00:07:48 -!- Pthing has quit (Remote closed the connection). 00:14:14 -!- FireFly has quit ("Later"). 00:16:13 Flot;;; 00:21:04 -!- Batmanifestdesti has joined. 00:21:45 * Batmanifestdesti made a program in Piet that prints "Hi!" :D 00:24:28 -!- BeholdMyGlory has quit (Remote closed the connection). 00:33:32 -!- oklokok has quit (Read error: 110 (Connection timed out)). 00:35:45 Man, this place is hppin' tonight 00:35:51 hoppin'* 00:36:09 totally 00:36:27 oh yeah, I got the go-ahead to make Piet forums! 00:36:28 :D 00:36:33 piet.forumotion.com 00:36:43 Piet forums. 00:36:49 yup 00:36:52 You really are a LOLCoder at heart. 00:36:56 That's the dumbest idea I've heard today. 00:36:58 And I hear a lot of dumb ideas. 00:37:01 how's that? 00:37:15 it's a lot like photobucket, only the pictures are programs! :D 00:37:40 (a) forums suck, (b) the potential userbase is about 3 people, (c) "general discussion"? fuckin' serious? why does everything have a useless forum? (d) that analogy is as bad as xkcd's tvtropes/rickrolling one. 00:37:45 Get off my lawn. Damn kids. 00:38:04 Oops, I just broke your rule #1. 00:38:08 how do forums suck, and chats not? 00:38:11 (Either once or twice depending on how you define swearing...) 00:38:42 Batmanifestdesti: they rarely conduct information and are mostly vents for useless crap. ok, so is IRC, but it's realtime, where such crap is expected 00:39:23 11. No demeaning other ideas/beliefs/religions: If you're a televangelist that wants everyone to be saved, then that's jolly for you, but don't go around telling people that they're going to burn in the fiery pits just because that are a Muslim, or Australian, or something else. This rule is related to flaming. Please don't confuse this with "don't post about anything related to your religion or lack thereof," because sometimes these conversations can be 00:39:25 fun and enlightening, all that this means is that if John or Jane Doe mentions that they're Mormon, don't send angry posts demanding to know why they joined that "devil church." 00:39:29 i can totally see people evangelizing on forums about Piet. 00:39:30 totally 00:39:45 it happens everywherew 00:39:46 [if I'm offending you, take comfort in the fact that I'm the grumpiest person in here.] 00:40:13 on a forum about DOOM modding, a guy demanded to know why mormons, muslims, and buddhists existed 00:40:39 ... 00:40:54 Because God made them in his image! 00:40:59 -!- upyr[emacs] has quit (Remote closed the connection). 00:41:03 God has some serious cognitive dissonance goin' on. 00:41:10 which just goes to show you that religious or ideological discrimination happens everywhere 00:41:35 What about me? I wantonly bash every religious ideology that exists. 00:41:39 Is it okay if I'm fair and balanced? 00:41:41 no, not that 00:41:49 I was referencing the doom thing 00:41:50 I feel rejected. 00:41:56 I know. 00:42:00 I wasn't replying. 00:42:04 oh 01:12:13 did you know that there's a channel for sudoku? XD 01:12:20 it's empty right now 01:12:54 Good. 01:12:57 It deserves to be. 01:13:05 lol 02:34:44 -!- dbc has quit (Client Quit). 02:37:38 -!- dbc has joined. 02:37:46 hi! 03:00:31 -!- upyr[emacs] has joined. 03:00:36 hi! 03:08:38 hi 03:09:31 -!- sgeo has changed nick to Sgeo. 03:19:02 -!- upyr[emacs] has quit (Remote closed the connection). 04:11:07 -!- augur has quit (Read error: 104 (Connection reset by peer)). 04:35:58 -!- augur has joined. 04:47:40 hi augur 04:47:49 hello Batmanifestdesti 05:06:23 -!- pikhq has joined. 05:08:13 -!- pikhq has quit (Client Quit). 05:08:51 -!- pikhq has joined. 