P
20:51:51 <ais523_> Gregor: how does it work again? I think I asked you earlier
20:52:03 <Gregor> ais523_: I put some small description on the wiki.
20:52:13 <Gregor> Which I almost spelled "weeki" for some reason.
20:52:15 <elliott> DOES EVERYONE LIKE MY TEMPLATE
20:52:28 <Gregor> elliott: YA DER TEMPLATE IS DAH SEXENFEUGEL
20:52:37 <elliott> grr, quintopia is violating style standards on [[BF Joust]]
20:52:53 <ais523_> Gregor: why would you switch to a philip strategy when you thought the opponent was already on your flag?
20:52:58 <elliott> and capitalising the first letter in brainfuck
20:53:02 <ais523_> on the assumption that there are no large decoys?
20:53:10 <ais523_> elliott: not at the start of the sentence?
20:53:20 <elliott> [[Possibly a better name than "Poke", which makes the purpose of the strategy more clear, is "Echolocation," since it locates an opponent on the tape in the same way that an ultrasonic ping from a bat lets it determine the distance to its enemy.]]
20:53:20 <ais523_> go fix the excessive italicising while you're at it
20:53:29 <elliott> (--quintopia, [[BF Joust strategies]])
20:53:35 <Gregor> ais523_: In my experience my poke was usually fast enough that philip would be faster to get their flag down.
20:53:44 <elliott> ais523_: OK, but NNPOV OR isn't :P
20:53:45 <Gregor> ais523_: Since I could /skip/ their big decoys.
20:53:54 <ais523_> POV, I'm not sure about, we don't have a ruling eiher way on that
20:53:55 <fizzie> To pre-empt any "you're just stealing from others" objections, I restarted the evolver from "first principles" (read: wiki "bfjoust strategies" examples), but it hasn't innovated anything especially good (yet).
20:54:14 <ais523_> Gregor: ah, on the basis that no fast rush uses a reverse decoy setup?
20:54:29 <Gregor> ais523_: Yeah ... which of course it could, but they don't :P
20:54:33 <elliott> oh wow, the italicisation really is bad
20:54:41 <elliott> any objections to me just reverting to the previous version? this is unreadable
20:54:51 <elliott> (the previous version isn't very good, but it's better than this)
20:55:11 <ais523_> elliott: I like the new page organisation, but not the actual content
20:55:29 <ais523_> but I'm hardly wiki dictator, that's Graue's job
20:55:33 <ais523_> so it's nice to get more opinions
20:55:38 -!- BeholdMyGlory has joined.
20:55:43 <Gregor> It needs some polish, but is still better than it was.
20:55:56 <ais523_> it needs de-myspacification
20:56:10 <elliott> I would <i>prefer</i> someone just <i>reorganise the previous page</i> rather than All Of This and <i>NOTE THIS IMPORTANT NOTE</i>
20:56:32 <elliott> You're an American, you don't have an allergic reaction to bludgeoning-by-text.
20:56:41 <ais523_> Gregor: hmm, your extremely complex metastrategy, what sort of extremely-conventional offset clear rush does it use after waiting 1000 cycles?
20:57:06 <ais523_> it can't be an extremely extremely conventional 2-cycler, or it wouldn't beat the majority of defence programs on long tapes
20:57:34 <Gregor> ais523_: ([(+)*32[-]]>+[(+)*32[-]]>-)*11
20:57:38 <Gregor> It is pretty damned conventional.
20:57:57 <ais523_> hmm, something's wrong here
20:58:06 <ais523_> which I should go analyse in more detail
20:58:19 <Deewiant> Prediction: ais fixes a bug and hits 70 score
20:58:45 -!- Behold has quit (Ping timeout: 246 seconds).
21:02:01 <fizzie> I should have some sort of dead-code removal from my evolver-kludge (or alternatively just build in some bias for shortness), it keeps building more and more complicated programs, which probably don't make any sort of sense at all.
21:02:07 <fizzie> !bfjoust evo_4 http://p.zem.fi/evo-4
21:02:36 <EgoBot> Score for fizzie_evo_4: 0.0
21:02:44 <Deewiant> fizzie: >*9 followed by >*21 followed by dead code
21:03:05 <fizzie> (Alternatively I guess I could always just do what everyone else seems to be doing, i.e. think, but that doesn't sound very reasonable.)
21:03:45 <Deewiant> My and your best scores prove that the more you rely on machines the worse you'll do and vice versa
21:05:15 <Gregor> Real men just get onto the tape and toil with their bare hands.
21:05:27 <fizzie> Well, out of pure curiosity, let's have that initial bit as evo_4.
21:05:33 <fizzie> !bfjoust evo_4 (-->)*9([(-)*128.>]+>)*21
21:05:47 <EgoBot> Score for fizzie_evo_4: 33.3
21:05:48 <fizzie> That at least shouldn't crash-and-burn.
21:06:07 <Sgeo> That does better than my mind can do
21:06:26 <Deewiant> Better than expected considering it looks like it runs off the end of the tape often :-P
21:06:31 <ais523_> fizzie: hmm, that's just a turtle that sets decoys
21:06:36 <Sgeo> Although that looks like a pblished strategy
21:06:44 <Sgeo> piblished. For pibs
21:06:48 <ais523_> I'm surprised it's doing that well given the amount of anti-turtleage around atm
21:07:10 <ais523_> oh, it also leaves a trail, although I'm not sure if that matters
21:07:30 <elliott> lance.c:65: error: variably modified ‘tape’ at file scope
21:07:49 <elliott> cell_t tape[max_prog - 1];
21:07:49 <elliott> err, how did that happen :-D
21:08:01 -!- cheater- has quit (Ping timeout: 240 seconds).
21:08:57 -!- oerjan has joined.
21:09:26 <Deewiant> The Internet is closed. Please go away.
21:09:46 -!- ais523__ has joined.
21:09:53 -!- ais523_ has quit (Disconnected by services).
21:09:55 -!- ais523__ has changed nick to ais523_.
21:10:25 -!- cheater- has joined.
21:10:27 <oerjan> 15:27:19 <Gregor> oerjan: Why are you not here now :P
21:10:40 <oerjan> BECAUSE THE HOUSE MAIN ROUTER DONE BROKEN
21:11:22 <elliott> lance.c:75: error: variably modified ‘nest_stack’ at file scope
21:11:22 <elliott> I wonder what on earth that /means/
21:11:25 <oerjan> I SENSE AN EVIL LACK OF SYMPATHY
21:11:51 <elliott> oerjan: stop losing internet like that
21:11:59 <Sgeo> When I ran out of Internet access in the past, I'd go on Active Worlds
21:12:11 <elliott> it's ok, i would probably die after two hours of no internet
21:12:20 <elliott> oerjan: we've been on somewhat of a bf joust kick.
21:12:31 <oerjan> elliott: one more day and i would have considered it
21:12:44 <elliott> oerjan: what, spontaneously dying?
21:12:53 <Sgeo> oerjan, is your Internet access permanently fixed?
21:13:16 <elliott> come to think of it, last time you could have just used the freenode webchat, sloppy!
21:13:21 <ais523_> Gregor: your program only beats defend9.75 because of egojoust limitations
21:13:22 <oerjan> elliott: ah yes. which reminds me that one of my last messages before i left was griping whether you actually would support my (({{}})*)* notation properly
21:13:22 * elliott swats oerjan [INSERT AMUSING ASCII DEPICTION]
21:13:32 <elliott> oerjan: egojsout also supports it
21:13:38 <elliott> (Gregor's fancy debugger thing)
21:13:50 * ais523_ catches elliott and oerjan in the same butterfly net -----------------\XXX/
21:13:55 <Sgeo> ais523_, what limitations?
