←2021-05-30 2021-05-31 2021-06-01→ ↑2021 ↑all
00:00:54 <shachaf> int-e: Should I be crediting the distinguished point method of finding collisions to Pollard at all?
00:01:21 <shachaf> Pollard's rho is a very different algorithm (for factoring) that can't use distinguished points.
00:07:12 <esolangs> [[Esolang:Sandbox]] M https://esolangs.org/w/index.php?diff=83722&oldid=83707 * PythonshellDebugwindow * (+61) Make better title
00:13:43 <esolangs> [[Esolang:Sandbox]] M https://esolangs.org/w/index.php?diff=83723&oldid=83722 * PythonshellDebugwindow * (+301) .
00:15:51 <salpynx> :( I have just realised my WIP omega-word language that was meant to be sub-TC and in the category of FSM + some strange but still limited infinite extension is a super set of oracle machines pretty much by definition, so is apparently super-TC
00:17:20 <salpynx> at least, unlike other haha-halting problems langs on the wiki, this thing can at least implement and play real low level FSMs with minimal syntax.
00:18:57 <b_jonas> shachaf: but isn't it like you have two ways to find repeats, the slow and fast movement that Pollard traditionally uses and the values with a hash ending in zero that you traditionally use for finding hash collisions, but in theory you can use either for a prime factoring algorithm similar to Pollard or for finding hash collisions?
00:21:41 <shachaf> How would you use it for factoring?
00:21:59 <shachaf> You can't tell if a point is distinguished.
00:50:53 <b_jonas> why not? just hash the points and distinguish them if the hash is low enough
00:51:06 <b_jonas> I could be wrong here
00:51:32 <b_jonas> I have implemented Pollard rho prime decomposition once but don't quite remember how it works, and hash collision finding very few times
01:07:45 <shachaf> You don't know the points, though.
01:07:58 <shachaf> I mean, you know the points mod n, but you care about the points mod p, and you don't know p.
01:08:03 <shachaf> Or am I confused?
01:08:25 <b_jonas> oh! yes, that's how Pollard worked
01:08:30 <b_jonas> then what I said is stupid
01:08:38 <b_jonas> sorry
01:09:21 <b_jonas> anyway, I have a very different programming question
01:12:01 <b_jonas> you know how in C++ or rust, you can use constructors and destructors to implement an interface for reference-counted smart pointers such that they automatically increment the refcount when you copy the pointer and decrement it when you get rid of it
01:12:10 <b_jonas> this works well and is proven technology
01:14:19 <b_jonas> you can also do something similar, but not quite the same, for a garbage collector. there the problem is that the collector has to be able to access not only structure on your heap but also your local variables on the stack. there are two methods for this: conservative garbage collectors that just scan the stack and assume that any word might be a pointer, in which case they have to be able to validate
01:14:25 <b_jonas> what numbers are valid pointers, and normal ones where you have to keep track of which variables are pointers
01:16:07 <b_jonas> my question is, how much harder does such a natural interface get between moving/generational garbage collectors compared to garbage collectors where the pointer value never has to change? I have seen two interfaces for moving garbage collectors: the one in Lua where you can effectively only store pointers only on the Lua stack and you have to index them, and the mzscheme one that I haven't looked at
01:16:13 <b_jonas> for a long time and can't remember how it works
01:18:39 <b_jonas> ghc has a garbage collector too, but no proper interface for extensions where they can easily own pointers. they can create stable pointers, or store pointers in normal haskell algebraic type values, but neither is really convenient from C code.
01:19:48 <b_jonas> ruby 1.6 has a conservative garbage collector so you can just use plain pointers, as long as you don't try xor trick on them or store them anywhere other than on ruby's heap or the stack
01:20:15 <b_jonas> I don't know if recent ruby 2 still has that
01:20:23 <b_jonas> they have changed a lot of details about the core
01:20:56 <b_jonas> well, they also changed a lot about the language
01:22:07 <b_jonas> perl uses refcounting of course
01:22:51 <int-e> shachaf: in the papers I've seen it used, the primary reference for the distinguished point method is van Oorschot+Wiener
01:24:05 <int-e> But obviously Pollard was a big inspiration.
01:24:57 <shachaf> Aren't the cycle-finding algorithms (tortoise-hare etc.) a bigger inspiration? The Pollard method doesn't even require a collision, I think, just a cycle, right?
01:25:18 <b_jonas> ah yes, tortoise and hare, that's what it's called!
01:25:20 <shachaf> Another question: Can Pollard rho factoring be parallelized at all?
01:26:07 <b_jonas> shachaf: dunno but it's probably not worth because if you want something more complex than pollard, there are much faster prime factoring algorithms these days
01:26:13 <shachaf> What's the best cycle-finding algorithm for the Pollard rho thing? Is it some variant of Brent's algorithm?
01:26:56 <int-e> shachaf: sure, it was meant for cycle-finding, but the rho clearly has a collision in it.
01:27:31 <shachaf> Hmm, does it work if you start inside a cycle?
01:27:57 <int-e> Parallelizing pollard rho... no, since there's no useful notion of distinguished points since every value is a mix of something modulo p and something modulo q.
01:28:09 <int-e> And the whole idea is that those cycle independently
01:28:16 <shachaf> Right, that's what I was thinking.
01:28:54 <int-e> hopefully; if you start in cycles of equal length both modulo p and q then it doesn't work
01:29:55 <int-e> So it only parallelizes in the bad way where N processors result in an sqrt(N) expected speedup.
01:30:41 <shachaf> Right. Whereas the distinguished point method is approximately the best case for a parallel algorithm that needs to share memory at all.
01:30:47 <b_jonas> hmm, I wrote a pollard rho prime factorizer for the Sharp EL-5120 calculator, but I can't find it in my records of EL-5120 programs. but then, I wrote way more programs for that than I have records for, in particular I have written a rudimentary minesweeper program that I don't have the source code for.
