00:00:55 https://github.com/wopot/jsgifkeylogger/blob/master/gifjs.asm cute but I don't totally see the security relevance 00:01:07 i do like the use of assembler to build binary files that aren't machine code at all 00:01:49 aw, it's not instructions 00:02:22 this is only for tests. 00:02:30 i like the misspelling 00:02:44 wehre 00:02:47 *where 00:02:57 heigth; also, good 00:03:32 it was mistake 00:03:44 i, too, was mistake once 00:04:35 like it could be used to exploit sites with a very specific combination of file uploading, restricted user-specified HTML, and path and URL whitelisting 00:04:45 a combination that doesn't seem very likely to come up in practice 00:04:49 but maybe I'm missing something 00:04:53 anything's possible online, right 00:07:25 kmc: that reminds me of the thing a while back where I tried one of those "smallest possible executable" things 00:07:33 someone had posted their own attempt so I spent a few hours poking at it and trying things 00:07:40 it was fun laying out the ELF header manually 00:07:42 "turns out you can get linux to execute a negative amount of bytes" 00:07:49 pff XD 00:07:59 fancy 00:08:02 i wonder if you can exploit a filesystem bug like that 00:08:07 like if they used signed types accidentally 00:08:13 so you can convince it you have a -1 byte file 00:08:13 I wonder if it's remotely convenient to write a bitstream using an assembler, though 00:08:22 and it'll read backwards because of the way they wrote their loop or something 00:08:23 like writing bytes is just "db whatever" and so on 00:08:27 but like, a bitstream would be trickier 00:08:27 so you can store the file contents in the metadata 00:08:37 imo someone should make this happen 00:08:41 oh gosh I'm thinking of writing a variable-length coder in yasm macros someone -stop me- 00:08:46 :D 00:08:51 * Bike stares impassively 00:09:00 i will not stand in the way of this 00:09:34 Fiora: Can I ask why? 00:09:36 Fiora: at least you're in the right channel 00:09:41 Mind, this is probably awesome. 00:09:47 :) 00:10:18 um. because like, the gif thing above 00:10:23 the idea of laying out a file in an assembler, right? 00:10:37 but like, assemblers only let you write bytes, not bits, usually 00:10:37 Oh. Hah. 00:10:44 at least I think? I might be wrong 00:10:59 That is an amusing way of distributing a binary. 00:11:59 nasm lets you use binary literals. 00:17:19 Amusing thought: you could actually have a GPL-compliant program distributed only as a binary. 00:17:27 If you actually wrote it that way. 00:17:30 Fiora: in nasm, %assign (etc) should be enough to let you write bitwise macros that output bytes as you go along 00:21:19 -!- nooga_ has quit (Ping timeout: 276 seconds). 00:21:24 Fiora: in any case, do it 00:24:04 http://privatepaste.com/094d840fe8 okay I have no idea if this is right but it runs and the first 4 hex digits seem to be right I think 00:24:33 *Nice*. 00:25:19 oh geez. now I'm imagining someone, like, writing a JPEG compressor in nasm macros 00:25:58 (probably faaar beyond me though <.<) 00:27:14 that is fantastic 00:30:19 I think it only handles codes up to 32 bits long 00:30:31 actually I have no idea how nasm represents variables 00:30:40 like gosh what happens when you try to calculate something bigger than int_max 00:30:47 -!- Phantom_Hoover has quit (Read error: Connection reset by peer). 00:31:09 -!- Phantom_Hoover has joined. 00:31:17 turns out it just uses GMP 00:31:19 "for speed" 00:32:39 oh gosh it is actual bignums 00:32:52 or, like, (200000000000000000000000000000<<1) does resolve to 400000000000000000000000000000 00:32:58 wow. 00:33:05 i don't know what to believe any more 00:33:19 1238273829327893192*973289472384729384729348923 resolves to 1205198882014363608197849789575338441444232216 00:33:28 I am kind of um. wow 00:33:57 > 1238273829327893192*973289472384729384729348923 == 1205198882014363608197849789575338441444232216 00:33:58 True 00:34:03 Why wouldn't it use bignums? 