21:14:05 <elliott> oerjan: also we decided to unify * and %, but I'm reversing that decision because exponential programs could be useful, I guess
21:14:08 <oerjan> elliott: i'll assume you mean according to my "obvious" semantics :)
21:14:21 <elliott> oerjan: technically Gregor counts up both ways, but he does it differently
21:14:23 <ais523_> Sgeo: it crashes if you repeat a long program containg loops several times
21:14:36 <elliott> I wonder why Herobrine died
21:14:39 <ais523_> so I had to reduce a loop count I originally had at 125 to 9
21:14:41 <elliott> (oerjan: don't logread from my links any more)
21:14:51 <elliott> I possibly won't bother putting it back up
21:14:54 <oerjan> elliott: um counting up both ways means having to modify the counter stack when you jump between left and right parts...
21:15:08 <ais523_> it didn't matter against any programs that existed at the time
21:15:14 <oerjan> and that's true even with single ({})*
21:15:24 <elliott> ins_t nest_stack[max_prog - 1];
21:15:30 <elliott> lance.c:75: error: variably modified ‘nest_stack’ at file scope
21:15:36 <elliott> oerjan: he does it some different way, as i said, ask him
21:16:00 <oerjan> Sgeo: i assume relatively permanently, i just managed to plug in the new router that arrived in door-to-door mail today
21:16:12 <Sgeo> oerjan, huzzah!
21:16:19 <Sgeo> (Since when do I say huzzah?)
21:16:29 <oerjan> Sgeo: SINCE YOU BECAME A BEE
21:16:45 <Deewiant> elliott: That doesn't work with a const, it has to be a #define or such
21:16:48 <Sgeo> But I don't wanna go extinct!
21:16:58 -!- Behold has joined.
21:17:32 <elliott> Deewiant: but, but, it's static const!
21:17:36 <elliott> and with a literal initialiser
21:18:09 <ais523_> anyway, it did highlight a bug in defend9.75 too, which is trivially fixable
21:18:18 <oerjan> <elliott> (oerjan: don't logread from my links any more) <-- what? :(
21:18:33 <elliott> oerjan: Herobrine pinged out and for some reason didn't recover
21:18:40 <elliott> despite recovering successfully from the same only hours before
21:18:43 <elliott> it didn't even log itself dying
21:18:45 <ais523_> note: on work computers, always remember to go to pastie.org, not pastie.com
21:18:47 <elliott> I haven't investigated yet
21:18:55 <Sgeo> ais523_, I'm afraid to ask
21:19:02 <elliott> ais523_: "The Leading Nipple Covers Site on the Net"... err, wow.
21:19:12 <elliott> "can u watch tv shows on this thing?
21:19:12 <elliott> hello? srry im new in this
21:20:13 -!- BeholdMyGlory has quit (Ping timeout: 250 seconds).
21:20:16 <elliott> is there a way to jump to a line in emacs without using annoying prefix arguments?
21:20:25 <Ilari> Heh. Reminds me of whitehouse.gov(SFW)/whitehouse.com(NSFW).
21:20:35 <ais523_> then enter the line number
21:20:38 <oerjan> elliott: that ping out was only today was it? so i can still look at earlier days?
21:21:00 <ais523_> Ilari: it hasn't happened to me, but python.org(SFW)/python.com(NSFW) has caught out some of my friends
21:21:04 <ais523_> oerjan: it was quite a while ago
21:21:18 <elliott> oerjan: last logged date is the 14th and it's missing most of the talk on that day
21:21:18 <oerjan> elliott: oh. well i guess it will be obvious when the logs end.
21:21:20 <elliott> oerjan: the 13th is OK though
21:21:24 <ais523_> someone here with adblock, what's the syntax for raw links in pastie.org?
21:21:36 <elliott> why can't someone without adblock answer?
21:21:41 <ais523_> pastie here is currently trying and failing to load an advert, and the raw link isn't loading because it's later on the page than the advert is
21:21:41 <elliott> ais523_: I highly suggest you do not use pastie, however
21:21:44 <elliott> they deleted all my pastes ages back
21:21:49 <elliott> without ever deleting any before
21:21:52 <ais523_> it's only meant to be temporary
21:21:52 <elliott> or saying they'd delete any
21:21:58 <ais523_> I just want to feed a program to egobot
21:22:04 <elliott> ais523_: there are no ads on pastie. oh, there's one, it seems.
21:22:06 <ais523_> and this isn't my computer
21:22:09 <elliott> a really really tiny one, and just for their web host
21:22:21 <elliott> ais523_: I'll check, though
21:22:25 <ais523_> and yet it's preventing me viewing the raw link, without guessing the syntax
21:22:40 <elliott> ais523_: http://pastie.org/pastes/1572313/text
21:22:46 <elliott> ais523_: found your paste :P
21:22:51 <ais523_> ah, it loaded eventually, after about a minute
21:22:52 <elliott> ais523_: also, why can't you use sprunge?
21:22:58 <ais523_> !bfjoust defend9.75 http://pastie.org/pastes/1572313/text
21:23:08 <oerjan> <elliott> aha! admission!
21:23:22 <ais523_> it means I can't expect to have my usual software installed
21:23:22 <elliott> ais523_: does it not have curl?
21:23:27 <elliott> no, but you can always check
21:23:32 <elliott> curl ships with every stock Linux distro
21:23:53 <elliott> ais523_: err, that machine doesn't have bash?
21:24:06 <elliott> I doubt a Linux machine without a bash-similar shell could even run
21:24:18 <ais523_> hmm, seems it does have curl
21:24:40 <ais523_> elliott: if you try to chsh bash, it tells you not to try because it causes logins to fail
21:24:59 <EgoBot> Score for ais523__defend9_75: 0.3
21:25:04 <elliott> ais523_: meh, that just means it has login scripts written in tcsh
21:25:26 <ais523_> yep, same thing that caught out impomatic
21:25:31 <elliott> ais523_: tl;dr sprunge. It's the only pastebin that you can trust to not do absolutely idiotic crap.
21:25:43 <elliott> ais523_: hmm, I should put up an html form pointed at sprunge.us somewhere
21:25:46 <elliott> so it can be used from anywhere
21:25:50 <ais523_> larger question: how does it get any points at all when the first character is <
21:25:55 <ais523_> elliott: don't, it'd just be spammed
21:26:06 <ais523_> i.e. password or something
21:26:06 <elliott> ais523_: err, how would it be spammed?
21:26:09 <elliott> I'd only link it from here
21:26:11 <ais523_> then tell everyone you know the password
21:26:13 <elliott> and I doubt spambots will look at our logs
21:26:22 <ais523_> I think they would, they look at everything
21:26:33 <elliott> ais523_: I'd make it require JS, then
21:27:04 <oerjan> 18:24:24 <ais523_> what does (a{b}c{d}e)%2 expand to?
21:27:05 <oerjan> 18:24:35 <elliott> aabccdee
21:27:12 <oerjan> that's not _my_ interpretation
21:27:26 <oerjan> mine would be aabcdecabcdee
21:27:54 <elliott> oerjan: we've decided that multiple {}s in one % is ridiculous
21:27:57 <oerjan> although maybe yours fits better into your implementation, mine would require a different system that keeps much more information on the stacks
21:27:58 -!- ais523_ has changed nick to ais523.