01:33:08 <b_jonas> also a program that implements McCulloch's 2nd machine, but that's sort of useless because it only handles values up to between 10 and 12 digits long (12 digit should work I think, but I'm not sure, because there's some peculiar behavior that makes anything longer than 10 digits hard to handle, specifically if you subtract two numbers and the exponent of the result would be ten orders of magnitude less
01:33:14 <b_jonas> than the inputs than the result is forcibly zero, even though there are two significant digits that the operation could return)
01:41:57 <int-e> shachaf: oh, btw, that f(x) = g(y) problem and the associated trick of combining them into a single function also comes up in discrete logs: let f(x) = x*g and g(x) = x*g^floor(sqrt(p)). Start with the unknown value, and for all x/y values computed, keep track of the corresponding exponents of g multiplied in.
01:42:21 <int-e> shachaf: I suspect that has a name attached to it too, but I'm too lazy to check right now.
01:44:26 <b_jonas> I do have the source of my two trial division prime factorization programs for the EL-5120, and they're called FELB.4+ and FELB.5 so they aren't the first such programs that I wrote
01:44:35 <b_jonas> I don't know why I don't have the pollard one preserved
01:45:17 <b_jonas> ISTR that was by far the best prime factorizer I wrote for that calculator
01:45:59 <b_jonas> unless perhaps this FELB.5 is a Pollard program and I just have no idea how it works
01:46:29 <b_jonas> but that seems unlikely because I'm pretty sure "if =fpart(V/Wgoto 5;if =fpart(V/(W+2goto 5;W=W+6" is trial division
01:48:17 <b_jonas> apart form the two factorizers, I have a quadratic equation solver and the maze game
01:49:53 <esolangs> [[User:Salpynx/Galveston]] https://esolangs.org/w/index.php?diff=83724&oldid=83667 * Salpynx * (+3441) /* Computational class */ musings, trying to classify this thing
01:53:31 <int-e> shachaf: Ah that doesn't work yet, since a g^n = a g^m doesn't tell us anything about a. https://en.wikipedia.org/wiki/Pollard%27s_rho_algorithm_for_logarithms is what I had in mind and adds a third function into the mix: squaring modulo p.
01:54:16 <int-e> But in any case *that* trick can be attributed to Pollard, maybe? Or maybe it's older than that.
01:55:44 <int-e> As usual, it's used for a slightly different purpose... there's no attempt to obtain cross-collisions.
01:58:19 <b_jonas> int-e: TAOCP might answer that about the history
01:58:31 <int-e> true
02:00:16 -!- Noisytoot has quit (Quit: ZNC 1.8.2 - https://znc.in).
02:01:06 -!- Noisytoot has joined.
02:04:49 <esolangs> [[Esolang:Sandbox]] https://esolangs.org/w/index.php?diff=83725&oldid=83723 * ColorfulGalaxy (disambiguation) * (+136)
02:12:58 <int-e> Hmm, he doesn't really do discrete logs, it seems. There's a HM30 exercise which at a glance seems to be the fully fledged Adleman's index calculus approach.
02:13:53 <int-e> He doesn't seem to do cycle finding much either... there's Floyd's algorithm as an exercise.
02:14:05 <int-e> So no, I don't think Knuth is helpful in this case.
02:15:48 -!- Noisytoot has quit (Quit: ZNC 1.8.2 - https://znc.in).
02:16:34 -!- Noisytoot has joined.
02:16:54 <int-e> One thing I do love about Knuth is that he's meticulous in preparing his indices. So it's relatively quick and easy to check these things with high confidence.
02:31:56 -!- Noisytoot has quit (Quit: ZNC 1.8.2 - https://znc.in).
02:32:31 -!- dcristofani has joined.
02:33:11 -!- Noisytoot has joined.
02:50:53 -!- Noisytoot has quit (Quit: ZNC 1.8.2 - https://znc.in).
02:51:05 -!- Noisytoot has joined.
03:05:34 <zzo38> Is there a video codec for making screenshots of computer programs including games with 2D graphics?
03:06:05 <keegan> why should a special codec be needed?
03:10:56 <int-e> keegan: same reason why we tend to use png for screenshots rather than jpg: artifacts
03:12:36 <keegan> i see. so you want a lossless video codec? those exist
03:15:11 <keegan> or you want it to be lossy but in a different way? that's trickier
03:15:55 <int-e> probably specially tuned for the characteristics of applications (less sure about 2D video games, apart from the idea that copying rectangles from one frame to the next is probably very useful)
03:16:20 <int-e> but it's not my question.
03:17:52 <nakilon> zzo38 "video" codec for making "screenshots"? I guess ffmpeg can make png screenshots in a loop
03:18:12 <int-e> remote desktop servers/clients would be the first place where I'd look for this kind of thing
03:18:35 <nakilon> since when does #esolangs learned to make screenshots in png though
03:19:26 <keegan> int-e: good idea
03:19:40 <shachaf> You're confused. It's #esoteric that you have a feud about screenshots or encodings or whatever in, not #esolangs.
03:19:42 <keegan> I'm sure someone has already made tools to record and play back VNC streams
03:21:45 -!- dyeplexer has joined.
03:23:57 -!- Lord_of_Life has quit (Ping timeout: 268 seconds).
03:24:11 -!- Lord_of_Life_ has joined.
03:25:15 -!- Lord_of_Life_ has changed nick to Lord_of_Life.
03:30:17 -!- dcristofani has quit (Ping timeout: 264 seconds).
03:39:50 -!- mnrmnaugh has quit (Ping timeout: 264 seconds).
04:02:00 <zzo38> Yes I did mean copying rectangles can be very useful, possibly also fonts, icons, etc that may reappear in another frame later after it is absent for some frames, or multiples of the same icon or font, too
04:02:38 -!- hendursaga has quit (Remote host closed the connection).
04:03:05 -!- hendursaga has joined.
04:18:35 <esolangs> [[Special:Log/newusers]] create * Doridian * New user account
04:24:02 -!- Lykaina_ has changed nick to Lykaina.
04:25:21 <esolangs> [[Esolang:Introduce yourself]] https://esolangs.org/w/index.php?diff=83726&oldid=83702 * Doridian * (+195)
04:26:20 -!- Lykaina has left.
04:43:56 -!- xkapastel has quit (Quit: Connection closed for inactivity).
