00:03:33 if you tall likk thii, all worr can be redd to fouu lett. 00:05:20 um... shouldn't that be "cann", "beee" and "tooo"? 00:05:50 oh and alll, ifff (stands for: if and only iff) 00:06:08 night 00:06:44 -!- oerjan has joined. 00:07:19 AnMaster: only words >=4 letters 00:08:06 ah 00:08:08 still, I like ifff 00:08:12 iffffffffffffffffffff 00:09:02 indeed! :D 00:12:09 -!- KingOfKarlsruhe has quit (Remote closed the connection). 00:13:02 I like using punctuation at random."“»›‹¡¿‽’—– 00:13:19 ‽‽‽‽‽“º-ß•ªˆ´•¶þ̂‡†›‹̂†‡°̂*(~↙‽↘⁴⁵⁴⁵ 00:13:35 Touché. 00:13:45 (douché) 00:14:29 New favourite word: Douché 00:14:42 French 00:14:42 [edit]Verb 00:14:43 douché 00:14:45 Past participle of doucher. 00:14:47 "To shower" 00:14:54 pikhq: you've showered? 00:14:57 *you 00:15:07 LMAO 00:19:06 * oerjan tends to smirk whenever he sees the company name Deloitte Touche 00:19:18 the reason, now, should be obvious. 00:20:09 LMAO 00:21:03 * oerjan didn't know about the tohmatsu at the end. it doesn't really help. 00:23:30 Touché, douch险®. 00:25:01 * oerjan notes that he is not alone 00:25:32 Indeed. Ørjan. 00:25:54 * oerjan gives pikhq a spelling bee medal 00:26:25 SPËLLÏNG BËË! 00:26:54 * pikhq has been abusing the Compose key 00:28:52 Höw rúdê 00:31:06 Ï’ḿ ābüßïñg tḧẽ őþtıøñ ké¥ 00:55:29 -!- Patashu has joined. 01:06:20 -!- nescience has joined. 01:07:20 -!- inurinternet has quit (No route to host). 01:14:50 in this bfjoust thing, values are 8 bits right? (256 = 0) 01:15:03 yes 01:15:11 nescience: http://codu.org/eso/bfjoust/report.txt 01:15:16 and !bfjoust name program 01:15:21 omit your name, it adds it automatically 01:15:35 nescience: program sources at http://codu.org/eso/bfjoust/in_egobot/ 01:15:41 i submitted it already 01:15:43 :> 01:15:43 ah 01:15:45 i didn't notice :) 01:15:50 but i apparently miscalculated 01:15:57 i need more details on how [] works 01:16:03 ehm 01:16:09 [a] is while (*ptr) a 01:16:16 as in, while the current cell is not 0, run a 01:16:20 my understanding was that when it reaches ] it checks if 0 01:16:23 nescience: a loop takes two iterations, iirc 01:16:29 so [a][a][a] 01:16:38 so if it was 0 when it executes [, it will still execute the code inside the brackets, yes? 01:16:39 that is, ] jumps back to [ if the cell isn't 0, I believe 01:16:42 nescience: no 01:16:45 whoops 01:16:48 :P 01:16:50 both [ and ] could be said to check 01:17:02 annoying, ok trying again 01:17:03 nescience: let's say your a runs 3 times before setting it to 0 01:17:05 the cycles used are: 01:17:13 [a][a][a]b 01:17:15 where b is the program after 01:17:17 I believe. 01:17:24 that's only relevant if you're counting cyclse 01:17:25 *cycles 01:17:38 if [ checks, that explains what happened 01:17:42 fixing now :> 01:17:52 so wait 01:17:55 if your poiner is on a 0 01:18:06 and your code is [+], for example 01:18:10 -!- Corun has changed nick to Corun|away. 01:18:12 nothing gets incremented, and it takes one cycle? 01:18:18 ehird: not two iterations, the ] jumps to _after_ the [ 01:18:27 oerjan: oh, right 01:18:32 nescience: it'd use [a]a]a]b 01:18:38 * nescience nods 01:18:38 nescience: and correct 01:18:48 ok 01:18:55 i can remove one wait instruction at least then :) 01:19:04 if your code is [-] and your pointer is 3, [-]-]-] is executed 01:19:21 i'm just wanting to know about the case where pointer is 0 atm 01:19:34 nescience: then [ is executed 01:19:44 but not +] 01:20:06 and the 0-check, for the purposes of both programs executing at once, happens before instructions are executed 01:20:28 nescience: here's some exact semantics for you 01:20:37 note that these all happen on the tick - nothing happens before or after a tick 01:20:48 [ if the current cell is 0, jump to the instruction after the matching ]. otherwise, nop. 01:20:56 ] if the current cell is 0, nop. otherwise, jump to the instruction after the matching [. 01:21:03 thanks 01:21:12 welcome 01:21:47 knowing my luck the 'fixed' version will probably score worse, but we'll see 01:21:49 -!- Corun|away has changed nick to Corun. 01:22:01 nope, did better :> 01:22:19 and beats defend6, like it was supposed to 01:22:19 nescience: btw, you should try it in channel 01:22:23 that chart is hard to read 01:22:23 as it reports the score immediately 01:22:28 ah 01:22:40 it only runs one matchup? 01:22:41 and it saves us going to the file list :-) 01:22:45 nescience: right, then updates the report 01:23:06 since the board is random, that seems like a bad idea 01:23:15 nescience: 'the board is random'? 01:23:15 err, the "tape" 01:23:17 ah 01:23:19 not really 01:23:23 it doesn't change the scores much. 01:23:26 no, wait 01:23:27 the length of the playing field 01:23:28 nescience: it runs many matchups 01:23:32 but only (yourprog,*) 01:23:36 the actual report has to run (*,*) 01:23:39 although it -does- cache 01:23:40 mk 01:23:51 thought that would be how it works, but i don't get the +/- then 01:23:58 ehird: it didn't cache before today, at least 01:24:02 i guess it just tells you overall if you won >50% or something? 01:24:03 reruns all matches 01:24:08 nescience: + is "won against this program" 01:24:11 - is "lost against this program" 01:24:18 0 is "draw" 01:24:21 space is "this IS that prorgam!" 01:24:24 oerjan: it doesn ow 01:24:34 nescience: a random tape length is picked for each matchup 01:24:54 yes, the "won" or "lost" was apparent, but made it seem like each pair only battled once 01:24:59 nescience: they do 01:25:00 which, for a random length field seems like a bad idea 01:25:04 ah. 01:25:05 why? 01:25:09 it doesn't matter in 90% of cases 01:25:14 nescience: it's just to stop this degenerate strategy: 01:25:18 because either 1) random length doesn't matter, in which case it shouldn't be random 01:25:19 (>)*tapelength[-] 01:25:29 nescience: which is provably perfect for a given constant tapelength N 01:25:31 or 2) it matters, which means it affects the outcome and you can get "lucky" 01:25:42 yeah, i follow 01:25:52 nescience: you can get lucky ONCE by guessing the tape length correctly in advance. However, that length will be different for every other matchup. 01:26:02 i don't mean get lucky that way 01:26:04 And the probability is incredibly high that it will fail hard on the others. 01:26:10 i mean, some strategies are likely to be faster or slower than others 01:26:20 and thus the tape length might cause a given matchup to win or lose 01:26:22 nescience: Yes, but the statistical difference is incredibly minimal. 01:26:37 it doesn't seem fair to give only one chance if the random selection is what determines the outcome 01:26:39 * nescience shrugs 01:26:40 +/- 10 points in one uncommon case in my tests. 01:26:44 er, score 01:26:57 which is just 10% error, really quite irrelevant 01:27:02 nescience: remember that whenever someone adds a new warrior, 01:27:04 well since there's <10 points difference between the top entries.. :P 01:27:06 it will battle against every one 01:27:08 with a new tape length 01:27:23 nescience: remember, that was an odd cas 01:27:23 e 01:27:29 I imagine +/- 3-5 points is the most common case 01:27:32 it's Good Enough 01:27:33 yes, i'm doing a lot of remembering :) 01:27:37 if caching is working, then it would be feasible to battle _all_ possible tape lengths 01:27:42 and if #1 is -5 and #2 is +5? 01:27:50 and take the average 01:27:54 oerjan: that would be pretty nice 01:28:11 oerjan: that would be over 10x slower 01:28:17 i didn't really consider that the randomness is so narrow that an exhaustive test is feasible 01:28:18 as there is over 10 variations in tape length 01:28:21 it's slow enough as is 01:28:24 ANYWAY 01:28:27 This really doesn't matter. 01:28:29 seemed pretty fast to me :P 01:28:38 do like corewars stuff and queue up the entries, who cares about instant results! 01:28:44 hehe 01:28:52 ehird: 10..30 isn't it? so only 20 times, which is half made up for by the hill now caching the rest 01:28:57 nescience: it took ~25-35 seconds to generate the report previously 01:28:59 which is lame 01:29:09 (w/o caching) 01:29:20 ok i guess that's not so nice 01:29:47 25-35 seconds to run 10 battles 01:29:55 so it'd take a minute to run 20, w 01:29:59 not that bad :P 01:30:01 !bfjoust et_tu_brute (>)*9((-)*128.>)*20 01:30:06 Score for ehird_et_tu_brute: 12.0 01:30:13 hm that did really terribly 01:30:16 -7 points 01:30:19 I wonder why? 01:30:36 It goes the minimum length, then bruteforcedly decrements 128, sleeps a cycle, then steps onward. 01:30:51 !bfjoust et_tu_brute (>)*9((-)*128.>)*21 01:30:53 Score for ehird_et_tu_brute: 12.0 01:30:55 ehird: because you don't stop when it gets to 0? 01:31:02 run off the end 01:31:07 no no no 01:31:09 maybe 01:31:23 nescience: it decrements 128 (the default tape value, and nobody seems to tamper with it), then sleeps one cycle 01:31:27 yeah, i can see it 01:31:34 so if it hits the opponents' flag, it wins immediately 01:31:37 as long as it's 128 01:31:46 seems like some sort of bug 01:31:55 oerjan: stop when it gets to 0? I don't follow 01:32:00 it sleeps one cycle after decimating 01:32:01 well, defend6 will beat it 01:32:07 so it wins, if it's got the opponent's flag 01:32:15 and it's 128 beforehand 01:32:15 i haven't read all the rest but most of them don't "turn back" and do anything 01:32:16 ehird: it will be horrible for all spots that _aren't_ 128 01:32:25 yeah, i guess it just takes too long 01:32:29 oerjan: ah, probably 01:32:29 since it doesn't skip over the 0s fast 01:32:40 * ehird looks up the "if not 0" thing on the wiki 01:32:51 oh? i'm interested 01:32:58 i was just pondering that earlier 01:33:07 hm... 01:33:14 nescience: http://esolangs.org/wiki/Brainfuck_algorithms#if_.28x_.3D.3D_0.29_.7B_code_.7D 01:33:16 you need two temporaries 01:33:18 which is a pain 01:34:04 ooh 01:34:25 nescience: http://esolangs.org/wiki/Brainfuck_algorithms#x_.3D_not_x_.28boolean.2C_logical.29 01:34:28 perhaps 01:34:48 argh 01:34:50 it's inefficient 01:35:05 i kinda like my solution, it exploits defend6 and makes it kill itself :> 01:35:33 i score less than impomatic's though 01:35:40 and i have 2 losses and a tie, while his has 3 losses 01:35:42 i r confus? 01:36:28 !bfjoust mi_fili (>)*9(+[--[(-)*127]]>)*21 01:36:42 !bfjoust et_tu_brute (>)*9(>[-]>[-]<<[>+>+<<-]>[<+>-]+>[<->[-]]<[<(-)*128.>-])*19 01:36:44 Score for ehird_et_tu_brute: 12.0 01:36:49 haha bollocks. 01:36:50 Score for oerjan_mi_fili: 14.9 01:36:54 oh wait 01:36:57 I BEAT OERJAN YAY 01:36:58 what is this stuff on the wiki 01:37:05 i'm not checking it it's _already_ 0 01:37:05 nescience: whatya mean 01:37:05 temp0[-] 01:37:13 In the interest of generality, the algorithms will use variable names in place of the < and > instructions. Temporary cells are denoted "temp". When using an algorithm in a program, replace the variable names with the correct number of < or > instructions to position the pointer at the desired memory cell. 01:37:14 Example: 01:37:16 If "a" is designated to be cell 1, "b" is cell 4, and the pointer is currently at cell 0, then: 01:37:18 nescience: that 01:37:24 so, if you say x=cell 0, temp0=cell 1, temp1=cell 2 01:37:31 x temp0 x temp1 01:37:32 would be 01:37:34 ah 01:37:36 (starting on cell 1) 01:37:42 <><>> 01:37:44 !bfjoust mi_fili2 (>)*9([+[--[(-)*127]]]>)*21 01:37:47 (>)*9(>[-]>[-]<<[>+>+<<-]>[<+>-]+>[<->[-]]<[<(-)*128.>-])*19 01:37:48 should be 01:37:48 i follow 01:37:49 Score for oerjan_mi_fili2: 41.0 01:37:54 huh 01:37:59 O_O 01:38:06 woot 2nd place 01:38:08 ZR 01:38:10 :P 01:38:21 * nescience studies the competition 01:38:24 "Go right 9 times. cell = (cell == 0) (using temps cell+1, cell+2). If it is not zero, decrement 128. 01:38:25 hm wait 01:38:27 I may have it the other way around 01:38:28 whatever 01:38:44 http://codu.org/eso/bfjoust/in_egobot/impomatic_kicks_ehird.bfjoust 01:38:48 nescience: oerjan: try and beat t hat one 01:38:51 it's really good for some reason 01:39:13 it was just designed to beat http://codu.org/eso/bfjoust/in_egobot/ehird_flux_a_counteracting_monomorphism_cocktails.bfjoust, which is my copy of impomatic_flux.bfjoust swapping + and - 01:39:24 !bfjoust creep (>-+[-].++>-+[-].--)*15 01:39:25 Score for nescience_creep: 64.0 01:39:44 wait what just happened 01:39:51 824.69-5nescience_creep.bfjoust 01:39:57 ha 01:39:59 also i think i musta had a bug 01:40:02 nescience: tape length dependent? 01:40:05 no 01:40:13 that as the 2nd plac entry, just changed to decs instead of incs 01:40:20 'cause impomatic's leaves cells with 1 01:40:28 lemme think for a sec 01:40:52 i dunno, putting it back :P 01:40:58 !bfjoust creep (>-+[+].++>-+[+].--)*15 01:41:00 Score for nescience_creep: 24.7 01:41:04 !bfjoust flux_a_revolutionary_perspective_and_et_tu_brute (>)*8(>[(+)*128.])*21 01:41:06 the bot says the old score 01:41:07 Score for ehird_flux_a_revolutionary_perspective_and_et_tu_brute: 24.0 01:41:07 not the new one 01:41:17 now i was at 4th plac 01:41:18 nescience: it takes a little to generate the report. 01:41:20 on the same hill 01:41:20 refresh 01:41:28 talk about random lengths not affecting it eh? :) 01:41:38 !bfjoust flux_a_revolutionary_perspective_and_et_tu_brute (>)*8(>[(+)*255.])*21 01:41:40 Score for ehird_flux_a_revolutionary_perspective_and_et_tu_brute: 24.0 01:41:55 i wonder, maybe something's just wrong 01:42:03 !bfjoust flux_a_revolutionary_perspective_and_et_tu_brute (>)*8([(+)*128.]>)*21 01:42:05 Score for ehird_flux_a_revolutionary_perspective_and_et_tu_brute: 24.0 01:42:10 !bfjoust flux_a_revolutionary_perspective_and_et_tu_brute (>)*8([(+)*128..]>)*21 01:42:12 Score for ehird_flux_a_revolutionary_perspective_and_et_tu_brute: 24.0 01:42:18 i assumed that submitting it would give the same result against the same hill where it was already sitting there 01:42:21 true/not true? 01:42:28 nescience: wut 01:42:29 ? 01:42:56 i'm referring to toggling between 2nd and 4th(5th? didn't get a very good look) place with the same code 01:43:04 !bfjoust hangup >(+)*20001 01:43:12 nescience: odd. oh well 01:43:18 Score for ehird_hangup: 6.0 01:43:20 mwahaha, hangup is the slow 01:43:24 wow that's bad 01:43:28 lol 01:43:36 nescience: submit a warrior - any warrior - just to get rid of hangup :D 01:43:57 !bfjoust hangup2 >(-)*20001 01:44:00 * nescience grins 01:44:06 ha 01:44:12 Score for nescience_hangup2: 6.0 01:44:16 and it replaces me! 01:44:17 !bfjoust chopsueyside (-)*128 01:44:21 Score for ehird_chopsueyside: 6.0 01:44:24 baha 01:44:26 *bahaha 01:44:31 i bet 6.0 is the worst possible score 01:44:39 GregorR-L: why does it add .0 and .00 if it's always whole? 01:45:43 !bfjoust creep (>--[+]+>--[+]-)*15 01:45:44 Score for nescience_creep: 60.0 01:45:51 didn't do much for it did it :P 01:46:09 !bfjoust risktaker (>)*20[>[-].+] 01:46:09 oh crap, i wanted to do slightly different 01:46:13 but i guess we're bein all spammy in here so 01:46:13 Score for ehird_risktaker: 6.0 01:46:30 hahaha 01:46:34 nescience: what the hell is up with my progs :) 01:46:41 also i broke the exploitation of defend6 again 01:47:32 !bfjoust defend6_a_parody_or_just_plain_ripoff_question_mark http://pastie.org/490821.txt?key=w47xqyyhmkklmhegqzsog 01:47:37 !bfjoust creep (>-+[-]+>-+[+]-)*15 01:47:39 Score for nescience_creep: 52.0 01:47:47 Score for ehird_defend6_a_parody_or_just_plain_ripoff_question_mark: 76.0 01:47:55 IDScorePtsProgram 01:47:55 276.006ehird_defend6_a_parody_or_just_plain_ripoff_question_mark.bfjoust 01:47:55 wtfwtf 01:47:56 772.004impomatic_kicks_ehird.bfjoust 01:47:58 862.004impomatic_shortsword.bfjoust 01:48:00 158.001ais523_defend6.bfjoust 01:48:02 nescience: i just swapped + to - 01:48:04 in defend6 01:48:06 :-D 01:48:07 i guess i should just stick to not changing anything 01:48:08 AND NOW I AM KING OF THE HILL!!!! 01:48:11 har 01:48:20 i didn't know it took a url 01:48:20 nescience: try and beat both defend6 and defend6_a_parody_or_just_plain_ripoff_question_mark 01:48:21 >:) 01:48:27 luls 01:48:28 ok 01:48:28 and yeah, that's how ais523 entered defend6 01:48:31 since it was too long 01:48:34 first let me find my working code 01:48:44 01:45 nescience: !bfjoust creep (>--[+]+>--[+]-)*15 01:48:44 01:45 EgoBot: Score for nescience_creep: 60.0 01:49:02 fun fact: flux_a_counteracting_monomorphism_cocktails did quite well 01:49:05 when impomatic_flux was in the hill 01:49:07 I swapped + and - 01:49:10 just like I did with defend6 01:49:13 it's a sure-fire strategy! 01:49:30 but i'm writing my warrior to be sign ambiguous 01:49:31 :> 01:50:59 oh, i looked at the code 01:51:00 you dick :P 01:51:22 sry 'bout that 01:51:25 hehe 01:51:28 mah internet did broke 01:51:31 no worries, i will just submit a different one 01:51:36 that does something different! 01:51:40 ofc since I'm on a bouncer you all didn't notice 01:51:46 thus you, nescience, think I am referring to bf joust! 01:51:57 17:49:31 :> 01:51:57 17:50:59 oh, i looked at the code 01:51:59 17:51:00 you dick :P 01:52:01 didn't notice that 01:52:03 but er I just said I did that : 01:52:05 D 01:52:07 aha? 01:52:07 :D 01:52:14 yeah, but i forgot quite what it entailed 01:52:24 my solution can't be as elegant now 01:52:34 or i could just submit two eaters 01:52:34 :P 01:52:41 how do I look at the current hill's status? 01:52:43 !bfjoust im_a_creep_im_a_weirdo_im_a_ripoff_lolradiohead (>--[-]->--[-]+)*15 01:52:46 Patashu: http://codu.org/eso/bfjoust/report.txt 01:52:52 !bfjoust creep (>-+[+]++>-+[+]--)*15 01:52:53 Score for nescience_creep: 27.0 01:53:01 Score for ehird_im_a_creep_im_a_weirdo_im_a_ripoff_lolradiohead: 26.0 01:53:04 hahaha 01:53:05 !bfjoust inverse_creep (>+-[-]++>+-[-]--)*15 01:53:10 Score for nescience_inverse_creep: 58.0 01:53:12 nescience: that's not inverse 01:53:14 the last one should be + 01:53:15 like mine 01:53:17 no 01:53:37 whatever, it doesn't matter about that part 01:53:48 you swapped it wrong :) 01:54:11 hehe, one beats defend6 and one beats parody 01:54:17 i wonder if i can do em both at once though 01:54:31 the elegant part is i didn't have to delay every step 01:54:31 !bfjoust (>>[-])*200 01:54:31 Use: !bfjoust 01:54:37 !bfjoust naive (>>[-])*200 01:54:39 since i took advantage of it sitting on a 1 01:54:40 Score for comex_naive: 15.0 01:54:48 i just made mine loop around the "long way", giving it time to suicide 01:55:03 !bfjoust naive2 (>[-])*200 01:55:07 but i can't very well do that in both directions 01:55:07 Score for comex_naive2: 5.0 01:55:18 so i'd have to delay at every step, not just the one before the winning spot (vs that warrior) 01:55:39 or time it so that it doesn't matter 01:55:40 !bfjoust naiverush (>)*9([-]>)*21 01:55:42 maybe i should do that 01:55:46 Score for Patashu_naiverush: 9.0 01:55:57 !bfjoust naive2 (>-[-])*200 01:55:58 Score for comex_naive2: 5.0 01:56:05 !bfjoust naive2 (>--[-])*200 01:56:06 Score for comex_naive2: 5.0 01:56:08 !bfjoust mover_is_a_groover_not_naiver_but_a_cleaver (>[-].)*30 01:56:10 !bfjoust naive2 (>--[-])*0 01:56:12 Score for comex_naive2: 5.0 01:56:18 !bfjoust rushpolarity (>)*9([-]>[+]>)*11 01:56:25 oh, I forgot to change the name 01:56:26 freaks :P 01:56:29 Score for Patashu_rushpolarity: 13.9 01:56:29 Score for ehird_mover_is_a_groover_not_naiver_but_a_cleaver: 13.2 01:56:30 !bfjoust naive3 (>--[-])*0 01:56:33 comex: stop that 01:56:34 use the same name 01:56:37 otherwise you cluter the hill 01:56:37 poor bot 01:56:39 *clutter 01:56:44 01:56 EgoBot: Score for ehird_mover_is_a_groover_not_naiver_but_a_cleaver: 13.2 01:56:47 sweet i lost awesomely 01:56:51 whoa 01:56:52 12 programs 01:56:56 ehird: then why does an empty script have a score of 5 01:57:05 comex: because some other ones are even stupider? 01:57:10 !bfjoust rushpolarity >++++++>------(>)*7([-]>[+]>)*11 01:57:11 Score for Patashu_rushpolarity: 13.9 01:57:13 !bfjoust naive2 < 01:57:14 Score for comex_naive2: 6.2 01:57:18 WAT 01:57:31 !bfjoust rushpolarity >++++++>------(>)*7(+[-]>-[+]>)*11 01:57:32 Score for Patashu_rushpolarity: 42.1 01:57:33 comex: off end of tape = draw 01:57:34 i believe 01:57:36 aha 01:57:37 :) 01:57:37 !bfjoust naive4 (>-->-[-])*0 01:57:44 !bfjoust frenchie (-)*128 01:57:44 sorry, mistake 01:57:48 !bfjoust naive2 (>-->-[-])*200 01:57:49 Score for ehird_frenchie: 5.4 01:57:50 Score for comex_naive4: 5.4 01:57:50 Score for comex_naive2: 6.2 01:57:53 5.4 01:57:54 beat that fuckers 01:57:57 !bfjoust naive2 (>-->--[-])*200 01:57:58 Score for comex_naive2: 6.2 01:58:03 if I submit one with a name and the previous one with that name got a higher score, it stays right? 