←2023-08-20 2023-08-21 2023-08-22→ ↑2023 ↑all
00:30:47 -!- razetime has quit (Ping timeout: 245 seconds).
00:31:31 -!- razetime has joined.
00:35:51 -!- razetime has quit (Ping timeout: 245 seconds).
00:36:13 -!- razetime has joined.
00:44:27 -!- zzo38 has joined.
00:57:24 -!- razetime has quit (Ping timeout: 246 seconds).
00:57:55 -!- razetime has joined.
01:42:11 -!- Melvar has quit (Ping timeout: 260 seconds).
01:54:16 <zzo38> Is there a name of the sequence of the run length of Thue-Morse?
01:56:41 -!- Melvar has joined.
02:04:28 -!- ais523 has quit (Ping timeout: 248 seconds).
02:11:42 -!- siesta has joined.
02:20:14 <esolangs> [[Mic]] N https://esolangs.org/w/index.php?oldid=113993 * D * (+1561) Created page with "[[Mic]] is a [[stack]]-based [[esoteric programming language]] created by [[User:D]], loosely based on the concept of primitive recursive functions. == Data Structures == The primary data structure is a numeric stack, but there's also an "accumulator" (which, despite the name, i
02:22:17 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=113994&oldid=113993 * D * (+5)
02:23:00 -!- razetime has quit (Ping timeout: 245 seconds).
02:23:37 -!- razetime has joined.
02:45:38 <esolangs> [[Mic]] https://esolangs.org/w/index.php?diff=113995&oldid=113994 * D * (-164) Actually, I'll just make it Turing-incomplete anyway
02:46:03 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=113996&oldid=113995 * D * (-95)
02:46:42 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=113997&oldid=113996 * D * (-14)
02:50:56 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=113998&oldid=113997 * D * (+235)
02:51:10 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=113999&oldid=113998 * D * (-42)
02:54:07 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114000&oldid=113999 * D * (+261)
02:57:13 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114001&oldid=114000 * D * (+19)
02:58:00 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114002&oldid=114001 * D * (+16)
02:59:10 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114003&oldid=114002 * D * (-8)
03:01:42 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114004&oldid=114003 * D * (-46)
03:02:43 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114005&oldid=114004 * D * (+0)
03:05:27 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114006&oldid=114005 * D * (-9)
03:06:39 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114007&oldid=114006 * D * (+14)
03:13:49 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114008&oldid=114007 * D * (+133)
03:14:33 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114009&oldid=114008 * D * (+185)
03:45:06 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114010&oldid=114009 * D * (+872)
03:45:35 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114011&oldid=114010 * D * (+0)
03:46:01 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114012&oldid=114011 * D * (-33) /* Primitive Recursion */
04:23:46 -!- razetime has quit (Ping timeout: 245 seconds).
04:58:36 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114013&oldid=114012 * D * (+38)
05:01:58 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114014&oldid=114013 * D * (+153)
05:02:21 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114015&oldid=114014 * D * (-51)
05:24:56 <siesta> guys please look at my paint script it is in python3 https://github.com/rald/paint-script
05:25:15 <siesta> can you consider it an esolang?
05:36:01 -!- bgs has joined.
06:24:17 -!- razetime has joined.
07:07:01 <river> https://mysterymath.github.io/simple_cpu/ https://laughtonelectronics.com/Arcana/One-bit%20computer/One-bit%20computer.html
07:07:16 <river> siesta: looked. yes
07:08:43 <siesta> http://fria.bsdforall.org/images/fac0.jpg
07:09:03 <siesta> http://ix.io/4E98
07:09:26 <siesta> i did it turtle graphics in paint script
08:00:22 -!- Sgeo has quit (Read error: Connection reset by peer).
08:05:39 -!- siesta has quit (Quit: ...zzzZZZ).
08:05:53 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114016&oldid=114015 * D * (+6)
08:08:21 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114017&oldid=114016 * D * (-5) /* Looping Commands */
08:27:57 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114018&oldid=114017 * D * (+196)
08:29:31 <esolangs> [[Mic]] M https://esolangs.org/w/index.php?diff=114019&oldid=114018 * D * (+0)
08:35:41 -!- razetime has quit (Ping timeout: 246 seconds).
08:42:37 <esolangs> [[Catshark]] N https://esolangs.org/w/index.php?oldid=114020 * D * (+816) Created page with "[[Catshark]] is a [[Joke language list|Joke language]] based on [[Deadfish]]. Instead of one accumulator, Catshark has two accumulators. The language is called Catshark because catsharks have two heads. (So you'll have to eat two raw fishheads per fish.) == Commands == Ther
08:42:51 <esolangs> [[Catshark]] M https://esolangs.org/w/index.php?diff=114021&oldid=114020 * D * (+23)
08:43:15 <esolangs> [[Catshark]] M https://esolangs.org/w/index.php?diff=114022&oldid=114021 * D * (-1)
08:56:30 -!- razetime has joined.