05:11:01 <esolangs> [[Channeler]] N https://esolangs.org/w/index.php?oldid=83727 * Doridian * (+4890) Created page with "'''Channeler''' is an [[esoteric programming language]] created by [[User:Doridian]]. It is based around the idea of messaging channels (essentially function calls). '''Chann..."
05:17:13 -!- mnrmnaugh has joined.
05:34:50 -!- salpynx has quit (Quit: Ping timeout (120 seconds)).
05:37:20 -!- salpynx has joined.
05:42:06 <esolangs> [[Channeler]] https://esolangs.org/w/index.php?diff=83728&oldid=83727 * Doridian * (+4)
05:48:47 <pikhq> shachaf: i'll have a feud about screenshots wherever i feel like tyvm
06:01:17 <shachaf> Make kittens, not feuds.
06:07:44 <esolangs> [[User:Salpynx/Galveston]] https://esolangs.org/w/index.php?diff=83729&oldid=83724 * Salpynx * (+338) /* External resources */ something I need to finish reading. Good because it talks about more than just accepting automata
06:10:23 <int-e> shachaf: https://nitter.fdn.fr/RabbitEveryHour
06:11:40 <shachaf> It's too bad cats are cuter than rabbits, since they're carnivores and all.
06:15:37 <int-e> pfft, https://nitter.fdn.fr/RabbitEveryHour/status/1398172245344796674#m
06:16:15 <esolangs> [[Line]] https://esolangs.org/w/index.php?diff=83730&oldid=83716 * ColorfulGalaxy (disambiguation) * (+34) /* Conditional turn */
06:18:08 <esolangs> [[Line]] https://esolangs.org/w/index.php?diff=83731&oldid=83730 * ColorfulGalaxy (disambiguation) * (+156) /* Brainfuck compatibility */
06:23:00 <shachaf> Do you like Wegman-Carter MACs?
06:54:42 -!- dcristofani has joined.
07:01:02 -!- tromp has joined.
07:06:22 <esolangs> [[Line]] https://esolangs.org/w/index.php?diff=83732&oldid=83731 * ColorfulGalaxy (disambiguation) * (+62) Recategorization
07:14:54 -!- arseniiv has joined.
07:26:18 -!- dcristofani has quit (Ping timeout: 268 seconds).
07:41:07 -!- imode has quit (Ping timeout: 268 seconds).
07:52:45 <b_jonas> shachaf: rabbits are bred for eating. we don't eat their eggs or milk, they don't pull carts or herd sheep or guard your house, their hide isn't too useful either. there's a limit on how cute you want an animal like that to be, above that people wouldn't be willing to eat them.
07:56:13 <esolangs> [[Talk:Line]] N https://esolangs.org/w/index.php?oldid=83733 * ColorfulGalaxy (disambiguation) * (+122) Created page with "==Crossover== The multiplication program mentioned signal crossover. Could you tell me how it works? <pre> | -+- |</pre>"
07:59:10 <esolangs> [[Line]] https://esolangs.org/w/index.php?diff=83734&oldid=83732 * ColorfulGalaxy (disambiguation) * (+41) Recategorization
07:59:11 <b_jonas> whereas we don't eat cats, they're bred to hunt rats and other pest animals. and I guess that's the evolutional reason why they offer you animals that they hunt: that lets the breeders actually evaluate how successful each cat is in at hunting. either that, or the egyptian gods did give some rewards for sacrificing lots of very low XP animals.
07:59:27 <shachaf> I don't think eating mammals is good.
07:59:37 -!- ais523 has joined.
07:59:53 <ais523> zzo38: there's https://wiki.multimedia.cx/index.php?title=DosBox_Capture_Codec which was designed for making video recordings of 2D video game gameplay
08:00:17 <ais523> the format seems to have a few shortcomings, but it's supported widely enough that you can, e.g., upload a ZMBV-encoded video to YouTube and it understands it
08:00:34 <ais523> (and it replays in my computer's video player, too)
08:01:07 -!- salpynx has quit (Quit: Ping timeout (120 seconds)).
08:02:42 <ais523> I think it would be possible to do better using some sort of sprite+background encoding, like most 2D computer games actually use
08:02:52 <ais523> for communicating with the graphics card
08:03:13 -!- salpynx has joined.
08:07:38 -!- ais523 has quit (Quit: quit).
08:08:44 <esolangs> [[Talk:Line]] https://esolangs.org/w/index.php?diff=83735&oldid=83733 * Relt * (+119)
08:08:54 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
08:09:40 -!- hendursaga has quit (Ping timeout: 252 seconds).
08:12:59 -!- Sgeo has quit (Quit: Leaving).
08:16:50 -!- hendursaga has joined.
08:17:49 -!- tromp has joined.
08:23:21 <esolangs> [[Line]] https://esolangs.org/w/index.php?diff=83736&oldid=83734 * Relt * (-225)
08:49:56 -!- hanif has joined.
09:19:18 -!- salpynx has quit (Quit: Connection closed).
09:41:02 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
09:45:11 -!- tromp has joined.
10:43:14 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
11:47:18 -!- Thelie has joined.
11:49:14 <fizzie> b_jonas: I don't think it's likely true, but a cat book suggested the thing about bringing you little offerings (often half-alive) is because they've noticed you kind of suck at the hunting part, and are trying to sort of get you started.
11:50:46 -!- hanif has quit (Ping timeout: 252 seconds).
12:19:13 <arseniiv> :D
12:22:23 <esolangs> [[Esolang:Sandbox]] M https://esolangs.org/w/index.php?diff=83737&oldid=83725 * PythonshellDebugwindow * (+11) /* OK */ /* OK */
12:22:39 -!- hanif has joined.
12:24:05 <esolangs> [[Esolang:Sandbox]] M https://esolangs.org/w/index.php?diff=83738&oldid=83737 * PythonshellDebugwindow * (+76) /* OK */ /* OK */ /* OK */
12:24:34 <esolangs> [[Esolang:Sandbox]] M https://esolangs.org/w/index.php?diff=83739&oldid=83738 * PythonshellDebugwindow * (+23) ;
12:39:10 -!- hanif has quit (Ping timeout: 252 seconds).
13:00:16 <nakilon> now I wonder about rabbit milk
13:10:10 -!- hanif has joined.
