00:00:04 I also suspect it wouldn't be capable of enchanting anything 00:00:36 if it were an artifact enchantment - equipment, I guess it'd be possible to use the enchant ability to equip it to something as it was cast? not sure though 00:01:38 -!- earenndil has changed nick to Elronnd. 00:03:06 ais523: If it wasn't an Aura? Why would that be possible? 00:03:46 the rules for something that's both aura and equipment (which is possible!) were changed recently 00:03:57 to make equipping and enchanting basically synonyms 00:04:05 in order to get around some rules issues 00:04:05 But only an aura can target. 00:04:18 Oh, you said both aura and equipment. 00:06:35 yes 00:07:29 How is it possible? 00:09:07 -!- adu has joined. 00:10:32 liquimetal coating + bludgeon brawl + something that's naturally an Aura 00:11:29 Oh, but that only works when it's a permanent. 00:11:33 I thought you meant as a spell. 00:11:55 Or e.g. as a card you return from the graveyard. 00:12:26 ah right, I don't think you can do that 00:12:40 yet, at least 00:14:00 Are there any cards with effects like that? 00:15:20 allowing cards to be cast as another card type? not as far as I know 00:15:36 err, subtype, at least 00:15:40 not sure if enchanted evening works on spells 00:15:43 it works on most things 00:16:12 We should have an mtg card bot here. 00:17:04 we should 00:17:07 does Freenode have one? 00:17:18 I'm not sure how you can tell, possibly by guessing the name 00:17:44 EFNet has one. 00:19:04 I've asked in #mtg 00:20:25 /ne 00:21:04 -!- tromp_ has joined. 00:21:28 helronnd 00:22:06 Elronnd: How are you today? 00:22:27 hey hppavilion[1] 00:22:28 alright 00:22:35 Elronnd: Excellent 00:22:39 My feet are kinda messed up from dancing last week 00:22:42 Ouch 00:22:48 I'm trying not to put too much weight on them 00:23:13 What would be a good, strange basis for a declarative language? 00:23:38 I'm afraid I don't know what a "declarative language" is 00:23:50 ^prefixes 00:23:50 Bot prefixes: fungot ^, HackEgo `, EgoBot !, lambdabot @ or ?, thutubot +, metasepia ~, idris-bot ( , jconn ) , blsqbot ! 00:23:58 Elronnd: Like Haskell or Prolog 00:24:06 As opposed to Python or C, which are imperative 00:24:58 Haskell isn't high enough on my list of things to look at that I've looked at it yet 00:25:04 Elronnd: Ah 00:25:26 Elronnd: The classic explanation is that in declarative programming, you specify /what/ to do instead of /how/ to do it 00:25:33 Elronnd: Thue is declarative 00:25:38 Elronnd: Brainfuck is imperative 00:25:47 From what I've heard, it's really cool thougyhthough 00:26:16 Elronnd: It is 00:27:40 Elronnd: Declarative languages are usually based on something mathematical- Haskell is λ-calculus, Prolog is formal logic, Thue is semi-thue systems. 00:28:00 Ah, okay 00:28:04 The new colorless mana symbol is scow. 00:28:21 `scow 00:28:23 /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: scow: not found 00:28:58 Elronnd: So what I was asking is what would be an interesting thing to do for a declarative language 00:29:40 Use math-style notation 00:30:07 Elronnd: Perhaps 00:31:12 -!- \oren\ has quit (Ping timeout: 265 seconds). 00:31:32 -!- tromp_ has quit (Remote host closed the connection). 00:32:00 Elronnd: I did think of a language based on the idea of a really complicated calculator 00:32:34 -!- \oren\ has joined. 00:32:55 Elronnd: stdout<<="Hello, world!" 00:37:48 Realistically, in math, that would just be "Hello, world!" 