09:04:04 <esolangs> [[Catshark]] https://esolangs.org/w/index.php?diff=114023&oldid=114022 * D * (-70)
09:04:23 <esolangs> [[Catshark]] M https://esolangs.org/w/index.php?diff=114024&oldid=114023 * D * (+19)
09:06:29 <esolangs> [[Catshark]] M https://esolangs.org/w/index.php?diff=114025&oldid=114024 * D * (-58)
09:53:05 -!- ais523 has joined.
10:01:04 -!- Thelie has joined.
10:19:41 <esolangs> [[Addition Automaton]] https://esolangs.org/w/index.php?diff=114026&oldid=113990 * Ais523 * (+225) link to the original CGCC posts
10:21:53 -!- SGautam has joined.
10:31:24 <esolangs> [[Befunge]] M https://esolangs.org/w/index.php?diff=114027&oldid=113967 * None1 * (+58) /* 99 Bottles of Beer */
11:28:48 -!- zzo38 has quit (Ping timeout: 246 seconds).
11:33:54 -!- wib_jonas has joined.
11:36:47 <esolangs> [[Catshark]] M https://esolangs.org/w/index.php?diff=114028&oldid=114025 * D * (-19) You don't need halt to be TC
11:45:57 <esolangs> [[The Waterfall Model]] M https://esolangs.org/w/index.php?diff=114029&oldid=70335 * Ais523 * (-1) /* ZISC interpretation */ fix typo
11:46:31 <esolangs> [[Catshark]] M https://esolangs.org/w/index.php?diff=114030&oldid=114028 * D * (+585)
11:50:36 <esolangs> [[Catshark]] M https://esolangs.org/w/index.php?diff=114031&oldid=114030 * D * (+142)
11:50:52 <esolangs> [[Catshark]] M https://esolangs.org/w/index.php?diff=114032&oldid=114031 * D * (+22)
11:52:56 -!- razetime has quit (Ping timeout: 260 seconds).
11:55:29 <esolangs> [[Catshark]] M https://esolangs.org/w/index.php?diff=114033&oldid=114032 * D * (+497) /* Implementations */
11:55:46 <esolangs> [[Catshark]] M https://esolangs.org/w/index.php?diff=114034&oldid=114033 * D * (-5) /* Computational class */
11:56:59 <esolangs> [[Catshark]] M https://esolangs.org/w/index.php?diff=114035&oldid=114034 * D * (+50)
11:57:07 <esolangs> [[Catshark]] M https://esolangs.org/w/index.php?diff=114036&oldid=114035 * D * (+1) /* Computational class */
12:19:31 -!- razetime has joined.
12:20:39 <esolangs> [[User:None1]] https://esolangs.org/w/index.php?diff=114037&oldid=113976 * None1 * (+66)
12:41:19 <esolangs> [[Special:Log/upload]] upload * None1 * uploaded "[[File:BinaryLanguage.png]]": Official logo of BinaryLanguage
12:47:41 <esolangs> [[BinaryLanguage]] N https://esolangs.org/w/index.php?oldid=114039 * None1 * (+4183) Created page with "{{infobox proglang |name=BinaryLanguage [[File:BinaryLanguage.png|thumb|center|Official logo of BinaryLanguage]] |author=[[User:None1|None1]] |year=[[:Category:2023|2023]] |paradigms=unknown |typesys= |memsys=[[:Category:Cell-based|Cell-based]] |class=[[:Category:T
12:49:02 <esolangs> [[User:None1]] https://esolangs.org/w/index.php?diff=114040&oldid=114037 * None1 * (+71) /* My Esolangs */
12:49:32 -!- razetime has quit (Ping timeout: 245 seconds).