13:12:00 <nakilon> looks like there is not only nitter for twitter written in nim but also invidious for youtube written in crystal
13:13:42 -!- Thelie has quit (Remote host closed the connection).
13:13:46 <nakilon> chrome doesn't open https://nitter.net/
13:16:38 <nakilon> and since "As of September 1st 2020, invidio.us has closed down" looks like there is no other domain that would be always up according to https://stats.uptimerobot.com/89VnzSKAn (those green are just mirror lists)
13:17:41 <esolangs> [[Talk:Line]] M https://esolangs.org/w/index.php?diff=83740&oldid=83735 * PythonshellDebugwindow * (+68) Unsigned
13:26:00 <nakilon> someone should make an instance that would consume APIs of all those instances rotating them when they are down
13:29:16 <nakilon> also two days ago I've learned there is RSS in Youtube
13:30:29 <esolangs> [[Special:Log/newusers]] create * Arcane * New user account
13:34:49 <esolangs> [[Esolang:Introduce yourself]] https://esolangs.org/w/index.php?diff=83741&oldid=83726 * Arcane * (+167) /* Introductions */
13:35:19 <esolangs> [[User:Arcane]] N https://esolangs.org/w/index.php?oldid=83742 * Arcane * (+12) Created page with "Hello There!"
13:35:45 -!- fungot has quit (Remote host closed the connection).
13:39:32 -!- fungot has joined.
13:39:50 <fizzie> fungot: You okay in there?
13:39:50 <fungot> fizzie: i think it was
13:39:54 <fizzie> Sounds good.
13:42:53 <hanif> ^style
13:42:53 <fungot> Available: agora alice c64 ct darwin discworld enron europarl ff7 fisher fungot homestuck ic irc* iwcs jargon lovecraft nethack oots pa qwantz sms speeches ss wp ukparl youtube
13:53:49 -!- ircseeker has joined.
14:19:04 <nakilon> how are styles being added?
14:19:07 -!- Lord_of_Life has quit (Excess Flood).
14:19:20 <nakilon> there should be a Terry A. Davis style
14:19:33 -!- Lord_of_Life has joined.
14:21:17 <nakilon> I see some mention in repo but I see no examples in it, I guess it's a server configuration
14:21:22 <arseniiv> nakilon: fizzie markov-chains a corpus in an unspecified way and then feeds it to fungot through a big alchemical funnel. I didn’t say anything
14:21:22 <fungot> arseniiv: they are? i see a wiki as a collaborative enterprise when quality emerges out of iterative editing by various contributors. it's not like i'm going to an oriental philosophy class, only in a very very subset of) bf.
14:26:25 -!- Lord_of_Life has quit (Read error: Connection reset by peer).
14:26:53 -!- mistbreeze has quit (Remote host closed the connection).
14:27:21 <nakilon> arseniiv damn mages
14:27:21 <fizzie> The *process* itself isn't unspecified, it's all defined step-by-step in https://github.com/fis/fungot/blob/master/varikn/readme.txt
14:27:21 <fungot> fizzie: there have been some problem in my c/ c++ compiled scheme would integrate much better with an expression that the evaluator just needs to know what is a explorbation?
14:32:26 -!- slavfox has quit (Ping timeout: 268 seconds).
14:33:03 -!- dyeplexer has quit (Ping timeout: 268 seconds).
14:33:36 -!- Hooloovoo has quit (Read error: Connection reset by peer).
14:38:50 -!- HackEso has quit (Remote host closed the connection).
14:44:09 -!- ircseeker has quit (*.net *.split).
14:44:09 -!- mnrmnaugh has quit (*.net *.split).
14:44:10 -!- sprock has quit (*.net *.split).
14:44:10 -!- dbohdan has quit (*.net *.split).
14:44:10 -!- hanif has quit (*.net *.split).
14:44:10 -!- hendursaga has quit (*.net *.split).
14:44:10 -!- Noisytoot has quit (*.net *.split).
14:44:10 -!- cd has quit (*.net *.split).
14:44:10 -!- jix has quit (*.net *.split).
14:44:10 -!- integral has quit (*.net *.split).
14:44:10 -!- perlbot has quit (*.net *.split).
14:44:10 -!- jinn has quit (*.net *.split).
14:44:10 -!- fizzie has quit (*.net *.split).
14:44:10 -!- ski has quit (*.net *.split).
14:44:10 -!- pikhq has quit (*.net *.split).
14:44:10 -!- lifthrasiir has quit (*.net *.split).
14:44:10 -!- mich181189 has quit (*.net *.split).
14:44:10 -!- relrod has quit (*.net *.split).
14:44:10 -!- keegan has quit (*.net *.split).
14:44:10 -!- shikhin has quit (*.net *.split).
14:44:10 -!- oren has quit (*.net *.split).
14:44:10 -!- Taneb has quit (*.net *.split).
14:44:10 -!- sknebel has quit (*.net *.split).
14:44:10 -!- FireFly has quit (*.net *.split).
14:44:10 -!- lambdabot has quit (*.net *.split).
14:44:10 -!- zegalch has quit (*.net *.split).
14:44:10 -!- sebbu has quit (*.net *.split).
14:44:10 -!- kluk has quit (*.net *.split).
14:44:10 -!- archenoth has quit (*.net *.split).
14:44:10 -!- V has quit (*.net *.split).
14:44:10 -!- Melvar has quit (*.net *.split).
14:44:10 -!- ProofTechnique has quit (*.net *.split).
14:44:10 -!- citrons has quit (*.net *.split).
14:44:11 -!- MrAureliusR has quit (*.net *.split).
14:44:11 -!- nakilon has quit (*.net *.split).
14:44:11 -!- Soni has quit (*.net *.split).
14:44:11 -!- fungot has quit (*.net *.split).
14:44:11 -!- APic has quit (*.net *.split).
14:44:11 -!- b_jonas has quit (*.net *.split).
14:44:11 -!- joast has quit (*.net *.split).
14:44:11 -!- op_4 has quit (*.net *.split).
14:44:11 -!- velik has quit (*.net *.split).
14:44:11 -!- arseniiv has quit (*.net *.split).