21:28:03 <ais523> !bfjoust defend9.75 http://sprunge.us/cJII
21:28:06 -!- ais523 has changed nick to ais523_.
21:28:09 <elliott> oerjan: but we've also decided that (x(a{b{c}d}e)%my)%n is OK
21:28:14 <elliott> the innermost {} associates with the outermost ()%
21:28:31 <elliott> note: this is slightly insane, but used in the preferred version of a recent good contester
21:28:34 <elliott> (one of Gregor's philip programs)
21:28:37 <oerjan> elliott: yeah that's my basic (({{}})) idea
21:28:42 <elliott> oerjan: you don't follow the hill itself right?
21:28:56 <elliott> if you do, uh, everything's been shaken up, we have a 30 Kio program that ruled for ages
21:29:06 <elliott> hybrid attack/defence, first of its kind
21:29:15 <elliott> now the hill is populated with Gregor's
21:29:15 <elliott> 3 6 53.32 20.08 Gregor_furry_furry_leather_discipline_girls.bfjoust
21:29:17 <elliott> 5 7 48.46 18.63 Gregor_furry_furry_nipple_clamp_girls.bfjoust
21:29:17 <oerjan> elliott: however you said something about (({})*)%, which in my view is almost the same as that multiple {}'s thing
21:29:22 <elliott> with two of Deewiant's rushes and... some other program, at the top
21:29:36 <oerjan> the only difference is that there aren't several different "centers"
21:29:42 <ais523_> elliott: it's just a rush, enhanced with most of the things that make rushes better
21:29:53 <oerjan> elliott: yeah that too would require much more stack information
21:29:59 <elliott> so basically, dumb rushes are beating highly advanced partially-computer-generated programs.
21:30:02 <EgoBot> Score for ais523_defend9_75: 42.6
21:30:27 <oerjan> (which means i have a sort of idea how it _could_ be done. i vaguely recall Icon uses its stack in a similarly weird manner
21:30:57 -!- cheater- has quit (Quit: Leaving).
21:31:16 <elliott> quick, someone push good_vibrations off
21:31:22 <oerjan> (allowing programs lower in the stack to run while the stack below is still filled, but any new recursion is put on top of the stack)
21:31:22 <ais523_> I don't think so, maybe it stayed constant
21:31:28 <elliott> aww, interior_crocodile_alligator sunk a lot
21:31:40 <elliott> on the report.txt that I had loaded
21:31:41 <Deewiant> I think allegro just uses different clears and more early decoys
21:31:50 <ais523_> perhaps it loses on one more cycle length to a fastish defend, then
21:32:13 <Deewiant> It's faster and shouldn't die to vibrators at least as easily
21:32:25 <oerjan> <elliott> oerjan: you don't follow the hill itself right? <-- no, i don't have much understanding of the grits of the actual _game_ :D
21:34:03 -!- Lymia_ has joined.
21:34:37 <ais523_> haha, I fixed the bug in one place but not another
21:36:37 <oerjan> <elliott> Gregor: (...{ goes up, and *leaves it on the stack*.
21:36:53 <oerjan> once again, i said _not_ to leave it on the stack then.
21:37:09 -!- Sgeo has quit (Ping timeout: 240 seconds).
21:37:18 <oerjan> when finishing the going up, of course
21:37:22 <ais523_> elliott: please finish lance so I can get defend9.75 to beat all the furry furry girls
21:37:36 <elliott> ais523_: I'm workin' on it
21:37:38 <ais523_> it's close atm, having only narrow losses
21:37:56 <elliott> then Gregor does the right thing, I guess
21:38:00 <elliott> oerjan: Gregor doesn't actually use a stack
21:38:00 -!- Lymia__ has quit (Ping timeout: 276 seconds).
21:38:03 <elliott> his counter is kept in the ( instruction
21:38:11 <ais523_> but loses on long tapes due to me needing to cut loops short in order to avoid crashing egojoust
21:38:20 <elliott> ais523_: just do (x)*n(x)*m
21:38:23 <oerjan> elliott: what _do_ you interpret (a(b{c{d}e}f)%m g)%n as?
21:38:38 <oerjan> elliott: yeah that's also possible
21:38:42 <ais523_> elliott: I don't think it'd help
21:38:48 <elliott> oerjan: err, I can't expand it off the top of my head, but the outer {} is bound to the inner ()%, and the inner {} is bound to the outer ()%, is the egojsout interpretation
21:39:03 <ais523_> there's a threshold betweens "runs quickly" and "doesn't run at all", and I fear it's the point at which egojoust runs out of memory for the expanded program
21:39:13 <ais523_> in which case no amount of synonyming would help
21:40:09 <oerjan> elliott: the thing is that the inner { needs to reach the stack element the outer ( put there, which is akward if the outer { doesn't pop _its_ stack element
21:41:02 <oerjan> although you could probably keep track of it still. in fact that multiple side-by-side {} would probably require it.
21:41:37 <ais523_> oerjan: (a(b{c{d}e}f)%2g)%2 = (abbc{d}effg)%2 = abbcabbcdeffgeffg
21:41:49 <elliott> op_chars[parser->dest[parser->nest_stack[parser->nest_ptr]].op],
21:42:21 <oerjan> ais523_: yeah that should be correct
21:43:17 -!- iconmaster has joined.
21:43:51 <elliott> elliott@elliott-MacBookAir:~/Code/lance$ ./lance '[' '.'
21:43:52 <elliott> Syntax error: Unmatched [ (on line 1, at column 2)
21:44:00 <elliott> ugh, I had better store the original line/column pair in the nesting structure
21:44:13 <elliott> unclosed X gets error at end of file
21:44:55 <elliott> iconmaster: that's not a fast way to commit suicide
21:45:11 <EgoBot> Score for iconmaster_kill_self: 13.6
21:45:52 <Gregor> It's ~256 cycles either way
21:46:07 <Gregor> !bfjoust slowly_kill_self [-(.)*100]
21:46:35 <EgoBot> Score for Gregor_slowly_kill_self: 9.0
21:47:14 <EgoBot> Score for iconmaster_fastfail: 0.0
21:47:46 -!- pikhq_ has joined.
21:48:01 <oerjan> <Grulf> oh noes, wheres oerjan
21:48:03 -!- pikhq has quit (Ping timeout: 260 seconds).
21:48:06 <elliott> !bfjoust i_hope_you_fuck_up_my_tape_before_i_finish_my_deed (-)*128
21:48:17 <EgoBot> Score for elliott_i_hope_you_fuck_up_my_tape_before_i_finish_my_deed: 7.6
21:48:50 <Gregor> !bfjoust spin_cycle [(+)*100000]
21:48:56 <Lymia_> !bfjoust zero-point-one (>-)*10([-]>)*2
21:49:10 -!- Lymia_ has changed nick to Lymia.
21:49:14 -!- Lymia has quit (Changing host).
21:49:14 -!- Lymia has joined.
21:49:15 <elliott> oerjan: inventor of some language
21:50:11 <EgoBot> Score for Gregor_spin_cycle: 11.6
21:50:21 <oerjan> ah Calculon, which i proofread a bit
21:50:39 <EgoBot> Score for Lymia__zero-point-one: 0.0
21:51:09 <Lymia> !bfjoust zero-point-one (>+)*10(>[-])*2
21:51:15 <elliott> Syntax error: Unmatched [ (on line 1, at column 1)
21:51:21 <EgoBot> Score for Lymia_zero-point-one: 0.0
21:51:25 -!- augur has joined.