00:34:14 oh, huh 00:34:20 It works in yasm 00:34:21 but not in nasm 00:34:28 in nasm it overflows and complains that my big constsant thing won't fit in 64 bits 00:34:30 isn't that the reason nasm isn't cross-platform or sth 00:34:31 in yasm it works okay 00:34:34 no wait i'm thinking of floats 00:34:47 that is super weird 00:36:42 what's going on with those integers there i don't understand 00:37:03 in yasm it just kind of does what you'd expect and works in nasm it overflows ? 00:37:32 oh, i missed the multiplication. 00:41:08 -!- zzo38 has joined. 00:42:21 * Fiora peeeeers at the code, grepgregperpegrep 00:42:25 struct yasm_intnum { 00:42:25 union val { 00:42:25 long l; /* integer value (for integers <32 bits) */ 00:42:25 wordptr bv; /* bit vector (for integers >=32 bits) */ 00:42:25 } val; 00:42:27 enum { INTNUM_L, INTNUM_BV } type; 00:42:30 }; 00:42:32 wow. 00:42:33 -!- Koen_ has joined. 00:42:34 nice 00:42:42 yasm, scales to 1024-bit cpus 00:42:43 what 00:42:52 what's a wordptr 00:43:04 Bike: uint32_t* 00:43:14 gregperp 00:43:22 oh. wouldn't that be a vector of words 00:43:29 ("whatever") 00:43:36 http://guest.engelschall.com/~sb/download/ and it uses this thing for bignum math? 00:43:41 or um, the thing labelled bit-vector 00:43:42 typedefs of pointers are the devil hth :'( 00:43:49 Well, actually, it might be a vector of "words" in the x86 sense. 00:43:55 Making that a uint16_t* 00:44:11 The use of a camel in combination with Perl and Perl-related topics 00:44:12 is a trademark of O'Reilly & Associates, Inc. 00:44:13 It's funny that the system word on my system is a "quad word". 00:44:53 That's Intelese and/or Microsoftese for you. 00:44:55 Wait, so how's it know how many words it's using. 00:45:12 Bike: ... :( 00:45:21 how many bytey things. 00:45:26 i'm not good with computers sorry 00:45:33 Maybe the first word is a count? 00:45:46 shachaf: Intelese really. 00:45:54 it's a pointer to a Bit-Vector object, I think 00:45:56 Bike: could be 0 terminated 00:45:57 bit-vector is a library 00:45:59 And really makes sense from a legacy point of view. 00:46:02 oh ok 00:46:04 or more like a single c file apparently @_@ 00:46:15 that's c libraries for you 00:46:21 x86 asm is for a system with a 16-bit word. 00:46:28 I'm guessing the author didn't want any dependencies for yasm? still wow 00:46:34 well it's not like naïve bignums are hard right 00:46:43 so how often do bignums come up in macroassembly 00:46:51 Bike: Merely tedious if you don't care much about performance. 00:46:52 I... I have no idea @_@ 00:46:59 I don't think I ever touched them before this 00:47:07 but I'm probably not like a good sample or anything 00:47:09 pikhq: the algorithms i learned in elementary school are good enough for my PC, dammit! 00:47:10 naıve bignums 00:47:24 Fiora: maybe they come up in intermediate results and you don't notice. 00:47:31 this might bother me now 00:47:55 but then it'd break on nasm 00:48:14 right? 00:48:16 intel syntax is weird 00:48:24 yeah, i guess it would. 00:48:37 elliott: btw wouldn't 0 terminated be pretty bad for bignums since most of the algorithms go right to left 00:48:54 0 initiated....................... 00:48:55 maybe you could store them backwards. 00:49:09 0 terminated? 00:49:14 for bignums 00:49:23 0 terminated would be awful. 00:49:31 And probably unworkable. 00:49:59 Seeing as 32 0b0s in a row is quite reasonable for sufficiently large numbers. 00:50:11 wow i didn't even think of that wtf 00:50:30 And no, switching to BCD won't help there. 00:50:38 "exactly what i was thinking" 00:50:43 Well. Actually, 8 bits per digit BCD works just fine. :P 00:50:58 pikhq: the idea is you increment by one 00:50:59 wordptr is unsigned int * btw 00:51:02 let's just write numbers in unary 00:51:06 or just reserve one bit as a tag bit or w/e 00:51:13 Heck, at that point you might as well use ASCII. :P 00:51:26 ASCII: best bignum format? 