14:44:11 -!- river has quit (*.net *.split).
14:44:11 -!- simcop2387 has quit (*.net *.split).
14:44:11 -!- Cale has quit (*.net *.split).
14:44:11 -!- j4cbo has quit (*.net *.split).
14:44:11 -!- Bowserinator has quit (*.net *.split).
14:44:11 -!- zgrep has quit (*.net *.split).
14:44:12 -!- Trieste has quit (*.net *.split).
14:44:12 -!- Deewiant has quit (*.net *.split).
14:44:12 -!- iovoid has quit (*.net *.split).
14:44:12 -!- myname has quit (*.net *.split).
14:44:12 -!- leah2 has quit (*.net *.split).
14:44:12 -!- int-e has quit (*.net *.split).
14:44:12 -!- lucky has quit (*.net *.split).
14:44:12 -!- zzo38 has quit (*.net *.split).
14:44:12 -!- shachaf has quit (*.net *.split).
14:44:12 -!- dnm has quit (*.net *.split).
14:44:12 -!- user3456 has quit (*.net *.split).
14:49:30 -!- hanif has joined.
14:49:30 -!- HackEso has joined.
14:49:30 -!- Hooloovo- has joined.
14:49:30 -!- dyeplexer has joined.
14:49:30 -!- slavfox has joined.
14:49:30 -!- mistbreeze has joined.
14:49:30 -!- Lord_of_Life has joined.
14:49:30 -!- fungot has joined.
14:49:30 -!- hendursaga has joined.
14:49:30 -!- arseniiv has joined.
14:49:30 -!- Noisytoot has joined.
14:49:30 -!- APic has joined.
14:49:30 -!- sprock has joined.
14:49:30 -!- dbohdan has joined.
14:49:30 -!- river has joined.
14:49:30 -!- simcop2387 has joined.
14:49:30 -!- FireFly has joined.
14:49:30 -!- sknebel has joined.
14:49:30 -!- lambdabot has joined.
14:49:30 -!- nakilon has joined.
14:49:30 -!- archenoth has joined.
14:49:30 -!- Bowserinator has joined.
14:49:30 -!- leah2 has joined.
14:49:30 -!- cd has joined.
14:49:30 -!- zegalch has joined.
14:49:30 -!- lucky has joined.
14:49:30 -!- int-e has joined.
14:49:30 -!- zgrep has joined.
14:49:30 -!- Soni has joined.
14:49:30 -!- b_jonas has joined.
14:49:30 -!- Trieste has joined.
14:49:30 -!- V has joined.
14:49:30 -!- jix has joined.
14:49:30 -!- integral has joined.
14:49:30 -!- zzo38 has joined.
14:49:30 -!- perlbot has joined.
14:49:30 -!- Melvar has joined.
14:49:30 -!- jinn has joined.
14:49:30 -!- ProofTechnique has joined.
14:49:30 -!- joast has joined.
14:49:30 -!- shachaf has joined.
14:49:30 -!- fizzie has joined.
14:49:30 -!- dnm has joined.
14:49:30 -!- citrons has joined.
14:49:30 -!- ski has joined.
14:49:30 -!- Cale has joined.
14:49:30 -!- pikhq has joined.
14:49:30 -!- j4cbo has joined.
14:49:30 -!- lifthrasiir has joined.
14:49:30 -!- sebbu has joined.
14:49:30 -!- kluk has joined.
14:49:30 -!- op_4 has joined.
14:49:30 -!- MrAureliusR has joined.
14:49:30 -!- oren has joined.
14:49:30 -!- velik has joined.
14:49:30 -!- relrod has joined.
14:49:30 -!- mich181189 has joined.
14:49:30 -!- shikhin has joined.
14:49:30 -!- Taneb has joined.
14:49:30 -!- keegan has joined.
14:49:30 -!- user3456 has joined.
14:49:30 -!- Deewiant has joined.
14:49:30 -!- iovoid has joined.
14:49:30 -!- myname has joined.
14:55:00 -!- mnrmnaugh has joined.
15:00:06 <nakilon> the only quotes list I've found contains 177 and it does not contain random ones that I find elsewhere (
15:02:00 <fizzie> The department VariKN is from, which I think at the time was probably the Department of Information and Computer Science at the Helsinki University of Technology, but both of those names have changed since then.
15:02:26 <fizzie> http://hack.esolangs.org/repo/file/tip/quotes is what `quote looks at in.
15:02:28 <fizzie> Off for a bit.
15:03:45 <nakilon> I mean terry's quotes
15:04:35 <esolangs> [[Deadfish]] https://esolangs.org/w/index.php?diff=83743&oldid=82781 * Tux1 * (+374)
15:09:30 <esolangs> [[Channeler]] https://esolangs.org/w/index.php?diff=83744&oldid=83728 * Doridian * (+430) Add memory cell specifications
15:10:58 -!- hanif has quit (Ping timeout: 252 seconds).
15:18:23 -!- imode has joined.
15:20:29 <nakilon> the second most annoying bug that I can't do anything about is in Github repo Insights tab the Network tab doesn't show the tree until you refresh the page
15:21:45 <nakilon> it's been years already and is not my local problem because I started seeing it on machine in the office
15:31:01 -!- xkapastel has joined.
16:06:09 -!- xkapastel has quit (Quit: .).
16:17:06 -!- hanif has joined.
16:25:46 -!- esolangs has joined.
16:25:46 -!- ChanServ has set channel mode: +v esolangs.
16:45:17 -!- arseniiv has quit (Ping timeout: 252 seconds).
16:52:28 -!- arseniiv has joined.
17:08:44 -!- tromp has joined.
17:21:26 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
17:23:17 -!- tromp has joined.
17:28:04 <nakilon> "I only maintain the accursed thing, I don't actually read it. " (c) Chris Pressey about Trefunge-98
17:28:24 <nakilon> *Trefunge-98 specification
17:31:11 <b_jonas> fizzie: yes, that's the traditional explanation (giving you food as an offer of friendship and to help you). nor is that exclusive with the evolutionary perspective really.
17:31:39 -!- dcristofani has joined.
17:33:48 <b_jonas> nakilon: .pl is a confusing extension that means either perl or prolog. you can use .pm as extension for your perl files instead if that's confusing. pascal does not use .pl extension.