21:51:50 <EgoBot> Score for iconmaster_super_fail: 3.1
21:52:00 <elliott> Lymia: egobot rounds, try report.txt
21:52:03 <elliott> for that extra decimal digit!
21:52:19 <EgoBot> Use: !bfjoust <program name> <program> . Scoreboard, programs, and a description of score calculation are at http://codu.org/eso/bfjoust/
21:52:32 <elliott> !bfjoust for_want_of_a_better_name (>)*9(-)*128
21:52:48 <Gregor> elliott: 10 is the only tape length dood
21:52:54 <EgoBot> Score for elliott_for_want_of_a_better_name: 0.4
21:53:18 <Gregor> !bfjoust other_side_sui_cide (>)*30
21:53:26 <EgoBot> Score for Gregor_other_side_sui_cide: 0.0
21:54:00 <elliott> ais523_: anti-defence strategy idea
21:54:17 <Lymia> !bfjoust zero-point-one >+(>)*8([-]>)*2
21:54:19 <elliott> ais523_: look for the first non-zero cell, assume it's a tripwire, use this fact to skip it, and also to get some idea of where the flag is
21:54:27 <elliott> (because tripwires aren't likely to be placed too far from the opponent's flag)
21:54:31 <ais523_> elliott: that's what tripwire avoiders do
21:54:43 <elliott> At least I'm as smart as whoever came up with that.
21:54:55 <EgoBot> Score for Lymia_zero-point-one: 2.0
21:54:55 <EgoBot> Score for iconmaster_less_fail: 3.0
21:55:00 <ais523_> although it's often a decoy rather than a tripwire
21:55:21 <Lymia> !bfjoust zero-point-one (>+)*2(>)*7((-)*128>)*2
21:55:30 <EgoBot> Score for Lymia_zero-point-one: 0.0
21:55:39 <elliott> ais523_: The difference being that clearing decoys just wastes time, whereas clearing tripwires is downright dangerous.
21:55:40 <Gregor> !bfjoust flag_is_tripwire [](+)*100000
21:55:54 <ais523_> elliott: indeed; although wasting time is often /also/ outright dangerous
21:55:55 <EgoBot> Score for Gregor_flag_is_tripwire: 3.1
21:56:09 <ais523_> and typical tripwires are not normally last, but buried in a maze of decoys
21:56:25 <ais523_> e.g. defend9.75 changes the value of 7 cells and its flag; three of them are tripwires
21:56:40 <elliott> ais523_: indeed, but wasting time destroying decoys doesn't matter if your program is a defence and you don't hit a tripwire
21:56:54 <elliott> (unless your rush is so slow that the defence program decides to go attack)
21:57:14 <elliott> hmm, attackers have turned into "rushers"
21:57:19 <elliott> despite the fact that a rush is just a fast attack
21:57:21 <EgoBot> Score for iconmaster_dumb: 0.0
21:57:47 <Gregor> !bfjoust this_vibrators_got_four_d_batteries (-)*128+[[(---+++)*100000]+]
21:58:15 <EgoBot> Score for Gregor_this_vibrators_got_four_d_batteries: 13.8
21:58:16 <Lymia> !bfjoust zero-point-one (>+)*2(>)*7([-]>)*2
21:58:30 <EgoBot> Score for Lymia_zero-point-one: 0.2
21:58:36 <Deewiant> !bfjoust möbius (>(+>->)*14(+<-<)*14<)*10000
21:58:42 <EgoBot> Score for Deewiant_m__bius: 0.0
21:59:10 <ais523_> Lymia: trying to get 0.1 exactly?
21:59:20 <Deewiant> !bfjoust möbius (>(+>->)*4(+<-<)*4<)*10000
21:59:22 <ais523_> elliott: terminology's been around ever since slowrush
21:59:33 <elliott> ais523_: yep; slowrush is a bit of a contradiction in the name, though
21:59:38 <EgoBot> Score for Deewiant_m__bius: 0.3
21:59:44 <Lymia> Gregor, damn your huge hill.
21:59:52 <Lymia> How am I supposed to evolve agienst that.
21:59:55 <elliott> ais523_: exercise: figure out whether there are any attack programs that aren't rushes
22:00:00 <elliott> Lymia: fizzie did quite well :P
22:00:12 -!- Mathnerd314 has joined.
22:00:22 <ais523_> elliott: ones that use a forwards tripwire for synchronisation, possibly
22:00:22 <Deewiant> His result was pendolino with a few + added :-P
22:00:45 <elliott> Deewiant: Err, his latest submission was from the example programs.
22:00:47 <elliott> And got something like 33.
22:00:48 <ais523_> btw, I think that if you're trying to counter a particular program, evolving based on the program itself is a good idea
22:00:51 <Gregor> His SOURCE was pendolino, wannit?
22:01:03 <ais523_> e.g. try to make a few changes in order to win the mirror match, at the cost of everything else
22:01:30 <Lymia> !bfjoust zero-point-one (>+)*3(>)*7((-)*128>>)*2
22:01:51 <elliott> ais523_: It is a belief of mine that there exists an algorithm that takes a BF Joust program and returns a program which beats it on all polarities and tape lengths.
22:02:01 <EgoBot> Score for Lymia_zero-point-one: 0.0
22:02:02 <Deewiant> elliott: Yes, that was the careless clear example with a -- trail at the start and a + trail at the end
22:02:10 <EgoBot> Score for iconmaster_dum2: 3.1
22:02:20 <Gregor> elliott, ais523_: Since runtime is limited, there should be by definition.
22:02:22 <Lymia> !bfjoust zero-point-one (>+)*3(>)*7((-)*128>)*2
22:02:29 <elliott> Gregor: Oh, I was ignoring that pesky issue.
22:02:35 <ais523_> I mean, brute force is one such algorithm
22:02:38 <elliott> Gregor: Also I think polarity/tape lengths make it slightly less than trivial.
22:02:40 <elliott> Because you can't detect that.
22:02:44 <EgoBot> Score for Lymia_zero-point-one: 0.0
22:02:45 <Lymia> elliott, question is.
22:02:46 <elliott> (But you can know how the program behaves on all the different ones)
22:02:56 <Lymia> Do you have to solve the halting problem as a part of it?
22:03:06 <Lymia> !bfjoust zero-point-one (>+)*2(>)*9((-)*128>)*2
22:03:08 <elliott> The halting problem is never the answer to any serious question.
22:03:18 <elliott> (Apart from maybe "What dun that Turing fellow")
22:03:20 <EgoBot> Score for Lymia_zero-point-one: 0.0
22:04:02 <Lymia> elliott, eh, well.
22:04:44 <Lymia> I can see it working, it's just that I have no clue how it would work.
22:04:55 <elliott> The nonnull attribute specifies that some function parameters should be non-null pointers. For instance, the declaration:
22:05:11 <elliott> Lymia: there are a finite number of BF Joust programs
22:05:12 <Lymia> Mostly because if you modify your warrior to behave differently, the enemy has a good chance of changing it's behavior in response too.
22:05:16 <elliott> because the cycle count is finite
22:05:27 <elliott> therefore, iterate through every BF Joust program until you find one that beats your input on all tape lengths and polarities
22:05:46 <elliott> (Note: This would be hideously impractical.)
22:05:58 -!- jcp has quit (Ping timeout: 272 seconds).