00:37:57 not in a calculator, though 00:38:58 Elronnd: I'm stuck on how to allow things like 99 bottles of beer on the wall without variables 00:39:18 The EFNet bot gives both cards and rules. 00:39:37 And can look up rules by keyword and so on. 00:39:45 But the code is not available. 00:40:06 hppavilion[1]: recursion 00:40:16 -!- haily_ has joined. 00:40:18 Elronnd: No functions are in the language 00:40:41 Elronnd: That's the strange part; a program in the language is just a big expression 00:40:57 that's math for ya 00:41:07 Hi 00:41:45 `welcome haily_ 00:41:46 haily_: Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: . (For the other kind of esoterica, try #esoteric on EFnet or DALnet.) 00:42:44 hi 00:42:47 `relcome yaily_ 00:42:50 yaily_: Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: . (For the other kind of esoterica, try #esoteric on EFnet or DALnet.) 00:42:57 My name is Haily? 00:43:05 I m a nerd 00:43:19 but now I go to retire 00:43:23 cause Im old 00:43:36 -!- rodgort has quit (Quit: Leaving). 00:43:54 Mine Backbone looks like 00:44:09 some bombs in Irakwar 00:44:25 broken but never fall? 00:47:51 -!- elias1 has joined. 00:48:18 -!- haily_ has quit (Quit: :-)). 00:48:28 <\oren\> `welcome elias1 00:48:30 elias1: Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: . (For the other kind of esoterica, try #esoteric on EFnet or DALnet.) 00:49:38 -!- lambda-11235 has joined. 00:49:44 -!- elias1 has quit (Read error: Connection reset by peer). 00:55:06 -!- rodgort has joined. 01:03:11 -!- tromp_ has joined. 01:04:26 -!- tromp_ has quit (Remote host closed the connection). 01:06:08 -!- idris-bot has quit (Ping timeout: 256 seconds). 01:06:36 -!- tromp_ has joined. 01:08:02 -!- Melvar has quit (Ping timeout: 250 seconds). 01:09:46 -!- Melvar has joined. 01:11:02 -!- oerjan has joined. 01:12:20 hppavilion[1]: that's easy 01:13:05 hppavilion[1]: (map display-bottles (iota 99)) 01:13:53 (iota 99) is ((99 s) k) 01:14:07 I don't want to mentally figure out what giving 99 s as an argument even does 01:14:59 ais523: does that return a list with 1 2 3 ... 01:15:19 no, I'm misinterpreting Iota as the esolang (and its defining operation) 01:15:32 -!- Sprocklem has joined. 01:15:32 (map display-bottles (list-reverse (iota 99))) 01:15:36 and thus 99 as the church numeral (which would be the only sensible way to define it in esolang) 01:15:38 I totally got it wrong 01:15:40 *in iota 01:15:52 (iota 99) is unsurprisingly ι99 in APL 01:15:52 http://srfi.schemers.org/srfi-1/srfi-1.html#iota 01:16:15 except it's probably a custom iota character in APL, and not the greek one I used 01:16:46 ooooo or (iota 0 99 -1) 01:17:13 Don't you mean (iota 99 99 -1) ? 01:17:18 maybe 01:17:43 I've never used all three arguments before 01:18:05 Underlambda has a "u" command 01:18:30 if you give it an integer as an argument, it produces a list from 1, 2, etc., up to that integer (inclusive) 01:18:53 it also works on non-integers, in which case its behaviour is defined but mostly bizarre 01:20:15 hmm 01:21:09 for example, (~)u^ will append the length of a list, plus one, to that list 01:21:16 causing the list to end with its new length 01:33:46 -!- Tod-Autojoined has changed nick to TodPunk. 01:36:50 Is Ruler-and-SuperCompass construction any more powerful than normal ruler-and-compass? 01:37:12 what's a supercompass? 