01:58:08 !bfjoust naive2 (>(-)*128)*200 01:58:09 Score for comex_naive2: 6.2 01:58:10 Patashu: no 01:58:12 !bfjoust frenchie (-)*128 01:58:14 Score for ehird_frenchie: 5.5 01:58:15 !bfjoust beatinit [-] 01:58:20 45.50-10ehird_frenchie.bfjoust 01:58:22 that's just beautiful 01:58:22 k 01:58:28 beautifully terrible 01:58:29 Score for nescience_beatinit: 11.0 01:58:34 !bfjoust rushpolarity >++++++>------(>-)*7(+[-]>-[+]>)*11 01:58:34 WAT 01:58:35 nescience: takes longer to suicide 01:58:36 Score for Patashu_rushpolarity: 24.0 01:58:39 !bfjoust defense >[([+]+)*200] 01:58:40 oh 01:58:42 [-]-]-]-] etc 01:58:43 wait 01:58:43 !bfjoust rushpolarity >+++++>-----(>)*7(+[-]>-[+]>)*11 01:58:45 whereas mine does ------------------------ 01:58:45 Score for Patashu_rushpolarity: 46.9 01:58:48 !bfjoust beatinit < 01:58:49 nice 01:58:50 Score for nescience_beatinit: 11.0 01:58:52 WAT 01:58:55 !bfjoust defense >[([+]+)*200] 01:58:57 nescience: < = draw 01:58:59 unmatched ] = draw 01:58:59 etc 01:59:04 no 01:59:07 Score for comex_defense: 5.8 01:59:07 Score for comex_defense: 5.8 01:59:08 win > draw > loss 01:59:08 it only draws if they both lose 01:59:12 !bfjoust rushpolarity >++++>----(>)*7(+[-]>-[+]>)*11 01:59:13 i got -11 01:59:14 Score for Patashu_rushpolarity: 24.8 01:59:15 :) 01:59:19 all losses 01:59:19 !bfjoust rushpolarity >+++++>----(>)*7(+[-]>-[+]>)*11 01:59:20 how the hell is this calculated 01:59:21 Score for Patashu_rushpolarity: 25.9 01:59:24 fjear 01:59:25 !bfjoust rushpolarity >+++++>-----(>)*7(+[-]>-[+]>)*11 01:59:26 Score for Patashu_rushpolarity: 59.3 01:59:29 hmm 01:59:34 comex: Wins against bad opponents count less for your score. 01:59:36 those decoys are really important haha 01:59:39 Points is just wins - losses. 01:59:45 !bfjoust rushpolarity >+++++>----->+(>)*6(+[-]>-[+]>)*11 01:59:46 But score is better. 01:59:47 Score for Patashu_rushpolarity: 6.2 01:59:52 !bfjoust rushpolarity >+++++>----->+++++(>)*6(+[-]>-[+]>)*11 01:59:54 Score for Patashu_rushpolarity: 19.8 01:59:56 As you have to beat good warriors to get a high score. 02:00:01 !bfjoust rushpolarity >+++++>-----(>)*7(+[-]>-[+]>)*11 02:00:03 Score for Patashu_rushpolarity: 13.6 02:00:03 nobody told me that 02:00:07 nescience: now I did 02:00:11 huh 02:00:13 so i guess if i beat both defend6 and its inverse i can get good points 02:00:18 I changed it back to an old setup and it's scoring much lower 02:00:20 haha 02:00:22 Patashu: nescience: http://codu.org/eso/fyb/SCORES 02:00:23 i could just resubmit until i get lucky 02:00:25 >:) 02:00:27 !bfjoust creep (>+-[-]++>-+[+]--)*15 02:00:28 Score for nescience_creep: 30.9 02:00:29 !bfjoust defense (>-)*50 02:00:30 it has one for each of them 02:00:31 Score for comex_defense: 5.8 02:00:31 applies to FYB but bfjoust is the same system 02:00:37 !bfjoust defense (>->+)*25 02:00:39 Score for comex_defense: 5.0 02:00:45 why 02:00:50 it has one for each of them 02:00:53 wait 02:00:53 !bfjoust creep (>+-[-]++>-+[+]--)*15 02:00:55 !bfjoust defense (>->+)*20 02:00:55 Score for nescience_creep: 29.0 02:00:56 Score for comex_defense: 5.0 02:01:00 !bfjoust rushpolarity >+++++>-----(>)*7(++[-]>--[+]>)*11 02:01:01 02:00 comex: why 02:01:02 Score for Patashu_rushpolarity: 33.0 02:01:04 you go off the edge of the tape 02:01:07 and never get a flag 02:01:10 how big is the tape? 02:01:12 !bfjoust rushpolarity >+++++>-----(>)*7(+++[-]>---[+]>)*11 02:01:12 comex: 10-30 02:01:13 randomly 02:01:14 Score for Patashu_rushpolarity: 42.0 02:01:15 oh 02:01:17 so different from agora 02:01:21 yes... 02:01:23 and . is a nop 02:01:23 !bfjoust defense (>->+)*5 02:01:25 Score for comex_defense: 5.5 02:01:27 !bfjoust defense (>->+)*10 02:01:28 Score for comex_defense: 5.5 02:01:29 !bfjoust rushpolarity >+++++>-----(>)*7(++++[-]>----[+]>)*11 02:01:31 we should get rid of - 02:01:31 Score for Patashu_rushpolarity: 43.2 02:01:32 and just have + 02:01:35 to avoid trivial swaps 02:01:37 lol you guys quit fucking up the report 02:01:41 i wanna see if i got lucky or not 02:01:49 damn, nope 02:01:49 !bfjoust (+)*127(>[-])*30 02:01:49 Use: !bfjoust 02:01:53 !bfjoust creep (>+-[-]++>-+[+]--)*15 02:01:55 Score for nescience_creep: 38.0 02:01:55 wow mine is doing really well o.O 02:01:56 !bfjoust naive (+)*127(>[-])*30 02:01:57 Score for comex_naive: 15.0 02:02:02 !bfjoust rushpolarity_with_reversed_polarity >----->+++++(>)*7(----[+]>++++[-]>)*11 02:02:03 !bfjoust naive (-)*127(>[-])*30 02:02:04 Score for comex_naive: 7.5 02:02:08 !bfjoust naive (+)*127(>[+])*30 02:02:09 Score for comex_naive: 7.5 02:02:15 !bfjoust naive (+)*127(>[--])*30 02:02:17 Score for comex_naive: 7.5 02:02:18 comstop it. 02:02:19 ...... 02:02:20 !bfjoust naive (+)*127(>[-])*30 02:02:20 comex: stop it. 02:02:21 Score for comex_naive: 7.5 02:02:22 STOP IT 02:02:28 why did the exact same program get a different score 02:02:30 Score for ehird_rushpolarity_with_reversed_polarity: 44.6 02:02:36 comex: randomized tape lengths 02:02:43 because the lengths are random, so sometimes it might win or lose depending on that 02:02:50 ha 02:02:51 !bfjoust naive (>)*29(-)*128 02:02:52 yes, but if you do, your program sux 02:02:52 i got one that tied them both 02:02:53 Score for comex_naive: 7.0 02:02:56 hey I have an idea 02:02:58 but i don't want that 02:03:01 !bfjoust creep (>+-[-]++>-+[+]--)*15 02:03:01 to avoid trivial swaps but keep the functionality of + and - differing 02:03:02 Score for nescience_creep: 29.0 02:03:03 BEAT THEM BOTH 02:03:03 !bfjoust naive (>)*20[-](>)*9(-)*128 02:03:04 Score for comex_naive: 7.0 02:03:12 how about at the start of each match it randomly swaps or doesn't swap all your +s and -ses? 02:03:30 Patashu: meh 02:03:34 no luck 02:03:36 !bfjoust naive (>)*20--[+](>)*9(-)*128 02:03:36 oh well 02:03:37 Score for comex_naive: 8.0 02:03:46 well it would make a program and the program with opposite +-s functionally equivalent 02:03:51 all of a sudden i have a bunch of losses 02:03:51 !bfjoust naive (>)*20(--[+]>)*200 02:03:53 Score for comex_naive: 8.0 02:03:55 !bfjoust rushpolarity >+++++>-----(>)*7(+++++[-]>-----[+]>)*11 02:03:56 Score for Patashu_rushpolarity: 60.0 02:04:08 !bfjoust am_i_lame_enough_to_rip_off_shortsword_yes (>-->++)*2(>)*6([+[-]]>)*20 02:04:13 oh aah I see why this works 02:04:17 Score for ehird_am_i_lame_enough_to_rip_off_shortsword_yes: 7.0 02:04:19 !bfjoust rushpolarity >+++++>-----(>)*7(++++++[-]>------[+]>)*11 02:04:20 !bfjoust naive (+)*127(>------)*20(--[+]>)*200 02:04:20 haha 02:04:21 Score for Patashu_rushpolarity: 57.0 02:04:21 Score for comex_naive: 13.5 02:04:24 IDScorePtsProgram 02:04:25 482.007ehird_defend6_a_parody_or_just_plain_ripoff_question_mark.bfjoust 02:04:29 still chillin' at the top w/ my ripoff 02:04:29 !bfjoust naive (+)*127(>------)*20(--[+]>++[-])*200 02:04:30 Score for comex_naive: 13.5 02:04:35 comex: *200? 02:04:39 laffo 02:04:39 you're not even thinking about this logically 02:04:40 also 02:04:46 comex: the flag has to stay 0 for two generations 02:04:47 "[-]-" 02:04:48 Fail. 02:04:52 beating all but 1, 3, 4... 02:04:53 It will never obliterate a flag. 02:04:54 not fail 02:04:54 also a change from agora 02:04:55 use . to nop 02:04:58 I should probably actually read the rules 02:04:59 ] will nop 02:05:03 according to what you said earlier 02:05:05 nescience: oh, true 02:05:08 still 02:05:11 so the - will take it to 0 02:05:13 (1 round) 02:05:17 then the ] will nop 02:05:19 (2 rounds) 02:05:23 !bfjoust naive (+)*127(>------)*20(--[+]>++[-])*10. 02:05:24 Score for comex_naive: 40.1 02:05:29 that's why i removed by .'s 02:05:31 nescience: true 02:05:32 wat 02:05:42 !bfjoust rushpolarity >+++++>-----(>)*7(+++++++[-]>-------[+]>)*11 02:05:44 you shoudl make it take longer 02:05:44 Score for Patashu_rushpolarity: 54.3 02:05:47 do the exhaustive battle 02:05:54 and don't let people submit more than one while it's running 02:05:58 make em sit back and think for a while 02:06:00 instead of spamming :P 02:06:01 !bfjoust rushpolarity >+++++>-----(>)*7(++++++++[-]>--------[+]>)*11 02:06:03 Score for Patashu_rushpolarity: 59.3 02:06:29 !bfjoust (>)*14[>[-].+] 02:06:30 Use: !bfjoust 02:06:33 !bfjoust an_thing (>)*14[>[-].+] 02:06:40 Score for ehird_an_thing: 26.0 02:06:53 this is fun 02:07:24 !bfjoust rushpolarity >++++++>------(>)*7(++++++++[-]>--------[+]>)*11 02:07:26 Score for Patashu_rushpolarity: 59.0 02:07:26 ehird: It didn't occur to me that it's always whole, it is because there are 10 programs ... 02:07:40 GregorR-L: not true 02:07:44 most often we have 11 02:07:45 0-11 02:07:48 and sometimes we've had 12 02:07:50 !bfjoust rushpolarity >++++++>------(+>)*7(++++++++[-]>--------[+]>)*11 02:07:52 Score for Patashu_rushpolarity: 51.9 02:07:55 it doesn't seem to trim well if people submit a lot of proposals at once 02:08:00 *warriors 02:08:02 It shouldn't always be whole then, as in that example :P 02:08:03 atm: 02:08:03 638.89-3impomatic_dumb.bfjoust 02:08:04 538.89-3impomatic_chrome.bfjoust 02:08:06 217.28-6ehird_an_thing.bfjoust 02:08:09 GregorR-L: Yes, but make it always stay at 10... 02:08:12 !bfjoust rushpolarity >+++++>-----(>)*7(++++++++[-]>--------[+]>)*11 02:08:13 Score for Patashu_rushpolarity: 42.0 02:08:32 wait 02:08:33 ehird: It removes them after evaluating. 12 would be from two added at once, 11 is the "norm" 02:08:34 are you starting on cell 0 02:08:34 or 1? 02:08:50 like, track is 10-30 cells and you're on cell 1 right? 02:08:52 (Now that I'm thinking about the actual results of this deletion code :P ) 02:08:58 Patashu: 10-30 cells total 02:09:01 k 02:09:05 but we number them from 0-29 02:09:23 !bfjoust rushpolarity (>)*7>++++++>------(++++++++[-]>--------[+]>)*11 02:09:25 Score for Patashu_rushpolarity: 58.0 02:09:47 !bfjoust rushpolishmenarity (>)*8>++++++>------(++++++++[-]>--------[+]>)*11 02:09:55 Score for ehird_rushpolishmenarity: 28.0 02:09:58 !bfjoust kekeke (>+>-)*4(>[[[-]]])*20 02:10:00 !bfjoust rushpolishmenarity (>)*8>++++++>------(++++++++[-]>--------[+]>)*10 02:10:02 !bfjoust rushpolarity >++++++>------(>)*7(++++++++[-]>--------[+]>)*11 02:10:02 Score for ehird_rushpolishmenarity: 28.0 02:10:03 Score for Patashu_rushpolarity: 23.0 02:10:06 huh 02:10:07 \o/ 02:10:08 losing suddenly haha 02:10:11 i beat you :P 02:10:22 Score for nescience_kekeke: 21.0 02:10:47 only two scores >50 now 02:10:50 !bfjoust kekeke (>+>-)*4(>[[[+]]])*20 02:10:50 defend6 and my ripoff 02:10:51 Score for nescience_kekeke: 21.0 02:10:55 with defend6 winning slightly 02:10:56 the original 02:11:04 lols 02:11:10 !bfjoust rushpolarity >+++++++++>---------(>)*7(+++++++[-]>-------[+]>)*11 02:11:12 Score for Patashu_rushpolarity: 39.5 02:11:12 didn't help it any :) 02:11:19 !bfjoust rushpolarity >+++++++++>---------(>)*7(++++++++[-]>--------[+]>)*11 02:11:21 Score for Patashu_rushpolarity: 60.5 02:11:24 oo 02:11:32 i wonder though why it loses to defend6 02:11:38 the extra brackets shoul... ah 02:12:19 !bfjoust now_is_the_time_for_all_bad_men_to_come_backwards_in_time_for_their_country [(-)*128+] 02:12:20 actually I should clean the code up 02:12:26 Score for ehird_now_is_the_time_for_all_bad_men_to_come_backwards_in_time_for_their_country: 29.0 02:12:52 !bfjoust rushpolarity >(+)*9>(-)*9(>)*7((+)*8[-]>(-)*8[+]>)*11 02:12:53 hey 02:12:54 Score for Patashu_rushpolarity: 50.0 02:12:56 only inverse creep beats my program 02:13:09 that's because it exploits "your" program 02:13:13 defend6 draws though 02:13:14 !bfjoust kekeke (>+>-)*4(>[-](.)*132)*20 02:13:15 nescience: ;P 02:13:16 :P 02:13:16 Score for nescience_kekeke: 19.8 02:13:17 I'm going now 02:13:18 bye 02:13:30 bam, beats defend now 02:13:42 AND parody 02:13:45 do i get candy? 02:13:49 !bfjoust rushpolarity >(+)*9>(-)*9(>)*7((+)*7[-]>(-)*7[+]>)*11 02:13:51 Score for Patashu_rushpolarity: 45.7 02:14:06 !bfjoust rushpolarity >(+)*9>(-)*9(>)*7((+)*9[-]>(-)*9[+]>)*11 02:14:08 Score for Patashu_rushpolarity: 25.9 02:14:26 oo 02:14:30 now it says all but 1,2 beaten 02:14:33 which are the defend6s 02:14:35 how do they work? :o 02:14:43 read them 02:15:46 hey if I have symbol*number without brackets 02:15:48 does that compile fine? 02:17:41 !bfjoust creep >+>->+>->+>->+(>-++-.[+]++>-++-.[+]--)*15 02:17:42 Score for nescience_creep: 12.0 02:18:18 !bfjoust creep >+>->+>->+>->+(>-++-(.)*132[+]++>-++-(.)*132[-]--)*15 02:18:19 Score for nescience_creep: 32.5 02:18:32 beats both defends still hehe 02:18:34 but too slow 02:19:37 -!- psygnisfive has joined. 02:20:32 wow haha 02:20:33 top of the hill 02:22:57 !bfjoust rushpolarity >(+)*9>(-)*9(>)*7((+)*9[-][>](-)*9[+][>])*11 02:22:59 Score for Patashu_rushpolarity: 75.3 02:23:19 lol broke it 02:23:23 huh 02:23:26 it said I got 75.3 score 02:23:31 but when I look at hte report I lost every one 02:23:31 that was your last score 02:23:33 refresh the table 02:23:39 nope 02:23:41 lol 02:23:53 that was your score before submitting 02:23:56 !bfjoust rushpolarity >(+)*9>(-)*9(>)*7((+)*9[-][>](-)*9[+][>])*11 02:23:58 Score for Patashu_rushpolarity: 0.0 02:23:59 okay 02:24:01 see 02:24:02 ah I see 02:24:05 why does it do that 02:24:07 it's broken 02:24:10 !bfjoust rushpolarity >(+)*9>(-)*9(>)*7((+)*9[-]>(-)*9[+]>)*11 02:24:12 Score for Patashu_rushpolarity: 0.0 02:24:29 haha 02:24:31 !bfjoust rushpolarity >(+)*9>(-)*9(>)*7((+)*9[-]>(-)*9[+]>)*11 02:24:33 Score for Patashu_rushpolarity: 48.1 02:24:36 there we go 02:24:50 wonder why the [>] made it worse... 02:25:00 oh wait 02:25:06 I'd need a reverse [] to do what I was thinking of 02:25:10 lol :) 02:25:12 yep 02:25:31 loop while cell under pointer IS zero 02:27:13 !bfjoust rushpolarity >(+)*10>(-)*10(>)*7((+)*10[-]>(-)*10[+]>)*11 02:27:15 Score for Patashu_rushpolarity: 48.1 02:27:37 !bfjoust rushpolarity >(+)*9>(-)*9(>)*7((+)*9[-]>(-)*9[+]>)*11 02:27:39 Score for Patashu_rushpolarity: 54.3 02:27:53 hmm 02:28:11 !bfjoust rushpolarity >(+)*8>(-)*8(>)*7((+)*8[-]>(-)*8[+]>)*11 02:28:13 Score for Patashu_rushpolarity: 54.3 02:28:37 !bfjoust rushpolarity >(+)*8>(-)*8(>)*7((+)*9[-]>(-)*9[+]>)*11 02:28:39 Score for Patashu_rushpolarity: 42.0 02:29:04 !bfjoust rushpolarity >(+)*8>(-)*8(>)*7((+)*10[-]>(-)*10[+]>)*11 02:29:06 Score for Patashu_rushpolarity: 50.6 02:29:37 !bfjoust rushpolarity >(+)*10>(-)*10(>)*7((+)*10[-]>(-)*10[+]>)*11 02:29:39 Score for Patashu_rushpolarity: 34.6 02:30:15 !bfjoust rushpolarity >(+)*10>(-)*10(>)*7((+)*6[-]>(-)*6[+]>)*11 02:30:17 Score for Patashu_rushpolarity: 60.5 02:30:33 !bfjoust rushpolarity >(+)*10>(-)*10(>)*7((+)*7[-]>(-)*7[+]>)*11 02:30:35 Score for Patashu_rushpolarity: 48.1 02:30:55 !bfjoust rushpolarity >(+)*10>(-)*10(>)*7((+)*8[-]>(-)*8[+]>)*11 02:30:57 Score for Patashu_rushpolarity: 60.5 02:31:56 !bfjoust rushpolarity >(+)*10>(-)*10(>)*7((+)*9[-]>(-)*9[+]>)*11 02:31:57 Score for Patashu_rushpolarity: 60.5 02:32:19 :o 02:32:48 where'd everyone go anyhow 02:34:11 !bfjoust rushpolarity >(+)*10>(-)*10>(-)*5(>)*6((+)*9[-]>(-)*9[+]>)*11 02:34:13 Score for Patashu_rushpolarity: 75.3 02:34:34 no that's worse, hmm 02:34:59 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-]>(-)*9[+]>)*11 02:35:01 Score for Patashu_rushpolarity: 49.4 02:35:25 !bfjoust slowpoke http://pastie.org/490848.txt 02:35:28 Score for nescience_slowpoke: 8.5 02:35:33 lol 02:35:35 must be broken 02:35:37 ouch 02:35:44 is that algorithmically generated? 02:35:51 no 02:36:13 oh 02:36:14 haha 02:36:17 i got my numbers wrong 02:36:20 lol 02:36:24 it shouldn't be <*8 >*8 02:36:27 bfjoust magnum opus 02:36:28 it should be <*8 >*9 02:36:49 this is for my bfjoust master's degree 02:38:00 oh, things are fixed? 02:38:07 neat! 02:38:27 yes 02:38:28 go wild 02:38:35 !bfjoust irritating [>[-]-] 02:38:42 Score for coppro_irritating: 21.5 02:38:49 just one glitch 02:38:54 the score it gives is for the LAST time you submitted a program 02:39:00 i.e. one submission delay 02:39:26 !bfjoust slowpoke http://pastie.org/490854.txt 02:39:28 Score for nescience_slowpoke: 8.5 02:39:41 !bfjoust irritating [>[-]->[+]+] 02:39:42 Score for coppro_irritating: 21.5 02:39:44 lol well it still beats defendd 02:39:56 haha 02:40:01 what is it meant to do? 02:40:09 ah, I see 02:40:11 beat defend, and hopefully also do something interesting 02:40:23 i have an idea 02:40:26 lol 02:40:32 but i guess it wouldn't matter 02:40:40 !bfjoust beat_defenders [>[[[[[[-]]]]]]] 02:40:51 Score for coppro_beat_defenders: 8.5 02:40:52 it doesn't 02:41:03 apparently not 02:41:19 you'll just tie or run off the end 02:41:35 gotta trip the tripwire and wait 02:41:39 or skip it somehow 02:41:46 only ones I'm not beating are the defend6s 02:41:48 so :C 02:42:05 i could be lame like ehird and just rip your code and then possibly fix it to beat them 02:42:05 :P 02:42:22 the thing i don't get is why you bother doing 10 +'s etc 02:42:24 unless it's to kill time 02:42:27 it's because 02:42:28 uh 02:42:30 when people set decoys 02:42:38 they're more likely to set them to low +- numbers than high ones 02:42:42 or wait... 02:42:44 hmmm 02:42:45 so? 02:42:52 that doesn't help you any, since you aren't looping 02:42:53 well it works better when I put them in! 02:43:04 like i said, must be killing time 02:43:06 wonder why though 02:43:16 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5([-]>[+]>)*11 02:43:18 Score for Patashu_rushpolarity: 76.0 02:43:19 i have an inkling of an idea 02:43:23 let's see what happens when I take them out then 02:43:30 yeah it gets much worse 02:43:36 let's replace them with an equivalent amount of .s 02:43:39 i meant on your decoys 02:43:46 see one thing that happens 02:43:50 if you have a set number of -s or +s 02:43:59 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-]>(-)*9[+]>)*11 02:44:01 Score for Patashu_rushpolarity: 21.0 02:44:02 and someone else is looping on the same position 02:44:13 if they dec to 0, you can put them past 0 02:44:19 and then they loop again 256 times 02:44:49 my decoys only bother setting once because they are only going to slow down loops 02:44:57 and for that you don't need more than one 02:44:59 hmm let's see 02:45:07 !bfjoust rushpolarity >+>->->+(>)*5((+)*9[-]>(-)*9[+]>)*11 02:45:09 Score for Patashu_rushpolarity: 60.5 02:45:35 I was under the impression that a high finite + or - for a decoy would stop constructs that go like [+[--[+++[----[... 02:45:38 i think you ought to study why it's working for you in detail, probably could learn something good 02:45:44 nope small decoys doesn't help it 02:46:03 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-]>(-)*9[+]>)*11 02:46:05 Score for Patashu_rushpolarity: 34.6 02:46:13 strange 02:46:16 this is ideal 02:46:20 beats all but defend6s 02:46:42 all you need to do to beat the defends is delay longer before moving on 02:46:57 you need to spend longer than 128 cycles at each location 02:47:04 but then I lose to everything else 02:47:05 assuming you zero them before moving 02:47:13 sucks eh? :) 02:47:23 maybe you should just not care about them, apparently tripwire is not that effective a strategy 02:47:23 well hmm... 02:47:26 unless it's all that's on the hill 02:47:27 8 and 9 beat defend6 a parody 02:47:34 but 8 and 9 do poorly in general 02:47:36 because i wrote them to 02:47:39 ya 02:47:42 creep did good earlier 02:47:45 i was 2nd 02:47:51 how do you make a defender? I wanna try 02:47:54 but even then it got 3 losses 02:48:02 what the defends are doing is 02:48:09 1) waiting at instruction 1 until it gets zeroed 02:48:16 2) going back to their flag and inc/dec'ing 128 times 02:48:26 3) then going back and forth across the tape alternately 02:48:50 they keep tweaking their flag, then attacking a new location 02:49:04 the goal is, presumably, to do like i said before 02:49:11 let the enemy's loop hit 0 and fall through 02:49:21 but inc/dec the flag on that same instruction so you don't lose 02:49:23 then they run off the end 02:49:37 so it starts 02:49:42 i wonder, the tripwire thing probably doesn't need to attack the opponent's flag at all 02:49:50 >[]<(+)*128 02:50:05 >+[] 02:50:12 oh yeah 02:50:23 what if say your opponent -s while you + 02:50:30 do you alternate +ing and -ing 02:50:48 infinite loop -> tie 02:51:06 !bfjoust matador >+[]<(+)*128(-)*128(>)*9(>[+])*21 02:51:10 Score for Patashu_matador: 0.0 02:51:12 nope 02:51:43 !bfjoust tripwire >+[]<(+)*192000 02:51:50 whoops i needed bigger number 02:51:53 Score for nescience_tripwire: 0.0 02:52:06 it ties the defends though lol 02:52:13 !bfjoust tripwire >+[]<(+)*384000 02:52:15 Score for nescience_tripwire: 0.0 02:52:15 should tie everything 02:52:23 !bfjoust morefail >+[]+[->[-]](>)*7[>[-]] 02:52:26 Score for coppro_morefail: 0.0 02:52:28 i woner why not 02:52:29 oops 02:52:33 !bfjoust matador >+[]<(+)*123456(-)*123456(>)*9(>[+])*21 02:52:35 Score for Patashu_matador: 0.0 02:52:35 !bfjoust morefail >+[]+[->[-]](>)*7[>[-].] 02:52:36 Score for coppro_morefail: 0.0 02:52:45 hmmm 02:52:52 !bfjoust matador >+[]<.(+)*123456(-)*123456(>)*9(>[+])*21 02:52:54 Score for Patashu_matador: 0.0 02:52:55 !bfjoust tripwire >(+)*384000 02:52:56 Score for nescience_tripwire: 8.5 02:52:58 !bfjoust matador >+[]<.(+)*123456.