05:19:27 hi again 05:28:02 * Batmanifestdesti is talking to a french person 05:29:07 Is it me? 05:29:44 depends on if you're on omegle 05:30:00 No. 05:30:07 oh, ok 05:32:53 * Sgeo tried Omegle once 05:32:58 Wish there was a way to get back in touch 05:33:04 it's being overrun by bots right now 05:33:06 I don't like losing touch with people, ever 05:33:09 Batmanifestdesti, o.O 05:33:31 almost every time I get on, I get bombarded by porn ads 06:25:05 aww 06:32:26 i once tried copy and pasting one omegle conversation to another, as some sort of omegle eavesdropping thing 06:32:39 took them about 3 minutes to start cybering 06:49:40 AnMaster: Yes, y=0 is indeed unreasonable; I think it should work (at least) for any x >= 0, y >= 1. Though you get overflowy problems if that (x+y-1) happens to be over your max-value, which won't happen for the divide-modulo-if thing. 06:55:24 -!- coppro has quit ("The only thing I know is that I know nothing"). 06:55:29 -!- CESSMASTER has quit ("Computer has gone to sleep"). 06:56:04 -!- CESSMASTER has joined. 07:00:10 -!- Sgeo has quit (Remote closed the connection). 07:04:53 -!- coppro has joined. 07:04:59 hi! 07:24:04 Hi 07:24:04 -!- augur has quit (Read error: 104 (Connection reset by peer)). 07:34:59 -!- augur has joined. 07:40:10 fizziew, overflow may be a real issue here. 07:40:23 so I'm sticking to the variant I decided on. 07:43:03 Hmm, Omegle. 07:43:40 Omegle is full of lulz 07:44:34 bbl 07:59:59 -!- clog has quit (ended). 08:00:00 -!- clog has joined. 08:05:26 -!- Batmanifestdesti has left (?). 08:29:06 -!- MigoMipo has joined. 08:51:59 -!- augur has quit (Remote closed the connection). 09:03:52 -!- FireFly has joined. 09:05:10 Well, that was fun. 09:06:06 Went on Omegle. The first guy I met was a Jewish guy in Israel; he told me about his experiences in the army and stuff like that. 09:06:16 It appears that we talked for over an hour. 09:06:46 We totaled about 400 lines. 09:10:52 -!- coppro has quit ("The only thing I know is that I know nothing"). 09:26:40 -!- atrapado has joined. 09:28:17 -!- GreaseMonkey has joined. 09:37:49 Incidentally, even the "a = x/y; b = x%y; return b ? a+1 : a;" you can write as "a = x/y; b = x-a*y; return b ? a+1 : a;" which has one division (well, modulo) switched to multiplication. Though of course optimizations; GCC 4.3 at least seems to be clever enough to get both the x/y and x%y from the same idiv opcode when optimizing. (It even does something extra-clever there.) 09:39:52 -!- upyr[emacs] has joined. 09:40:18 -!- BeholdMyGlory has joined. 09:43:08 why do some (all?) TFT monitors "buzz" with a high pitch when the brightness setting isn't at 100? 09:44:20 fizzie, except as I said I was using C syntax just for clarity. It is a lot simpler than trying to ask it in befunge syntax :P 09:45:52 even if you know befunge, not everyone do in the channel. Asking in a more "mainstream" language syntax when the question isn't directly related to the esolang in question seems to work best IME. 09:46:32 Oh. Though in that case the time spent in the additional "-" operation and moving the IP around and instruction-fetching might eat any speed gains you could concievably get by changing a / into a *. 09:47:01 fizzie, the fact that it is befunge is bound to slow it down (unless you start working on jitfunge again! ;P) 09:48:12 Ngh. Not before the things-with-deadlines are over, at least. 09:48:19 of course 09:48:41 fizzie, so around the age of 65 then? ;P 09:49:57 Maybe after I've gotten that master's thesis done. (So... Sep/Oct hopefully.) 