01:37:31 izabera: It draws ellipses instead of circles 01:37:39 (well, circles are a subset of ellipses, but...) 01:37:48 then no 01:38:22 izabera: Proof? 01:39:08 still algebraic, you don't ever get to pi 01:39:31 Ah 01:40:01 izabera: But perhaps you're able to move a circle with it or something, thus letting you trisect an angle? 01:40:36 your specs are a bit vague 01:41:21 izabera: You can draw an ellipse given any two points representing its focci. The focci may, of course, be the same 01:42:07 izabera: It's not a a tool one could easily construct to /look/ like a compass, of course. 01:42:24 what about moving things around? 01:43:22 izabera: That part I'm not sure about, but I just feel that it may be possible to use the ellipse to move a circle; I haven't formally tried it yet because I don't have any way to try it 01:43:35 I think I'll make a Tkinter-based application to try it out 01:43:40 (Or maybe Qt5) 01:43:47 try geogebra 01:44:42 \oren\: https://pbs.twimg.com/media/CbWvGzXUAAAy8MO.png GSUB at work! 01:50:50 izabera: Actually, that seems like a pretty good idea... 01:50:57 A S-construction tool 01:51:03 For some set S 01:51:37 e.g. Folding-and-compass or Ruler-and-protractor 01:52:07 (protractor is restricted so you can only make angles. You use it to create angles of a given size) 01:52:25 (or something less stupid) 01:52:47 you can trisect an angle with a ruler and a compass 01:52:51 the reason is that the ruler has lines marked on it 01:52:57 and even two marked lines is enough for the trisection to work 01:53:21 (another method you can do with just the ruler and compass is to use them at the same time, using the compass to mark a length along the ruler; the proof that you can't regards doing this as cheating) 01:53:46 err, s/ruler/straightedge/g in my most recent comment 01:54:02 Since some people use HackEgo in /msg, where the limit might be shorter. <-- istr HackEgo has an internal 350 limit before all that stuff is added on hth 01:54:16 oerjan: tdh 01:54:19 unexpectedly 01:59:30 `perl -e for ($i=0;$i<500;$i++) { print ($i%10), ; } 01:59:32 01234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789 02:00:05 yep, checks out 02:01:34 the amazing power of perl 02:02:59 `` printf %.s0123456789 {1..50} 02:03:00 01234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789 02:03:07 what did i win 02:03:53 izabera: I recently had to debug a program in an emergency, it was doing a series of parse tree transformations and I wanted to spot the point where they became incorrect 02:04:11 ok 02:04:12 it was dumping each stage but in a pretty sexp-like format (not actual sexps, but close) 02:04:25 and trying to wade through all the parens was basically impossible 02:04:33 so I wrote a Perl-oneliner to colourize matching parens to make it easier 02:04:42 (first I checked the repos but there didn't seem to be anything there) 02:04:48 neat 02:05:53 and where is this program? 02:07:20 in my bash history 02:07:20 I can drag it out if you like 02:07:28 perl -pe 's/[()]/$& eq "(" ? "\e[3".($x%6+1).($x++%12>5?";1":"")."m(\e[0m" : "\e[3".(--$x%6+1).($x%12>5?";1":"")."m)\e[0m"/ge' 02:07:52 it's not very neat or readable because it was written in like 5 minutes in a hurry, I stopped writing it as soon as I had something that worked 02:07:54 `` echo {,,}{,,}{,,}{0,1,2,3,4,5,6,7,8,9} 02:07:55 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 02:07:55 for my purposes 02:08:21 -!- nisstyre has quit (Ping timeout: 265 seconds). 