22:06:17 <Lymia> Nice understatement you have there
22:06:36 <Deewiant> There's only upwards of 7^10000 programs to go through
22:06:52 <iconmaster> Does a unmatched [ or ] cause EgoJoust to error? Or does it compensate somehow?
22:07:04 <Deewiant> Egojoust compensates somehow, the others error
22:07:13 <elliott> iconmaster: It nops. But no, don't do that.
22:07:44 <Lymia> !bfjoust quickest-loser ([)*1
22:07:44 -!- iconmaster has quit (Read error: Connection reset by peer).
22:07:47 <fizzie> Deewiant: For the record, I added in a slight bias for shorter programs; it got rid of the dead ballat at the end, but it hasn't innovated anything else than evo_4. Who knows, though, maybe tomorrow morning it has dreamed up a winrar.
22:07:54 -!- iconmaster has joined.
22:07:55 <elliott> Lymia: It's a nop in egojoust.
22:07:59 <EgoBot> Score for Lymia_quickest-loser: 3.1
22:08:11 <Deewiant> elliott: Oh, for some reason I actually thought it was 10000
22:08:25 <elliott> Nope, -1 just turns into that because Gregor am dum
22:08:38 -!- copumpkin has quit (Quit: Computer has gone to sleep.).
22:09:17 <iconmaster> How about a program that floods the hill with randomly generated programs?
22:10:10 <elliott> Ooh, I have the perfect name for such a program.
22:11:30 -!- jcp has joined.
22:11:40 <ais523_> the hill wouldn't be flooded, random programs would rarely get anywhere
22:11:49 <Lymia> Flood the hill with evolved programs?
22:12:02 <Lymia> As they are evolved agienst eachother...
22:12:03 -!- ais523_ has quit (Quit: Page closed).
22:12:08 <fizzie> That program is calles "fizzie". :p
22:12:26 -!- iconmaster has quit (Read error: Connection reset by peer).
22:12:33 -!- iconmaster has joined.
22:12:48 <fizzie> Maybe "4" does not quite yet equal "flood".
22:12:51 <Lymia> fizzie, how long does a generation normally take to process?
22:13:22 <elliott> He's using cranklance, which is like lance, but operated with a hand-crank.
22:14:14 <fizzie> Lymia: I haven't exactly timed it. Maybe around a minute or five? Depending on the parameters, of course.
22:15:10 <HackEgo> y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y \ y
22:15:17 <HackEgo> no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no \ no
22:15:56 <elliott> I HATE HATE HATE that a case can't point to a variable declaration
22:16:05 <fizzie> HackEgo: Are you saying 'no' just to be negative?
22:16:13 <HackEgo> yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no \ yes \ no
22:16:20 <HackEgo> yes \ no \ maybe \ yes \ no \ maybe \ yes \ no \ maybe \ yes \ no \ maybe \ yes \ no \ maybe \ yes \ no \ maybe \ yes \ no \ maybe \ yes \ no \ maybe \ yes \ no \ maybe \ yes \ no \ maybe \ yes \ no \ maybe \ yes \ no \ maybe \ yes \ no \ maybe \ yes \ no \ maybe \ yes \ no \ maybe \ yes \ no \ maybe \ yes \ no \ maybe \ yes
22:17:16 <oerjan> !sh ls interps | fmt -w400
22:17:17 <EgoBot> 1l 2l Makefile adjust axo befunge bf_txtgen bfjoust boof c-intercal cat cfunge clc-intercal dimensifuck egobch egobf fukyorbrane gcccomp gforth_quit ghc glass glypho kipple lambda lazyk linguine malbolge pbrain qbf rail rhotor sadol sceql trigger udage01 underload unlambda whirl
22:17:44 <elliott> wow that made an unholy noise
22:17:47 <elliott> as all the messages came in at once
22:18:02 <elliott> <EgoBot> # simply cheks when you make that these are installed, which is nice for fresh
22:18:19 <elliott> <EgoBot> .SUFFIXES: .c .cc .cpp .bin
22:18:28 -!- hiato has quit (Ping timeout: 240 seconds).
22:18:33 <elliott> also, s/bin/exe/ is the standard extension for such
22:19:12 <oerjan> !sh echo '(:aSS):aSS' | interps/underload
22:19:22 <EgoBot> /tmp/input.16487: line 1: interps/underload: is a directory
22:19:51 <oerjan> !sh ls interps/underload | fmt -w400
22:19:51 <EgoBot> underload.bin underload.c
22:20:06 <oerjan> !sh echo '(:aSS):aSS' | interps/underload/underload.bin
22:20:07 <EgoBot> Usage: derl (-o|-a) [inputfile]
22:20:33 <fizzie> Heh, that was amusing.
22:20:43 <elliott> variable: it's egobot's -o|
22:20:50 <oerjan> !sh echo '(:aSS):aSS' | interps/underload/underload.bin -o
22:20:50 <EgoBot> (:aSS):aSSAttempt to execute unknown command 10
22:21:03 <oerjan> !sh echo -n '(:aSS):aSS' | interps/underload/underload.bin -o
22:21:32 * variable reads this channel a lot - I should talk more
22:21:53 <elliott> wait we have actual lurkers? :)
22:22:16 * iconmaster talks a lot, and generally makes a fool outa himself
22:23:00 -!- fungot has joined.
22:23:58 <oerjan> !run ls interps/*/*.bin | fmt -w400
22:24:25 <oerjan> !run ls interps/*/*.bin | fmt -w400
22:24:37 <oerjan> isn't that legal syntax? :(
22:24:37 <elliott> iconmaster: not as much a fool as the rest of us make of ourselves on a regular basis :)
22:24:46 <elliott> !run ls interps/*/*; echo hi
22:25:01 <fizzie> I think I like that more.
22:25:06 <oerjan> !sh ls interps/*/*.bin | fmt -w400
22:25:10 <fizzie> It doesn't have them rude words.
22:25:21 <EgoBot> interps/1l/1l_a.bin interps/2l/2li.bin interps/adjust/adjust.bin interps/axo/axopp.0.1.0.bin interps/befunge/bef.bin interps/kipple/cipple.bin interps/malbolge/malbolge.bin interps/pbrain/pbrain.bin interps/sceql/sceql-0.1.bin interps/trigger/trigger.bin interps/udage01/switch.bin interps/underload/underload.bin interps/unlambda/unlambda.bin
22:25:50 <oerjan> bah there doesn't seem to be a uniform naming
22:25:57 -!- hiato has joined.
22:26:01 <elliott> oerjan: indeed, see the makefile for more
22:26:19 <elliott> fizzie: The palindromic one is surely even better, though.
22:26:21 * iconmaster is going to make RPOS v2 this weekend. Joy!
22:26:23 <oerjan> elliott: i am trying to make a program that can run another interpreter while cutting off the input
22:26:34 <oerjan> since EgoBot refuses to do so
22:26:45 <elliott> what do you mean by cutting off the input?
22:26:59 <oerjan> elliott: so that i can give EgoBot underload programs that produce an infinite line
22:27:08 <elliott> cutting off the output then
22:27:10 <elliott> why not just use fungot :-P
22:27:11 <fungot> elliott: you prolly have mzscheme running
22:27:38 <fungot> ^<lang> <code>; ^def <command> <lang> <code>; ^show [command]; lang=bf/ul, code=text/str:N; ^str 0-9 get/set/add [text]; ^style [style]; ^bool
22:27:50 <oerjan> elliott: fungot doesn't take urls
22:27:51 <fungot> oerjan: only i can use them, however, in book or lecture form. ( define-operation ( cell-contents cell))
22:27:56 <EgoBot> iconmaster is a great guy!