(-)*123456(>)*9(>[+])*21 02:52:58 lol 02:53:00 Score for Patashu_matador: 0.0 02:53:07 !bfjoust morefail >+[]+[->[-]](>)*7+[[-].>] 02:53:09 Score for coppro_morefail: 0.0 02:53:10 !bfjoust matador >+[]<(+)*123456.(-)*123456(>)*9(>[+])*21 02:53:12 Score for Patashu_matador: 0.0 02:53:15 please cut the spam out, jesus 02:53:19 at least stop and reflect on your scores 02:53:21 never! 02:53:21 -!- Gracenotes has joined. 02:53:25 remember the bot isn't reporting an accurate score 02:53:31 so you have to refresh the chart at least 02:53:32 there was an obvious bug in mine 02:53:35 and every time you submit it blanks the chart 02:53:45 damn bugs 02:53:57 it's probably my fault heh :P 02:54:00 running out the clock 02:54:15 hmm that last time it beat 10 and did a whole bunch of ties 02:54:20 rofl my new one can't win 02:54:23 it still loses 02:54:23 so why is it tieing then? it kills itself or it times out? 02:54:23 wtf 02:54:31 tie should be time runs out 02:54:38 oh, of course 02:54:41 the defends don't use loops 02:54:46 and anything that doesn't loop would win 02:54:48 silly me 02:54:56 !bfjoust matador >+[]<(+)*500.(-)*500(>)*9(>[+])*21 02:54:58 also i forgot something about my own attempt 02:54:58 Score for Patashu_matador: 9.0 02:55:12 that beats creep and kekeke 02:55:58 !bfjoust matador >+[]<(++-)*256.(--+)*256(>)*9(>[+])*21 02:56:00 Score for Patashu_matador: 14.0 02:56:07 oh-ho 02:56:11 beats rushpolarity XD 02:56:32 !bfjoust matador >+[]<(++-)*129.(--+)*129(>)*9(>[+])*21 02:56:34 Score for Patashu_matador: 41.0 02:56:41 no that does worse 02:56:48 !bfjoust matador >+[]<(++-)*500.(--+)*500(>)*9(>[+])*21 02:56:50 Score for Patashu_matador: 17.3 02:56:58 holy lol 02:57:00 that does real well 02:57:07 !bfjoust matador >+[]<(++-)*1000.(--+)*1000(>)*9(>[+])*21 02:57:09 Score for Patashu_matador: 55.6 02:57:17 !bfjoust tripwire http://pastie.org/490870.txt 02:57:19 Score for nescience_tripwire: 0.0 02:57:26 !bfjoust matador >+[]<(.+-)*1000.(.-+)*1000(>)*9(>[+])*21 02:57:28 Score for Patashu_matador: 48.1 02:57:28 yeah, didn't think it'd do much better 02:57:39 !bfjoust matador >+[]<(.+-)*1000(.-+)*1000(>)*9(>[+])*21 02:57:41 Score for Patashu_matador: 48.1 02:57:54 !bfjoust watch_killer [>[[-]+>[][+]]] 02:58:08 Score for coppro_watch_killer: 0.0 02:58:10 !bfjoust matador >+[]<(.+-)*512(.-+)*512(>)*9(>[+])*21 02:58:11 aw 02:58:12 Score for Patashu_matador: 39.5 02:58:29 !bfjoust matador >+[]<(.+-)*1000(.-+)*1000(>)*9(>[+])*21 02:58:31 Score for Patashu_matador: 39.5 02:58:39 !bfjoust matador >+[]<(.+-)*1000.(.-+)*1000(>)*9(>[+])*21 02:58:41 Score for Patashu_matador: 39.5 02:58:49 okay that . makes it win for some reason 02:58:55 must be a parity thing 02:59:07 likely 02:59:20 hmm I wonder 02:59:20 !bfjoust watch_killer [>[[-]+>[][+--]]] 02:59:22 Score for coppro_watch_killer: 0.0 02:59:24 still probably gonna be affected by randomness 02:59:27 !bfjoust matador >-[]<(.+-)*1000.(.-+)*1000(>)*9(>[+])*21 02:59:28 Score for Patashu_matador: 48.1 02:59:29 hrmm why's mine failing 02:59:44 what is that horrid thing doing 02:59:45 lol 03:00:31 !bfjoust watch_killer [>[[-](.)*1000]] 03:00:32 Score for coppro_watch_killer: 0.0 03:00:41 oh wait I see 03:00:42 duh 03:00:46 !bfjoust watch_killer [>[[-]+>[][+--]]+] 03:00:46 mixed up what [] does? 03:00:47 Score for coppro_watch_killer: 0.0 03:00:53 hrm 03:00:54 oo 03:00:55 !bfjoust watch_killer [>[[-](.)*1000]+] 03:00:56 it gets 31.5 now 03:00:57 Score for coppro_watch_killer: 31.5 03:01:00 there we go 03:01:01 don't forget to check the chart after you submit something 03:01:02 because 03:01:04 the score it gives you 03:01:05 oh yeah 03:01:07 right 03:01:11 was the score for the LAST program you submitted 03:01:12 !bfjoust tripwire >+[]<[]+(>[+](.)*128)*29 03:01:13 !bfjoust watch_killer [>[[-]+>[][+--]]+] 03:01:14 Score for nescience_tripwire: 45.0 03:01:14 Score for coppro_watch_killer: 37.0 03:01:29 basically it tricks defenders into killing themselves 03:01:30 can't believe i/nobody thought of that yet 03:01:40 coppro: yes, that's what i did like two hours ago :P 03:01:46 heh 03:01:51 the above should beat loopers and both defenders 03:01:58 but doesn't 03:01:59 bugs? 03:02:12 hmm... I've got a new one to try 03:02:14 i wish their numbers would stay the same 03:02:17 !bfjoust matador >-[]<[]+-++--+++---++++----(>)*9(>[+])*21 03:02:19 Score for Patashu_matador: 0.0 03:02:25 no good, ok 03:02:29 !bfjoust matador >-[]<(.+-)*1000.(.-+)*1000(>)*9(>[+])*21 03:02:30 Score for Patashu_matador: 58.0 03:03:02 hmm... it would be more interesting 03:03:10 if you had one 0 cell beyond each flag 03:03:40 suggest it 03:04:01 where? 03:04:11 http://esolangs.org/wiki/BF_Joust talk page 03:04:39 !bfjoust tripwire >+[]<[]+.+(>)*9(>[.+])*29 03:04:41 Score for nescience_tripwire: 0.0 03:04:49 would allow for more interesting possibilities, because then programs could identify flags other than by hoping they don't run off the end 03:04:58 hm, i think maybe the timing is not correct 03:05:06 coppro: it's called "joust" not "snipe" ;) 03:05:10 FULL SPEED AHEAD 03:05:30 nescience: you can, of course, trick your opponent still! 03:05:51 more importantly, it allows defensive strategies that don't need to use fixed numbers in the source 03:05:57 !bfjoust test [>[-]+] 03:06:01 Score for nescience_test: 35.0 03:06:15 yeah 03:06:17 timing thing 03:06:18 beats matador haha 03:06:19 wtf, that might be a bug 03:06:28 let's look at this closely 03:06:31 >+ sets it to 1 03:06:35 [] waits til its 0 03:06:41 from what was described to me before 03:06:46 no matter if it is executing [ or ] 03:06:51 [] loops if the cell is not 0\ 03:06:53 it will skip to < 03:06:55 when the cell is 0 03:07:03 then it waits on my fag 03:07:05 flag* 03:07:07 [] again 03:07:08 it will only halt if it encounters a -1 cell 03:07:19 when the flag becomes 0 it should execute + 03:07:23 and save m... ah 03:07:27 but the enemy...... no 03:07:34 'cause the enemy should be at that point executing ] 03:07:45 -]-]-]-]-] 03:07:50 or +]+]+]+]+] 03:07:53 so - makes it 0 03:07:57 it is 0 at the start of ] 03:08:00 loop falls through 03:08:03 I think I see why it beats matador 03:08:06 for my part, 03:08:15 it falls through a... i see 03:08:19 that seems non optimal 03:08:28 !bfjoust matador >-[]<(++-)*1000.(--+)*1000(>)*9(>[+])*21 03:08:30 Score for Patashu_matador: 53.1 03:08:40 so what happens is my test doesn't work until it has been 0 for 1 round 03:08:43 k, fixed the vulnerability 03:09:02 now it only beats creep and watch kille 03:09:03 +r 03:09:49 I have an idea to tweak rushpolarity now 03:10:06 !bfjoust tripwire >+[]<(.)*128+(>)*9(>[.+])*29 03:10:07 Score for nescience_tripwire: 8.0 03:10:18 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-].+>(-)*9[+].->)*11 03:10:20 Score for Patashu_rushpolarity: 71.0 03:10:24 lol test still beats it 03:10:25 hmm 03:10:32 k, does not help 03:10:39 oh right 03:10:40 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-]+>(-)*9[+]->)*11 03:10:41 timing mistake 03:10:42 Score for Patashu_rushpolarity: 49.4 03:10:49 hmm 03:10:53 !bfjoust tripwire >+[]<(.)*256+(>)*9(>[.+])*29 03:10:55 Score for nescience_tripwire: 8.0 03:10:58 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-][+]>(-)*9[+][-]>)*11 03:11:00 Score for Patashu_rushpolarity: 49.0 03:11:17 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5([-][+]>[+][-]>)*11 03:11:19 Score for Patashu_rushpolarity: 45.1 03:11:28 no that's not doing it at all haha 03:11:31 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-]>(-)*9[+]>)*11 03:11:33 Score for Patashu_rushpolarity: 29.6 03:11:34 back to the original spec 03:11:49 !bfjoust tripwire >+[]<(.)*256+(>)*9(>[.+])*29 03:11:50 Score for nescience_tripwire: 8.0 03:12:08 !bfjoust tripwire >+[]<(.)*257+(>)*9(>[.+])*29 03:12:09 Score for nescience_tripwire: 8.0 03:12:19 !bfjoust tripwire >+[]<(.)*255+(>)*9(>[.+])*29 03:12:21 Score for nescience_tripwire: 8.0 03:12:23 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-]+-++-->(-)*9[+]+-++-->)*11 03:12:25 Score for Patashu_rushpolarity: 60.0 03:12:38 hmm 03:12:39 gonna have to notepad this up and solve a simpler problem 03:12:47 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-]+-.++-->(-)*9[+]+-.++-->)*11 03:12:48 ah there we go 03:12:49 Score for Patashu_rushpolarity: 58.6 03:12:51 it beats test 03:12:52 lol 03:12:53 but nothing else 03:13:03 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-].+-++-->(-)*9[+].+-++-->)*11 03:13:05 Score for Patashu_rushpolarity: 45.1 03:13:06 i guess nobody uses a simple attack anymore 03:13:12 !bfjoust fooled_ya (>+>-)*4[>[-]+] 03:13:13 oo that's good 03:13:13 :) 03:13:16 Score for coppro_fooled_ya: 27.5 03:13:27 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-].+->(-)*9[+].+->)*11 03:13:28 Score for Patashu_rushpolarity: 68.5 03:13:33 that's not 03:13:44 oh right, need defense protection 03:13:44 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-].+-++--+++--->(-)*9[+].+-++--+++--->)*11 03:13:46 Score for Patashu_rushpolarity: 56.2 03:13:58 !bfjoust fooled_ya (>+>-)*4[>[[-]+]+] 03:13:59 Score for coppro_fooled_ya: 25.3 03:14:03 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-].+-.++-->(-)*9[+].+-.++-->)*11 03:14:04 Score for Patashu_rushpolarity: 49.4 03:14:05 !bfjoust fooled_ya (>+>-)*4[>[+[-]]+] 03:14:07 Score for coppro_fooled_ya: 25.3 03:14:29 hmm... that's better 03:14:32 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-].+-++-->(-)*9[+].+-++-->)*11 03:14:34 Score for Patashu_rushpolarity: 58.6 03:14:35 -!- inurinternet has joined. 03:14:37 how does the scoring system work? 03:14:44 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-].+-.++-->(-)*9[+].+-.++-->)*11 03:14:46 Score for Patashu_rushpolarity: 48.8 03:14:50 you get more points for beating programs with more points 03:15:04 !bfjoust fooled_ya (>(+)*32>(-)*32)*4[>[+[-]]+] 03:15:06 Score for coppro_fooled_ya: 29.6 03:15:11 haha matador's falling now that it doesn't beat rushpolarity any more 03:15:14 :') 03:15:21 i think the scoring system needs a little work 03:15:32 ._. 03:15:35 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5([-].+-.++-->[+].+-.++-->)*11 03:15:37 Score for Patashu_rushpolarity: 63.0 03:15:44 if you get points for beating something before its battle with you gets taken account of, then things are a little weird the first submit 03:15:45 damn that shortswod 03:15:47 i dunno how they work it out 03:15:48 *shortsword 03:16:03 I don't understand exactly how this game works well enough yet 03:16:07 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5(+[-].+-.++-->-[+].+-.++-->)*11 03:16:09 Score for Patashu_rushpolarity: 45.1 03:16:16 hm? How are parens and numerical symbols determined? 03:16:21 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-].+-.++-->(-)*9[+].+-.++-->)*11 03:16:22 Score for Patashu_rushpolarity: 45.1 03:16:23 ...is that just repeating text? 03:16:27 yes 03:16:31 (foo)*num = foofoofoofoo...num times 03:16:33 ah. mak sanse. 03:16:33 that's it 03:16:47 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-].+-.++-->(-)*9[+].+-.++-->)*11 03:16:48 !bfjoust tripwire >+[]<[-++-](>)*9(>-++-(.)*128[+])*29 03:16:49 (some{stuff}things)*2=somesomestuffthingsthings 03:16:49 Score for nescience_tripwire: 8.0 03:16:49 Score for Patashu_rushpolarity: 30.9 03:17:07 how are scores done? :) 03:17:12 hm 03:17:20 !bfjoust tripwire >+[]<[--](>)*9(>-++-(.)*128[+])*29 03:17:21 Score for nescience_tripwire: 0.0 03:17:22 you get points for a win 03:17:24 ha 03:17:25 and the more that program had 03:17:26 the more you get 03:17:38 interesting results 03:17:41 I have heard of BF jousting, but not the specifics 03:17:41 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*8[-].+-.++-->(-)*8[+].+-.++-->)*11 03:17:43 Score for Patashu_rushpolarity: 45.0 03:17:50 oh of course tripwire would always tie the defends 03:18:01 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-].+-.++-->(-)*9[+].+-.++-->)*11 03:18:03 Score for Patashu_rushpolarity: 36.4 03:18:03 !bfjoust tripstep [>[>[-]]+] 03:18:09 Score for coppro_tripstep: 37.0 03:18:37 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-].+.--++>(-)*9[+].+.--++>)*11 03:18:39 Score for Patashu_rushpolarity: 54.5 03:19:02 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-]+.--.++>(-)*9[+]+.--.++>)*11 03:19:04 Score for Patashu_rushpolarity: 47.5 03:19:14 aha 03:19:48 !bfjoust tripstep [>[[>[-].+]]+] 03:19:50 Score for coppro_tripstep: 43.0 03:20:19 !bfjoust test (>)*9(-++-(-)*70[+])*20 03:20:21 Score for nescience_test: 22.5 03:20:44 !bfjoust matador >-[]<(++-)*1000.(--+)*1000(>)*9(>[+][-])*21 03:20:46 Score for Patashu_matador: 29.0 03:21:06 !bfjoust matador >-[]<(++-)*1000(--+)*1000(>)*9(>[+][-])*21 03:21:07 Score for Patashu_matador: 29.0 03:21:30 !bfjoust matador >-[]<(++-)*1000.(--+)*1000(>)*9(>[+][-])*21 03:21:32 Score for Patashu_matador: 42.0 03:21:41 oop now it beats rushpolarity again 03:21:48 :> 03:21:58 !bfjoust test >----->+++++>----->+++++>----->+++++>-----(>-++-(-)*70[+])*20 03:21:59 Score for nescience_test: 7.5 03:22:02 whoops forgot to put > in it 03:22:24 !bfjoust matador >-[]<(++-+-)*1000.(--+-+)*1000(>)*9(>[+][-])*21 03:22:25 Score for Patashu_matador: 7.4 03:22:30 beats defends and some others now 03:22:32 heh 03:22:36 !bfjoust matador >-[]<(++-)*1000.(--+)*1000(>)*9(>[+][-])*21 03:22:38 Score for Patashu_matador: 44.4 03:22:40 better'n creep at least 03:23:01 !bfjoust matador >+[]<(++-)*1000.(--+)*1000(>)*9(>[+][-])*21 03:23:02 Score for Patashu_matador: 30.9 03:23:11 !bfjoust matador >-[]<(++-)*1000.(--+)*1000(>)*9(>[+][-])*21 03:23:12 Score for Patashu_matador: 30.9 03:23:28 hmm 03:23:39 it needs to do two fights against each program imho 03:23:43 !bfjoust tripstep [>[+[--[>[-].+]]<+[>[-].+].+]]+] 03:23:45 Score for coppro_tripstep: 0.0 03:23:49 :( 03:23:49 wow look at that thing 03:23:57 it needs indentation 03:24:04 haha 03:24:07 !bfjoust tripstep [>[+[--[>[-].+]]+[[-].>+].+]]+] 03:24:09 Score for coppro_tripstep: 0.0 03:24:23 hfm 03:26:10 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5([+[-]]+.--.++>[-[+]]+.--.++>)*11 03:26:12 Score for Patashu_rushpolarity: 55.6 03:26:22 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*9[-]+.--.++>(-)*9[+]+.--.++>)*11 03:26:24 Score for Patashu_rushpolarity: 30.9 03:26:36 !bfjoust playing_the_odds (>)*13(>(-)*128.)*15 03:26:38 Score for nescience_playing_the_odds: 27.0 03:26:45 hehe 03:26:57 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*8[+[-]]+.--.++>(-)*9[-[+]]+.--.++>)*11 03:26:59 Score for Patashu_rushpolarity: 55.5 03:27:01 !bfjoust playing_the_odds (>)*13(>(-)*128.)*15 03:27:03 Score for nescience_playing_the_odds: 27.0 03:27:12 hmm 03:27:33 !bfjoust playing_the_odds (>+>-)*6>(>(-)*128.)*15 03:27:33 !bfjoust rushpolarity >(+)*10>(-)*10>->+(>)*5((+)*8[+[-]]+.--.++>(-)*8[-[+]]+.--.++>)*11 03:27:34 Score for nescience_playing_the_odds: 40.7 03:27:35 Score for Patashu_rushpolarity: 51.9 03:28:29 !bfjoust playing_the_odds (>+++++>-----)*2(>)*9(>(-)*128.)*15 03:28:30 Score for nescience_playing_the_odds: 26.5 03:28:38 !bfjoust rushpolarity >(+)*10>(-)*10>-->++(>)*5((+)*8[+[-]]+.--.++>(-)*8[-[+]]+.--.++>)*11 03:28:40 Score for Patashu_rushpolarity: 54.3 03:28:45 ah well, second try was the best heh 03:29:15 ah well, second try was the best heh 03:29:18 !bfjoust playing_the_odds (>+++++>-----)*2(>)*9(>(-)*128.)*15 03:29:19 Score for nescience_playing_the_odds: 38.3 03:29:22 rolling for a better score! 03:29:29 win 03:29:31 :D 03:29:32 cache that pls 03:29:46 didn't quite beat both defends though 03:30:06 that'll be an amusing one to have on the hill 03:30:11 i wonder how long it'll stay 03:31:15 shortsword has staying power at least 03:31:25 here's a better one! 03:31:47 !bfjoust (>)*15([{-}])*100 03:31:47 Use: !bfjoust 03:31:52 !bfjoust dice!(>)*15([{-}])*100 03:31:53 Use: !bfjoust 03:31:55 !bfjoust dice! (>)*15([{-}])*100 03:32:04 Score for coppro_dice_: 0.0 03:32:09 ... 03:32:13 :P 03:32:41 the hill should have an "age" parameter ala corewars hills 03:46:11 wtf save page is failing miserably 03:46:55 keeps giving me an 'edit conflict' error 03:47:00 but there isn't one 03:47:54 on the esolang wiki? 03:48:19 add an edit comment 03:48:30 stupid wikis capitalizing my username 03:49:48 well, registering fied that 03:49:57 nescience: if that was you on the BF Joust talk page, please sign with ~~~~ 03:50:08 how come ais523 gets a lowercase nick and i don't :( 03:50:12 oerjan: i did 03:50:23 i didn't have an account before 03:51:20 are you Myndzi? 03:51:28 yes 03:51:42 were you the previous comment too? 03:51:44 it just looks wrong with a capital M >:( 03:51:46 no 03:51:52 ic 03:52:00 my edit was only the last paragraph 03:52:01 * oerjan will add an unsigned template 03:52:24 assuming i can get the damn molasses to load 03:52:59 coppro was that unsigned paragraph 03:53:25 oh, I must have not been logged in 03:53:27 oops 03:53:56 that was you? then maybe best you sign yourself 03:54:02 or wait 03:54:15 i can just change the nick 03:54:32 mine was the comment about going past the end 03:54:41 i don't really see how that would be useful though 03:54:47 since you can't simply test for 0 03:54:50 it doesn't gain you anything 03:55:26 coppro: actually i don't know your username 03:56:08 oh, dear, it appears I never signed up to esolang 03:56:11 that explains a lot 03:56:17 heh :D 03:56:32 >((o> lol fish submit that! 03:56:38 coppro: also, whether or not you are logged in you need to use a ~~~~ command to sign (it's in the button menu) 03:56:48 I did 03:58:23 nescience: dammit your decapitalization made an edit conflict :( 03:58:28 lol 03:58:44 now you know how i feel! :P 03:58:49 except there actually was one 03:58:51 o well 03:59:55 nescience: also the signature text can be set in preferences 04:00:24 you mean "nickname"? 04:00:29 yes 04:00:33 i thought that's what it was, but when i changed it, it didn't affect the page 04:00:37 that's why i edited it manually 04:00:37 * oerjan changed it to use an Ø 04:00:42 oh 04:00:52 nescience: it's not your username 04:01:07 ? 04:01:08 only used for signature i think 04:01:15 right 04:01:21 but my signature had still taken from before 04:01:26 thus i changed it too 04:01:38 oh of course 04:01:45 between wikis and muds lately, *sighs* 04:16:53 -!- Patashu has quit ("Patashu/SteampunkX - MSN = Patashu@hotmail.com , AIM = Patashu0 , YIM = Patashu2 , Googletalk = Patashu0@gmail.com ."). 04:16:56 -!- GregorR has quit (anthony.freenode.net irc.freenode.net). 04:17:38 -!- GregorR has joined. 04:18:00 !bfjoust test >+>->+>->+>->+>[[+]>-++] 04:18:01 Score for nescience_test: 38.5 04:19:38 !bfjoust test >+>->+>->+>->+>(-++-(-.)*130>)*20 04:19:40 Score for nescience_test: 0.0 04:20:43 !bfjoust test >(+)*50>(-)*50>(+)*30>(-)*30>(+)*10>(-)*10>+>(-++-(-.)*130>)*20 04:20:45 Score for nescience_test: 34.6 04:27:05 I'm glad to see bfjoust is being used. 04:27:06 -!- sebbu has joined. 04:27:32 Any further complaints from anybody? 04:27:51 You aren't creating a digital manifestation of God. 04:27:53 i think the game won't go all that far 04:28:14 it's about the best it can be in its current form, but there really aren't that many options 04:28:18 dunno though, could be wrong1 04:28:23 there are a number of comments on the wiki talk page 04:28:37 Then try FYB instead :P 04:29:06 heh heh 04:29:17 i'm not a huge fan of BF in the first place 04:29:37 When is bfjoust from ... does FYB predate it? 04:29:50 i wouldn't know 04:32:26 i wonder 04:32:40 in the true spirit of capture the flag, what would happen if you had to come back to your base and set your own flag back to 128? ;) 04:32:52 too complicated surely, also nobody would be able to keep track of their flag 04:36:35 !fyb minimangler +!> 04:36:53 Score for GregorR-L_minimangler: 0.0 04:37:01 Ouch :P 04:37:31 !fyb minimangler :+!>;* 04:37:33 Score for GregorR-L_minimangler: 0.0 04:37:37 *shrugs* 04:39:18 !fyb cheers :D 04:39:28 Score for oerjan_cheers: 4.5 04:39:34 lawl :P 04:41:32 :D 04:41:34 D: 04:43:03 -!- sebbu2 has quit (Read error: 110 (Connection timed out)). 04:44:51 -!- Corun has quit ("Leaving..."). 04:51:13 * coppro starts working on the Evil Calculus Book of Doom 04:54:58 itym the Evil Doom Calculus Book of the Apocalypse 05:04:19 sorry 05:04:39 so i have a dump of wikipedia 05:04:42 what should i do with it 05:04:43 stupid fact that I have to do tons of coursework even though I'm challenging the course 05:04:53 make a programming language using wikipedia links 05:07:13 bsmntbombdood: what kind of dump? 05:07:31 current pages in article namespace? 05:07:35 the current version of all namespace 1 pages 05:07:54 that's the talk namespace :) 05:08:05 nuh uh 05:09:11 article is the first namespace, but its numbering is 0 05:09:30 anyway. >_> 05:09:49 perhaps you could make a graph based on what articles link to each other? 