09:50:28 fizzie, what do you plan to do after that otherwise? Continue studying or getting a job? 09:50:35 -!- coppro has joined. 09:53:08 No clue, really. I think I'll at least apply for one of those graduate-school funding opportunities; I've sort of been promised employment here at the university if I get one of those places. Maybe otherwise too, but it's a bit unclear. The project from whose budget my salary is paid ends this year. 09:53:33 hm ok 09:53:49 Of course a real job would pay real money, instead of the pittance we get here. 09:53:59 But then I'd have to do real work too. :p 09:59:49 -!- augur has joined. 10:14:37 holy crap I got digested 10:16:20 The buzz is caused by audio-frequcy-range chopped electrical signals being converted somehow to sounds. 10:17:23 Those signals also supply power so they are rather powerful. 10:22:05 Not the worst device design screwup I have heard about... 10:27:38 The mechanism that dimming is usually implemented is that power electronics are used to chop the power supply and then the signal is lowpass-filtered somewhere. Unfortunately, the signal-to-sound conversion can happen before lowpass filtering or the lowpass filtering might be less than perfect. 10:27:38 -!- augur has quit (Read error: 104 (Connection reset by peer)). 10:32:22 -!- augur has joined. 10:36:01 -!- KingOfKarlsruhe has joined. 10:40:13 -!- augur has quit ("Leaving..."). 10:48:38 !befunge98 ffff***:*.a,@ 10:48:41 -1732076671 10:49:04 err 10:49:05 what? 10:49:22 oh 2^31-1 not 2^32... 10:49:28 Yes, it's signed. 10:49:37 why is it 32-bit btw hm 10:50:00 !befunge98 1y,a.@ 10:50:00 10 10:50:05 !befunge98 1y.a,@ 10:50:06 1 10:50:16 !befunge98 3y.a,@ 10:50:16 1128682830 10:50:46 cfunge indeed 10:50:58 !befunge98 2y.a,@ 10:50:58 4 10:51:12 !befunge98 4y.a,@ 10:51:13 41 10:51:35 0.4.1? 10:51:41 Guess so. 10:51:42 Deewiant, yes think so 10:51:52 which is last release too 10:51:52 Reasonably recent, then. 10:52:01 is this the befunge with the stack stack? 10:52:07 just missing a few revisions, including the CR correction bug. 10:52:12 coppro, yes 10:52:15 What do you return for 4y in bzr-built things? 10:52:16 :/ 10:52:25 s/correction bug/bug correction/ 10:52:52 fizzie, same, I have been too lazy to set up something in the build system to automatically find revision 10:54:50 Did you have a '93 compatibility mode in cfunge, or was that just efunge or something else? 10:55:09 -!- atrapado has quit ("Abandonando"). 10:55:39 fizzie, only in cfunge, but I never claimed it was very good. It only alters the semantics for spaces in strings. 10:55:42 to 93 ones 10:55:51 doesn't make instructions not in 93 reflect. 10:56:03 and I'm probably going to drop that mode. 10:56:31 !befunge98 aa*:*:*.a,@ 10:56:31 100000000 10:56:35 !befunge98 ee*:*:*.a,@ 10:56:36 1475789056 10:56:39 !befunge98 ef*:*:*.a,@ 10:56:39 1944810000 10:57:45 hm what is the best way to get 2^31-1 in befunge? 10:57:50 as in, shortest 10:57:50 Well, I don't think the 93 spec defined anything for undefined ops. 10:58:20 Put it in the playfield and use ...g for it, that's the shortest thing. 10:58:32 fizzie, right, but this is a one-time use so... 10:58:53 -!- KingOfKarlsruhe has quit (Remote closed the connection). 10:59:41 !befunge98 88*::*:**:1-+ 10:59:44 AnMaster: looking at the efunge code right now, I see you match [_|_] a lot - is that to avoid empty lists? 10:59:57 !befunge98 88*::*:**:1-+.a,@ 10:59:57 2147483647 11:00:03 That's 2^31-1. 11:00:11 I don't claim it's the shortest one, though. 11:00:13 coppro, in what specific place? 