02:08:40 `` echo {,,}{,,,}{,,,}{0,1,2,3,4,5,6,7,8,9} | tr -d \ 02:08:42 01234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789 02:08:52 mixing escape sequences and perl makes for very esoteric programs 02:09:13 `` printf %s {,,}{,,,}{,,,}{0..9} 02:09:15 01234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789 02:09:24 izabera: thanks 02:12:10 @bf ++++++++[>++++++>++++++<<-]>++[>.+.+.+.+.+.+.+.+.+.---------<-] 02:12:10 012345678901234567890123456789012345678901234567890123456789012345678901234... 02:12:15 oh come on lambdabot 02:12:24 !bf ++++++++[>++++++>++++++<<-]>++[>.+.+.+.+.+.+.+.+.+.---------<-] 02:12:24 01234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456 02:13:02 does it print too many? 02:13:22 since when did lambdabot do brainfuck? 02:13:46 ^bf >,[>,]<[[<]>[.>]<]!0123456789 02:13:46 012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456 ... 02:14:08 that's very cheating and unfair 02:14:26 I wasn't intending to cheat, more it was out of curiosity if I could find something that didn't hardcode the 0123456789 bit 02:14:37 to avoid the non-general .+.+.+.+ 02:14:43 ais523: It also does unlambda. 02:14:47 I wonder if it'd be shorter if it didn't use stdin; probably not 02:14:53 -!- TodPunk has changed nick to Tod-Home. 02:15:07 (shorter than the +.+.+.+. version, I mean) 02:15:22 oh i see, HackEgo was truncating it 02:15:47 !bf ++++++++++[>+++++>+<<-]>--[>[<.+>>+<-]>[<<->+>-]<<] 02:15:48 01234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456 02:16:14 Also D&D. 02:16:17 int-e: wow 02:16:20 o.o 02:16:21 @roll 2d0 02:16:21 unexpected 'd': expecting digit, operator or end of input 02:16:30 @roll d20 02:16:30 lambda-11235: 20 02:16:35 til there's a bf interpreter 02:16:38 i'll need a minute or two to understand how that works 02:16:46 @roll 2d20 02:16:46 Elronnd: 13+10 => 23 02:16:51 it took me a while to understand how it works but I understand itn ow 02:17:08 think of it like a Minsky machine, it makes things easier 02:17:13 (note that all the loops are balanced) 02:17:57 ais523: what took you a while to understand? 02:18:09 Elronnd: why it was repeating in sets of 10 02:18:19 izabera: it's an infinite loop, so cheating 02:18:27 ais523: ah 02:18:31 int-e: oh, I thought infinite was better 02:18:36 mine's an infinite loop too 02:19:21 ais523: but izabera's isn't, and spends 11 characters on that 02:19:37 thanks for noticing <.< 02:19:43 !bf ++++++++[>++++++<-]>++[.+.+.+.+.+.+.+.+.+.---------] 02:19:43 23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;23456789:;2345678 02:19:59 I wasn't treating this as a golf competition, more a code elegance competition 02:20:04 err, 13. 02:20:07 !bf ++++++++[>++++++<-]>[.+.+.+.+.+.+.+.+.+.---------] 02:20:08 01234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456 02:20:09 also I'm translating int-e's BF program into PMMN to see what my PMMN optimizer makes of it 02:23:58 what does it optimize? 02:24:01 apparently it segfaults 02:24:12 !bf ++++++++++[[>+++++>+>+<<<-]>-->[<.+>-]>] 02:24:12 01234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456 02:24:14 izabera: a lot of things, it tries to eliminate loops for example 02:24:17 -!- Phantom_Hoover has quit (Read error: Connection reset by peer). 02:24:30 int-e: *that* will take a while o.o 02:24:58 -!- andrew_ has joined. 02:25:11 since when did lambdabot do brainfuck? <-- since always? 02:27:01 -!- Tod-Home has quit (Quit: This is me, signing off. Probably rebooting or something.). 