05:10:05 xkcd :D 05:10:13 although there are already dumps that have that information 05:10:18 you could do lots of interesting stuff with that 05:10:33 you can download just the link graph actually 05:10:37 shortest path between two articles, for example 05:10:40 ›_› 05:10:49 （　≖‿≖） 05:11:24 hahaha 05:11:28 rofl xkcd 05:12:33 oh, really? 05:12:38 that good 05:12:41 *looks* 05:13:19 heh. heh. heh heh. 05:14:27 bsmntbombdood: there's always Markov chains 05:14:34 you'll have to strip out formatting 05:14:43 lots of training text... 05:14:47 that just gives you pagerank 05:14:58 oh, markov chains of text 05:16:37 run Flesch-Kincaid maybe 05:17:17 "The logistics of who can get drunk are nontrivial." 05:17:38 hmmm 05:17:41 i need a drunk 05:17:46 I disagree. 05:17:54 I think you've had quite enough already 05:18:37 bsmntbombdood: schorry, i cant 'elp chu there 05:19:30 hm. I should try implementing Flesch-Kincaid in Haskell 05:19:40 kinky flesh 05:19:54 down boy 05:21:59 the syllable part may be a bit complicated 05:22:56 -!- oerjan has quit ("leaving"). 05:23:01 er 05:23:03 :.; 05:23:43 actually it's less complicated than I thought 05:23:57 http://flesh.cvs.sourceforge.net/viewvc/flesh/Flesh%201.5/FleshLogic.java?revision=1.3&view=markup#l_523 05:24:05 more of a guess... but seems accurate enough 05:24:15 might be worth it to have a special table for corner cases 05:24:47 it doesn't say whether or not this is a standard algorithm. rather specific to English anyway 05:25:26 broken link 05:25:45 wha 05:25:56 yeah... that's messed up. it worked a second ago 05:26:28 okay, should work now 05:26:50 actually, it does screw up quite a bit: http://flesh.cvs.sourceforge.net/viewvc/flesh/Flesh%201.5/CommonMistakes.java?revision=1.3&view=markup 05:27:19 needs more English knowledge 06:01:32 you can get dictionary files that break it down 06:03:04 Flesch-Kincaid sounds like a bunch of bs anyway 06:03:17 who cares how long sentences or words are 06:03:45 Floridan insurance brokers, per Wikipedia 06:05:42 insurance and florida, that's all you need to know about their position 06:34:38 -!- GregorR-L has quit (Remote closed the connection). 06:44:05 -!- GregorR-L has joined. 07:01:37 -!- psygnisfive has quit ("Leaving..."). 07:27:17 -!- Sgeo has quit (Read error: 110 (Connection timed out)). 07:37:35 -!- coppro has quit (Read error: 60 (Operation timed out)). 07:41:32 -!- GregorR-L has quit ("Leaving"). 07:59:59 -!- clog has quit (ended). 08:00:00 -!- clog has joined. 08:55:25 -!- lereah_ has joined. 09:13:01 -!- GregorR-L has joined. 09:39:50 -!- Gracenotes has quit ("brb"). 09:42:44 -!- Gracenotes has joined. 09:58:01 -!- tombom has joined. 10:04:37 -!- Patashu has joined. 10:13:25 hmm, still on the top of the hill 10:18:27 The next challenge: DOMINATE the hill. 10:19:41 I have no idea how to beat the defense6s and still beat everything else 10:19:45 I haven't even seen their code 10:20:10 http://codu.org/eso/bfjoust/in_egobot/ 10:20:23 Uncommented though :P 10:20:40 oh 10:20:43 god damn that's long rofl 10:20:49 I need a decompiler 10:21:22 ais posted a link to a commented version ... 10:23:23 Ah, here 'tis 10:23:24 http://pastebin.ca/1435376 10:23:55 May 26 16:29:09 it attacks something in particular, which is the [-] or [+] loop 10:23:55 May 26 16:29:13 nearly all programs have one 10:23:55 May 26 16:29:38 and if they go into such a loop on its flag, then defend6 will keep them in it forever, whilst running off and sinking their flag 10:24:30 *zleep* 10:24:47 oh I see 10:24:54 and it keeps coming back incrementially to keep it off zero 10:25:16 so it has an ultra-huge source code since there's no construct that says 'repeat this but increase this number by x every time' 11:51:27 -!- oerjan has joined. 12:16:29 -!- oerjan has quit ("leaving"). 12:24:04 -!- M0ny has joined. 12:24:55 -!- Sgeo has joined. 12:52:55 -!- AnMaster has quit (Success). 12:57:06 -!- AnMaster has joined. 13:40:23 -!- Corun has joined. 14:13:19 -!- tombom has quit ("Peace and Protection 4.22.2"). 14:23:45 -!- MizardX has quit ("What are you sinking about?"). 14:39:36 -!- MizardX has joined. 14:47:15 -!- Corun has changed nick to Corun|away. 14:51:09 -!- Corun|away has changed nick to Corun. 15:34:59 -!- inurinternet has quit (Success). 15:49:47 -!- inurinternet has joined. 15:54:00 -!- impomatic has joined. 15:58:28 -!- lereah_ has quit ("Leaving"). 16:01:30 -!- Patashu has quit ("Patashu/SteampunkX - MSN = Patashu@hotmail.com , AIM = Patashu0 , YIM = Patashu2 , Googletalk = Patashu0@gmail.com ."). 16:43:04 -!- Gracenotes has quit (Success). 16:57:24 -!- BeholdMyGlory has joined. 17:00:50 -!- tombom has joined. 17:06:16 -!- fungebob has joined. 17:06:18 -!- FireFly has joined. 17:40:31 cool i still own bf joust 17:42:35 01:38 Patashu: just one glitch 17:42:35 01:38 Patashu: the score it gives is for the LAST time you submitted a program 17:42:37 01:39 Patashu: i.e. one submission delay 17:42:39 GregorR: GregorR-L: fix that 17:46:50 !bfjoust phantom (>)*10([+[--[+++[----[+++++[------[+++++++[--------[+++++++++[----------]]]]]]]]]].>)*20 17:46:56 Score for ehird_phantom: 41.5 17:47:06 not bad 17:47:08 not bad at all 17:47:28 !bfjoust phantom (>)*10(-[+[--[+++[----[+++++[------[+++++++[--------[+++++++++[----------]]]]]]]]]]>)*20 17:47:33 Score for ehird_phantom: 42.6 17:47:48 !bfjoust phantom (>)*10(-[+[--[+++[----[+++++[------[+++++++[--------[+++++++++[----------]]]]]]]]]].>)*20 17:47:49 Score for ehird_phantom: 42.6 17:48:35 Ehird: is that tailored for the decoys left by the programs on the current hill? 17:49:01 lol creep officially beats nothing except the defends now 17:49:04 impomatic: pretty much, it tries to find out whether it's big or small and efficiently zeroes it based on that 17:49:10 !bfjoust phantom (>)*10(+[-[++[---[++++[-----[++++++[-------[++++++++[---------[++++++++++]]]]]]]]]]>)*20 17:49:11 Score for ehird_phantom: 42.6 17:49:23 but amusingly, 'playing the odds' is in 4th 17:49:26 hmm flipping them does nothing 17:49:27 <3 cached scores 17:49:27 interesting 17:49:42 refresh page 17:49:47 remember it tends to show your last score 17:49:47 3 | - + + - - - - + - | 42.6| -3| ehird_phantom.bfjoust 17:49:50 that's what it was before 17:49:50 not the score after submission 17:50:03 i see 17:50:11 !bfjoust phantom (>)*10(-[+[-[+[-[+[-[+[-[+[-]]]]]]]]]]>)*20 17:50:12 Score for ehird_phantom: 42.6 17:50:16 not all that surprising i guess 17:50:17 I was wondering if it's possible to detect certain opponents and use a tailored attack. 17:50:17 ouch 17:50:20 that put it down to 14.2 17:50:33 speed, i guess 17:50:52 makes me wonder why so many +s before looping is effective though 17:50:55 i guess it sets decoys of its own 17:51:09 btw don't you want >*9? 17:51:15 if it's 10 units long, it's 9 to the enemy flag 17:51:19 oh 17:51:20 right 17:52:03 !bfjoust phantom (>-)*3(>+)*3(>)*3(+[-[++[---[++++[-----[++++++[-------[++++++++[---------[++++++++++]]]]]]]]]]>)*20 17:52:05 Score for ehird_phantom: 42.6 17:52:13 back to 42.6 17:52:28 5 losses, 3 wins, one draw 17:52:36 i think the strategy is essentially good, though 17:52:46 * ehird idea 17:54:40 !bfjoust phantom http://pastie.org/491591.txt?key=z90epgclqdxxiifgky728a 17:54:42 Score for ehird_phantom: 42.6 17:54:48 hahaha 17:54:51 that makes it worse 17:55:06 !bfjoust phantom (>-)*3(>+)*3(>)*3(+++++[-[++[---[++++[-----[++++++[-------[++++++++[---------[++++++++++]]]]]]]]]]>)*20 17:55:08 Score for ehird_phantom: 27.8 17:56:14 !bfjoust phantom (>-)*3(>+)*3(>)*3([-[(+)*2[(-)*3[(+)*4[(-)*5[(+)*6[(-)*7[(+)*8[(-)*9[(+)*10]]]]]]]]]]>)*20 17:56:16 Score for ehird_phantom: 36.4 17:56:20 this game is finicky 17:56:29 hmm doing it with repeats actually slowed it down 17:56:49 Patashu kept getting very different results by varying constants that didn't seem they should make that much difference 17:56:54 shrug 17:57:12 eh, the spec said repeats don't take any cycles 17:57:30 !bfjoust phantom (>)*9([(-)*32[(+)*32[(-)*32[(+)*32]]]])*20 17:57:32 Score for ehird_phantom: 36.4 17:57:33 just macros.. should be the same 17:57:36 unless it's randomness 17:57:39 * ehird crosses fingers 17:57:45 nescience: I probably mistyped 17:57:49 3 | - 0 0 - - - - - 0 | 8.0| -6| ehird_phantom.bfjoust 17:57:51 :D 17:58:58 !bfjoust phantom (>)*9([(-)*16[(+)*32[(-)*48]]][-])*20 17:59:00 Score for ehird_phantom: 8.0 17:59:14 !bfjoust phantom (>-)*3(>+)*3(>)*3(+++++[-[++[---[++++[-----[++++++[-------[++++++++[---------[++++++++++]]]]]]]]]]>)*20 17:59:16 Score for ehird_phantom: 8.0 17:59:17 stick to what works. 17:59:43 !bfjoust phantom >->->->+>+>+>>>(+++++[-[++[---[++++[-----[++++++[-------[++++++++[---------[++++++++++]]]]]]]]]]>)*20 17:59:45 Score for ehird_phantom: 33.5 17:59:51 wtf 17:59:59 gogo random! 18:00:03 nescience: the repeating one for the start got 33.5 18:00:06 but this one gets 20.4 18:00:07 !bfjoust phantom >->->->+>+>+>>>(+++++[-[++[---[++++[-----[++++++[-------[++++++++[---------[++++++++++]]]]]]]]]]>)*20 18:00:09 Score for ehird_phantom: 20.4 18:00:12 let's see if it's actually random 18:00:17 yep, it is 18:00:19 oh well 18:00:23 impomatic: viper? 18:00:27 look at 'playing the odds' 18:00:27 are you submitting stuff? do it in channel :-P 18:00:29 that's proof 18:00:44 i submitted it a couple times till i got a couple decent cached results 18:00:44 :P 18:00:53 nescience: that's called being a jerk 18:00:54 :) 18:00:55 this hill is too noisy for such fine tuning 18:01:09 if something was done to even out the randomness 18:01:18 say, multiple samples *ahem* like i was mentioning yesterday... :P 18:01:20 ehird: I didn't want to flood the channel ;-) 18:01:30 impomatic: but I don't notice new programs :-) 18:01:35 nescience: yeah I agree now. 18:01:36 i bet the max cycles probably doesn't need to be so high either 18:01:40 it does 18:01:44 defend6 needs it iirc 18:02:00 i don't know 18:02:07 i wonder how many cycles it'd take to complete at 30 cells 18:02:18 i'm sure someone can caluclate but i don't have the required consecutive time atm 18:02:22 or draw shortcutting at least 18:02:28 detecting two [] loops, for examle :P 18:02:41 or a [+] and [-] on the same cell 18:02:47 i have 21 programs here. they are provably optimum for their tape length. 18:02:50 should I submit them all? :P 18:03:02 Yes :-) 18:03:43 okie dokie 18:03:52 !bfjoust viper >--->+++>---(>->+)*3(-[-[-[-[-[-[-[-[-[-[+]]]]]]]]]]>)*20 18:03:54 Score for impomatic_viper: 42.0 18:04:23 !bfjoust bugger1 >>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:04:24 !bfjoust bugger2 >>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:04:27 !bfjoust bugger3 >>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:04:30 !bfjoust bugger4 >>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:04:33 !bfjoust bugger5 >>>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:04:34 !bfjoust ferret >(-)*9>(+)*9>---(>->+)*3((-)*10[+]>)*20 18:04:35 Score for impomatic_ferret: 66.0 18:04:36 !bfjoust bugger6 >>>>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:04:37 ...lol wut 18:04:39 !bfjoust bugger7 >>>>>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:04:42 !bfjoust bugger8 >>>>>>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:04:45 !bfjoust bugger9 >>>>>>>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:04:48 !bfjoust bugger10 >>>>>>>>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:04:48 what'd you say about being a dick? :P 18:04:51 !bfjoust bugger11 >>>>>>>>>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:04:54 !bfjoust bugger12 >>>>>>>>>>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:04:57 !bfjoust bugger13 >>>>>>>>>>>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:05:00 !bfjoust bugger14 >>>>>>>>>>>>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:05:03 !bfjoust bugger15 >>>>>>>>>>>>>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:05:06 !bfjoust bugger16 >>>>>>>>>>>>>>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:05:09 !bfjoust bugger17 >>>>>>>>>>>>>>>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:05:12 !bfjoust bugger18 >>>>>>>>>>>>>>>>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:05:15 !bfjoust bugger19 >>>>>>>>>>>>>>>>>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:05:18 !bfjoust bugger20 >>>>>>>>>>>>>>>>>>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:05:21 !bfjoust bugger21 >>>>>>>>>>>>>>>>>>>>>>>>>>>>>--------------------------------------------------------------------------------------------------------------------------------[-] 18:05:26 nescience: impomatic told me to :) 18:05:50 * ehird twiddles thumbs and waits for the report to update 18:06:06 poor bot 18:06:36 DOO DOO DOO DOO DEE DAA DOO DOO DOOOOOOO 18:06:39 * ehird pokes EgoBot 18:06:58 nescience: technically, I have to submit them all infinite times to be sure of their tape lengths being correct 18:07:02 since it's random 18:07:09 so they'll probably all do awfully 18:07:30 indeed 18:07:47 BUT DAT'S OKAY 18:07:55 * ehird kicks EgoBot 18:07:57 Score for ehird_bugger4: 22.4 18:07:57 Score for ehird_bugger19: 5.7 18:07:57 Score for ehird_bugger11: 18.3 18:07:57 Score for ehird_bugger3: 40.9 18:07:57 Score for ehird_bugger10: 4.8 18:07:58 Score for ehird_bugger21: 5.3 18:08:00 Score for ehird_bugger16: 7.4 18:08:02 Score for ehird_bugger20: 0.0 18:08:04 Score for ehird_bugger1: 29.4 18:08:04 40.9 for bugger3? 18:08:06 Score for ehird_bugger13: 8.1 18:08:08 Score for ehird_bugger17: 12.1 18:08:09 That's pretty good, like. 18:08:10 Score for ehird_bugger14: 6.9 18:08:11 Guess I got lucky. 18:08:12 Score for ehird_bugger8: 12.5 18:08:14 Score for ehird_bugger12: 8.3 18:08:16 Score for ehird_bugger15: 3.7 18:08:18 Score for ehird_bugger9: 12.6 18:08:20 Score for ehird_bugger18: 4.7 18:08:20 nescience: look at the report 18:08:22 Score for ehird_bugger7: 18.8 18:08:22 quick 18:08:24 Score for ehird_bugger6: 21.1 18:08:24 before it trims them 18:08:26 Score for ehird_bugger2: 42.2 18:08:27 Yeah, lucky score :-) 18:08:27 http://codu.org/eso/bfjoust/report.txt 18:08:28 Score for ehird_bugger5: 25.2 18:08:31 hahaha 18:08:32 it's huge 18:08:42 :D 18:08:43 they're all there 18:08:44 haha 18:08:45 nice 18:08:57 2343.678ehird_defend6_a_parody_or_just_plain_ripoff_question_mark.bfjoust 18:08:57 1342.1710ehird_bugger2.bfjoust 18:09:05 i'd say that's pretty good! 18:09:15 huh creep only has 4 losses now 18:09:19 nescience: haha 18:09:23 the power of buggery 18:09:39 but ferret is at the top :P 18:09:42 impomatic: is your ferret designed to be anti-bugger? 18:10:01 i don't think anything can be anti-bugger 18:10:13 nescience: it can, though 18:10:18 it only decrements 128 times, then [-]s 18:10:22 so, if you set your flag to 255 18:10:27 it has to loop 128 iterations 18:10:33 ofc, it'll still *win* 18:10:35 it'll just win slower 18:10:41 Yes, I starting coding when you'd pasted the first two. Didn't take long. 18:10:43 but correct, there is no way to actually beat it 18:10:48 i dunno how buggers do with themselves 18:10:54 ah 18:10:56 they win at the same time 18:10:57 cute 18:10:59 = draw 18:11:05 impomatic: how does yours work? 18:11:25 nescience: note - this is in the ideal case, when their tape length is the expected one 18:11:39 not that many got that 18:11:40 if any even did 18:11:57 nescience: in every other case, they just either: run right off the tape, or decrement a random cell 18:12:09 nescience: I assume ferret sets a bunch of cells 18:12:11 and then keeps them up 18:12:14 so that it [-]s forever 18:12:19 on a non-flag 18:13:10 ehird: just the usual build a decoy the go to the other end of the tape and start zeroing stuff 18:13:14 right 18:14:12 impomatic: submit ferret once more 18:14:14 well it doesn't really matter 18:14:19 and it'll trim most of them 18:14:20 unless the tape length is correct they lose 18:14:32 nescience: not if the other opponent suicides 18:14:52 impomatic: it only leaves a lot of >10/11 challengers if they're submitted at once 18:14:52 true enough i guess 18:14:55 if you add ferret, it'll trim 18:15:53 !bfjoust shadow (>(-)*9)*2(>-)*7(>(+)*10[-])*20 18:15:59 Score for impomatic_shadow: 58.0 18:16:13 tada 18:16:15 they're all gone 18:16:24 and parody rises once ore 18:20:58 laugh 18:21:08 -!- Corun has changed nick to Corun|away. 18:21:22 -!- Corun|away has changed nick to Corun. 18:21:56 -!- impomatic has left (?). 18:23:58 -!- AnMaster has quit (Connection timed out). 18:29:19 it's really hard to read which number is which warrior 18:29:39 and the numbers change because of the case sorting, too 18:29:57 what's the chance of reiterating the numbers immediately to the left of the warrior names or something? 18:30:28 would be nice 18:30:50 nescience: i think I'm just going to make my own variant instead of waiting for these to change :) 18:31:03 do you think I should have + or -? i.e. which? 18:32:27 it doesn't really matter 18:32:30 if you have to pick one, i mean 18:32:40 i was asking purely on aesthetic concerns :) 18:32:46 aha 18:32:51 i like - better! 18:32:52 - seems right 18:32:55 it looks like arrows ------> 18:32:57 CHARGE! 18:32:59 as the main objective is to deplete your opponents flag 18:33:03 nescience: haha 18:33:11 »»». 18:33:32 arrows or ... jesus, i can't think of the name of the things you joust with! 18:33:40 jousting sticks! 18:33:58 * pikhq should create a Brainfuck variant that uses », ›, and > for different things 18:34:07 nescience: ok: < move pointer closer to my flag > move pointer closer to their flag - decrease cell at pointer [ jump after matching closer if cell=0 ] jump after matching opener if cell!=0 . nop 18:34:36 lol jousting sticks 18:34:37 nescience: i think flags should start at 255 18:34:44 they only start at 128 in joust to make +/- not have an advantage 18:34:49 correct 18:35:01 but if you restrict it to only - it becomes about half as interesting 18:35:25 nescience: no, I think - vs + is boring. see defend6 vs parody 18:35:32 and my flipping of one of impomatics' did it to 18:35:33 o 18:35:36 it just makes it too trivial 18:35:40 as programs become tailored to one 18:35:44 and flipping them decimates them 18:36:17 rather, it means that nobody has got to the point where they write programs to take account of both polarities for example :P 18:36:22 lance! 18:36:24 that's the word i wanted 18:36:36 if you want to talk about trivial, what exactly are you going to write a program to do that only decrements? 18:36:46 nescience: doesn't rushpolarity take into account that? 18:36:54 i'm just not convinced it's an interesting aspect 18:37:19 * nescience shrugs 18:37:29 nescience: then how about thingy's idea 18:37:35 the polarity is flipped or not randomly before execution 18:37:37 actually, wait 18:37:40 same as with tape lengths 18:37:46 we do it for all (tape length,polarity) combinations 18:37:51 and take the mean 18:38:03 i'm going to write it in C, so it should be fast enough for that 18:38:09 nescience: sound good? 18:38:11 that's a possibility 18:38:32 i'm not sure how i feel about that, but it would at least get rid of the "swap the signs" bit 18:38:50 i guess it is true that you'll get the same program if you swap them 18:39:22 nescience: it forces you to take in account polarity, still 18:39:25 without being able to just swap things 18:40:16 i don't think so 18:40:26 well, i guess i see what you mean 18:40:33 in that you still can benefit from "figuring out" which way to go 18:40:54 lifthrasiir: ehm, so north korea just said they're going to nuke the shit out of your country 18:40:58 try not to die 18:41:12 nescience: yeah 18:41:18 nescience: you have to handle both +ers and -ers 18:46:50 -!- jix has joined. 18:47:38 nescience: definitely, though, randomness is the bad 18:54:57 -!- AnMaster has joined. 18:57:50 -!- jix_ has quit (Read error: 110 (Connection timed out)). 19:08:46 -!- KingOfKarlsruhe has joined. 19:10:36 -!- sebbu2 has joined. 19:15:46 OK, it's implementation time. 19:16:07 that's totally out of character for them 19:16:12 don't koreans play zerg?! 19:16:26 BADUM TISH! 19:18:27 OK, it's implementation time. <-- of what 19:18:44 (as you can see above I was disconnected) 19:18:44 AnMaster: ~20 lines up. 19:18:54 http://tunes.org/~nef/logs/esoteric/09.05.27 19:18:56 You're welcome. 19:20:07 ehird, a language based on magnetic tape? 19:20:17 ... 19:20:37 ehird, from reading ~20 lines up in the log :P 19:20:47 Mmmmmmmmmmmmmmmmmmmmyeeeeeeeeeeeeeeeeeeeno. 19:22:59 nescience: how many cycles do you think I should allow? 19:24:36 -!- sebbu has quit (Success). 19:25:19 ehird, theory 2: a variant of bfjoust 19:25:27 * ehird clap. clap. clap 19:25:33 You found the page up button! 19:25:49 -!- ais523 has joined. 