11:00:23 start() 11:00:48 coppro, Ah yes indeed, That is after all user provided data, better make sure it is reasonable :P 11:01:10 !befunge98 "PXIF"4(n284*1-R1-.a,@ 11:01:10 2147483647 11:01:14 * coppro starts poking around efunge, which has to be better-coded than his horrible Erlang messes 11:01:19 Deewiant, core instructions only :) 11:01:45 coppro, I was learning erlang as I coded it. So some parts are pretty bad 11:02:00 some of that code I wouldn't write today 11:02:00 well, you have -specs on yours, for starters 11:02:19 I was learning Erlang too, and I haven't actually gotten back to it. 11:02:21 coppro, yes, because I always preferred detecting bugs before I run the app. 11:03:37 fingerOpStacks()? 11:03:38 and if that isn't possible, at least catch them early on. Which might explain some matches like {_X,_Y} = Coord 11:03:39 and such 11:03:44 coppro, defined in some *.hrl 11:03:46 with -type 11:03:48 ah 11:03:49 I think 11:04:02 oh, yeah, that's where I'm looking 11:04:04 sorry 11:04:07 should be more specific 11:04:11 wondering what that is for 11:04:50 !befunge98 '@::*:**2*1-.a,@ 11:04:51 2147483647 11:04:53 Oh, and if you can assume ASCII, you can trivially replace the 88* with '@ to save one character... 11:05:07 Deewiant: I hates you! I had written that much and then you ruined it. 11:05:23 Oh, I didn't even notice your solution earlier 11:05:54 I use :1-+ instead of that 2*1- but it's otherwise pretty close. 11:05:55 Well, at least I'm being a bit more different in that I do 2*1- instead of :1-+ 11:05:59 NGH! 11:06:02 :-D 11:06:18 Stop being such a thought-vampire. 11:06:21 (Thought-pire?) 11:06:34 Thompire 11:07:05 heh 11:07:09 If you don't mind using a bit of nontextual data you can do '<256> 11:07:14 But I guess you do so I won't bother 11:07:56 Also <256> is tricky if you only have a non-locale-aware [0, 255]-bytes-as-input thing. 11:07:58 coppro, ah yes it is for fingerprint semantics stacks 11:08:06 A "thompire" sounds like one of those block-like things that drop from the ceiling and squish Mario, except edgier. 11:08:11 which is when you do like "PXIF"4( 11:08:15 Oh true, 256 is too big 11:08:18 Well, '<128> anyway 11:08:45 '<255>1+ might be better though, not sure 11:08:48 actually my input needs to be valid utf-8 as well as ASCII 11:09:05 so I'll go for '@::*:**2*1-.a,@ 11:09:05 Well, the '@ is rather nice. 11:09:13 Certainly it's more interesting than a silly 88*. 11:10:01 The :/* sequence in 8:*::*:**:1-+ is aesthetically pleasing, though. 11:10:20 !befunge98 "@@@@@@"***\+*.a,@ 11:10:21 1073745920 11:10:24 Darn 11:11:10 Oh, right, hmph. 11:11:28 !befunge98 "@@@@@@"+****1-.a,@ 11:11:28 2147483647 11:12:17 -!- coppro has quit (Remote closed the connection). 11:12:19 coppro, did that explanation help? I'm afraid you have to know funge-98 to understand fingerprints. 11:12:23 gah bad timing 11:12:32 -!- GreaseMonkey has quit ("HydraIRC -> http://www.hydrairc.org <- Nobody cares enough to cybersquat it"). 11:12:42 !befunge98 " @@@"+*:**1-.a,@ 11:12:43 2147483647 11:12:50 !befunge98 '~.a,@ 11:12:50 126 11:12:53 !befunge98 '@.a,@ 11:12:54 64 11:12:56 hm 11:13:12 !befunge98 '~2+:::***.a,@ 11:13:13 268435456 11:13:23 !befunge98 '~2+:::****.a,@ 11:13:23 0 11:13:26 !befunge98 '~2+::::****.a,@ 11:13:26 0 11:13:29 -!- coppro has joined. 11:13:30 err 11:13:35 wraparound? 