17:37:25 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
17:38:11 -!- dcristofani has quit (Ping timeout: 265 seconds).
18:06:07 <fizzie> Pascal uses .pas, was it?
18:06:38 <esolangs> [[NDBall]] https://esolangs.org/w/index.php?diff=83745&oldid=82580 * Tux1 * (-136)
18:07:25 <esolangs> [[NDBall]] M https://esolangs.org/w/index.php?diff=83746&oldid=83745 * Tux1 * (+0)
18:08:05 <esolangs> [[NDBall]] M https://esolangs.org/w/index.php?diff=83747&oldid=83746 * Tux1 * (+1) /* First Method: Point */
18:10:36 <esolangs> [[User talk:]] N https://esolangs.org/w/index.php?oldid=83748 * Hypocritical * (+5) Created page with "hello"
18:14:10 <b_jonas> fizzie: yes, usually
18:14:22 <esolangs> [[Channeler]] https://esolangs.org/w/index.php?diff=83749&oldid=83744 * Doridian * (+287) Add lowercase h for numeric channels
18:15:52 -!- hanif has quit (Quit: quit).
18:16:48 -!- Sgeo has joined.
18:17:30 <esolangs> [[Channeler]] https://esolangs.org/w/index.php?diff=83750&oldid=83749 * Doridian * (-281) Remove jump opcode, not in the spirit of this really
18:19:49 -!- dyeplexer has quit (Remote host closed the connection).
18:19:49 -!- tromp has joined.
18:24:56 <esolangs> [[User:PythonshellDebugwindow/Sandbox]] M https://esolangs.org/w/index.php?diff=83751&oldid=75210 * PythonshellDebugwindow * (+142) /* 1 */
18:26:11 -!- Guest81 has joined.
18:26:33 -!- Guest81 has quit (Client Quit).
18:26:41 <nakilon> I don't see an article on this one https://github.com/Property404/hdbf
18:30:31 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
19:01:29 <esolangs> [[Channeler]] https://esolangs.org/w/index.php?diff=83752&oldid=83750 * Doridian * (+98) Forgot documenting pow
19:13:15 -!- mistbreeze has quit (Read error: Connection reset by peer).
19:14:25 <esolangs> [[123]] https://esolangs.org/w/index.php?diff=83753&oldid=67469 * Bangyen * (+85) /* External resources */
19:29:01 -!- tromp has joined.
19:30:10 -!- tromp has quit (Client Quit).
19:45:51 <fizzie> 20:41 !irc.invalid pounce is GPLv3 fwee softwawe ^w^ code is avaiwable fwom https://git.causal.agency/pounce
19:45:54 <fizzie> I'm potentially having some second doubts about this bouncer, but maybe I'll still give it a go.
19:46:06 <fizzie> (Trying to migrate away from bip.)
19:47:25 <esolangs> [[Language list]] https://esolangs.org/w/index.php?diff=83754&oldid=83708 * Aspwil * (+8) /* Non-alphabetic */
19:52:58 <esolangs> [[Language list]] https://esolangs.org/w/index.php?diff=83755&oldid=83754 * Aspwil * (+97) /* Non-alphabetic */
19:53:39 <esolangs> [[Language list]] https://esolangs.org/w/index.php?diff=83756&oldid=83755 * Aspwil * (+1) /* Non-alphabetic */
19:56:58 -!- Thelie has joined.
20:11:28 <esolangs> [[NDBall]] https://esolangs.org/w/index.php?diff=83757&oldid=83747 * Tux1 * (-222) formalized
20:18:09 <oren>
20:21:53 <esolangs> [[Channeler]] M https://esolangs.org/w/index.php?diff=83758&oldid=83752 * Doridian * (+0) Fix wrong channel name
20:23:39 <esolangs> [[Channeler]] M https://esolangs.org/w/index.php?diff=83759&oldid=83758 * Doridian * (+52)
20:23:48 <esolangs> [[Channeler]] M https://esolangs.org/w/index.php?diff=83760&oldid=83759 * Doridian * (-107)
20:24:45 <esolangs> [[Channeler]] https://esolangs.org/w/index.php?diff=83761&oldid=83760 * Doridian * (+133) Restore accidental removal
20:48:37 -!- dcristofani has joined.
20:55:27 <esolangs> [[Special:Log/newusers]] create * Loris redstone * New user account
20:56:40 <esolangs> [[Esolang:Introduce yourself]] https://esolangs.org/w/index.php?diff=83762&oldid=83741 * Loris redstone * (+38)
21:18:19 <imode> thinking about petri nets.
21:28:37 -!- tromp has joined.
21:29:42 -!- dcristofani has quit (Ping timeout: 264 seconds).
21:35:47 <Taneb> imode: petri nets are neat
21:41:14 <imode> they are. but they're pretty limited.
21:42:13 <esolangs> <oren> Oh my god, I just want the CSS incantation for "centred horizontally but aligned to the pixel grid so its not blurry"
21:43:07 <esolangs> <oren> -moz-stop-blurring-my-text: yes;
21:43:54 <oren> oh neat it gets echoed to here
21:52:27 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
22:01:10 <imode> what would an analogue to the pi calculus as the turing machine is to lambda calculus?
22:01:44 -!- tromp has joined.
22:11:25 <river> nondeterministic turing machine maybe?
22:18:30 <arseniiv> something with many submachines interacting?..
22:20:04 <imode> it's interesting that there's no direct analogue.
22:20:10 <arseniiv> though like in a normal TM, all the lambdas don’t correspond to anything specially defined inside it, then maybe π processes shouldn’t give submachines too
22:20:39 <arseniiv> maybe register machines or other formalisms would pave the way more easily?
22:21:36 <arseniiv> and then it might become obvious if there’s a natural Turing-like analogy and then what it is
22:21:58 <zzo38> I think that objects (including text) should always be drawn aligned to the pixel grid, regardless of the CSS. Non-blurry text should be the user setting, probably.
22:22:06 <imode> I was thinking in that sense. but encoding data structures in register machines is tedious.
22:22:22 <imode> and rather computationally expensive.