22:29:01 <oerjan> !addinterp ulcut sh interp/underload/underload.bin | head -c5000
22:29:02 <EgoBot> Interpreter ulcut installed.
22:29:10 <fungot> ^<lang> <code>; ^def <command> <lang> <code>; ^show [command]; lang=bf/ul, code=text/str:N; ^str 0-9 get/set/add [text]; ^style [style]; ^bool
22:29:21 <elliott> well, it has that str stuff.
22:29:25 <elliott> ask fizzie if it does HTTP.
22:29:28 <EgoBot> /tmp/input.17170: line 1: interp/underload/underload.bin: No such file or directory
22:29:50 <oerjan> !addinterp ulcut sh interp/underload/underload.bin -o | head -c5000
22:29:51 <EgoBot> There is already an interpreter for ulcut!
22:29:58 <EgoBot> Interpreter ulcut deleted.
22:29:59 <oerjan> !addinterp ulcut sh interp/underload/underload.bin -o | head -c5000
22:30:01 <EgoBot> Interpreter ulcut installed.
22:30:05 <EgoBot> /tmp/input.17295: line 1: interp/underload/underload.bin: No such file or directory
22:30:20 <elliott> I'm pretty sure it does http
22:34:39 <EgoBot> Interpreter ulcut deleted.
22:34:49 <fizzie> The str stuff can be used for longer stuff though, right.
22:35:31 <EgoBot> userinterps: Users can add interpreters written in any of the languages in !help languages. See !help addinterp, delinterp, show | !userinterps. List interpreters added with !addinterp.
22:35:37 <EgoBot> perl (sending via DCC)
22:36:27 <oerjan> hm that one definitely can take either stdin or file
22:36:30 <EgoBot> languages: Esoteric: 1l 2l adjust asm axo bch befunge befunge98 bf bf8 bf16 bf32 boolfuck cintercal clcintercal dimensifuck glass glypho haskell kipple lambda lazyk linguine malbolge pbrain perl qbf rail rhotor sadol sceql trigger udage01 underload unlambda whirl. Competitive: bfjoust fyb. Other: asm c cxx forth sh.
22:36:40 <fizzie> The ^help text seems to be lying somewhat, str:n doesn't work everywhere.
22:37:16 <Lymia> !c int main() {printf("test");}
22:38:23 <oerjan> !addinterp test sh echo
22:38:23 <EgoBot> There is already an interpreter for test!
22:38:43 <elliott> Ouch, lance 2 is pretty slow.
22:38:48 <oerjan> !addinterp test sh echo
22:38:48 <EgoBot> There is already an interpreter for test!
22:39:02 <oerjan> must be something other then...
22:39:04 <EgoBot> That is not a user interpreter!
22:39:12 <oerjan> !addinterp test2 sh echo
22:39:15 <EgoBot> Interpreter test2 installed.
22:39:45 <oerjan> hm that looks like something taking stding
22:39:53 <EgoBot> Interpreter test2 deleted.
22:40:28 <oerjan> !addinterp ulcut sh interp/underload/underload.bin -o
22:40:28 <EgoBot> Interpreter ulcut installed.
22:40:34 -!- iconmaster has quit (Quit: Rooms • iPhone IRC Client • http://www.roomsapp.mobi).
22:40:35 <EgoBot> /tmp/input.17885: line 1: interp/underload/underload.bin: No such file or directory
22:41:07 <EgoBot> Interpreter ulcut deleted.
22:41:26 <oerjan> !addinterp ulcut sh interps/underload/underload.bin -o | head -c5000
22:41:27 <EgoBot> Interpreter ulcut installed.
22:41:37 <EgoBot> Attempt to execute unknown command 10
22:41:51 <Lymia> Where does the argument go in anyways.
22:42:06 <Lymia> That is, the program being submitted for running.
22:42:23 <oerjan> Lymia: i think it's sent on stdin for userinterps
22:42:48 <oerjan> but on files for languages
22:43:04 <elliott> Why is this program slower...
22:43:27 <oerjan> also why does it keep adding a newline
22:43:29 <Lymia> !addinterp runc sh cat > tmp.c | gcc -O3 -o tmp tmp.c | ./tmp
22:43:30 <EgoBot> Interpreter runc installed.
22:43:38 <Lymia> !runc int main() {printf("test");}
22:43:39 <EgoBot> /tmp/input.18166: line 1: tmp.c: Permission denied
22:44:00 <oerjan> Lymia: there is already a c language in EgoBot btw
22:44:22 <elliott> comedy of the day: "-Ofast" is an actual gcc option
22:44:28 <EgoBot> Interpreter runc deleted.
22:44:58 <oerjan> Lymia: no changing files directly allowed, iirc
22:45:06 <olsner> elliott: my compiler will have -Olsner
22:45:16 <elliott> olsner: turns on lsner optimisations?
22:45:28 <olsner> yep, lsner, that's right
22:45:42 <Lymia> !c int main() { printf("test or something"); }
22:45:42 <oerjan> !sh interps/underload/underload.bin
22:46:12 <oerjan> !sh interps/underload/underload.bin fnord
22:46:12 <EgoBot> Usage: derl (-o|-a) [inputfile]
22:46:22 <fizzie> Least-Squares Nonlinear Equation R-something?
22:46:33 <olsner> yuck, debugging... as soon as you find yourself debugging a C program you know that you've already failed and might as well start over from the beginning and try to get it right this time
22:46:52 <Lymia> !c void boom () { register int esp asm ("esp"); esp = 50; printf("Byebye, %u!",esp); } int main() { boom(); return 1 }
22:47:09 <oerjan> !sh echo -n '(:aSS):aSS' | interps/underload/underload.bin -o
22:47:30 <Lymia> !c void boom () { register int esp asm ("esp"); esp = 50; printf("Byebye, %u!",esp); } int main () { boom(); return 1; }
22:47:32 <EgoBot> ./interps/gcccomp/gcccomp: line 52: 18490 Segmentation fault /tmp/compiled.$$
22:48:48 <fizzie> There *are* easier ways to segfault.
22:49:00 <olsner> I wonder what happens if you do something like !c int main() { kill(getppid()); }
22:49:00 <Lymia> !c int nya[1024]; void mungTheStack() { register int *esp asm ("esp"); memcpy(esp,nya,1024*sizeof(int)); } int main () { mungTheStack(); }
22:49:03 <EgoBot> ./interps/gcccomp/gcccomp: line 52: 18598 Segmentation fault /tmp/compiled.$$
22:49:03 <Lymia> I"m just messing around
22:49:10 <EgoBot> Attempt to execute unknown command 10
22:50:12 <fizzie> s/1024*sizeof(int)/sizeof nya/ -- even when messing around, no need to look all ugly.
22:50:25 <Lymia> !c int main () { main(); }
22:50:29 <EgoBot> ./interps/gcccomp/gcccomp: line 52: 18705 Segmentation fault /tmp/compiled.$$
22:50:46 <EgoBot> Interpreter ulcut deleted.
22:52:46 <oerjan> !addinterp test2 sh wc
22:53:10 <EgoBot> Interpreter test2 installed.
22:55:34 <EgoBot> Interpreter test2 deleted.
22:56:46 <elliott> Syntax error: Unmatched ( (on line 1, at column 1), closed with [ (on line 1, at column 2)
22:56:51 <elliott> GOOD ENOUGH ERROR REPORTING, DEEWIANT?!