02:27:53 More precisely, since september 2006 02:28:08 http://sprunge.us/ghMc 02:28:34 inc_by(0, 10); while (dec(0)) { inc_by(1, 5); inc(2); } dec(1); dec(1); while (dec(1)) { inc(1); while (dec(2)) { /* PMMN non-destructive output code, not in original */ while (dec(1)) { inc(9); inc(8); } while (dec(9)) { inc(1); } inc(8); output(8); inc(1); inc(3); } while (dec(3)) { dec(1); inc(2); } } 02:28:39 that's the original program 02:29:00 oerjan: I was going to say that you of all people should know the answer to that. 02:29:04 But I was thinking of unlambda. 02:29:08 hmm, it didn't optimize out a useless jump 02:29:24 int-e: oh, I thought infinite was better <-- i only did finite because i vaguely recalled HackEgo might not handle infinite output without newlines. 02:29:48 oerjan: actually lambdabot seems to have trouble with that 02:29:55 @bf ++++++++++[[>+++++>+>+<<<-]>-->[<.+>-]>] 02:30:24 it looks like it doesn't figure out yet that (1) never changes 02:30:40 however, it has got the program down to the minimum number of loops without unrolling 02:31:41 fwiw, the PMMN optimizer so far is available via "darcs clone http://nethack4.org/media/ssapmmn" 02:31:57 nothing yet understands its output format though, you just have to read it by hand 02:32:16 @bf ++++++++++[[>+++++>+>+<<<-]>-->[<.+>-]>.] 02:32:28 hmm, no, newlines don't help. 02:32:36 !bf ++++++++++[[>+++++>+>+<<<-]>-->[<.+>-]>.] 02:32:37 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 \ 0123456789 02:32:41 But I was thinking of unlambda. <-- yep. although they didn't actually _tell_ me they'd used my code, so i'm not sure exactly when i discovered it. 02:33:27 oerjan: 2006-03-15 Add the unlambda plugin 02:33:47 it's a bit weird since i'm pretty sure there were other unlambda interpreters in haskell when i wrote mine 02:34:31 Plugin `bf' failed with: <> 02:35:08 I had one in 2000, but probably it wasn't public... since I really wanted to write one in C. 02:35:39 heh 02:36:00 by 2000 hadn't even learned haskell 02:36:04 * history: * 13-10-2000: initial version, in Haskell * 14-10-2000: initial version in C 02:36:10 good times 02:36:20 *+i 02:36:31 i think march 2006 was just before the end of my big internet break 02:36:59 -!- nisstyre has joined. 02:37:17 anyway, sleep. 02:37:50 or maybe it was the year i did learn it. not before, anyway. 02:38:27 anyway, dishes 02:41:16 `` bc <<< 2^1234-1 | tr -d \\n\\\\ 02:41:19 29581122460809862906004469571610359078633968713537299223955620705065735079623892426105383724837805018644364775907095599312082089933038176093702721248284094494136211066544377518349572681192920386118201521832389207735598339319120892886765265599360248790311370854940266862452110061179427034023276609931709804888749380902312739825386061877261903500988327 02:41:21 factor it 02:41:25 -!- AlexR42 has joined. 02:42:23 hackego cut off part of it 02:42:33 (2^1234-1) = ((2^617)^2) - (1^2) = (2^617-1)(2^617+1) 02:42:49 it probably factors further though 02:44:17 it's unfair, my computer didn't know about that 02:44:42 oerjan: I printed out R5RS and Haskell98 in 2000 for my summer reading 02:44:48 u always cheating >:( 02:45:17 adu: how about harry potter 02:45:26 izabera: I've never read it 02:45:43 * izabera deletes adu from her friends on myspace 02:45:55 too busy reading user manuals 02:46:08 izabera: this is the reason that mersenne primes always have a prime in the exponent 02:46:52 come to think of it it /definitely/ factors further; one of (2^617-1), (2^617+1), and 2^617 has got to be divisible by 3 02:46:55 and it's not going to be 2^617 02:47:42 wtf 02:48:05 2^617 is not divisible by 3 02:48:26 yes 02:48:31 that's what he said 02:49:21 izabera: I haven't been on myspace in 15 years 02:49:33 i've never been on myspace 02:50:58 -!