00:51:47 utf-32 is better 00:51:59 finally, a use for COMBINING CYRILLIC MILLIONS 00:52:17 reminds me of how intel added an example code bit in their new optimization manual for converting 64-bit integers into base 10 (for like, printf) 00:53:25 more sophisticated than mod by 10, i assume 00:53:55 using avx2 and stuff 00:54:12 convert between bases in O(1) time!! 00:54:22 "The AVX2 version of numeric conversion across the dynamic range of 3/9/17 output digits are approximately 23/57/54 cycles per input, compared to standard library implement ion’s range of 85/260/560 cycles per input." 00:54:31 geeeez 00:54:58 damn. 00:55:07 they're doing "vectorized montgomery reduction" 00:56:05 so I guess, like, they're calculating (x%10),(x%100),(x%1000) and so on in simd? 00:56:08 or something 00:56:12 I don't know what montgomery reduction is 00:57:04 "In arithmetic computation, Montgomery reduction is an algorithm introduced in 1985 by Peter Montgomery that allows modular arithmetic to be performed efficiently when the modulus is large (typically several hundred bits). 00:57:04 " 00:57:57 -!- sprocklem has quit (Remote host closed the connection). 00:58:20 -!- Phantom_Hoover has quit (Read error: Connection reset by peer). 00:58:25 that is big 00:58:53 pretty big 01:42:36 today i saw a tree near my house that was like a whole ecosystem 01:43:15 tree, two different plants growing on the tree, moss, lichen, algae, tiny Mycena mushrooms, ants, red spider, snail, black spider weaving a web attached to the snail 01:43:33 and a kmc 01:43:45 i wasn't 'on' the tree elliott 01:44:28 i thought you were all about being "on" the trees 01:44:30 if you know what i mean 01:44:33 i'm talking about drugz 01:44:39 i think the snail wasn't dead, just too slow for the spider to notice 01:44:40 damn i was hoping dendrophilia 01:44:43 elliott: ah ok 01:44:51 snail defense mechanism 01:45:13 maybe the spider and snail have developed a mutualistic partnership, where the snail makes the web move approprirately 01:45:13 it's always drugz 01:45:19 snails are adorable 01:45:24 yes 01:45:25 http://cuteoverload.com/2009/02/18/junior-snailio/ 01:45:34 awwwwwwwwwwwwwwwwwwww 01:45:49 hey kmc why does terminal.app often think links include some of the timestamp on the next line 01:45:54 oh my gosh that is so adorable 01:45:55 and more generally lines contain multiple lines 01:45:59 and then when i copy paste them the \ns are gone 01:46:03 w/ mosh 01:46:13 don't know 01:46:15 I put the sequent calculus of Turing machine: http://zzo38computer.org/tex/turing_sequent.tex Please tell me if there is a mistake. 01:46:18 it's amusing when like, it rains here and all the snails come out, and you walk outside and find them all having sex or something 01:46:20 thmc 01:46:26 wow that size difference 01:46:29 ('cause they're snails) 01:46:41 horny sluts the lot 01:46:42 yeah it rained today and there were lots of snails out 01:46:44 also mushrooms 01:47:00 there was this really big snail crawling up the wall the other day like, actually really fast (for a snail) 01:47:05 I wonder if it would be a bad idea to link work stuff here once it's public 01:47:09 and I was imagining it making cute little grunting noises each time it pushed up 01:47:11 we saw some of these: http://en.wikipedia.org/wiki/Nidulariaceae 01:47:20 only they were smaller and yellower 01:47:21 Fiora: haha 01:47:44 it's cool how they climb, like, just as fast as they crawl horizontally 01:48:04 there's also this one snail that lives on the wall outside my door. she wakes up every time the sun hits and moves a little bit to the side to stay in the shade 01:49:10 (another great thing about snails: you can use any pronoun you want to refer to them and you'll never be wrong) 01:50:31 :) 01:51:22 http://upload.wikimedia.org/wikipedia/commons/a/a2/Snail_sanctuary%5E_-_geograph.org.uk_-_215817.jpg oh gosh 01:51:28 -!- Bike has quit (Ping timeout: 268 seconds). 01:52:11 wow 01:52:59 SgeoBot: i promise to not try very hard to hack your company's websites 01:53:46 -!- Nisstyre-laptop has joined. 