22:57:13 <elliott> Syntax error: Unmatched ( (on line 1, at column 1), closed with ] (on line 1, at column 2)
22:57:18 <elliott> Deewiant: WOULD YOU LIKE THE LINE UNDERLINED
22:57:30 -!- iconmaster has joined.
22:57:40 <elliott> Deewiant: PERHAPS A TWO-PRONGED ARROW POINTING TO BOTH RELEVANT CHARACTERS?
22:57:51 <elliott> Including a 2D line for all those multi-line cases.
23:00:11 <fizzie> Perhaps you should produce a data: URL that encodes a PNG where you've printed the offending code with some fancy highlights and lens flares.
23:00:22 -!- Mathnerd314 has quit (Ping timeout: 240 seconds).
23:00:28 -!- cheater99 has joined.
23:00:49 <elliott> fizzie: Or open the user's browser auto-magically.
23:00:57 <Deewiant> Produce a data: URL with befunge code that draws the offending code using TURT
23:01:47 <elliott> So how's CCBI's error-reporting :)
23:02:06 <Gregor> !bfjoust nonsense [[(-)*256(+)*256]+]
23:03:16 -!- pikhq has joined.
23:03:30 -!- pikhq_ has quit (Ping timeout: 255 seconds).
23:03:56 -!- poiuy_qwert has joined.
23:05:20 <elliott> Deewiant: But Befunge has no errors :P
23:06:01 -!- oerjan has quit (Remote host closed the connection).
23:06:11 -!- oerjan has joined.
23:06:52 <EgoBot> Score for Gregor_nonsense: 11.9
23:06:52 <Deewiant> CCBI :: Infinite loop detected!
23:06:53 <Deewiant> Detected by Funge-Space at (112,0) with delta (1,0):
23:06:53 <Deewiant> IP found itself whilst processing spaces.
23:07:03 <elliott> Deewiant: STANDARDS INCOMPLIANT
23:07:14 <Deewiant> It has the -I option if you really want to infloop
23:08:01 -!- Slereah has quit (Ping timeout: 272 seconds).
23:09:07 <Deewiant> I think the only other one is in STRN.G
23:09:25 <Deewiant> Well, other than the various equivalent situations in which infloops can occur for IPs
23:10:41 <Deewiant> The static area even makes the error messages ugly, it'd detect it at (0,0) otherwise :-/
23:11:21 <elliott> Deewiant: I will pay you £5 if you remove the static area and implement some fancy just-as-fast-but-elegant thing in the next CCBI :-P
23:11:43 <Deewiant> There is no just-as-fast, that's the problem :-P
23:12:57 <elliott> Deewiant: Didn't you say some k-d tree thing would help? :-P
23:13:02 -!- Slereah has joined.
23:13:17 <Deewiant> That just helps when you have multiple boxes
23:13:35 <elliott> Deewiant: I refuse to believe that a static area is necessary to be competitive with cfunge.
23:14:38 <Deewiant> It doesn't change the fact that if (x>=0&&y>=0&&y<128&&x<128) return array[y*128 + x] is faster than if (x>=ox&&y>=oy&&y<ey&&x<ex) return array[y*(ey-oy+1) + x]
23:15:49 <elliott> $ ./lance '(>)*9([-]>*21' ''
23:15:49 <elliott> Syntax error: Unmatched ( (on line 2, at column 1)
23:16:08 <Deewiant> ... and stuff like that is by far the bottleneck?
23:16:20 <elliott> Deewiant: On benchmarks that are how artificial, exactly? :-P
23:16:28 <elliott> I'm sure there's some other part of CCBI less optimised.
23:16:32 <Deewiant> On benchmarks it's the other stuff
23:16:48 <Deewiant> Which is why 2.0 didn't have a static area
23:17:51 <Deewiant> I'm sure there's room for improvement in other areas but if it's in the low 10% and you're more than 10% slower it really doesn't matter
23:18:10 <elliott> Deewiant: How much is the difference? It surely can't be more than 0.05s on anything. It fundamentally violates my worldview that such simple differences in conditionals and arithmetic could be a deal-breaker.
23:18:16 <elliott> Are you sure D overhead isn't hurting you? :-P
23:18:44 <Deewiant> elliott: On fun_got's underload interpreter on some program the difference was like 20s to 16s or something
23:18:59 <elliott> Deewiant: "fun_got"; that kind of stuff violates channel rules I'm sure.
23:19:00 <Deewiant> Given that it was mostly just tight loops in the initial area
23:19:06 <Deewiant> elliott: I didn't want to ping it :-P
23:19:09 <elliott> fizzie: Please tell Deewiant that talking about fungot behind its back is a bannable offence.
23:19:10 <fungot> elliott: c has an ignore restarter, restart/ ignore do? the symbols look familiar ( p,q,r) but i can
23:19:16 <elliott> It has the right to express its views, after all.
23:19:46 <fizzie> elliott: But we all do that about you! Uh, I mean, never mind.
23:19:50 <elliott> Deewiant: Generalisation of my UTTER GENIUS THREADED CODE IDEA: Do magic so that tight loops don't /need/ to look at fungespace like that.
23:20:28 -!- nik340 has joined.
23:20:35 -!- nik340 has left (?).
23:21:48 <elliott> Hmm, what's up with the parser stack here.
23:24:23 <elliott> Syntax error: Unmatched ( (on line 2, at column 1)
23:24:23 <elliott> WHAT DEFINITION OF LINE ARE YOU USING HERE
23:24:39 -!- FireFly has quit (Quit: swatted to death).
23:25:19 -!- Sgeo has joined.
23:27:06 -!- iconmaster has quit (Ping timeout: 240 seconds).
23:27:30 -!- Behold has quit (Remote host closed the connection).
23:32:36 <elliott> there are definitely no newlines here.
23:33:13 -!- tswett has joined.
23:34:19 <Sgeo> Tk actually looks native on Windows nowadays?
23:34:28 <Sgeo> Am I in some kind of alternate reality?
23:35:04 <pikhq> Tk has always looked native on not-UNIX.
23:35:38 <pikhq> And it used to look native there.
23:37:11 <elliott> pikhq: Tk isn't native on OS X.
23:37:17 <elliott> But yeah, on Windows, it looks fine. See: IDLE.
23:37:24 <elliott> (It's native-ISH on OS X.)
23:37:31 <elliott> (But not nearly native enough.)
23:37:47 <Sgeo> elliott, OSX is not not-UNIX
23:37:54 <pikhq> Ah, right, it was merely "made an attempt at native" on OS X.
23:38:18 <elliott> Gregor: So... with your model, ( knows where the { is, and ) knows where the } is, right?
23:38:21 <pikhq> I think it was actually totally native on classic Mac OS.
23:38:21 <elliott> And everything knows where ( is.
23:38:29 <pikhq> And certainly was on random PDAs it was ported to.
23:38:35 <Gregor> elliott: Yeah. In fact, {}) all know where ( is, and ( stores where all the others are.
23:38:36 <fizzie> I don't think Tk looked very good on Windows either; wasn't the whole point of Tile that it looks more native on Windows?
23:38:46 <elliott> Gregor: Toooo muuuuuch indirectionnnnnnnnnnnnnn
23:38:55 <pikhq> fizzie: No, the whole point of Tile was to make it look right on X.
23:38:58 <elliott> All I know is IDLE looks perfect on Windows.
23:39:04 <elliott> As in is-natively-rendered.