- AlexR42 has quit (Quit: My Mac has gone to sleep. ZZZzzz…). 02:51:54 -!- AlexR42 has joined. 02:57:36 -!- AlexR42 has quit (Quit: Textual IRC Client: www.textualapp.com). 03:07:24 -!- adu has quit (Quit: adu). 03:07:52 ais523: also, because 2 divides 1234 03:07:59 `factor 1234 03:08:02 1234: 2 617 03:08:17 i guess that's it. is 2^617-1 a prime? 03:08:48 2^617+1 probably isn't. 03:09:03 > (2^617+1)`mod` 3 03:09:13 mueval-core: Time limit exceeded 03:09:19 oh come on, it's not that big 03:09:21 > (2^617+1)`mod` 3 03:09:50 lambdabot must be overworked 03:10:34 2 == -1 (mod 3), so 2^617 == -1 (mod 3). so it's 2^617+1. 03:12:22 2^617-1 is not on the list of known mersenne primes, so it's presumably checked and found not. 03:12:37 @bot 03:12:37 :) 03:12:43 huh 03:12:55 lambdabot never answered back on the last one 03:13:05 > (2^617-1)`mod` 5 03:13:20 > "hi" 03:13:27 @bot 03:13:37 :) 03:13:50 there's nfw lambdabot should have trouble calculating that 03:14:01 :t (2^617-1)`mod` 5 03:14:19 and int-e went to bed 03:14:39 `` bc <<< 2^617-1 03:14:40 54388530464436950905813832350972787438550335255248068935623079751721\ \ 32452975126965649024023195947885249426733939164170397148972417563722\ \ 13155348458256985448390483221335442656288489603071 03:14:55 `` bc <<< 2^617-1 | tail -1 03:14:56 13155348458256985448390483221335442656288489603071 03:14:58 > (2^617-1)`mod` 5 03:15:17 ok not divisible by 5 03:15:36 > "hi" 03:15:50 it seems like the @run plugin is having trouble. 03:15:50 mueval-core: Time limit exceeded 03:16:02 > "hi" 03:16:10 -!- adu has joined. 03:16:59 `` bc <<< (2^617)%5 03:17:00 /hackenv/bin/`: eval: line 4: syntax error near unexpected token `(' \ /hackenv/bin/`: eval: line 4: `bc <<< (2^617)%5' 03:17:04 damn 03:17:28 int-e: what is the date format MM-DD-YYYY? 03:18:06 oerjan: No good, even haskell's lambdabot is having trouble. 03:18:20 * #haskell 03:18:33 lambda-11235: um they're the same bot hth 03:18:56 `` bc <<< '(2^617)%5' 03:18:57 2 03:19:04 `` bc <<< 'scale' 03:19:05 0 03:19:07 good 03:19:11 `` bc <<< '(2^617)%7' 03:19:12 4 03:19:28 hm 7 is a mersenne prime so no point in checking it 03:19:32 `` bc <<< '(2^617)%11' 03:19:33 7 03:19:36 `` bc <<< '(2^617)%13' 03:19:37 6 03:19:43 `` bc <<< '(2^617)%17' 03:19:44 2 03:19:47 `` bc <<< '(2^617)%19' 03:19:47 13 03:20:24 gah loops in bash... 03:20:35 or bc, for that matter 03:21:22 `` bc << '2^2^2^2' 03:21:24 /hackenv/bin/`: line 4: warning: here-document at line 4 delimited by end-of-file (wanted `2^2^2^2') 03:21:36 oerjan: Can't other IRC channels run their own instances of the bot? 03:21:39 `` bc <<< '2^2^2^2' 03:21:41 65536 03:22:18 lambda-11235: well sure, but not with the same nick on the same network 03:25:07 `` bc <<< 'for (i=23;;i++) { if (2^617%i==0) { print i; break; } }' 03:25:08 32 03:25:15 huh 03:25:24 oh duh 03:25:32 `` bc <<< 'for (i=23;;i++) { if (2^617%i==1) { print i; break; } }' 03:25:38 59233 03:25:45 yay 03:26:00 * adu gives oerjan a golden cookie 03:26:10 perhaps should have added 2 instead of incrementing 03:26:15 *munch* 03:26:27 -!- madbr has joined. 