22:23:26 <arseniiv> tried generative art through SVG and Python with package svgwrite to lessen boilerplate. Made some fractal-y thing though not too interesting
22:24:22 <arseniiv> imode: that’s why I came up with generalized Minsky machines wink wink
22:24:48 <arseniiv> okay I’m going to sleep
22:25:00 <imode> lol.
22:25:33 -!- salpynx has joined.
22:25:36 <imode> sweet dreams.
22:25:59 <arseniiv> (but seriously. using terms represented as trees should be not as expensive as reparsing natural numbers)
22:26:00 <arseniiv> thanks
22:26:51 <imode> I've been trying to figure out what constitutes a model of "mobile agents", where there are some "places" that pieces of self-contained code can move between in order to interact with their local environment.
22:28:31 <imode> but there's this separation between the "places" that things exist at and the processes that are actually moving between them.
22:29:58 <imode> the processes need some kind of behavior, and a way to interact with their surrounding environment, but this means that the behavior of the process (its state) is all munged along a barrier that separates the inside of the process from the outside.
22:30:54 -!- arseniiv has quit (Ping timeout: 264 seconds).
22:31:30 <imode> imo "places" and "agents" should probably be the same kind of thing. but then you deal with the idea of communication between them, and all that other jazz, and it gets confusing, because you need multiple different rules that govern the inside and outside of agents.
22:32:21 <zzo38> In Magic: the Gathering, the text box may include ability text, flavor text, ability words, and reminder text; only the ability text is the part of the AST. What might be useful is also "descriptive text", which is not ability text but describes either some part of the AST which cannot easily be represented otherwise (e.g. if the mana cost or subtypes won't fit in those spaces) or something external to the game which normally affects it
22:33:38 <zzo38> Descriptive text is not a characteristic-defining ability, is not the "text box" part of the AST, and applies before anything else; it might define part of the card's initial text.
22:34:21 <imode> I guess my thinking is that the process/place split is inelegant because what goes on inside a process is not an extension of what goes on outside a process, and vice versa.
22:34:25 <zzo38> They wanted to reprint a card with a color indicator which is all five, but they didn't know how to do that and decided to write an ability to add that instead.
22:35:07 <zzo38> Another way would to be draw all five color indicators, but if that won't fit, then this idea (I don't know how it would be formatted) would make it so without being errata.
22:37:50 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
22:43:08 <imode> I have a feeling that the ideas I'm thinking of lie somewhere within dataflow models, state machines, and ring networks, but I can't prove it.
22:44:32 <salpynx> if you mean abstract spaces, the agents would 'extrude' the topology they inhabit by the possible relationships they have with each other.
22:45:12 <imode> right, but there always needs to be some kind of static background to determine what connections/relationships are possible.
22:45:15 <salpynx> ('extrude' is probably the wrong word, what is the term?) The topology would emerge from their behaviour
22:45:24 <imode> generative communication.
22:46:18 <imode> I imagine that at any point, where agents know and understand how to communicate with one another, there's a certain configuration of agents laid against a static background.
22:46:44 <imode> where that background contains the topology for all possible connections. in situations where all agents can communicate with eachother, you have the complete graph.
22:46:47 <salpynx> you could have dummy agents that enforce some particular structure and are just there to route
22:46:54 <imode> yup.
22:47:41 <imode> but what those agents are made of, their behavior, that's an entirely separate "theory" from how they behave and interact at the macro-scale.
22:47:52 <imode> it's like taking a look inside a cell vs. how that cell interacts with others. it's two separate models.
22:47:58 <imode> unifying it into one is difficult.
22:50:45 <salpynx> if you mean physical places, that's set by your hardware configuration, and you just have to accept that :)
22:51:06 <imode> nah. you're aiming for a model of computation.
22:51:20 <imode> abstract ones.
22:51:34 <imode> the rules for an agent moving around on a network are separate from the ones that allow an agent to compute things.
22:51:58 <esolangs> [[Channeler]] M https://esolangs.org/w/index.php?diff=83763&oldid=83761 * Doridian * (-25) Actually error on invalid ops
22:51:59 <imode> I guess the best analogy would be... if I shift to the left at a particular point on a turing machine tape, I end up in another "place".
22:52:15 <salpynx> sounds like is an addressing problem? How to locate = how to address?
22:52:17 <imode> say another computer over the network. but it'd be a logical place.
22:52:45 <imode> your program needs to travel along some topology to interact with another program and then carry some result back, for example.
22:53:33 <imode> what does your program carry on its back. nothing? then you'll have to leave your state "on the ground", per se, and come back to it at a particular point.
22:54:47 <imode> "travel in direction X until you see a box Y. then, pick up the box and head the opposite of direction X until you reach the place you started. put the box down and repeat."
22:54:51 <imode> hm.
22:55:38 <salpynx> In another tab I'm looking into the Cantor pairing function, to fold 2d coords into 1d, and was going to extend that to higher dimensions, each folded down... was wondering if there are other mappings to fold down dimensions
22:55:58 <imode> space-filling curves are probably what you're looking for.
22:58:20 <salpynx> comments above reminded me I was going to check whether that hdbf (hyperdimensional bf) link nakilon posted earlier had a sensible and consistent topology, i.e could you loop in an arbitrary 2d plane and end up in the same place.
22:59:03 <salpynx> hdby also made me wonder how you could 'order' dimensions, to incr and decr dimension in any kind of objective order
22:59:07 <salpynx> hdbf
23:00:42 <nakilon> IIRC salpynx https://www.reddit.com/r/programming/comments/4f78by/hyperdimensional_brainfuck_brainfuck_with/ this thread also had two links to other brainfuck variations
23:01:34 <esolangs> [[Channeler]] https://esolangs.org/w/index.php?diff=83764&oldid=83763 * Doridian * (+6) Fix comments to say more than just space is allowed for int termination
23:02:39 <imode> ohh yeah.
23:02:51 <imode> interdimensional travel lmao.
23:03:42 <salpynx> navigating on an arbitrary graph is nice because you don't have to worry about how many dimensions there are, as many as you need, but the problem is then you can't model using any predefined knowledge about what to expect if you move in a particular path
23:03:57 <esolangs> [[Language list]] https://esolangs.org/w/index.php?diff=83765&oldid=83756 * Doridian * (+16) Add Channeler to language list
23:04:28 <imode> yeah.. the complexity explodes.