01:53:56 I said yesterday that I would put the sequent calculus representation of Turing machine into the computer on the weekend; well, I did it early. Can you please review it, to tell me if you found a mistake, or another comment/question/complaints? 02:01:59 wow. snails mate after firing a "love dart" into the other. 02:02:00 http://en.wikipedia.org/wiki/Love_darts 02:02:52 Hot. 02:03:20 wow, snail courtship lasts up to 6 hours @_@ 02:04:49 -!- hagb4rd has joined. 02:05:44 Ohhh my 02:11:04 -!- Bike has joined. 02:11:09 itt snail porn 02:11:19 every few weeks my computer forgets that wlan0 exists 02:11:24 and i have to do a hard reset to fix it 02:11:26 animal mating is way more interesting than human mating okay :< 02:11:29 i hate computers. they make me miss snail porn 02:12:02 Bike: Maybe you should try to fix it some other way, if it does that there is probably something wrong with it? 02:12:04 Fiora: I dunno, there's a lot of depth to human mating. 02:12:35 hm what is the deepest human mating incident 02:12:37 zzo38: yes, there's something wrong with it and i don't know how to fix it because the solution isn't as simple as ./configure && make && make install which is the limit of my capabilities 02:12:44 oerjan: Define "deep". 02:12:48 oerjan: some engineers in the kola borehole 02:12:54 the joke is depth, 02:13:03 Deep Throat is pretty deep. 02:13:25 that reminds me of a great quote about bird semen 02:14:05 `addquote that reminds me of a great quote about bird semen 02:14:09 1051) that reminds me of a great quote about bird semen 02:14:30 -!- Nisstyre has quit (Quit: Leaving). 02:14:36 -!- Nisstyre-laptop has changed nick to Nisstyre. 02:15:26 pikhq: is... that a pun 02:15:28 -___- 02:15:40 Fiora: Yes. 02:16:05 pikhq, I am listening to the slow version of Magia using SennHeiser headphones 02:16:50 Fiora: I've got about half a degree in puns. 02:18:26 i've got about 451 02:18:39 -!- Frooxius has quit (Ping timeout: 252 seconds). 02:18:51 -!- Frooxius has joined. 02:18:52 Which is roughly the temperature parts of my home town are at right now. 02:18:58 * Bike looks for quote, finds detailed drawing of bird soliciting for sex 02:20:32 -!- SgeoBot has changed nick to Sgeo. 02:22:41 i don't think i have what it takes to be a zoology fieldworker. 02:23:04 Does anyone? 02:23:15 disturbingly enough, yes. 02:25:21 oh, hey, i found what i was thinking of. "My career in sperm competition has been a rollercoaster ride, energised by a number of particularly special moments." 02:25:30 me too 02:25:48 the essay this is from is titled "Undiminished passion" 02:28:34 I think I know which phone I want 02:28:38 Not certain though 02:30:04 :t (<~) 02:30:05 Not in scope: `<~' 02:30:05 Perhaps you meant one of these: 02:30:05 `<' (imported from Data.Ord), `<=' (imported from Data.Ord), 02:31:08 pikhq: so like, an associate's? 02:33:16 -!- Frooxius has quit (Ping timeout: 276 seconds). 02:42:31 Fiora: Sure. 02:46:21 -!- Frooxius has joined. 02:49:54 -!- Koen_ has quit (Quit: The struct held his beloved integer in his strong, protecting arms, his eyes like sapphire orbs staring into her own. "W-will you... Will you union me?"). 03:03:58 What do you expect (<~) to be? 03:07:00 back to pdfs: should i keep the crazy-ass ultrafinitist zeilberger pdf 03:07:22 zzo38: something from lens. 03:08:09 but then i heard lambdabot doesn't have lens at the moment for a stupid safehaskell reason. 03:08:20 Bike: yes 03:08:25 ultrafinitists are the best 03:09:26 REAL (i.e. discrete) analysis is conceptually simpler than traditional 'real' (continuous) analysis, and of course is much truer. But it is, on the whole, technically more difficult. Hence 'Naked Brain' humans had no choice but to pursue the latter kind. 03:10:00 what's discrete analysis in this context 03:10:49 Analysis for crazy people. 