19:25:53 hi ais523 19:25:56 hello ais523 19:26:03 I'm plotting revolution by writing a better bf joust ;) 19:26:06 hello ehird, AnMaster 19:26:11 ehird: as in, better rules, or better interp? 19:26:14 ais523: both 19:26:16 the interp can definitely be improved 19:26:29 it's varying randomly according to tape length [implementation issue], just flipping +/- can change the hill wildly [design issue] 19:26:35 but if you want to improve the rules, /please/ do it a different way from zzo38 19:26:36 and it's really slow [implementation] 19:26:53 for the former, run all tape lengths and take the average (requires fixing #3) 19:27:02 for the latter, add another variable: is_plus_and_minus_flipped 19:27:05 and do the solution to #1 for it too 19:27:09 oh, ok 19:27:13 for the latter latter, write it in C and make it superfast 19:27:14 so you aren't changing the language rules, but the tournament rules 19:27:19 I agree with you on that 19:27:23 ais523: well, the flipping +/- thing 19:27:35 either you randomly flip one program, or you don't 19:27:36 means that you can't just rip off someone else's program, flip +/- and dominate the hill 19:27:41 ais523: actually, no 19:27:45 well, yes 19:27:48 ais523: but not randomly 19:27:53 ais523: just like how you try all tape lengths and take the averag 19:27:54 e 19:27:54 do both 19:27:58 exactly 19:28:05 try every combination, take the average score 19:28:13 that means the same program will always have the same score 19:28:55 ais523: how did zzo do it? 19:29:25 ehird: having , read from the opponent's . output 19:29:28 and a couple of other dubious changes 19:29:38 ais523: that change is very zzo 19:29:41 link to his variant? 19:29:47 ehird: http://esolangs.org/wiki/Talk:BF_Joust 19:29:57 grr, wiki seems to be slow 19:30:09 ais523: btw, you know how it has to stay 0 for two cycles? 19:30:12 yes 19:30:21 that means that [-] works if it isn't tampered with. [-]-]-]-(it's zero)](two cycles) 19:30:23 is that intentional? 19:30:27 i suppose so 19:30:33 but it was unintuitive to me at first 19:30:36 and I did [-]. 19:30:38 not quite, but when I noticed I realised it was correct, so didn't change it 19:30:39 (with the dot) 19:30:48 Safari can’t open the page “http://esolangs.org/wiki/Talk:BF_Joust” because the server unexpectedly dropped the connection. This sometimes occurs when the server is busy. Wait for a few minutes, and then try again. 19:30:52 * pikhq observes that [-]-] is an infinite loop 19:31:00 pikhq: not in BF Joust 19:31:11 pikhq: i'm counting cycles 19:31:11 ... WTF? 19:31:15 there is no way to write a definitively 100% infinite loop in BF Joust 19:31:18 [-] runs [-]-]-]-]-]-]-] 19:31:23 pikhq: not program 19:31:24 cycles 19:31:27 because the data on the tape might change due to the opponent messing with it 19:31:32 [ ticks, - ticks, ] ticks, - ticks, ] ticks 19:31:33 etc 19:31:40 also, what ais523 said 19:31:48 ais523: [[In case of a draw, figure out who would have lost using the old rule that whoever's flag is zero loses immediately (and determine the winner according to this rule). ]] 19:31:50 ok, that's just ridiculous 19:31:55 [[If both programs have ended and neither player's flag is zero, then both programs shall restart from the beginning with the tape pointer pointing to their own flag again like it was at the start, but using the current values on the tape instead of resetting them to zero. ]] 19:31:57 so's that 19:32:02 all of them are ridiculous :) 19:32:06 agreed 19:32:15 ais523: how could you even use ,/. 19:32:18 you couldn't use it to your advantage 19:32:22 because your opponent would always lie 19:32:30 ehird: I don't think they'd even be lying 19:32:34 oh? 19:32:41 given that there's no information as to what the random stream of data means anyway 19:32:45 true 19:32:55 it's like, if I say 6 202 34, am I lying? 19:33:03 ais523: Mu 19:33:20 [[[ and ] should take the value on the tape *after* the opponent has (potentially) modified it.] 19:33:21 ] 19:33:26 ais523: does it do it before modification? 19:33:30 atm, yes 19:33:30 two opcodes running at once is sticky 19:33:34 it takes the value at the start of the cycle 19:33:37 hmm 19:33:53 I experimented with before and after 19:34:00 ais523: ah, and before worked better? 19:34:07 and decided that although it made a difference, it didn't fundamentally change the nature of things 19:34:12 so left it where Kerim had it 19:34:23 http://esolangs.org/w/index.php?title=Byte_Syze&curid=2347&diff=14541&oldid=14540 19:34:26 lol wat 19:34:31 !bfjoust foo .[+.] 19:34:37 Score for pikhq_foo: 17.5 19:34:41 pikhq: that's just a suicider 19:34:43 a slow one at that 19:34:48 well, not always 19:34:52 Seems to be effective. 19:34:58 but I don't think any program will try and give your flag viagra— what? 19:35:00 How is that effective? 19:35:11 It has a positive score. :p 19:35:12 if we weren't caching scores, the top program gets ~85pts 19:35:16 pikhq: all scores are positive 19:35:17 pikhq: we changed the scoring 19:35:19 points are negative 19:35:21 scores not 19:35:21 Oh. 19:35:25 50% is average, 0% is terrible, 100% is perfect 19:35:49 ais523: i submitted (>)*N(-)*128[-] (but manually expanded) for all possible tape lengths today 19:35:54 !bfjoust [>+] 19:35:54 one even got ~40-something 19:35:55 Use: !bfjoust 19:35:57 :-D 19:36:01 Erm. 19:36:03 !bfjoust foo [>+] 19:36:04 Score for pikhq_foo: 20.5 19:36:04 ais523: also, btw 19:36:08 if you resubmit a program 19:36:10 you get its last score 19:36:15 you have to look at the report to get the proper one 19:36:15 another bug? 19:36:18 yep 19:36:21 ah, ok 19:36:23 ais523: why do you think i'm writing my own :) 19:36:31 ais523: also, well, taking the value at the start requires copying 19:36:33 doesn't it? 19:36:35 well 19:36:39 i guess no 19:36:39 since 19:36:42 with after-modification 19:36:45 you have to wait for your opponent 19:36:48 which is just as much fuss 19:36:54 it only requires copying for one of the programs 19:36:58 and it's as much fuss both ways round 19:37:08 yep 19:37:15 you could order the execution of the programs inside the interp in the way that avoids copying, but that's even fussier in other ways 19:38:28 ais523: do you think I should keep the 20k max cycles? 19:38:46 that's about right 19:38:50 it may take 3s to execute one combination for 20k cycles in your perl interp, but it'll probably be on the order of milliseconds for mine :) 19:38:51 I determined that number via experiment 19:39:04 as in, if you make it much lower, valid programs start drawing rather than winning 19:39:11 but then we have the reverse polarity 19:39:13 which is 3-bits: 19:39:19 0 (no flip), 1 (flip left), 2 (flip right) 19:39:22 hmm 19:39:28 is that 20**3 for the tape lengths and polarity 19:39:31 or 3**20? 19:39:33 20**3, I think 19:39:58 ais523: hm, oh dear 19:40:07 for all tape length combinations, and the three polarity states, 19:40:16 why three polarities? 19:40:16 ais523: is 8000 executions per (prog1,prog2) 19:40:18 you only need two 19:40:21 also 19:40:24 ais523: no i don't 19:40:25 no flip, 19:40:25 because flipping either program has the same effect 19:40:30 er, are you sure? 19:40:31 and flipping both has the same effect as flipping neither 19:40:31 always? 19:40:44 ehird: the values at the start are either 128 or 0 19:40:46 hmm, i'll take your word for the left-is-right 19:40:53 ais523: yeah, you're right 19:40:55 ais523: but... 19:40:58 which of left/right should I flip? 19:41:00 i don't wanna be biased :D 19:41:05 it makes no difference to the result 19:41:18 0 = no flip, 1 = flip, the 19:41:18 n 19:41:36 and you need 20 * 2 (no exponentiation) 19:41:39 well, 21 * 2 19:41:44 oh, heh 19:41:47 each of the 21 possible lengths in each of the 2 flip states 19:41:55 i'll finish the message I was typing anyway: 19:42:00 ais523: well, that's good; with 3 flip states, I have to run the match 8000 times— so if I take 25ms to run one iteration, that's 200 seconds! 19:42:02 ↑ disclaimer: wrong 19:42:45 200 seconds would be a bit slower than it currently is; OTOH, you'd have a fairer tournament 19:42:47 ais523: OK, if I take 25ms to run one iteration (reasonable, I think) and do all 21 tape lengths, and have 2 polarity states, a match will take 1.05 seconds 19:42:58 a full complete matchup of (prog1,prog2) 19:43:33 ais523: err, for 10 programs, you have how many pairs of (prog1,prog2)? my brain is in coding mode atm... 19:44:12 (10*9)/2 = 45 19:45:00 ais523: so, assuming I do no caching whatsoever, and have a 10-length hill (one less than the current hill; it's 0-10, I guess as a mistake) then it'd take 47.25 seconds to run all matches 19:45:08 fairly and deterministically, on every combination of tape length/polarity 19:45:10 that's not bad at all 19:45:14 and with some caching.. 19:45:16 *... 19:45:39 with some caching, you only need to run the 9 matches involving the current challenger 19:45:42 but I don't think that is a mistake 19:45:50 the hill atm is 10 returning programs, plus 1 newbie 19:45:54 ah 19:45:56 that makes sense 19:46:31 ais523: so 51.45 seconds then 19:46:33 + caching 19:46:47 ais523: 9.45 seconds, then, with caching 19:46:51 for a challenger 19:46:58 ais523: hmm... do you think 25ms for a match is over or underestimating? 19:47:00 i'm not sure 19:47:02 are you going to pre-expand programs? or expand on the fly to save memory? 19:47:13 some of them are likely to get very long with abbreviations removed 19:47:27 ais523: i don't care about memory usag 19:47:27 e 19:47:33 ais523: how long's defend6 expanded? 19:47:39 not more than a few kilobytes, I'd wager 19:48:02 yep, defend5 is possibly quite long though 19:48:20 ais523: megabytes? 19:48:28 codu.org runs on a VPS with 1GB of RAM 19:48:30 probably not 19:48:37 i'm sure a few megabytes is just fine 19:48:50 but the reason I didn't expand is that the way I programmed my interp, I'm O(n) below the optimum 19:48:59 so it's like I wrote a Mathematica program rather than a C program 19:49:10 you can make each cycle run in O(1) time, though, so the length of the program isn't an issue 19:49:26 ais523: I'm probably going to have a trivial optimization step 19:49:31 ++++++++++++++++++ -> +*foo 19:49:33 nothing fancy at all 19:49:40 although, wait 19:49:44 ais523: you know your (x)*foo thing? 19:49:47 it doesn't change cycles, does it? 19:49:47 yes 19:49:50 and no 19:49:51 you said it makes the program "run faster" 19:49:53 it's strictly abbreviations 19:49:55 i guess that's just interp time 19:49:57 it makes the interp run faster 19:49:59 ais523: ok, I won't do any optimization at all 19:50:01 in realtime 19:50:03 it's simplest 19:50:09 rather than the program in gametime 19:50:11 after all, a cycle will probably take a fraction of a millisecond 19:52:25 okay then, time to write the interpreter loop 19:54:33 ais523: are you sure, by the way, that 20 is a good variation in tape length? 19:54:37 I would go for 10 max 19:54:42 otherwise, strategies are too unpredictable 19:55:00 IMO tape length variation should just stop the lame strategy of (>)*constant mwahaha i'm at the opponent's flag! 19:55:08 ehird: I was wondering about that 19:55:20 I definitely want to keep the tape length shorter than 32 19:55:25 and I wanted the chance of a very short tape 19:55:34 but maybe 10-20, or even 10-15, would be a better variation 19:55:38 ais523: also, with the averaging of tape lengths I'm doing, there doesn't need to be enoguh variation 19:55:43 ais523: I'd go for 15-25 19:55:53 15 is just long enough to not be stuffy 19:56:01 and 25 is just short enough not to be massive 19:56:12 the idea is that an aggressive program starting >>>>>>>>>>> should be able to arrive before the opponent completes a complex decoy 19:56:15 at least some of the time 19:56:40 eh, I'll just keep 10-30 19:56:58 ais523: btw, how come it's 21 possibilities? 19:57:04 10-30 inclusive 19:57:17 oh, ofc 19:57:23 #define MAX_CYCLES 20000 19:57:23 #define MIN_TAPE_LENGTH 10 19:57:25 #define MAX_TAPE_LENGTH 30 19:57:27 configurability! 19:58:20 ais523: is \0 a valid char? :P 19:58:36 ehird: if it were, it would be a comment 19:58:43 I'd say that if you encounter a \0, treat it as a comment 19:58:50 but don't expend effort to make sure it doesn't get eaten before then 19:58:51 ais523: C byte arrays, yo. 19:59:05 I'll just let it be the end of the string like normal, nobody's gonna use it. 19:59:11 or you can just strip out the \0s when loading the program, if oy ucare 19:59:13 *you care 19:59:23 not particularly 19:59:37 ais523: how many nested loops do you think there might be? 19:59:48 defend5 has several thousand 19:59:55 using a ({})% expansion 19:59:56 ais523: hmm 20:00:04 ais523: is the C stack that big, d'you think? :) 20:00:13 probably, it depends on your OS 20:00:22 on Windows, the default stack is exactly 1 MiB, I think 20:00:25 * ehird runs a simple test program 20:00:26 I'm not sure what it is on Linux 20:03:24 * ehird runs stack-user 20:03:33 ais523: on OS X, it's many millions so far 20:04:23 hmm 20:04:30 maybe gcc removed the tail recursion 20:04:32 * ehird uses the result to make sure 20:04:34 How about running 'ulimit -s'? 20:04:59 8192, which is surprisingly small. But Deewiant, I want to know how many calls, on average. 20:05:03 Not the actual size. 20:05:14 Just find out the size of your stack frame? 20:05:19 How boring 20:05:24 Also, that's in kilobytes 20:05:28 So it's 8x the Windows size. 20:05:43 yes 20:06:08 ais523: does lostkng have more nesting than defend5, do you think? 20:06:16 also, 20:06:16 probably not 20:06:17 943219 20:06:17 zsh: segmentation fault ./lance 20:06:20 int stack_size(int i) 20:06:21 { 20:06:23 int foo = i; 20:06:25 printf("%i\n", foo); 20:06:27 return foo + stack_size(i+1); 20:06:29 } 20:06:31 so I should be fine 20:06:45 lostkng doesn't seem to nest deeply 20:06:47 261842 stack calls on my system. 20:07:01 ais523: lostkng's bundled interp uses recursion to parse loops, y'see 20:07:04 I'm just wondering if that's wise 20:07:09 gcc-bf nests a lot more, it's a bit deeper than 256 20:07:15 wait, 256*3 20:07:18 on a sufficiently long program 20:07:24 because it uses nested loops to do switch statements 20:08:02 Millions and counting with -Os. 20:08:28 pikhq: it may have optimized it into a loop 20:08:38 * pikhq will laugh if it craps out at INTMAX. 20:08:54 ehird: I bet it did at this rate. 20:09:10 i suppose the loop's smaller than the recursion 20:09:22 * pikhq does -S and looks at the generated assembly 20:09:48 Hmm. 20:10:08 ais523: going through an array's likely to be more efficient than traversing a linked list, isn't it? 20:10:21 (I'm thinking about speed upfront because of the large number of combinations I have to run each program through) 20:10:33 hmm, let me think 20:10:38 It got optimized into a silly loop. 20:10:40 they're both O(1) 20:10:52 ais523: not efficiency, speed 20:10:56 pikhq: heh, translate the loop to C? 20:10:57 so it depends on the individual opcodes 20:10:59 *to C 20:11:12 ais523: well, I'm thinking about, e.g. the cache 20:11:13 taking the next list element involves dereferencing a pointer 20:11:19 ais523: a linked list won't be together in memory 20:11:26 so jumps will be further away 20:11:27 ehird: could be, depending on how it's allocated 20:11:29 an array is all clumped together 20:11:32 but it's going to take up more memory 20:11:34 ais523: could be, but isn't guaranteed to be 20:11:36 so less will fit in the cache 20:11:38 that also 20:11:40 so I suspect the array will be faster 20:11:56 void stack_test(int i){while(1)printf("%i", i++);} 20:12:08 also, you can use a posix_foo function to tell the memory manager that you're about to read a load of memory sequentially 20:12:14 pikhq: that doesn't include a return value 20:12:21 which is what I did to attempt to force some recursion 20:12:21 maybe even vectorise it! 20:12:26 ais523: hahahahahano. 20:12:27 Ahah. 20:12:33 _posix_fadvise 20:12:41 yeah without a _ 20:12:42 whatever :P 20:14:24 hmm 20:14:35 * ehird thinks of names for + and - that don't relate to adding and subtracting, due to polarity issues 20:14:40 dink and donk‽ 20:14:52 clockwise and anticlockwise 20:15:06 ais523: but... that doesn't even make any sense :D 20:15:11 it does 20:15:16 because it's using modular arithmetic 20:15:19 heh 20:15:22 the numbers form a circle, not a numberline 20:15:27 [+]--+- "Loop dink end loop donk donk dink donk" 20:15:37 ais523: but if they're flipped, it'll be anticlockwise and clockwise 20:15:43 I want them to not require flipping when...flipped 20:15:47 as it's not flipping, just two different views 20:15:48 equal 20:17:21 ais523: no? 20:17:51 well, clockwise/anticlockwise depends on which end of the tape you're looking from 20:18:00 true. 20:18:04 but the problem is, with one polarity, one program's dink will be the other program's donk 20:18:11 with the other, they'll both dink the same way 20:18:12 ais523: nope, wrong 20:18:16 dink doesn't mean any particular way 20:18:17 it just means dink 20:18:18 same with donk 20:18:21 ehird: agreed 20:18:24 add, however, is presumably adding 20:18:27 but are you altering the programs when altering the polarity? 20:18:28 which would be false under other polarities 20:18:30 ais523: nope 20:18:38 ais523: the point is, it's defined as: 20:18:47 ehird: well, either one program's dink cancels out the other program's donk, or it doesn't 20:18:52 + If polarity is floob, increment. If polarity is boolf, decrement. 20:18:54 in one polarity, it does; in the other; it doesn't 20:18:58 so it's always dink/donk 20:19:29 the point is, changing the polarity has to reverse the meanings of +/- for exactly one program 20:19:45 ais523: the meanings don't change 20:19:51 the definitions of the operations change 20:20:00 well, ok 20:20:01 ais523: that's like saying that the meaning of [ is reversed if the cell is zero 20:20:06 reverse the definitions for exactly one program 20:20:07 it's not; that check is part of the operation and doesn't change 20:20:14 no definition, meaning or anything 20:20:21 just a conditional in its definition always evaluates one w ay 20:20:22 *way 20:21:15 ok 20:21:26 just bear in mind, that with one polarity, dink in program 1 will do the same thing as dink in program 2 20:21:33 yes 20:21:35 with the other polarity, dink in program 1 will do the same thing as donk in program 2 20:21:40 ais523: i'm really just trying to be nice to the names dink/donk as they sound funny 20:21:43 ok 20:21:55 you may end up convincing me to use clockwise/counterclockwise instead, though :P 20:22:01 bikesheds hooray 20:22:06 nah, dink/donk is shorter 20:22:10 very true 20:22:24 ais523: > is march, < is retreat 20:22:37 heh, that sounds about right 20:23:24 . is rest 20:23:35 hmm 20:23:41 ais523: [ = consider, ] = reconsider? :P 20:23:47 just trying to go for an intercal vibe :) 20:23:59 that's quite a vibe 20:24:18 hmm 20:24:45 ais523: it doesn't really reflect the jumpyness of it, though, does it? 20:25:21 does it need to? 20:25:31 not particularly 20:25:35 i'm trying to come up with a military analogist 20:25:38 *analogy 20:25:48 ] = ONE_AND_A_TWO_AND_A is a bit too long. 20:28:51 ais523: advantage of linked list— easier to mark end of program 20:28:54 wait, no 20:29:07 just {a,b,c,NULL} 20:29:12 or the like 20:29:25 OP_EOF 20:31:26 ais523: Polarities: sieve, kettle. 20:33:49 fair enough 20:34:15 -!- Gracenotes has joined. 20:34:49 ehird, i was just thinking that if you do the swapping thing, 20:34:55 that's 4 * 20 fights per pair 20:34:58 if you do every combination 20:35:03 nescience: nope 20:35:03 2 20:35:04 ... but, effectively, you only need to do 2 20:35:08 'cause the others would be equivalent 20:35:13 nescience: ais523 told me this ages ago :) 20:35:14 ... yeah, so i guess you already did that :> 20:35:16 i had none, left, both 20:35:17 hehe 20:35:20 er 20:35:21 none, left, right 20:35:24 but ais523 told me left=right 20:35:26 i'm not reading, just occurred to me to mention it 20:35:26 so, yeah 20:36:36 -!- olsner has joined. 20:37:49 (i guess it's actually *21) 20:37:54 -!- MizardX has quit ("Dead pixels in the sky."). 20:37:55 yep 20:37:59 anyway, this is technically my work computer so i'm sporadic :P 20:38:09 you should get super tricky and run all of them in parallel! 20:38:10 hehe 20:38:14 lawl 20:38:30 nescience: iirc, codu only has one virtual cpu 20:38:39 so that wouldn't help too much 20:38:43 locally, though... 20:40:23 "In January 2003, the toothbrush was selected as the number one invention Americans could not live without, beating out the automobile, computer, cell phone, and microwave oven, according to the Lemelson-MIT Invention Index.[4]" 20:41:24 ais523: grr, variadic macros are C99 aren't they? 20:41:30 and gnuc89 20:42:00 yes 20:42:03 darn 20:42:08 although, you can often have the same effect by using a second pair of parens 20:42:10 even in C89 20:42:13 yes 20:42:18 because (a,b) is one param to a macro 20:42:21 ehird: This is the USA, not Britain. 