11:13:41 !befunge98 '~2+:::::****.a,@ 11:13:42 0 11:13:47 !befunge98 '~2+:::::*****.a,@ 11:13:47 0 11:13:51 !befunge98 '~2+::**.a,@ 11:13:52 2097152 11:13:54 !befunge98 '~2+::::**.a,@ 11:13:54 2097152 11:13:56 !befunge98 '~2+::::****.a,@ 11:13:56 0 11:14:05 !befunge98 '~2+:::::****.a,@ 11:14:05 0 11:14:06 hm 11:14:12 Deewiant, that seems a bit strange to me 11:14:15 !befunge98 '~2+:::::******.a,@ 11:14:15 0 11:14:17 !befunge98 '~2+:::::::::::******.a,@ 11:14:17 0 11:14:48 !befunge98 '~2+.a,@ 11:14:48 128 11:15:20 > 128^5 `mod` 2^32 11:15:22 0 11:15:37 If you want a differently shaped version, you can use 11:15:39 8#v:*:< 11:15:39 >**:1-+ 11:15:56 It runs that :*: part both ways, which is always nice. 11:16:08 Use 'v, that would be nicer 11:16:13 Instead of an ugly # 11:16:29 (Not sure if it can be made nice though) 11:16:35 But 'v is a ugly number. Well, I guess I could try. 11:16:35 heh 11:16:44 fizzie, is 'v a number? 11:16:50 Of course it is 11:16:53 !befunge98 'v.a,@ 11:16:54 118 11:16:55 oh wait, right 11:16:56 duh 11:24:37 I don't think I can make a 118-based calculation very pretty. Though if you can enter the block using the v direction, you can do something like 11:24:39 8 11:24:39 [:*:< 11:24:39 >**:1-+ 11:24:43 No #jumps there. 11:25:33 Maybe for symmetricity the lower > should be a [ too. 11:27:54 hm what is the smartest way in befunge to check that a possibly unsorted list on the stack contains exactly a given set of elements. For example the stack might have 5 7 3 9 7 7 and I want to check that there is exactly those elements on it, but ignoring order of them. 11:30:54 for each element of list { for each element2 of set { if element == element2 goto Next } fail Next: } success 11:32:17 Deewiant, Not sure how to translate that into befunge... 11:33:03 Depends on how/where your list/set are stored 11:35:38 > ;if no more list elems, success; ;get list elem; > ;if no more set elems, fail; ;get set elem; -#v_ ;same - jump to beginning of this line; 11:35:40 : parse error on input `;' 11:35:53 The v goes to get another set elem 11:52:14 -!- coppro has quit (Read error: 110 (Connection timed out)). 11:59:01 Deewiant, the length of the list is known 11:59:16 (sorry got a phone call so was away) 12:00:14 basically I want to verify in a test suite for ATHR that a specific call to an instruction pushes a specific list on the stack, but where the ordering is unknown. 12:04:18 ah I have an idea. 12:05:40 somewhat like mycorand I think 12:08:33 Doing it as above works 12:08:50 Just fill in the blanks (most of it)... 12:11:24 Deewiant, I suspect what I have in mind results in less code. Not sure though. 12:11:58 Whatever works 12:12:19 Though it doesn't work for the multiset case; if the list is (7, 7, 7) and the multiset contains (7, 8, 9) it will succeed. I'm not sure if that was the case with the question. 12:12:58 fizzie, that should fail 12:13:52 5 7 3 9 7 7 is ok, so is for example 3 5 9 7 7 7 or 7 9 3 7 5 7, but not 7 9 3 5 or 7 7 9 3 5 12:14:08 nor 9 9 9 7 7 7 3 5 12:15:11 Oh, multiset and not a set 12:15:53 Then just remove the element from the multiset whenever you find it 12:16:12 my idea is to use j to create a jump table, then increment a counter in funge space for each hit, then after all elements are processed, check that the counters are correct. 12:18:23 !befunge98 02-;<;j @,a.2 ;@,a.1 12:18:28 um what 12:18:34 !help 12:18:34 help: General commands: !help, !info, !bf_txtgen. See also !help languages, !help userinterps. You can get help on some commands by typing !help . 12:18:47 !befunge98 01-;<;j @,a.2 ;@,a.1 12:18:51 !befunge98 03-;<;j @,a.2 ;@,a.1 12:18:51 2 12:19:02 Deewiant, does that make any sense to you? 12:19:24 (does mycology test negative arguments to j?) 12:19:31 (or is this a real bug?) 12:19:56 Sure; unless you hit exactly that < character, you're going to hit one of the @s before outputting anything. 