03:26:29 hey 03:27:13 hey 03:27:32 looking at some c++ compiler output to see how much operation chaining it has 03:27:44 chaining? 03:27:46 looks like about 50% of ops could be chained 03:27:55 chained? 03:28:00 yeah, basically a series of operations that update the same register 03:28:18 example of chain: 03:28:18 079443AF subss xmm1,xmm3 03:28:18 079443B3 mulss xmm1,xmm4 ; chain 03:28:19 079443BE addss xmm1,xmm7 ; chain 03:28:19 079443C2 movss dword ptr [esi+2F4h],xmm1 ; chain 03:28:19 079443D2 mulss xmm1,xmm4 ; chain 03:28:50 5 consecutive operations involving the same 03:28:53 register 03:30:08 I'm trying to think how that might apply to "belts" 03:30:47 https://en.wikipedia.org/wiki/Mill_CPU_Architecture#The_Belt.2C_a_pipelining_register_system 03:31:13 afaik, the belt is essentially an improved VLIW 03:31:33 it could do the chain but it has to mix it up with a few other calculations to run in parallel 03:33:16 afaik the main problem that the belt fixes is that it makes it really easy to write 4+ values at the same time 03:33:26 since they just end up on contiguous belt spaces 03:33:57 instead of that cpuid clusterf**k 03:34:13 cpuid clusterfuck? 03:34:38 <\oren\> cpuid instruction returns data in like 4 registers, right? 03:35:03 \oren\: sometimes 2, I think, sometimes 4, iirc 03:35:15 that can't be good 03:35:40 and they have different meaning depepding on the input 03:36:03 <\oren\> apparently 3; ebs, ecx and edx 03:36:22 <\oren\> oh, no 03:36:25 well, I still think its a clusterf**k 03:36:33 <\oren\> it really does depend on what's in EAX 03:36:57 the belt basically fixes a sequence like add r4, r7, r6; sub r3, r11, r13; and r2, r21, r19; mul r5, r1, r20 03:37:09 <\oren\> the "processor brand string" is returned in EAX thru EDX 03:37:09 which you typically find on VLIW cpus 03:38:12 cpuid is basically a vector of bits the size of the known universe 03:39:54 it writes different parts of the cpuid to eax,ebx,ecx,edx depending on which part you ask for? 03:40:45 @tell b_jonas fungot, does the amplitude of gravity waves attenuate linearly or quadratically in distance? <-- *gravitational hth 03:40:45 oerjan: if the power, as duly noted. i've been up on the mountain is symbolic of the eternal quest for wisdom and understanding beyond what you or the weepies as their ruler now. 03:40:45 Consider it noted. 03:40:57 <\oren\> yes, and if EAX = 0x00000004 then it writes an ASCII string to EAX:EBX:ECX:EDX 03:41:44 <\oren\> er, wait? 03:41:49 lolol 03:41:49 <\oren\> WTF it's even worse 03:42:32 -!- hppavilion[1] has quit (Ping timeout: 256 seconds). 03:42:36 so microcoded 03:44:44 <\oren\> I don't see why they couldn't just make users detect cpu by seeing which instructions work? 03:45:17 \oren\: like adding something to cflags like a "not implemented bit"? 03:45:27 <\oren\> yah 03:45:54 <\oren\> the bit turns on if an instruction is maformed 03:46:53 \oren\: that would only work on archs with constant size opcodes 03:46:56 because doing an illegal instruction triggers an interrupt 03:47:09 which easily eats hundreads of cycles 03:47:15 \oren\: x86 doesn't have that 03:47:50 also flags are bad 03:48:02 <\oren\> or you could just make users execute instructions then catch the SIGILL? 03:48:42 \oren\: I like how you're thinking outside the box, but you need two paths either way 03:48:50 oren : ok but then where does windows restrart the thread? 03:49:14 it might as well be represented by an instruction that gives you data to test against 03:49:19