20:42:43 GregorR-L: in britain we never use toothbrushes 20:42:50 as evidenced by our TERRIBLE BRITISH ORAL HYGIENE 20:42:51 HA HA HA 20:43:22 I used to believe that was just a stereotype, until I went there on vacation :P 20:43:44 i am constantly amazed that people actually go here voluntarily 20:43:49 it's... not an exciting country 20:43:53 stereotypes are usually based in fact 20:43:59 i found the same thing about canadians and "eh" 20:44:25 nescience: correct, you blond black homosexual. quick, be blackface, dumb and feminine! 20:44:51 if you think about it though, stereotyping is a pretty critical function of the human brain 20:45:09 yes 20:45:09 imagine if you tried to be politically correct and assume that no ALL bears will maul you if you punch them in the face 20:45:10 :P 20:45:10 More accurately, categorization is critical, stereotyping is the unfortunate result of categorization. 20:45:21 i'm pretty sure political correctness only applies to humans 20:45:27 ehird: Not true. 20:45:33 o rly 20:45:52 Be kind to animal rights activists, they're people too! (And therefore other animals are people by proxy) 20:46:24 GregorR-L: It was fine until the parenthical, which is bullshit. 20:46:32 :P 20:48:06 ais523: nescience: Hokey dokey, interpreter loop time. 20:48:37 warrior_id_t run_match(ins_t *a, ins_t *b, polarity_t polarity, int tape_length). Maybe I need 7 more parameters. 20:49:04 ais523: (a,b)=(b,a) matchup in your interp, right? 20:49:26 ehird: yes 20:49:31 right. 20:49:34 and a program against itself always draws 20:49:59 ais523: draws have no effect on anything, right? 20:50:08 apart from draw count... 20:50:13 I don't know how that scoring system works 20:50:16 ais523: do you special-case that, btw? it seems like it may be quicker just to run it against itself 20:50:17 I wrote the interp, not the scorer 20:50:30 you'll need to ask GregorR about tournament infrastructure 20:50:58 ↑ 20:51:43 ais523: pingeriffic 20:52:14 ehird: as I said, I don't know 20:52:19 -!- tombom_ has joined. 20:52:21 20:50 ehird: ais523: do you special-case that, btw? it seems like it may be quicker just to run it against itself 20:52:26 ais523: that's part of your interp is it not? 20:52:28 or? 20:52:38 I don't check for both programs being identical 20:52:41 ah 20:52:42 OK 20:52:44 I thought you meant, checking for the same filename 20:52:47 -!- MizardX has joined. 20:52:54 I don't think GregorR's tournament driver pits programs against themselves, since they get spaces 20:52:58 instead of 0s (= draw) 20:53:29 Yeah 20:54:16 http://codu.org/eso/fyb/SCORES <-- score system 20:58:11 -!- tombom has quit (Read error: 60 (Operation timed out)). 20:58:12 -!- tombom_ has changed nick to tombom. 20:59:08 GregorR-L: well, if you categorized but didn't act on it, it'd be afwul useless, wouldn't it? :P 21:01:33 nescience: Categorizing is dividing people into groups, stereotyping is taking properties of some part of the group and blindly applying them to the entire group. Saying "All black people have dark skin" is categorization, since that's the definition of the category, saying "All black people steal cars" is stereotyping. 21:04:51 Stereotyping is prescribing to a group a property that is not inherent to it. 21:05:03 "All black people have dark skin" → having dark skin is inherent in being black 21:05:18 "All people who steal cars steal vehicles" → again, inherent 21:05:23 "All black people steal cars" → not 21:05:50 I believe that's what I just said. 21:06:08 GregorR-L: "stereotyping is taking properties of some part of the group and blindly applying them to the entire group" 21:06:11 not the same definition 21:06:17 mine is more rigorous and precise 21:06:58 I guess stereotyping can be based on properties that apply to nobody, like "All black people have antennae" or "All Jewish people came from the Mother Ship" 21:07:12 :-D 21:07:30 GregorR-L: "All Jewish people control the banks and are part of the New World Order" would be more of a real-world example 21:07:52 The first part is an example of my definition :P 21:08:48 Yes. 21:08:50 So's mine 21:08:57 [[Fabian Escalante, who was long tasked with protecting the life of Castro, estimated the number of assassination schemes or attempts by the CIA to be 638. Some such attempts allegedly included an exploding cigar, a fungal-infected scuba-diving suit, and a mafia-style shooting.]] 21:09:45 stereotypes that are nonsensical, unless they are a joke i guess, are not going to arise 21:10:20 it's more like applying an observed link to the whole 21:10:30 but it's considered "rude" if the observed link is negative :P 21:10:41 nescience: feel free to observe a Jew new world ordering 21:10:46 i'll wait here 21:11:28 lol. 21:11:39 i have observed no links between jews and conspiracy theories, sadly :P 21:11:53 you're not likely to catch hell if you say something like "damn, latinas are sexy!" 21:11:56 but it's equally a stereotype 21:11:57 * nescience shrugs 21:12:16 ais523: how many nested loops does defend5 have again? 21:12:38 I think it's a bit over 2000 21:12:43 "as a female member of latin descent, i am offended by your assertion that i must be sexy! take that back right now!" :> 21:14:34 sometimes I fly around in a spaceship... 21:14:36 anyway uhh, back to brainfuck? :P 21:14:43 ais523: :> 21:14:53 ROCKET OR SMILEY YOU THE PUBLIC DECIDE 21:17:07 aaaaaaaargh 21:17:19 my c function needs to return a tuple, now i have to pass a pointer 21:17:20 phooey :) 21:19:49 -!- M0ny has quit ("Read error: 182 (Connection reset by beer)"). 21:22:07 ais523: why are syntactically invalid programs admitted, btw 21:22:07 ? 21:22:10 I'd just drop them, 21:22:14 s/,$// 21:22:31 ehird: I exit with an error message, I think 21:22:37 but the tournament infrastructure doesn't notice 21:22:43 ais523: well, ] treats it all as draws 21:22:49 yes 21:22:55 it's interpreting the error as a draw 21:22:59 ah 21:23:28 you know you need to restructure when [[fprintf(stderr, "Unmatched ] in %s.\n", filename);]] ends up at column 91 21:29:39 re: the scoring system 21:29:56 you use the points for a program in calculating the points for a program 21:30:03 how exactly does that work? 21:30:11 nescience: points for score, no? 21:30:17 or score for score 21:30:27 The base score of any given program is the sum of: 21:30:29 * The values for the programs which it defeated as defined above. 21:30:32 * Half the values for the programs that it tied with as defined above. 21:30:46 which means, the first time you submit the program, it gets a better score than afterwards 21:30:59 nescience: it removes the program you submitted before running 21:31:00 I'm pretty sure 21:31:01 since afterwards, whatever programs it defeated are going to have less score 21:31:21 but it doesn't recalculate scores for everything on the hill? 21:31:27 or rather, i thought that's what the cache thing was about 21:31:33 It caches wins, not scores. 21:31:38 It always recalculates the scores. 21:31:40 i see 21:31:46 but then... 21:31:54 how do you calculate how many points ANY warrior has 21:31:58 even if you remove the new entrant 21:32:13 each time you recalculate the score of a given pre-existing warrior, its score is still dependent..? 21:32:28 Run it against every program currently on the hill, using the cache instead when available. 21:32:34 I have an issue 21:32:40 AnMaster: what with? 21:32:43 The score will change with every run, as it is hill-dependent. 21:32:44 sure, that gives you wins/losses 21:32:46 a good image format 21:32:48 for something 21:32:50 to be specific 21:32:51 AnMaster: png 21:32:55 but points is based on the points of the other programs 21:33:04 ehird: you don't know what sort of image it is 21:33:08 arguably, iso might be better 21:33:09 Score is based on points of the other programs. 21:33:10 i don't see how you can have them all in synch without some complicated math ala pagerank :P 21:33:11 ais523: heh. 21:33:12 scanned diagram. Must not be lossy. 21:33:14 but seriously. 21:33:18 In png the file is 70 MB 21:33:19 AnMaster: absolutely PNG 21:33:23 it's great for geometric shapes 21:33:24 ahh 21:33:27 there is a non-lossy format for scanned diagrams 21:33:28 Points is wins minus losses, score is based on that. 21:33:29 AnMaster: 70MB? there's nothing you can do 21:33:29 which is larger than the xcf at 65 MB 21:33:31 ... 21:33:32 it's gonna be gigantic anyway 21:33:33 the text ony said "value" 21:33:36 AnMaster: look at Wikimedia Commons for similar files 21:33:37 ... 5MB makes a difference to you? 21:33:38 and see what it uses 21:33:40 i thought it was referring to teh value of those programs' scores 21:33:46 AnMaster: just go w/ PNG 21:33:53 nescience: Nono, that's their value as in how much they're worth. 21:33:54 ehird, that isn't as good for scans 21:33:54 it's simple and near the best you'll prolly get 21:34:01 AnMaster: how large is it after running it through pngcrush? 21:34:02 ehird, also it is 1200 dpi, full A4 21:34:05 based on the number of wins they have 21:34:08 AnMaster: 5MB is insignificant at 65-70MB size. 21:34:11 nescience: Yeah 21:34:14 just use png 21:34:17 well at least that can synch up, i see 21:34:21 nescience: I didn't want to make a recursive algorithm :P 21:34:26 ais523, that was *after* pngcrush, optipng and advpng 21:34:28 seems like it's partly responsible for some of the noise on the hill too :P 21:34:31 before it was well over 80 MB 21:34:48 if you only beat two programs, but those programs beat everything else... should your score be that high i wonder? 21:34:54 oh well, it's there for the exploitation! 21:35:03 AnMaster: So, you can store 65MB, but not 20MB more? 21:35:08 I seeeee. 21:35:14 ehird, wrong 21:35:19 I think 65 is too much too for this 21:35:22 nescience: I decree that it should be fairly high, but not as high as the program that beat all the others :P 21:35:31 AnMaster: Tough 21:35:36 1200dpi + A4 size = big. 21:35:40 since I need to upload it. 21:35:44 ehird, yes I know 21:35:53 AnMaster: er, how small is your upstream 21:36:01 ehird, 1 mbit 21:36:06 fast enough 21:36:14 ehird, on the paper 21:36:17 in practice? 21:36:24 ~600 kbps 21:36:30 ehm 21:36:34 AnMaster: 600 kbps ~= 1mbit 21:36:37 KB!=Kbit 21:36:45 Kilonibble 21:36:49 that was kbit 21:36:49 ... 21:36:51 AnMaster: oh. 21:36:54 600 kbit/sec 21:36:57 AnMaster: so, um 21:36:59 http://www.google.com/search?hl=en&safe=off&client=safari&rls=en-us&q=70+megabytes+%2F+600+kilobits+per+second&btnG=Search&aq=f&oq=&aqi= 21:37:02 AnMaster: 16 minutes? 21:37:10 You're fussing over 16 minutes for uploading a PNG? 21:37:14 ehird, something like that yeah. 21:37:17 Nothing better to do I see. 21:37:19 ehird, 5 png 21:37:21 all like this 21:37:30 so 16 * 5 21:37:39 AnMaster: So, you have 80 minutes to spend on finding an obscure format that saves 10-20MB, but you can't spend those 80 minutes uploading. 21:37:43 Bit ridiculous, that 21:37:52 GregorR-L: i agree. i guess there's enough wins to go around 21:37:55 ehird, wrong. I was just hoping someone here knew something good 21:37:57 ehird: it depends on how many of these images AnMaster needs to upload 21:38:07 ais523, as I said above. 5. At the moment 21:38:08 ais523: he said 5 21:38:10 Instead, let's waste 90 minutes arguing about whether obscure formats are worthwhile :P 21:38:11 might be more in the future 21:38:11 = 80 minutes in total 21:38:12 but with some tricksiness it might be possible to submit some warriors to act as stepping stones on each other 21:38:15 but not today 21:38:15 using 70MB PNGs 21:38:23 seems to me like the hill score should be based on a 0 sum approach somehow 21:38:25 @ 16 minutes / file 21:38:27 which is just fine 21:38:30 nescience: we were talking about doing that while the hill was upside-down 21:38:35 nescience: gimme an algorithm and it goes into my impl :) 21:38:38 in order to try to get good programs there 21:38:40 nescience: Mebbe, feel free to give me a new algo :P 21:40:09 ehird, ~75 kbyte/sec up would be 600 kbit/sec as far as I can tell. 21:40:22 AnMaster: you said you usually got 600 kbit/up 21:40:23 per sec 21:40:26 i calculated based on that 21:40:29 = 16 minutes / file 21:40:34 = 80 minutes total for 5 files 21:40:35 i can't quite wrap my head around what's going on with the current score 21:40:36 which isn't a problem 21:40:54 as far as how things affect each other i mean 21:40:57 so maybe it already is 21:41:00 ehird, and it might be more in the future. Just 5 today. Probably 20 more next week 21:41:15 AnMaster: 5 hours 21:41:17 one off 21:41:20 is this seriously a problem? 21:41:22 5.3333(...) 21:41:37 ehird, probably not a one-off! 21:41:50 AnMaster: so first it was 5. then 20. now it's 498357348957893457894353495? 21:41:55 what's the problem with basing it on most wins, for example? 21:42:12 i assume there's some behavior you want from giving more points for defeating "better" warriors 21:42:14 nescience: because then you can target bad programs 21:42:19 and lose against good ones 21:42:21 and rise to the top 21:42:22 which sucks 21:42:28 but 21:42:31 with the 11-length hill... 21:42:32 except, there won't be many bad programs on the hill after a while 21:42:35 so can everyone else(?) 21:42:40 all the ones there are pretty good, I imagine 21:42:48 exactly 21:42:49 I'll resubmit attack1, and I bet it'll do terribly 21:42:52 despite not being all that bad 21:43:02 !bfjoust attack1 [>[-]+] 21:43:08 Score for ais523_attack1: 13.5 21:43:08 :-D 21:43:15 not bad 21:43:18 531.50-2impomatic_mirage.bfjoust 21:43:19 113.50-8ais523_attack1.bfjoust 21:43:20 being the "best" in this setup no longer means having the best record/defeating the most enemies 21:43:24 lowest pts apart from it is 21:43:26 948.50-4nescience_creep.bfjoust 21:43:28 yep, lost to everything but creep 21:43:34 lol :> 21:43:40 still, not that terrible 21:43:45 and the setup here is such that i don't think it's really much different to target "better" programs than it is to target "worse" programs 21:43:46 !bfjoust attack_experimental >>>>>>>>>[>[-]+] 21:43:46 we've had 0.0 - 8.0 21:43:55 since there's not really much you can do to target any particular program 21:43:57 Score for ais523_attack_experimental: 15.5 21:43:57 Attaxperimental 21:44:00 except something like defend6, for example 21:44:02 very slightly better... 21:44:11 1 | - 0 0 - - - - - - + | 15.5| -6| ais523_attack_experimental.bfjoust 21:44:17 ais523: two draws and one win isn't too bad 21:44:31 how did that draw with defend6, I wonder? 21:44:52 !bfjoust revelation_fell_short (>)*9[[-]>+] 21:45:01 the - loop 21:45:02 Score for ehird_revelation_fell_short: 6.5 21:45:04 wait, nevermind 21:45:08 !bfjoust revelation_fell_short (>)*9[[+]>-] 21:45:08 it only incs 128 times 21:45:10 Score for ehird_revelation_fell_short: 6.5 21:45:19 AnMaster: so first it was 5. then 20. now it's 498357348957893457894353495? <-- stop being silly 21:45:20 ais523: see, I can do worse! 21:45:36 ehird: an idea regarding the polarity flip: 21:45:39 just it is 5 today. And "the customer might want more next week ;)" 21:45:45 what about taking either the best or the worst score for the attacker? 21:46:00 nescience: I think averaging all the combinations is petter 21:46:08 well, the average of a win and a loss is a tie 21:46:09 ah, I see 21:46:16 attack_experimental never goes into the [] loop 21:46:20 nor does revelation_fell_short 21:46:24 unless the tape length is exactly 10 21:46:36 but if polarity comes into play, either you will take advantage of it or not take advantage of it 21:46:36 !bfjoust attack_experimental >>>>>>>>>+[>[-]+] 21:46:38 Score for ais523_attack_experimental: 15.5 21:46:42 doing both and averaging is kina mediocre i guess 21:46:44 i dunno, just a thought 21:46:58 polarity? 21:47:03 reverse it! 21:47:59 and now it loses to defend6, as it ought to 21:47:59 well that's what i'm getting at 21:48:10 you can discourage taking advantage of polarity (and thus writing more generalized solutions) 21:48:15 by taking the worst score of the two 21:48:15 nescience: hmm 21:48:26 or you can promote taking advantage of it by doing the opposite 21:51:29 -!- BeholdMyGlory has quit (anthony.freenode.net irc.freenode.net). 21:51:30 -!- tombom has quit (anthony.freenode.net irc.freenode.net). 21:51:30 -!- ais523 has quit (anthony.freenode.net irc.freenode.net). 21:51:30 -!- Sgeo has quit (anthony.freenode.net irc.freenode.net). 21:51:30 -!- kerlo has quit (anthony.freenode.net irc.freenode.net). 21:51:31 -!- ineiros has quit (anthony.freenode.net irc.freenode.net). 21:51:31 -!- Slereah has quit (anthony.freenode.net irc.freenode.net). 21:51:33 -!- Asztal has quit (anthony.freenode.net irc.freenode.net). 21:51:33 -!- ehird has quit (anthony.freenode.net irc.freenode.net). 21:51:33 -!- rodgort has quit (anthony.freenode.net irc.freenode.net). 21:51:33 -!- Gracenotes has quit (anthony.freenode.net irc.freenode.net). 21:51:33 -!- GregorR has quit (anthony.freenode.net irc.freenode.net). 21:51:34 -!- lifthrasiir has quit (anthony.freenode.net irc.freenode.net). 21:51:34 -!- FireFly has quit (anthony.freenode.net irc.freenode.net). 21:51:34 -!- fungebob has quit (anthony.freenode.net irc.freenode.net). 21:51:34 -!- fungot has quit (anthony.freenode.net irc.freenode.net). 21:51:34 -!- MizardX has quit (anthony.freenode.net irc.freenode.net). 21:51:34 -!- Deewiant has quit (anthony.freenode.net irc.freenode.net). 21:51:34 -!- fizzie has quit (anthony.freenode.net irc.freenode.net). 21:51:34 -!- jix has quit (anthony.freenode.net irc.freenode.net). 21:51:35 -!- nescience has quit (anthony.freenode.net irc.freenode.net). 21:51:35 -!- Ilari has quit (anthony.freenode.net irc.freenode.net). 21:51:35 -!- pikhq has quit (anthony.freenode.net irc.freenode.net). 21:51:35 -!- comex has quit (anthony.freenode.net irc.freenode.net). 21:51:35 -!- Leonidas has quit (anthony.freenode.net irc.freenode.net). 21:51:35 -!- sebbu2 has quit (anthony.freenode.net irc.freenode.net). 21:51:35 -!- AnMaster has quit (anthony.freenode.net irc.freenode.net). 21:51:36 -!- Corun has quit (anthony.freenode.net irc.freenode.net). 21:51:36 -!- Dewi has quit (anthony.freenode.net irc.freenode.net). 21:51:37 -!- olsner has quit (anthony.freenode.net irc.freenode.net). 21:51:37 -!- bsmntbombdood has quit (anthony.freenode.net irc.freenode.net). 21:51:38 -!- EgoBot has quit (anthony.freenode.net irc.freenode.net). 21:51:39 -!- inurinternet has quit (anthony.freenode.net irc.freenode.net). 21:51:39 -!- KingOfKarlsruhe has quit (anthony.freenode.net irc.freenode.net). 21:51:39 -!- ski__ has quit (anthony.freenode.net irc.freenode.net). 21:51:39 -!- dbc has quit (anthony.freenode.net irc.freenode.net). 21:51:40 -!- GregorR-L has quit (anthony.freenode.net irc.freenode.net). 21:51:40 -!- mtve has quit (anthony.freenode.net irc.freenode.net). 21:52:14 -!- impomatic has joined. 21:52:14 -!- MizardX has joined. 21:52:14 -!- tombom has joined. 21:52:14 -!- olsner has joined. 21:52:14 -!- Gracenotes has joined. 21:52:14 -!- ais523 has joined. 21:52:14 -!- sebbu2 has joined. 21:52:14 -!- KingOfKarlsruhe has joined. 21:52:14 -!- AnMaster has joined. 21:52:14 -!- jix has joined. 21:52:14 -!- FireFly has joined. 21:52:14 -!- fungebob has joined. 21:52:14 -!- BeholdMyGlory has joined. 21:52:14 -!- inurinternet has joined. 21:52:14 -!- Corun has joined. 21:52:14 -!- Sgeo has joined. 21:52:14 -!- GregorR-L has joined. 21:52:14 -!- GregorR has joined. 21:52:14 -!- nescience has joined. 21:52:14 -!- bsmntbombdood has joined. 21:52:14 -!- Slereah has joined. 21:52:14 -!- pikhq has joined. 21:52:14 -!- ehird has joined. 21:52:14 -!- Dewi has joined. 21:52:14 -!- comex has joined. 21:52:14 -!- lifthrasiir has joined. 21:52:14 -!- kerlo has joined. 21:52:14 -!- Deewiant has joined. 21:52:14 -!- Ilari has joined. 21:52:14 -!- ineiros has joined. 21:52:14 -!- fungot has joined. 21:52:14 -!- fizzie has joined. 21:52:14 -!- Leonidas has joined. 21:52:14 -!- Asztal has joined. 21:52:14 -!- rodgort has joined. 21:52:14 -!- EgoBot has joined. 21:52:14 -!- dbc has joined. 21:52:14 -!- mtve has joined. 21:52:14 -!- ski__ has joined. 21:52:44 you can do a[bd]cd 21:52:44 Can I nest () inside {}? 21:52:44 AnMaster: If polarity is sieve, + (dink) increments and - (donk) decrements for both warriors. If polarity is kettle, the same applies for one warrior, but the other warrior has the operations reversed - dink (+) decrements and donk (-) increments. 21:52:44 impomatic: I think so 21:52:44 AnMaster: what we're talking about with "polarity" .. 