03:11:32 Df(x) := (f(x+h) - f(x))/h "where h is the Fundamental mesh size, a Mathematical Universal constant, that unlike Planck's constant, we will never know, but it is very tiny." 03:11:42 … 03:12:00 c'mon zeilberger isn't really crazy 03:12:02 ok i see what you mean. 03:12:03 i mean he's not a crank 03:12:21 elliott are you reading this 03:12:25 yes i've read it 03:12:33 this is time cube level discourse here 03:12:37 it's like a PLANK LENGTH but for the PLATONIC REALM 03:12:40 i mean obviously he's not a crank 03:12:45 but this is fucking nuts 03:13:07 ultrafinitism is not really that nuts 03:13:13 i mean it's not like mathematicians don't go off on platonic type philosophising all the time 03:13:21 yeah but do they do it in allcaps 03:13:23 ultrafinitism is 100% nuts though 03:13:24 where the information content amounts here to "we work with h as a free variable" 03:13:27 "just sayin'" 03:14:29 "fundamental mesh size" eughs me out 03:15:06 The main stumbling block in the further development of the theory of Discrete Analytic Functions is the fact that the property of being discrete-analytic is not preserved under multiplication. But using the discrete Leibniz rule one can express the derivative of a product, and then the product is "almost analytic". So I am sure that the full arsenal of continuous complex analysis can be discretized, but the details might be too complicated fo 03:15:44 btw finite calculus is extremely cool 03:16:16 What is a "fundamental mesh size"? 03:16:22 are the details too complicated for humans 03:16:47 That "Df(x) := (f(x+h) - f(x))/h" doesn't seem to make much sense to me 03:16:48 zzo38: the difference between two points on the real line. 03:17:27 Bike: O, OK, well, it isn't an actual number that can be measured, I think. 03:17:44 its like a derivative but with a philisophical twist 03:18:11 hm i wonder if i'm sleeping 03:18:16 no 03:18:17 yes this is a dream 03:18:27 dont listen to bike he is the Deciever 03:18:37 i meant i wonder if i'm sleeping tonight but ty 03:18:45 Bike: /nick Deciever 03:19:17 no 03:19:35 how about nick Deceiver 03:19:52 Neo-Pythagoreanism or: Anaxagoras deserved to be drowned 03:20:36 If that is how the "fundamental mesh size" is defined, then it can't be a real number. 03:21:17 Bike: i'm reading the book are you proud of me 03:21:22 Yes. 03:21:23 mnoqy won't read it with me because he's ""busy""" :-| 03:21:42 It is utter nonsense to say that √2 is irrational, because this presupposes that it exists, as a number or distance. The truth is that there is no such number or distance. What does exist is the symbol, which is just shorthand for an ideal object x that satisfies x² = 2. 03:21:47 hellooooo platonism 03:22:05 hello......... 03:22:26 -fidget- 03:22:35 -roll- 03:22:41 Well, I suppose whether or not it exists as a distance in physical space is really a different question, although it does mathematically exist, regardless. 03:23:12 Blessed are the Δifference Equations for They Shall Inherit Math 03:24:49 Whether or not there is such a number, depends on what number system it is in; since it is irrational, that means the rational numbers has no square root of 2, but the real numbers does have. 03:24:51 Andrew Wiles' alleged 'proof' of FLT, while a crowning human achievement, is not rigorous, since it uses continuous analysis, which is meaningless. 03:24:51 'sorry wiles' 03:25:13 using the word 'alleged' *and* putting "proof" in scarequotes 03:25:24 Why are you writing such nonsensical things? 03:25:40 Because I think mocking elliott is funny. I'll stop. 03:25:54 are you making fun of elliott because of ultrafinitism? 03:26:16 Well, mostly because this guy specifically is nutso. 03:26:28 The only other ultrafinitist I know of is the one from the anecdote, so he's cool. 03:27:51 that's a great anecdote 03:27:58 if it's the one anecdote about ultrafinitism that i know of 03:28:11 i like that anecdote too 03:28:20 [assuming it's the one i know of] 03:29:14 how many anecdotes about ultrafinitists can there possibly be, i wonder 03:29:19 My philosophy is that mathematics is the real reality, not the physical universe and so on. 03:29:21 finitely many 03:29:26 well played. 