12:20:12 fizzie, but shouldn't -2 to j hit the < 12:20:18 (that was my plan) 12:20:36 I guess it depends on where/when the delta is added. I don't remember the spec by heart. 12:21:18 fizzie, "Negative values are legal arguments for j, such that 04-j@ is an infinite loop." is all the spec says 12:22:04 "If there is a 1 on the stack, j will work like # does." In this case it sort of makes sense that you need a -3 to hit the <, i.e. the next instruction is k*delta from the position after the j, not the j itself. 12:22:09 !befunge98 04-ja,@ 12:22:24 (that differs with regards to wrapping yes) 12:22:34 (in fact, j when wrapping is quite a headache 12:22:36 ) 12:22:45 !befunge98 ;@,ax';06-j'ya,@ 12:22:55 !befunge98 ;@,ax'<;06-j'ya,@ 12:23:09 FATAL: Out of memory at [/home/egobot/egobot.hg/multibot_cmds/interps/cfunge/cfunge-src/src/stack.c:90]: 12:23:17 Deewiant: You're not actually printing that x, you know. 12:23:31 Just the newline. 12:23:49 Oops 12:23:55 !befunge98 ;@,,ax'<;06-j'ya,,@ 12:24:01 Blargh 12:24:06 Whatever :-P 12:24:45 I get 'x' here 12:24:49 after a newline 12:24:55 Asztal, using what interpreter? 12:24:59 mine 12:25:05 Asztal, which one is that? 12:25:12 !befunge98 ;@,a,x'<;06-j'ya,,@ 12:25:12 x 12:26:10 anyway a program with exactly 04-j@ does not loop forever in cfunge *debugs* 12:26:31 AnMaster: Mycology does test it. 12:26:39 AnMaster: it's called stinkhorn now. to be honest, it's not very good, but it should support j properly 12:26:42 Deewiant, I get all GOOD in mycology 12:26:56 Then you're doing something weird :-) 12:27:17 !befunge98 ;@,a,x'<;06-j'y,a,@ 12:27:17 x 12:27:22 That works, at least 12:27:23 Deewiant, it is related to wrapping 12:27:45 There's no wrapping in that code 12:28:11 Deewiant, no I meant in 04-j@ 12:29:42 Anyway, I'm not quite sure what the jump table is for in that stack-multiset-element thing. If your elements are small integers (like what you'd think is the case if you're using a jump table there) then you could directly use them as x offsets to increment counters in the playfield, then compare the row-with-counters to another row which holds the desired multiset in that representation. 12:30:06 fizzie, x offsets? 12:30:14 I guess you meant p 12:30:21 x-dimension offsets, for p/g. 12:30:26 yeah 12:32:02 AnMaster: There is no wrapping in that code. 12:32:52 Deewiant, sure the test in mycology isn't buggy in some way? 12:33:47 How would I know, haven't looked at it in months 12:34:05 I doubt it, given that other interpreters are happy with it 12:34:22 Deewiant, btw: http://pastebin.ca/1496943 12:35:02 Deewiant, oh and mycology says: 12:35:04 GOOD: 04-j jumps backward the right number of cells 12:35:15 08:43 AnMaster: why do some (all?) TFT monitors "buzz" with a high pitch when the brightness setting isn't at 100? 12:35:16 only low quality ones 12:35:34 Mycology's "04-j" test might not be at the left border of the file, though. 12:35:40 AnMaster: Like said, you do something weird. 12:35:55 09:14 coppro: holy crap I got digested 12:35:56 wat 12:35:59 fizzie, yes, as far as I understand j there crosses the border 12:36:22 AnMaster: No, it does not. 12:36:24 AnMaster: There is no wrapping in that code. 12:36:32 Deewiant, 01-j for example would as far as I understand hit the - right? 12:36:46 No, it would hit the j itself. 12:36:50 hm ok 12:36:53 Because "0jx" hits the x. 12:37:03 that explains a bit of what is happening 12:37:08 !befunge98 0j'a,a,@ 12:37:08 a 12:37:18 right 12:37:24 !befunge98 02-j'a,a,@ 12:37:24 a 12:38:07 i had a dream where AnMaster was talking about how Kucinich personally tutored him or something. 