21:52:44 it's meant to work, I'm not sure if it does though 21:52:44 er, yeah. 21:52:44 because I've never tested 21:52:45 AnMaster: This stops people just copying a program and swapping + and - to get an advantage over [-] or [+], etc. but still makes you take into account the other program doing either and both. 21:52:45 err 21:52:50 ehird, are these terms from intercal? 21:52:52 ;P 21:52:55 AnMaster: No, I made them up. 21:52:57 ehird: defend6 is robust against swapping + and -, by the way 21:52:59 (dink and donk and so on) 21:52:59 ais523: yep 21:53:00 ais523: only if d is 0 21:53:16 nescience: no, I mean d never finishes 21:53:18 ais523: but my parody changes its +s or -s (I forget which) to the other, and beat it for a long time 21:53:19 it's the rest of the program 21:53:19 by about 10 score 21:53:21 and c executes regardless, not just "else" 21:53:26 ah 21:53:28 372.005ehird_defend6_a_parody_or_just_plain_ripoff_question_mark.bfjoust 21:53:28 263.004ais523_defend6.bfjoust 21:53:29 still does 21:53:31 aha 21:53:41 ehird: interesting, that shouldn't make a difference 21:53:46 ais523: with my new system, swapping + and - has no effect 21:53:53 unless it interacts differently with one of impomatic's anti-defend6 programs 21:53:55 ais523: and possibly it's just luck of the draw; it loses some other times 21:53:57 ais523: it does, because it depends on the other programs on the hill 21:53:58 also, yes 21:53:59 I think 21:54:00 right 21:54:15 if you just have a simple [+] loop it will behave one way, and a [-] loop differently 21:54:26 when i wrote creep it was 2nd for a while 21:54:27 how does defend6 work? 21:54:29 beat defend6 every time 21:54:43 but that was because it loope dso as to wait 256 instructions when it hit your tripwire 21:54:45 AnMaster: it tries to trap the opponent in a [+] or [-] loop over its flag 21:54:45 AnMaster: it traps a [-] or a [+] 21:54:52 AnMaster: it waits until it's 0, then increments/decrements it 21:54:55 so that the loop runs once again 21:54:55 ehird reversed the "polarity" and i couldn't do that both ways 21:54:56 iirc 21:55:01 ehird, interesting 21:55:01 so it required other strategies to beat them both 21:55:08 which i pulled off, but were less successful against the rest of the hill 21:55:08 nescience: heh, that trick; it's what fool1 used, too 21:55:17 on the other hand, that's an insanely bad strategy against an attack program 21:55:18 AnMaster: so basically, if you try and just [-] or [+] out the opponents' flag, it infinite loops you then goes and rips off your flag 21:55:30 ais523: what does fool do? 21:55:33 that's why I think BF Joust is stone/paper/scissors in defence/fool/attack strategies 21:55:34 thus my comment that the reason your defend is on top currently is because beating it doesn't give much score so nobody's doin it :P 21:55:35 ehird, hm ok... 21:55:44 ehird: walks very slowly along the tape, trying to trick defensive programs into suiciding 21:55:48 by changing their flag very slowly 21:55:50 ais523: heh 21:55:57 ais523: defense1 just puts up fakes, right? 21:56:14 ehird: no, defense1 is a specific anti-attack1 program 21:56:17 ah 21:56:21 so no fakers in the base? 21:56:31 it lets its flag go to 0 for one cycle 21:56:34 then increases it again 21:56:40 and lets the opponent fall off the end 21:56:40 it'd be a very distorted rock/paper/scissors 21:56:52 defend5 works like that, but for a choice of opposing strategies 21:57:00 programs like defend only work in a niche, right now you've got that niche :P 21:57:06 Unmatched ] in butt. 21:57:08 harumph 21:57:14 but impomatic proved with shortsword that a program incapable of falling off the end is not hard to write 21:57:23 :-) 21:57:28 so defend6 has to be a little more active in its defensive play 21:58:07 it can still win defend1-style, but against shortsword and similar programs to it it locks them in place while grabbing the enemy flag a bit at a time 21:58:16 ais523, btw... that -*- mode string in gcc-bf output 21:58:21 how do you highlight bf 21:58:25 esolangs.el 21:58:26 his thingy 21:58:32 yep, I have emacs modes for esolangs 21:58:36 which I've pasted in the past 21:58:40 ais523, I can't think of a useful way to highlight bf 21:58:43 hmm I wonder why it thinks I have an unmatched butt 21:58:46 AnMaster: you can highlight comments 21:58:51 and stray BF chars inside comments 21:58:52 also, constants 21:58:57 like [-]+++++++ 21:59:07 and I colour <> differently to +- differently to [] differently to ,. 21:59:11 to make the program a bit faster to read 21:59:14 ais523 what about [-]>++<++++++++ ? 21:59:27 AnMaster: [-] is a constant 0 21:59:39 the +++++++ at the end is highlighted as addition, though 21:59:46 it's a syntax highlighter, not an optimising interpreter 21:59:49 fair enough 21:59:56 ais523, where can I find this thing 22:00:02 and does it highlight b98 too? 22:00:04 AnMaster: pastebin in logs 22:00:06 and no, it doesn't 22:00:13 Befunge is a real pain to syntax-highlight 22:00:17 ais523, which languages then apart from bf 22:00:30 Thutu, and Unlambda 22:00:56 ais523, you could highlight it mostly based on which character it is. You could handle string mode easily enough for the cardinal directions 22:01:00 it also compiles Relambda to Unlambda, a bit buggily 22:01:16 AnMaster: can you tell whether an individual Befunge command is part of a string or not? 22:01:18 it might be both 22:01:19 hmm it's parsing really weird 22:01:26 ais523, indeed it might 22:01:43 ais523: Thue too 22:01:45 ais523, and you could partially by tracing program and using that info? 22:01:46 no? 22:01:54 ehird: stub code for that, but IIRC I didn't finish it 22:01:55 AnMaster: it can be used as both, always 22:01:56 so... 22:02:02 ais523: i've used it, I'm sure 22:02:04 ehird, yes of course 22:02:05 AnMaster: I'm beginning to worry that you don't understand the concept of syntax highlighting 22:02:21 ais523, it wouldn't be CLASSICAL syntax highlighting 22:02:22 syntax highlighting != flymake 22:02:33 ais523, "flymake"? 22:02:35 ... 22:02:37 how can you use emacs 22:02:39 and not know what flymake is? 22:02:49 ehird: admittedly, I don't /use/ flymake, but I know what it is 22:03:14 "flymake not found in emacs info pages"? 22:03:23 or rather "no flymake in index" 22:03:36 Info-index: No `flymake' in index <-- there. See it ehird? 22:03:49 .. 22:04:03 This is an excellent time to brb, as my stupidity-thermometer was about to blow. 22:04:03 → 22:04:04 ehird: defend is a good example of why giving the *better* score might be advantageous 22:04:10 (specifically, the difficulty of overcoming the tripwire strategy and still being viable in other contexts) 22:04:32 not really too solidly sure on that one though 22:04:45 I fail to see what ehird meant with that 22:05:01 Anyway if emacs docs doesn't list it. it probably isn't there in my emacs version... 22:05:11 maybe it is gnu emacs specific? 22:05:19 * AnMaster is currently using µemacs 22:05:27 ais523, so tell me what flymake is 22:05:48 AnMaster: it repeatedly runs make in the background 22:05:52 so you get error messages on the fly 22:06:08 interesting 22:07:34 but not very useful. Since it would generate error messages a lot when editing. For example something like typing "if (foo) bar();" (without quotes!) isn't valid at all until completely typed in C at least. 22:07:46 is that one make for each key-press? 22:07:52 it would be rather slow if so 22:08:01 I think these problems have been thought about 22:08:03 but I don't use it 22:08:11 I wouldn't use it either. 22:10:48 ais523, the wiki page mentions that "Note that the commands are specified slightly more precisely than in standard Brainfuck; in particular, the timing surrounding [ and ] is relevant in BF Joust." but doesn't actually describe this timing... 22:11:11 AnMaster: yes it does 22:11:27 they both take one tick, and use the value at the start of the cycle, and jump to the command after the matching pair 22:11:45 hm... 22:12:16 why this interest in joust btw... 22:12:26 also what does the word "joust" mean 22:12:31 if anything 22:13:08 AnMaster: it's a sport that hasn't been seriously played for several thousand years 22:13:23 it involves two people on horseback riding at each other with lances 22:13:26 ah 22:13:29 ehird: did you see my comment about timing in the wiki talk page? 22:13:29 and trying to hit each other's shields 22:13:30 turnering in Swedish 22:13:36 what do you think about that for your interpreter? 22:14:29 ais523, several thousands isn't right. More like several hundred. 22:17:05 ah, yes 22:17:09 Hell, jousting only began in the second millennium CE :-P 22:18:32 ehird: have you seen the trailer for Enigma, btw? 22:18:36 it's so bad it's hilarious 22:18:59 why does a certain level of jpeg compression make an image look (badly) photoshopped? Hm... I mean photoshopped as in "made by combining two images" 22:19:29 like this one: http://en.wikipedia.org/wiki/File:Joust1c.jpeg 22:20:11 -!- cowb0y has joined. 22:20:26 http://retrocode.blogspot.com/2009/05/bf-joust-king-of-hill.html <-- was it someone in here who wrote that? 22:21:07 -!- cowb0y has left (?). 22:21:35 * impomatic owns up 22:21:42 ah... 22:22:06 that explains why the blog seems to focused on war programs... 22:22:23 :-) 22:22:33 * AnMaster wonders if you could make a successful pacifist program in joust 22:22:40 "pacifist"? 22:22:50 ais523, misspelled? 22:22:56 no 22:23:01 I just don't know what it would mean in this context 22:23:07 ais523, nor do I! 22:23:11 Makes the other program kill itself? 22:23:13 but it sounds interesting. 22:23:19 Deewiant, maybe 22:23:22 Deewiant: like defence1? 22:23:37 Possiblyy what's defence1 22:23:38 demonstrations against war should be involved 22:23:40 s/yy/y;/ 22:24:12 Corewars - UN Edition 22:24:14 :D 22:24:57 -!- sebbu has joined. 22:25:23 -!- BeholdMyGlory has quit (Remote closed the connection). 22:25:39 -!- jix has quit ("leaving"). 22:25:52 Deewiant: it defends its flag, trying to get the opponent to fall off the end 22:26:20 Sounds as close to pacifist as you can get 22:27:55 Is there some kind of time limit? 22:28:40 yes 22:28:45 and after that, it's a draw 22:29:19 Okay, so then you could make even more of a pacifist 22:30:47 !bfjoust farmer1 [>+*128>+*128<[-]>[-]<<] 22:30:49 Score for Deewiant_farmer1: 0.0 22:30:50 Or something 22:31:18 Hmm, did I mess that up, seems like it should tie against the defence-types 22:31:31 Or do they advance steadily as well 22:32:06 Deewiant: defend6 doesn't 22:32:12 advance, that is 22:32:16 I mean, it starts >+[] 22:32:25 ah, your syntax is wrong 22:32:28 it's (+)*128 22:32:30 not +*128 22:32:40 What did that do, then 22:32:46 Just comments? 22:32:50 I'm not sure 22:32:53 probably it broke the interp 22:32:53 !bfjoust farmer1 [>(+)*128>(+)*128<[-]>[-]<<] 22:32:55 Score for Deewiant_farmer1: 0.0 22:32:55 which would explain the low score 22:32:55 Heh 22:33:03 Well, no difference 22:33:07 Deewiant: off the left end 22:33:10 after one iteration 22:33:11 wait, no 22:33:14 I miscounted the > and < 22:33:16 Shouldn't 22:33:25 um 22:33:30 hm 22:33:32 Deewiant: let me run it against defend6 locally 22:33:35 to see what happens 22:33:40 By all means 22:33:57 ais523, you said it broke the interpreter? Huh. 22:34:03 that doesn't make sense 22:34:24 Muah, I won http://www.irregularwebcomic.net/cgi-bin/poll.pl 22:34:27 not that the program seems to make sense either 22:35:01 I never voted 22:35:03 Deewiant: timeout and draw 22:35:05 as expected 22:35:14 oh, of course 22:35:17 So the scoring is just such that it gets 0? 22:35:18 bug in the channel reporter 22:35:20 if you replace a program 22:35:24 it reports its old score, not a new one 22:35:31 !bfjoust farmer1 [>(+)*128>(+)*128<[-]>[-]<<] 22:35:33 Score for Deewiant_farmer1: 0.0 22:35:38 Always the original 22:35:38 Meh 22:35:43 what about calling it farmer1a then? 22:35:49 !bfjoust farmer1-fixed [>(+)*128>(+)*128<[-]>[-]<<] 22:35:56 and the report has gone suspiciously blank 22:36:01 ... 22:36:13 Hyphens not allowed in names? :-P 22:36:20 !bfjoust farmer1_fixed [>(+)*128>(+)*128<[-]>[-]<<] 22:36:21 Deewiant, probably not 22:36:28 !help 22:36:28 Supported commands: addinterp bf_txtgen bfjoust daemon daemons delinterp fyb help info kill mush userinterps 1l 2l adjust asm axo bch bct befunge befunge98 bf bf16 bf32 bf8 bfbignum boolfuck c chiqrsx9p choo cintercal clcintercal cxx dimensifuck echo echo_sh forth glass glypho google hello kipple lambda lazyk linguine malbolge ook pbrain perl qbf rail rhotor rot13 sadol sceql sh show slashes test trigger udage01 underload unlambda whirl yodawg 22:36:41 !hello h 22:36:41 http://codu.org/eso/bfjoust/report.txt is blank... 22:36:46 I wonder what's happening now? 22:36:51 !underload (test)S 22:36:52 wait a bit and refresh 22:36:58 I think EgoBot's mostly dead 22:36:58 that's what it's been like when things were going 22:36:58 !befunge "foo">:#,_@ 22:37:03 even though it's responding to !help 22:37:10 i seem to hae submitted a broken script but i can't see why 22:37:10 !daemons 22:37:11 Running daemons: bottles butt 22:37:37 Score for Deewiant_farmer1-fixed: 0.0 22:37:37 !underload (test)S 22:37:38 Score for Deewiant_farmer1_fixed: 0.0 22:37:38 test 22:38:04 Deewiant: it seems that the scoring method uses doesn't give you points if you don't get any wins 22:38:04 Still 0, though 22:38:05 Oh well 22:38:10 you got many draws 22:38:15 and a few losses 22:38:17 but no win 22:38:18 *wins 22:38:22 it was all draws before 22:38:45 wtf creep is on top now 22:38:50 Hmmnh, that scoring system sucks :-P 22:39:00 look at all those ties 22:39:01 something broke 22:39:05 oh right 22:39:06 farmer 22:39:07 0 is a tie? 22:39:08 too many ties 22:39:25 defend has hit its downside 22:39:41 Can those extraneous farmers be removed? 22:39:42 nescience: well, I'll probably make a defend7 at some point 22:39:50 Deewiant: they will be 22:39:53 It'd be better to run each pair of warriors at every possible tape length and base the score on that. 22:39:56 next time a program is submitted 22:39:58 ais523: Manually, by someone else? :-P 22:39:59 -!- sebbu2 has quit (Connection timed out). 22:40:04 Deewiant: no, automatically 22:40:16 !bfjoust attack1 [>[-]+] 22:40:18 Score for ais523_attack1: 13.5 22:40:30 see, gone 22:40:41 Oh, they're not permanent by default? 22:40:51 impomatic: ehird is doing that 22:41:01 Deewiant: only the 10 best programs from each run are maintained 22:41:07 Ah 22:41:10 although your three farmers were placed into the same run, it seems 22:41:25 If it doesn't finish a run before you add a new one, the new run will have two added instead of one. 22:41:28 ok, it is still tying everything 22:41:30 what's the deal 22:41:40 sorry for the repeated submissions, i kept missing the refresh to see the chart 22:41:47 is there some error i am getting? 22:42:37 nescience: you have to wait for the report to regenerate 22:42:43 i did 22:42:45 it wasn't on the report 22:42:47 i submitted, 22:42:50 now it is, but all ties 22:42:58 nescience: all ties normally indicates a bug in the program 22:43:00 or a typo 22:43:03 strikes me as highly improbable, but i can't see a problem with the program 22:43:07 as in, invalid syntax 22:43:09 (>(-)*9)*2>(>-)*4(>[(>(+)*10[-])*20])*20 22:43:12 do you see one? 22:43:53 not an obvious one 22:43:57 let me run it through my local interp 22:44:16 -!- tombom has quit ("Peace and Protection 4.22.2"). 22:44:44 bug in the interp, it doesn't like the syntax (>[(>(+)*10[-])*20])*20 22:44:50 I think it doesn't like [] inside () 22:45:00 try using (>[(>(+)*10[-])*20]{})%20 instead 22:45:25 what's %20 about 22:45:35 nescience: it's the nesting abbreviation 22:45:38 like *, but []-aware 22:45:55 it means "20 times, replace the {} with the contents of these parens" 22:45:55 mmk 22:45:59 as in, it's recursive 22:46:01 i see 22:46:08 so (a{}b)%5 = aaaaabbbbb 22:46:47 Does this work? (a{}b{}c)%5 = aaaaabbbbbccccc 22:47:01 What would (a{b}c{d}e)%5 be 22:47:17 didn't seem to like it any better 22:47:21 impomatic: no, that makes no sense 22:47:31 you should write it (a{(b)*5}c)%5 22:47:46 Deewiant: only one {} block inside each ()% 22:47:58 ais523: Yeah, it was at impomatic 22:47:58 i'll have to mess with it later 22:48:13 i don't have the time to concentrate on it >:( 22:48:31 or i guess i could just manually expand it 22:50:08 %20? 22:50:28 AnMaster: it's an improved, []-aware, version of runlength compression 22:50:33 read the wiki article 22:50:40 ah 22:50:42 found it 22:51:05 ais523, no way I will implement that in in-between btw 22:51:12 err, you don't have to 22:51:16 good 22:51:20 optimising BF Joust programs makes no sense as it is 22:51:27 because all tape elements are volatile 22:51:36 you can't even guarantee memory is at 0 after [-] 22:51:45 ais523, well I thought you might use it in normal bf 22:51:47 ahah 22:51:48 it worked 22:51:48 that compression I mean 22:51:58 (and some BF Joust programs specifically check for 0 after [-], as it happens) 22:52:18 nescience: wow, that's a good program 22:52:30 credit goes to impomatic though 22:52:30 what program... 22:52:39 AnMaster: nescience_shade 22:52:41 i just combined his current best code with an idea he put in my head :P 22:52:47 ais523, the source? 22:53:06 I mean, this is fairly pointless and boring if I can't see the source. 22:53:06 AnMaster: http://codu.org/eso/bfjoust/in_egobot/nescience_shade.bfjoust 22:53:14 ah 22:53:18 wow that is large 22:53:31 see above 22:53:39 i manually expanded it because the interpreter failed on the condensed code 22:53:40 nescience: I don't get how that beats defend6 22:53:42 * ais523 runs locally 22:53:55 it skips your tripwire 22:54:08 and, usefully, inverse defend6's too 22:54:15 -!- nooga has joined. 22:54:22 it'll run right off the end for any program that doesn't have decoys though 22:54:44 but hey, can't have it all 22:55:32 you're right, the tripwire never changes 22:56:05 that's brilliant, assuming that the opponent has decoys, and skipping them 22:56:11 clearly what defend6 needs is... a decoy decoy! 22:56:27 -!- oerjan has joined. 22:57:01 hey, some hats are missing on http://choosemyhat.com/index.php 22:57:18 (compared to a few months ago at least) 22:57:20 hi oerjan 22:57:32 * oerjan gets spam from the Google Kingdom of Thailand, and thinks google branding may be going a bit overboard 22:57:36 hi AnMaster 22:57:37 then i'll skip two! 22:57:37 :P 22:57:50 People don't actually read choosemyhat.com 22:58:25 GregorR-L, didn't notice the notice at the top 22:58:33 oerjan: Google branded spam? 22:58:46 !bfjoust defend7 http://pastebin.ca/raw/1436788 22:58:56 Score for ais523_defend7: 68.0 22:58:59 not bad 22:59:01 GregorR-L, I was planning to vote for http://codu.org/hats/Tricorn-med.jpg 22:59:49 GregorR-L: the text at the top may be too long, everyone knows AnMaster cannot read more than three lines 23:00:16 and straight to the top of the table 23:00:19 oerjan, I'm trying to do like haskell for everything 23:00:32 only program that beats it is impomatic_creep 23:00:50 ais523: the spam text started with the line "Google Kingdom of Thailand" 23:01:17 so no one is going to ask what I meant with that... 23:01:28 it seems 23:01:38 oerjan: yes, I guessed 23:02:08 AnMaster: BWAHAHAHA 23:02:16 oerjan, huh? 23:02:31 AnMaster: we wickedly refuse to ask 23:02:35 :/ 23:02:38 VERY WELL! 23:02:44 I shall then tell you anyway! 23:02:50 AAAAAAAAAAAAAAAAA 23:02:56 Lazy evaluation of everything! 23:03:03 OMF 23:03:04 that is the key to getting ehird irritated. 23:03:12 nooga: HUM? 23:03:12 lazy read scrollback 23:03:14 and so on 23:03:43 of course it is, *verbose* lazy evaulation 23:03:51 we notice 23:03:55 22:05 AnMaster: maybe it is gnu emacs specific? 23:03:55 22:05 AnMaster is currently using µemacs 23:03:56 you think you're so clever, don't you... 23:04:05 ehird, what 23:04:26 ehird, checked in gnu emacs too and couldn't find it 23:05:05 22:13 nescience: ehird: did you see my comment about timing in the wiki talk page? 23:05:06 yes 23:07:46 -!- Patashu has joined. 23:09:23 14:53:06 I mean, this is fairly pointless and boring if I can't see the source. 23:09:26 original bf joust worked sourceless 23:09:28 which was fine 23:09:47 ehird, you are basing that comment on false premises. 23:09:52 You assume I want to "play the game" 23:09:53 I don't 23:10:02 i don't give a fuck what you want to do, AnMaster 23:10:04 I'm interested in reading the source and seeing how it works though 23:10:06 never have, never will 23:10:35 but you keep seeming to believe I do... 23:10:39 ehird, my comment is still true. Since I said was referring to myself. 