23:05:02 <esolangs> [[Program Number System]] N https://esolangs.org/w/index.php?oldid=83766 * Aspwil * (+1903) Created page with "'''Program Number System''' or PNS is a system designed by ~~~ to take every useable program in a language and turn it into a unique number, that can then be converted back. t..."
23:06:25 <salpynx> ... unless your edges are labelled with some kind of indicator of pre-defined notation, so there's a sub graph with a fixed and well defined topology
23:06:53 <imode> which at that point, why not go the reverse and embed whatever graph you want within a static topology.
23:06:53 <esolangs> [[User:Aspwil]] https://esolangs.org/w/index.php?diff=83767&oldid=77508 * Aspwil * (+88)
23:09:08 <esolangs> [[Program Number System]] https://esolangs.org/w/index.php?diff=83768&oldid=83766 * Aspwil * (+29)
23:09:27 <salpynx> I wonder if you can get the agents to fill in the gaps and find more direct routes by adding intermediate nodes, they can fill in the gaps and resolve inconsistencies?
23:10:20 <salpynx> Get the agents to decide if it's a ring or a hypercube or whatever, so you don't have to!
23:12:24 <imode> that's kind of the idea! the trick is picking a background. I figure something like a ring is good enough, but you have to push the power of calculating that curve onto the agents themselves.
23:14:07 <esolangs> [[Channeler]] https://esolangs.org/w/index.php?diff=83769&oldid=83764 * Doridian * (+182) Add better cat program
23:14:27 <esolangs> [[Channeler]] M https://esolangs.org/w/index.php?diff=83770&oldid=83769 * Doridian * (+7)
23:18:05 <esolangs> [[Channeler]] M https://esolangs.org/w/index.php?diff=83771&oldid=83770 * Doridian * (+46)
23:18:12 -!- Thelie has quit (Remote host closed the connection).
23:19:19 <salpynx> If agents could have any number of undirected labelled and directed (next, previous + id tag ) labelled edges you could create any topology, just with the optional notation of next, previous, and room for tagged dimensions, as well as 'wormhole' connections that bypass static topologies, then the network could sort itself out, resolve ambiguities,
23:19:20 <salpynx> and try optimisations
23:20:01 <salpynx> .. if you had some clever rules
23:24:06 <salpynx> hmm, all the agents would have to understand that the arbitrarily tagged 'dimension' x had evolved into a ring to use it consistently -- where would they store that knowledge, and how to share it, in order to make it consistent?
23:25:44 <imode> ding.
23:26:25 <salpynx> agents need to be able to model their environment, if you want them to modify their environment, explore it, and make community decisions to improve it, you're getting into AI territory
23:26:36 <esolangs> [[Program Number System]] https://esolangs.org/w/index.php?diff=83772&oldid=83768 * Aspwil * (+687)
23:27:34 <salpynx> That'll be why static topologies are good for dumb agents
23:30:46 <imode> if you pick a singular topology, I guarantee you there will be some kind of gradient related to how much state you allow individual elements of that topology to store and compute.
23:31:06 <imode> cellular automata are on one end. some form of mobile processes are on the other end.
23:32:44 <salpynx> Yes, even CA cells have an environment model (static)
23:33:22 <salpynx> fongot, how do you model your enviromment?
23:33:37 <salpynx> fungot, how do you model your enviromment?
23:33:37 <fungot> salpynx: the masamune! how can i do for you? thanks! it's about the missing queen. she still looks so much like leene, that they will take you to your place of execution?! strange, but!?
23:34:35 <salpynx> oh, someone changed the model (I think?), I was expecting something more likely to be on topic.
23:35:34 <b_jonas> welcome, oren
23:36:21 <esolangs> [[Talk:Uyjhmn n]] https://esolangs.org/w/index.php?diff=83773&oldid=65787 * Monochromeninja * (+111) /* File Extension */ new section
23:36:34 <imode> CA cells are finite state machines in nature, but depending on the CA, they either rely on the multiplicity of nearby types of cells, or on the specific arrangement of their neighbors.
23:37:19 <imode> so they can examine their locality, but they cannot be separated from their locality. if you're gonna program anything that travels, you need that to be a part of the rule set that _everything_ embodies.
23:37:40 <imode> meaning that every cell needs to understand how a data element can move left or right.
23:38:19 <imode> I wonder what would happen if you traveled slightly towards the "mobile processes" end of the gradient.
23:38:52 <imode> you'd probably get something like a turing machine.
23:40:09 <imode> I'm imagining that "transporting your state" at that point is like moving and managing a physical asset.
23:40:39 <imode> gotta move things one by one, and build barriers that prevent others from viewing what you don't want them to view.
23:43:54 <salpynx> I had an idea of making 'smart' CA agents as a kind of game, a cell could be an agent containing arbitrary code to interact with a CA API to read adjacent cells, modify adjacent cells, and set its next position, so it could try to survive and potentially build things to protect itself (or hunt opponent agents)
23:44:56 <imode> you'll find that that's what dave ackley's movable feast machine is.
23:47:34 <salpynx> all the cells are smart in that though aren't they? I wonder if that'd help to have some number of simple FSM agents that could be marshalled by smarter agents with higher level programming and goals
23:47:44 <imode> not all of them are smart.
23:48:04 <imode> each cell _can_ contain an agent.
23:48:15 <imode> and can spawn agents adjacently.
23:48:21 <imode> but they can't materialize things out of nowhere.
23:51:27 <salpynx> has anyone extended the idea of cellular automatons to graph automatons? (nodular automatons??) They could do 2d, 3d, but also less structured things. Would be hard to visualise, I imagine.
23:51:59 <imode> cellular automata on a graph?
23:53:04 <salpynx> yes, they'd need to extend their playfield, maybe that's not as easy to translate as I first thought
23:53:31 <imode> you could absolutely define CAs on arbitrary graphs, but they do need to be static.
←2021-05-30 2021-05-31 2021-06-01→ ↑2021 ↑all