23:39:07 <pikhq> fizzie: On Windows it just used native widgets.
23:39:09 <fizzie> "# Native look and feel under Windows XP
23:39:10 <Gregor> Wooooh, my favorite episode of Seinfeld is on.
23:39:12 <fizzie> # Native look and feel under other Windows versions"
23:39:15 <elliott> Gregor: But ( just needs to know where { is, and ) where } is, right?
23:39:20 <pikhq> fizzie: Very odd. It used native widgets.
23:39:23 <elliott> Does } know anything? Apart from where ( is?
23:39:24 <fizzie> That is what Tile lists as their features.
23:39:28 <Gregor> elliott: Yeah. And the counts.
23:39:32 <elliott> fizzie: Perhaps IDLE uses Tile then.
23:39:36 <Sgeo> Why are we talking paste tense?
23:39:50 <elliott> Gregor: Right, well, everything knows where ( is. But I'd rather that ) know where } is directly.
23:39:53 <elliott> Since ( doesn't need to know that.
23:40:00 <elliott> Which means I need MORE FIELDS YAY
23:40:22 <oerjan> 00:50 elliott> Syntax error: Unmatched ( (on line 2, at column 1)
23:40:22 <oerjan> 00:50 elliott> WHAT DEFINITION OF LINE ARE YOU USING HERE
23:40:40 <oerjan> my guess is that it is giving the position of the eof instead of the original (
23:41:01 <elliott> it was that i forgot a break; in the ) case, so it fell through to the \n case
23:41:09 <elliott> a wonderful, stupid combination
23:41:58 <fizzie> In crank, there's two fields; "match" points ( to ) and back, { to } and back; "inner" points from ( to { and back, } to ) and back. I'm not sure if I use the "match" fields at all after parsing is done, though.
23:42:15 <elliott> fizzie: You have to; ( on )*0 jumps to the ).
23:42:22 <elliott> Unless you SNEAKILY ELIDE IT; which I probably should.
23:42:31 <fizzie> Yes, well, for inner-less loops, sure.
23:42:33 <pikhq> Yup, Tk for Windows has always *literally used native Win32 widgets*.
23:42:44 <pikhq> It functions as a wrapper around Win32 on Windows.
23:43:36 <pikhq> Tile, it seems, just lets you use different themes, including using Windows native widgets if you so wish.
23:46:22 <oerjan> elliott: } during runtime doesn't need to know anything if you do things the way i imagine
23:46:38 <elliott> oerjan: that's true, using a stack, but I'm not using a stack
23:46:41 <fizzie> "In the current version, the default
23:46:41 <fizzie> font for Tk buttons, labels, entries etc. is "MS Sans Serif" 8 pt regardless of what font the user
23:46:44 <fizzie> configured in Windows; this makes Tk programs look slightly (but noticably) different from native
23:46:48 <fizzie> Windows dialog applications."
23:46:50 <elliott> oerjan: I'm storing the iteration in the (, because egojsout does it that way, and because my last attempt was broken
23:46:55 <fizzie> Could have been bugs like that, I guess.
23:47:11 <oerjan> elliott: ok without a stack it _also_ doesn't need to know anything
23:47:24 <elliott> oerjan: it needs to know where ( is
23:47:28 <elliott> so it knows where to put the 0 or whatever
23:47:41 <oerjan> elliott: no it doesn't, because { should already have set it to 0
23:47:56 -!- pikhq has quit (Ping timeout: 240 seconds).
23:48:10 -!- pikhq has joined.
23:48:55 <oerjan> elliott: { and ) both need to know the location of the counter, one of them, ) it seems in this case, needs to know the maximum value for it
23:49:20 <oerjan> and { needs to know the position of (, and ) needs to know the position of }
23:49:31 <Gregor> "Guys and dolls, isn't that a lavish Broadway musical?" "It's guys and dolls, not guys and guys."
23:49:42 <oerjan> ( needs to know the position of the counter and the maximal value
23:50:00 <fizzie> oerjan: cranklance uses a stack of repetition-counter values; ( pushes old to keep it safe, inits to N; { decrements; pops if zero, jumps otherwise; } again pushes the old value and initializes to 1; ) increments and then either jumps back or pops. Is that the same as your way?
23:50:19 <elliott> oerjan: incidentally I've been learning that making everything elegant and structs and functions slows things down even if you inline them :/
23:50:22 <elliott> mostly for max-cycle programs
23:50:38 <elliott> Gregor: I need your EXPERT OPINION.
23:50:49 <Gregor> In my opinion, I am an expert.
23:50:53 <Gregor> There, now you have my expert opinion.
23:51:07 <elliott> Gregor: Did it take the gcc developers longer to write the error for "case 'x': type var = val;" than it would have for them to automatically transform it into "case 'x': ; type var = val;"?
23:51:21 <elliott> This would also have achieved: NOT LITTERING CODE WITH SEEMINGLY-MEANINGLESS ";"S
23:51:50 <fizzie> I'm sure it's THE LAW.
23:51:51 <Gregor> That's a bad place to put a type declaration anyway, if you need types local to a case you should put them in a block.
23:51:54 <oerjan> fizzie: apart from the "old value" stuff which i don't see the point of
23:52:28 <Gregor> s/types local/variables local/
23:52:38 <elliott> <Gregor> That's a bad place to put a type declaration anyway, if you need types local to a case you should put them in a block.
23:52:42 <elliott> Maybe if C didn't make that hideous, I would.
23:52:46 <elliott> <fizzie> I'm sure it's THE LAW.
23:52:52 <fizzie> oerjan: The "old value" is the previous value in the "current counter" variable { and ) manipulate. I guess you could treat it as the top of the stack.
23:52:53 <Gregor> case foo: { int x; /* OH NOSE */ }
23:53:03 <elliott> Gregor: Yes, that is indeed ugly.
23:53:14 <elliott> Gregor: Of course were the syntax:
23:53:20 <elliott> switch (foo) { case ('x') { ... } }
23:53:31 <elliott> then there would be no problem. Also, Duff would have been stopped before he could propagate his insanity.
23:53:39 <elliott> fizzie: You could do what C# does, have "continue;".
23:54:06 <fizzie> Gregor: I used to do that sort of thing before my editor started indenting it all funny.
23:54:19 <oerjan> fizzie: oh i see you are not considering the "current value" as on the stack. well that should be equivalent to considering it so, in which case this is all equivalent to my idea i think
23:54:39 <Gregor> fizzie: My editor (VIM!) does it right :P
23:55:26 <fizzie> Gregor: Curiously enough, this happened about the same time as my latest Vim to Emacs flop. Hmm, I wonder... nah, can't be related.
23:55:52 <oerjan> elliott: incidentally if you store the counter and maximal value at (, then only ) needs to keep track of more than one of the others, i think ( both } and ) )
23:56:45 <oerjan> well ( and } need to know none of the others, { needs to know (
23:57:19 <oerjan> er *i think ( both } and ( )
23:57:58 <elliott> "Oh, I want to use vi, but I'd rather it be at least half as bloated as Emacs..."
23:58:04 <elliott> "Think you could arrange that for me?"
23:58:29 <elliott> ((((DELETED X-RATED SCENE "Also, it would be great if it told me about Ugandan children every time I started it up. Then I'd get good, bloated software, *and* awareness!"))))
23:59:57 <fizzie> Oh, the current splash (at least in this build) doesn't mention Uganda at all. Feels weird. (I've been ignoring it so haven't noticed.)