03:29:29 i think lecturing the guy for ranting and mocking people he disagrees with is a bit silly 03:29:33 zzo38: platonist!! hides in shelter 03:29:36 since have you noticed how much everyone ever does that 03:29:48 they don't do it quite so amusingly 03:29:52 it's not like any of it was submitted to a real journal or anything 03:29:57 Bike: O, what is *your* philosophy of this? 03:30:20 of math? i suppose i'm a materialist. 03:30:54 Bike: Can you please be a bit more specific? 03:31:22 my philosophy of math is "philosophy of math is silly" 03:31:37 Also a good philosophy. 03:31:52 But why not: I don't think mathematical objects "exist" in any meaningful or usefully talked about sense. Math is just a kind of useful thinking. 03:32:28 sounds like formalism to me 03:32:55 yeah, something like that. 03:33:02 so i'll probably go full platonic at some point derp. 03:33:10 ~consistency~ 03:33:28 my philosophy of things existing is "talking about things existing is silly" 03:33:56 i think hilbert was kind of wack tho. 03:33:58 so there's that. 03:34:09 mnoqy philosophy: Q: what does it mean for something to exist / A: dont ask that question shut up shut up shut up 03:34:09 well your philosophy of things existing doesn't really exist, so there 03:34:28 mnoqy: would you call yourself a sillyist 03:34:36 @quote saulgorn formal 03:34:36 No quotes for this person. Are you on drugs? 03:34:38 !! 03:34:40 @quote formal 03:34:40 Raguel says: does it unsettle anyone that there isn't a formal semantics for haskell? its like someone got drunk, invented a gadget, then when he woke up, couldn't quite draw a blueprint 03:34:40 cool 03:34:42 help 03:34:43 @quote formalist 03:34:43 SaulGorn says: A formalist is one who cannot understand a theory unless it is meaningless. 03:34:46 thachaf 03:34:48 thambdabot 03:34:53 thanks saul gorn 03:34:58 is that shachaf 03:35:12 no 03:35:18 Shaul Chorn 03:35:26 helpful 03:35:49 Can you understand my implementation of a Turning machine in the sequent calculus? 03:35:57 nope. 03:36:07 maybe 03:36:16 Did you read it? Did you find a mistake within it? 03:36:27 no 03:36:34 mnoqy: but seriously philosophy of math has /some/ use, if only to motivate weirdos to make non-classical logics etc 03:37:07 Bike: ok so long as they need it to motivate them....i do like those non-classical logics 03:37:09 maybe the mathematics establishment should just recruit in asylums to skip the middlemen 03:37:22 good idea 03:37:26 Well, yes perhaps it can motivate some things. 03:37:29 non-classical logics are p. cool 'Bike stamp of approval' 03:37:57 Not only in mathematics, but I think this kind of motivation can be using in science too, in order to form hypotheses and so on. 03:38:37 mnoqy's philosophy of mathematics allergy reminds me of the philosophy of mathematics allergy i used to have 03:38:55 what happened 03:38:56 where's shachaf 03:38:58 i also used to hate talking about anything i completely arbitrarily categorised as "politics" 03:39:04 ``i grew out of both'' 03:39:05 /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: `i: not found 03:39:06 haha same 03:39:15 now i rant about the NSA on twitter so that went well 03:39:18 it was the "same kind of thing" i think 03:39:21 srry 14 yo me 03:40:06 hm i wonder if i can apply this "useful crazies" philosophy to psychology stuff none of you give a damn about 03:40:11 i've forgotten how old bike is HELP 03:40:13 i require more research on this 03:40:18 `run printf "#!/bin/sh\n\necho 'DID YOU REALLY'\n" > 'bin/`i' && chmod +x 'bin/`i' 03:40:19 bike is old iirc 03:40:21 No output. 03:40:24 ``i grew out of both'' 03:40:26 DID YOU REALLY 03:40:26 so old 03:40:34 kmc: awesome 03:41:24 i haven't read any [psychologist's name] directly yet, i don't know if he was crazy. i know chomsky is kind of wack and skinner is super wack. everybody's wack 03:41:43 chomsky aint whack :| 03:41:48 hi augur 03:41:49 chomsky is an original gangsta 03:41:53