12:38:19 ehird, who is that? 12:38:38 Dennis Kucinich, the only left-wing politician in America. 12:38:42 (Modulo hyperbole.) 12:39:10 Deewiant, the code for j is http://pastebin.ca/1496948 btw 12:39:28 wrote that long ago 12:39:31 You might get wrapping in "04-j" if you implement j so that when the IP is still at j you add that -4*delta, in which case you would go to the position one left of the 0, then rely on the normal +delta movement from there. And cared about the borders in the -4*delta part. 12:39:35 * AnMaster wonder what broke with wrapping 12:40:01 fizzie, correct, the ip is moved forward normally after always. 12:41:14 fizzie, it doesn't match how other instructions work, if you move before actually executing the side effects of the instruction 12:41:17 hm 12:41:39 Well, what does your ip_forward do when x=3, delta.x=-4, and the left border of the playfield is at x=0? 12:41:44 The spec is explicit in this case, though. 12:42:10 -!- sgeo has joined. 12:42:12 I'm changing it atm indeed 12:43:07 fizzie, hm... Don't ask me such tricky questions. But when delta isn't cardinal, wrapping is implemented as the wrapping algorithm in the spec (that is, reverse ip, go on to other edge, reverse again) 12:43:20 and -4 isn't cardinal 12:47:17 ok I changed behaviour, and now it gets into some infinite loop *sometimes* after "GOOD: z doesn't reflect", but not on all runs. Huh 12:48:10 -!- Judofyr has joined. 12:49:05 So... what I think happens there with "04-j@" is that when you call ip_forward with x=3, delta.x=-4, you'd be attempting to pass to the empty space and would do the backtrack-warping; setting delta.x=4 temporarily, and starting from x=3. Then you're already at the other edge (only whitespace from x=7 and so on) so you stop with the backtrack there. After that your j resets the delta to x=1, and execution proceeds to the @. 12:49:47 fizzie, quite likely 12:51:18 but the new behaviour of failing sometimes is rather weird. Especially since ? hasn't been used at all by that point. Valgrind reports no issues either. 12:57:56 -!- upyr[emacs] has quit (Read error: 54 (Connection reset by peer)). 12:58:40 -!- upyr[emacs] has joined. 12:59:30 oh found it 13:00:02 -!- MigoMipo has quit (Read error: 104 (Connection reset by peer)). 13:01:15 -!- Pthing has joined. 13:17:37 http://www.youtube.com/watch?v=cH6r2tIaRXU // needs hooking up to a visor 13:31:15 to send to ais523 when he comes on: wikipedians are arguing over whether to include the public domain rorschach test images on the article ... because it would make it harder for psychologists to use them, or something. http://en.wikipedia.org/wiki/Talk:Rorschach_test 13:32:05 meanwhile, http://pouletvous.files.wordpress.com/2009/04/rorschach_inkblots.jpg 13:34:55 not as though the test works any more than say, astrology :) 13:35:44 http://truthsilo.com/stds/rorschach.htm 13:37:57 -!- MigoMipo has joined. 13:40:13 "mentioning more than four sex images in the ten plates is diagnostic of schizophrenia" 13:40:36 #scheme people had that bot which understood "later tell [some interesting things]", which would then say ": [some interesting things]" at the point when joined or next said something. 13:42:55 so do we 13:43:08 lambdabot: tell ais523 wikipedians are arguing over whether to include the public domain rorschach test images on the article ... because it would make it harder for psychologists to use them, or something. http://en.wikipedia.org/wiki/Talk:Rorschach_test 13:43:12 @help 13:43:12 help . Ask for help for