23:10:49 that is not what you said. 23:10:49 I wasn't saying it was pointless for other people 23:10:51 only for me 23:11:02 !bfjoust defend8 http://pastebin.ca/raw/1436819 23:11:14 that one should beat defence programs /and/ most attack programs 23:11:42 but it takes ages to run 23:11:47 ais523, any chance of small and elegant programs ever being best? 23:11:54 haha 23:11:54 this is brainfuck 23:11:56 doesn't look like it does it 23:11:58 unlikely 23:12:02 rushpolarity... 23:12:05 nescience, yeah :/ 23:12:14 ais523: :P 23:12:16 AnMaster: programs can be big and still elegant 23:12:21 !bfjoust defend8mwahahaha http://pastie.org/492056.txt?key=srkdkdwiqowwbzvvyoow 23:12:24 well actually what we need 23:12:27 you can add as many decoys as you want, i can still take you out every time! :> 23:12:29 laf 23:12:29 is a preprocessor command that can take a variable 23:12:36 ehird: did you just reverse his +s and -s again? 23:12:36 ais523, sure. But I didn't say that. 23:12:39 and copies it x times, increasing the value y by 1 each time 23:12:40 nescience: maybe 23:12:48 (copies a section of code) 23:12:48 laf 23:12:53 now that there's 4 copies of defend on there 23:12:56 ehird: a +/- flipped version? 23:12:59 writing a straight up defend-killer is feasible 23:13:00 yep 23:13:11 ais523: boy this is slow 23:13:12 nescience: defend7 and defend8 both beat your decoy-skipper 23:13:26 ehird: yes, the program isn't handled well by the interpreter 23:13:56 ais523: of course, because i used a cheap trick to beat 6 23:14:05 but that's far from the only way to beat it 23:14:05 it was a clever trick, though 23:14:10 and agreed 23:14:12 what was that trick 23:14:14 defend8 also beats defend6 23:14:15 i've already beat it like 5 different ways 23:14:26 and now that there's so many.. time to see some other ways 23:14:28 AnMaster: deliberately not stepping on the tripwire 23:14:32 i wonder how to allow multichar ids in next sadol 23:14:35 if the program ever updates 23:14:40 $ time perl bfjoust.pl defend8.bj defend6.bj > /dev/null 23:14:41 ais523, how would it know where that tripwire was 23:14:41 real0m40.113s 23:14:43 user0m19.613s 23:14:44 sys0m13.169s 23:14:54 AnMaster: it can't be to the left of the flag 23:14:58 so it must be to the right 23:15:10 hm ok 23:15:14 however, defend7 has a decoy decoy 23:15:25 confusing 23:15:31 so that nescience's program skips the decoy decoy, and falls into the real decoy 23:15:56 ais523, and what about defend8 23:16:14 and when will that report be generated... :/ 23:16:21 AnMaster: it's like defend7, but after 2000 cycles it stops waiting and zooms off to attack the opponent instead 23:16:22 the interpreter seems horribly slow 23:16:25 and it is 23:16:31 that's why ehird's writing a faster one 23:16:40 with polarities! 23:16:52 dink donk march retreat consider reconsider rest eof 23:17:02 ais523, why is the old report zapped before the new one is ready 23:17:02 * ehird twiddles with the parser 23:17:05 it seems silly 23:17:48 !bfjoust simpletest (>[[-]])*29 23:18:16 I think !bfjoust broke 23:18:18 that one can only fall through if the instruction it is on is 0 twice in a row 23:18:24 no brainfuck program can take this long to write 23:18:25 defend all you want 23:18:25 :> 23:18:28 *run 23:18:37 max cycles is 384000 23:18:38 so ... 23:18:42 Patashu: it's running, just slowly 23:18:47 i dunno what the interpreter is written in but it could take a while 23:18:51 the problem is that it's trying to do regexen on a very long string 23:19:01 is it unpacking all the *s? 23:19:02 that interp is O(n) in the length of the program to run one cycle 23:19:03 which is bad 23:19:09 Patashu: no, the %s 23:19:13 it has to unpack them when they run 23:19:17 shouldn't it do it all first 23:19:19 then run the full string? 23:19:41 Patashu: to speed it up, it unpacks as it goes 23:19:49 the problem is that regex matching on the unpacked strings is very slow 23:19:58 -!- impomatic has left (?). 23:20:26 still, it should have finished by now 23:20:45 lol 23:20:49 even if each game takes 40 seconds, I'd expect it to have finished 23:20:50 good job you broke it! 23:20:58 it's only playing another programs, that's 400 seconds 23:21:05 or a little under 7 minutes 23:21:36 !befunge98 'A,@ 23:21:39 !help 23:21:39 Supported commands: addinterp bf_txtgen bfjoust daemon daemons delinterp fyb help info kill mush userinterps 1l 2l adjust asm axo bch bct befunge befunge98 bf bf16 bf32 bf8 bfbignum boolfuck c chiqrsx9p choo cintercal clcintercal cxx dimensifuck echo echo_sh forth glass glypho google hello kipple lambda lazyk linguine malbolge ook pbrain perl qbf rail rhotor rot13 sadol sceql sh show slashes test trigger udage01 underload unlambda whirl yodawg 23:21:39 if it's going to run ten games in a row 23:21:45 !daemons 23:21:45 Running daemons: bottles butt 23:21:49 it would be much more efficient to unpack it all at the beginning 23:21:50 !butt wut 23:21:55 ais523: heh, my interp doesn't even handle the abbreviations 23:21:58 that's for a frontend 23:22:00 Patashu: much of the program often never needs to unpack 23:22:10 ehird: my interp stores the program with the abbreviations in 23:22:10 !c int main(void){printf("HALDO!\n");return 0;} 23:22:11 it is dead. Borked. Out of order. Not working. 23:22:12 and unpacks on the fly 23:22:19 AnMaster: the same happened earlier 23:22:19 well 23:22:23 clearly it's not working very well <.< 23:22:23 and then all the results came at once 23:22:27 ais523: yep 23:22:32 when I submitted the 21 buggerers 23:22:32 ais523, very odd that it throttles then 23:22:36 I thought it was async 23:22:40 that was part of the point of it 23:22:48 i want some time to examine defend8 23:22:49 : 23:22:54 it doesn't just sit there like the others :P 23:23:00 "the 21 buggerers" sounds like a biblical story 23:23:00 not quite 23:23:04 it starts with [ 2000 times 23:23:10 lol 23:23:19 and has a copy of defend6 after each ] 23:23:19 ais523: how big is it, expanded? 23:23:23 (the defend7 setup is before that) 23:23:24 7 billion megabytes? 23:23:24 !c printf("WTF, Egobot?\n"); 23:23:33 i see 23:23:36 i didn't quite follow that to the end 23:23:40 ehird: a bit less than 1 MB 23:23:41 GregorR-L: !c is borken. 23:23:46 pikhq: NO. 23:23:47 *NO. 23:23:48 lol 23:23:49 pikhq: all EgoBot commands seem borken 23:23:49 *No. 23:23:52 EgoBot is borken 23:23:54 whilst bfjoust is running 23:23:56 !daemon 23:23:56 It's hung up doing something 23:23:57 Invalid invocation. 23:23:58 apart from help, daemon, etc 23:23:59 i was considering suggesting a "loop max N timse" 23:24:00 times* 23:24:03 !bf_txtgen this too? 23:24:03 GregorR-L: EgoBot is borken. 23:24:05 obviously that would be better than this 23:24:07 97 +++++++++[>+>+++++++>+++++++++++++>++++<<<<-]>>>-.------------.+.++++++++++.>----.<+.-----..<.<+. [120] 23:24:09 !underload (Test)S 23:24:09 no 23:24:11 not that one 23:24:16 AnMaster: only the interps, it seems 23:24:16 just the normal commands are 23:24:22 I know, I know, when the CPU is being raped by bfjoust nothing else can run. 23:24:25 ais523, which is the main feature of it 23:24:29 GregorR-L: nice foo 23:24:36 GregorR-L, you need to throttle it then 23:24:36 (nice(1)) 23:24:39 it's not very brainfucky but still 23:24:49 GregorR-L, I thought this slox already did that... 23:25:36 GregorR-L, I guess slox is broken 23:25:46 I guess slox doesn't do what you thought. 23:25:50 UNPOSSIBLE 23:26:23 i gues i still don't quite get how {} works 23:26:23 ehird, um. What would it do instead of throttling programs. 23:26:29 it repeats the part between ( and { X times 23:26:33 and the part between } and ) X times? 23:26:37 AnMaster: Limit memory usage? 23:26:38 Score for nescience_simpletest: 20.5 23:26:38 Score for ais523_defend8: 45.1 23:26:38 Score for ehird_defend8mwahahaha: 58.3 23:26:41 YAY 23:26:43 ais523: LOL 23:26:45 it did badly 23:26:49 defend8 only got 45.1? 23:26:52 ehird, could you please check what on earth you are talking about 23:26:56 ehird, you are so wrong. 23:26:57 ais523: I beat it by reversing polarity! 23:27:14 ah, it /lost/ to defend6 and defend7 23:27:17 Great Leader AnMaster says slox does one thing. Lo, his superlative opinion is the objective fact! 23:27:19 which is strange, as it beat them in my local test 23:27:26 ehird, I checked duh. 23:27:48 AnMaster: Slow down program != throttle program 23:27:50 's CPU usage 23:27:52 ehird, GO FUCKING READ THE WEBSITE BEFORE BEING STUID. 23:27:57 STUPID* 23:27:57 STUID 23:27:59 YOU'RE SO STUID 23:28:00 http://codu.org/slox.php 23:28:07 AAAAAAAAARGH ME AnMaster CRUSH PUNY HUMANLINGS 23:28:10 ehird, I corrected the typo before you replied. 23:28:23 -!- ais523_ has joined. 23:28:29 -!- ais523 has quit (Nick collision from services.). 23:28:35 -!- ais523_ has changed nick to ais523. 23:28:40 ehird, please see http://codu.org/slox.php, then come back again. 23:28:41 sorry about that, my grandmother turned off the router by mistake 23:29:04 AnMaster: 23:29:04 23:27 ehird: AnMaster: Slow down program != throttle program 23:29:05 23:27 ehird: 's CPU usage 23:29:10 I did, thank you very much. 23:29:24 ehird, so how would it fail in this case 23:29:50 23:27 ehird: AnMaster: Slow down program != throttle program 23:29:50 23:27 ehird: 's CPU usage 23:30:15 as far as I can remember from what GregorR said, it is used for all programs that egobot runs. It gives equal limited CPU time to each such program 23:30:33 the total cpu time for all programs is limited to a max 23:30:45 i musta broke simpletest 23:30:48 haha 23:30:48 I just tested: defend8 definitely beats defend6 23:30:49 creep still wins 23:30:51 at my end, at least 23:30:58 where's your defend now! 23:31:13 so non-egobot processes always have a lot 23:31:24 maybe it fails to use slox for joust 23:31:28 since it is a special process 23:31:31 or something like that 23:31:33 and creep beats defend7 too... 23:31:36 something weird is going on 23:31:42 is that the same creep, or an improved version? 23:31:50 check the source code for it 23:31:52 same creep 23:31:55 * ehird wonders wtf his parsing is doing 23:31:59 been sitting there about since i submitted it yesterday 23:32:24 it still has that 132-cycle wait 23:32:30 it parses [>[-]+] as >-]+] 23:32:30 it's not surprising 23:32:33 it would seem rather vulnerable to fast attack programs... 23:32:35 creep is a defend killer anyway 23:32:41 and gives jump locations to non-jumps 23:32:43 oh, it's the slow sucky one 23:32:47 and yes, creep is a paper strategy 23:32:48 that was made to counteract inverse defend 23:32:51 aha 23:33:02 if it's not waiting as long as you need it to sprinkle nops 23:33:03 defend programs are doing so well that anti-defence is making a comeback 23:33:03 works now 23:33:23 !bfjoust attack1 [>[-]+] 23:33:25 Score for ais523_attack1: 6.5 23:33:52 i'm not sure why the test doesn't beat the defends though 23:34:04 [[-]] should loop if it's not 0 twice in a row, correct? 23:34:08 aw 23:34:12 it beat creep but lost to everything else 23:34:12 rushpolarity finally got pushed off 23:34:15 survival of the fittest etc 23:34:20 nescience: not necessarily 23:34:31 oh, yes necessarily 23:34:34 I was wondering about 0, 1, 0 23:34:38 but yes, it needs two 0s in a row 23:34:44 so why doesn't it beat defend? 23:34:54 -!- coppro has joined. 23:35:00 so why doesn't what beat defend? 23:35:10 alrighty, ais523 23:35:18 I have all the parsing done for an efficient representation 23:35:23 so I'm going to write run_match() 23:35:30 ais523: two simple warriors, please? 23:35:38 ehird, what is this representation 23:35:38 [>[-]+] and... 23:35:47 AnMaster: the very trivial: 23:35:48 struct _ins_t { 23:35:49 op_t op; 23:35:51 ins_t *jump; /* NULL unless consider or reconsider */ 23:35:53 }; 23:35:53 ehird: [>+[]<(.)*258(+)*127] 23:36:00 ais523: i said simple 23:36:00 ehird, in an array? 23:36:01 like really simple 23:36:03 AnMaster: yes 23:36:17 [>[-]-.-.-.-.-.-] 23:36:32 why doesn't my test script beat defend 23:36:34 i probably did it wrong 23:36:35 1sec 23:36:36 !bfjoust fool2 [++[++[>+[(.)*5-]+>]-]-] 23:36:39 ehird, and "consider or reconsider" are strange names for loops? 23:36:54 OP_DINK, /* + */ 23:36:54 OP_DONK, /* - */ 23:36:55 OP_MARCH, /* > */ 23:36:57 OP_RETREAT, /* < */ 23:36:58 Score for ais523_fool2: 24.0 23:36:59 OP_CONSIDER, /* [ */ 23:37:00 ehird: this is #esoteric, get used to it 23:37:01 OP_RECONSIDER, /* ] */ 23:37:03 OP_REST, /* . */ 23:37:05 OP_EOF, /* (end of program) */ 23:37:07 AnMaster: you must realise that i _did_ name my two polarities sieve and kettle 23:37:09 please don't do that 23:37:09 ais523: defense1, then... 23:37:13 ... (>[[-]])*29 23:37:14 coppro: it was a few lines. 23:37:18 why doesn't that beat defend, was the question 23:37:18 ehird, true. You are going ICK style. 23:37:19 8 is a few? 23:37:23 ick* 23:37:25 it should work as far as i can see 23:37:25 AnMaster: no, intercal styl 23:37:26 e 23:37:30 coppro: Yes. 23:37:49 ehird, well. ick has some odd names internally too iirc. Not as many though. 23:38:19 ais523: ? 23:38:22 fungot, how are you? 23:38:22 AnMaster: ohno ohno 23:38:24 !bfjoust fool3 >>>>>>>>>[++[++[>+[(.)*5-]+>]-]-] 23:38:33 ehird: ick does have a few, mostly filenames 23:38:38 fungot, um... Are you Rincewind? 23:38:38 AnMaster: youi usuck,boeings better airbus is a writer and writer for snl and crackin' up watchin' this video..when suddenly the local nbc station froze up. 23:38:49 ais523: I just asked for defense1 23:38:55 nop... not discworld 23:38:55 ehird: I pasted it above 23:38:56 ^style 23:38:57 Available: agora alice c64 darwin discworld europarl ff7 fisher ic irc jargon lovecraft nethack pa speeches ss wp youtube* 23:39:00 [>+[]<(.)*258(+)*127] 23:39:01 ahaha 23:39:04 ais523: I cannot find it 23:39:07 err 23:39:09 okay, that's defense1? 23:39:10 ehird: see my last comment 23:39:12 what about fool1 23:39:12 yes, it is 23:39:13 ^style discworld 23:39:14 Selected style: discworld (a subset of Terry Pratchett's Discworld books) 23:39:18 i'm looking for something very trivial :) 23:39:19 fungot, ping 23:39:20 AnMaster: ' i think you'll find,' he said. " so i should get something.' now lobsang straightened up and stared straight ahead of him, pushing through the bushes, and he'd sat up all night," said 23:39:21 [>[-(.)*64]-] 23:39:22 that's fool1 23:39:25 and it is very trivial 23:39:34 fungot, make sense! 23:39:34 AnMaster: they did. look," said sergeant colon. " i don't think my calculations were that fnord" 23:39:42 heh 23:40:02 ais523: kay, [>[-................................................................]-] 23:40:11 Score for ais523_fool3: 19.0 23:40:22 fungot, . 23:40:24 hm 23:40:32 run_match(a, b, POLARITY_SIEVE, 20); 23:40:35 mind resetting it's limit? 23:40:42 !bfjoust fool4 (>(+)*6)*9[++[++[>+[(.)*5-]+>]-]-] 23:41:11 Score for ais523_fool4: 41.5 23:42:35 !bfjoust lazy (+-)*150000(>)*9([-]>)*21 23:43:12 ais523: is it a pain to synchronize the two programs? 23:43:14 in your interp 23:43:29 ehird: not really 23:43:40 !bfjoust simple [>[-]] 23:43:49 hmm 23:43:49 ais523: so how does [ work? 23:43:53 must be fighting defend8 ;) 23:43:57 do you just indescriminately save the current cell before doing anything? 23:44:04 *indiscriminately 23:44:08 ehird: that's how I implemented it 23:44:10 after all, it's only one int 23:44:16 yar 23:44:42 __asm__("nop"); 23:45:39 ehird, that seems wasteful. Wouldn't it be better to just directly skip to the other program 23:45:52 also gcc will optimise away that asm 23:45:59 you need a volatile asm to actually do it 23:46:33 __asm__ about to erupt("nop"); 23:46:37 -!- FireFly has quit ("Later"). 23:46:42 ehird, um? 23:46:46 Very volatile. 23:46:53 ... 23:46:54 hehe, bfjoust is stuck again 23:46:56 fight to the death with defend8 23:47:10 !bfjoust simpletest (>[[-].])*29 23:47:11 Score for nescience_simpletest: 20.5 23:47:26 huh? 23:47:26 hang on 23:47:30 ais523: what happens if a program ends? 23:47:32 that is 23:47:35 + vs [>[-]+] 23:47:39 !bfjoust lazy (+-)*180000(>)*9([-]>)*21 23:47:42 ehird, read on the wiki 23:47:49 AnMaster: lazy :) 23:47:51 I'm pretty sure it said 23:48:00 problem with +- is 23:48:04 ehird, hey I patented lazy evaluation over iIRC 23:48:06 IRC* 23:48:12 bah 23:48:19 I can't screw around without my instant results 23:48:22 eventually they'll get you to 0 23:48:24 you'd have more of a hope at being funny with that joke if you knew what lazy evaluation meant 23:48:33 and then execute the opposite of your code which is 0 twice 23:48:34 -!- coppro has quit ("The only thing I know is that I know nothing"). 23:48:35 * GregorR-L reappears. 23:48:37 aah 23:48:39 ehird, I know. And I'm making a parody on it 23:48:43 hmm 23:48:45 It is throttled, but bfjoust is only nice'd. 23:48:47 like if you are 1 and you dec, you become 0, then if they dec and you add, it's 0 again 23:48:48 AnMaster: that's like saying grapes are a parody of bread 23:48:49 oh, I remember how I was doing it with matador 23:48:49 ehird: if a program ends, it nops forever 23:48:50 ehird, I'm well aware it isn't exactly the same as this. 23:48:55 it was ++- for a while then --+ for a while 23:48:57 ais523: oh. 23:49:03 -!- coppro has joined. 23:49:04 same problem, less likely 23:49:12 there's nothing wrong with going in only one direction 23:49:18 switching direction is what causes a potential loss 23:49:19 ehird, don't be daft. They are a parody of apples. 23:49:46 apples are, admittedly, a parody of bred 23:49:48 bread* 23:49:50 argh, I need a type WARRIOR_A or WARRIOR_B. 23:50:08 ehird, bool 23:50:14 or something 23:50:14 nothx 23:50:15 -!- KingOfKarlsruhe has quit (Remote closed the connection). 23:50:19 typedef enum { 23:50:20 WARRIOR_A, 23:50:20 ehird, enum 23:50:21 WARRIOR_B, 23:50:23 yep 23:50:23 } warrior_id_t; 23:50:30 !bfjoust lazy ((+)*1024(-))*256(>)*9([-]>)*21 23:50:43 one has value 0, the other value 1 23:50:46 iirc 23:50:51 it doesn't like my program :) 23:50:57 it instantly told nescience what HIS program scored 23:50:59 favouritism! 23:51:10 Patashu: yours is probably longer than his 23:51:18 I thought it ran them sequentially 23:51:20 it's parallel? 23:51:20 either in time to execute, or in physical length 23:51:28 and it is sequential 23:51:35 but it has to run against all the programs there 23:51:42 some of which are slow against various other strategies 23:51:43 GregorR, there still? 23:51:49 i wish you guys would at least wait for the table to update before submitting revisions :\ 23:52:00 so you wish I would 23:52:02 GregorR, can you fix EgoBot so it doesn't stall every other interpreter if bfjoust runs 23:52:12 GregorR, properly rate limit, like other interpreters 23:52:40 I guess this is due to it being implemented as a special command 23:52:43 "the reward for a job well done is two more jobs..." 23:53:12 i wish you guys would at least wait for the table to update before submitting revisions :\ <-- agreed 23:53:34 Score for KingOfKarlsruhe_simple: 11.8 23:53:34 Score for Patashu_lazy: 11.8 23:53:34 Score for Patashu_lazy: 11.8 23:53:34 Score for Patashu_lazy: 11.8 23:53:38 or that someone fixed the bug with the table being cleared before the new results were read 23:53:40 oop here we go 23:53:40 ready* 23:54:01 Patashu, silly to resubmit it 23:54:10 lazy gets lots of ties :0 23:54:20 beats...fool4 and shade 23:54:57 fool4 is moving up the rankings... 23:55:08 and I don't know why 8 is doing so badly, I guess it's tape-length dependent 23:55:15 6 will beat it on a short tape, it'll beat 6 on a long tape 23:55:17 it still doesn't beat defend 23:55:19 i do not get this 23:55:45 nescience: what is simpletest? 23:55:45 .bfjoust is a very long extension... 23:55:50 you haven't pasted the source in-channel 23:55:54 AnMaster: ehird told be not to use .bj 23:56:01 ais523, why 23:56:03 blowjob? 23:56:05 :) 23:56:09 what Patashu said 23:56:12 Patashu, you have a dirty mind 23:56:21 ehird: my friends use it as an abbreviation for "bonjour" 23:56:23 AnMaster: and LOL means Lots of Love. 23:56:23 i did above 23:56:25 ais523: hahaha 23:56:28 "hi there! bj!" 23:56:29 (>[[-]])*29 23:56:34 ehird, it does? 23:56:42 AnMaster: Yep. And BJ means BF Joust. 23:56:55 i added a . in case it was some weird parser error 23:56:59 ehird, "bj" doesn't mean anything specific to me 23:57:25 * ehird thinks of a way to run some code twice in C without a temp var 23:57:50 ehird, repeat the statement twice? :P 23:57:58 AnMaster: more than one 23:58:17 -!- EgoBot has quit (Remote closed the connection). 23:58:20 -!- EgoBot has joined. 23:58:30 ehird, um. Like foo(); bar(); quux(); foo(); bar(); quux(); ? 23:58:34 macro 23:58:54 #define REPEAT_SHIT foo(); bar(); quux(); and then REPEAT_SHIT REPEAT_SHIT 23:58:54 AnMaster: like switch(){gigantic} 23:59:00 ah 23:59:00 nescience: it loses to defend6 the same way defend6 beats anything else with [-] in it 23:59:04 that could be messy in a macro 23:59:22 ehird, inline function? 23:59:32 meh :) 23:59:51 ehird, not sure if it would affect speed anyway...