12:50:03 <esolangs> [[Language list]] https://esolangs.org/w/index.php?diff=114041&oldid=113992 * None1 * (+21) /* B */ +[[BinaryLanguage]]
12:50:24 <esolangs> [[BinaryLanguage]] M https://esolangs.org/w/index.php?diff=114042&oldid=114039 * None1 * (+20) /* Example Programs */
12:51:04 <esolangs> [[Truth-machine]] https://esolangs.org/w/index.php?diff=114043&oldid=113684 * None1 * (+48) /* Binary lambda calculus */ +[[BinaryLanguage]]
13:01:30 <wib_jonas> is there such a thing as a language to define editable table views? eg. in SQL I can make a view like SELECT employee.EmployeeName, employee.JobTitleId, jobtitle.EnglishName FROM employee LEFT JOIN jobtitle ON employee.JobTitleId = jobtitle.JobTitleId; assuming JobTitleId is the primary key of jobtitle, and then display that view as a table on a
13:01:31 <wib_jonas> webpage. but suppose I want to make the jobtitle.EnglishName column editable on that webpage. there are at least two natural ways to do that: either if you edit then it UPDATEs EnglishName in the corresponding row of the jobtitle page, or looks up the new value in the EnglishName column and UPDATEs the employee.JobTitleId field to point to that
13:01:31 <wib_jonas> jobtitle row instead. what I'm asking for is a language, possibly esoteric, that lets me define not just the contents of the view but what happens when the user modifies columns in it, ideally in a way that's easier in the common case than having to define custom triggers for each column.
13:05:50 <ais523> I'm trying to find a halting algorithm which, given some positive integer x, finds a square-free number y > x such that y consists only of the digits 7 and 8
13:06:02 <ais523> like, a provably halting algorithm
13:06:13 <ais523> in practice these things are almost trivial to find, but it's hard to prove that they always exist
13:07:02 <ais523> (fwiw I suspect the 8 isn't even necessary and you can do it with just 7 – obviously you can't do it with just 8, because 4 will always be a factor)
13:08:04 <ais523> one potential thought: are repunits with prime numbers of digits always square-free? it feels like they probably are, but I can't prove it and it doesn't seem to be an existing result
13:08:44 <wib_jonas> ais523: I think that's equivalent to wanting a proof that there is an infinite number squarefrees with such pattern, because the algorithm is trivial
13:09:15 <ais523> wib_jonas: yes, the hard part is the proof
13:09:22 <ais523> if you can prove it's always possible, you can just look for them by brute force
13:10:17 <wib_jonas> I don't think so, because 11 in base 3 isn't squarefree
13:10:33 <ais523> oh, that's a good point
13:12:56 <ais523> although, every number from 3 upwards is 11 in some base
13:13:14 <ais523> still, there's no obvious reason why 2-digit repunits should be a special case – after all, 2 is prime
13:13:24 <ais523> so if that isn't a special case, larger numbers could potentially be square too
13:13:28 <ais523> err, non-square-free
13:15:53 <esolangs> [[BinaryLanguage]] M https://esolangs.org/w/index.php?diff=114044&oldid=114042 * None1 * (+5) /* Turing Completeness */ Fixed table
13:18:27 <esolangs> [[BinaryLanguage]] M https://esolangs.org/w/index.php?diff=114045&oldid=114044 * None1 * (+28) It is a Turing Tarpit
13:20:15 <ais523> OK, so 111 in base 18 is 343 = 7³
13:20:29 <esolangs> [[Turing tarpit]] https://esolangs.org/w/index.php?diff=114046&oldid=102879 * None1 * (+50) /* Survey */
13:22:10 <esolangs> [[User:XKCD Random Number]] https://esolangs.org/w/index.php?diff=114047&oldid=113514 * None1 * (+41) /* Setlang */ Added [[Self-modifying Brainfuck]] implementation
13:23:36 <wib_jonas> ais523: 11111 in base 2 is divisible by 3**2
13:23:47 <wib_jonas> I have more countrexamples if you want
13:23:48 <esolangs> [[User:XKCD Random Number]] https://esolangs.org/w/index.php?diff=114048&oldid=114047 * None1 * (+27) /* Implementations */ +[[BinaryLanguage]]
13:23:53 <ais523> isn't 11111 31?
13:24:06 <wib_jonas> uh
13:25:58 <ais523> anyway, 11111 in base 27 is 551881 which is 11²×4561
13:26:11 <ais523> (I have also been running a computer search for these things)
13:27:10 <ais523> `! jelly 1x5ḅ27ṄÆf
13:27:12 <HackEso> ​/hackenv/bin/!: line 4: /hackenv/ibin/jelly: No such file or directory
13:27:16 <ais523> aww
13:27:36 <wib_jonas> (11111 in base 3) is 11**2
13:28:09 <esolangs> [[User:XKCD Random Number]] https://esolangs.org/w/index.php?diff=114049&oldid=114048 * None1 * (+126) /* Implementations */
13:28:13 <ais523> looks like I will need a different approach to finding these numbers
13:29:35 <esolangs> [[User:XKCD Random Number]] https://esolangs.org/w/index.php?diff=114050&oldid=114049 * None1 * (+75) /* F!-- */ +[[Factor]]
13:29:59 -!- wib_jonas has quit (Quit: Client closed).
13:30:20 -!- wib_jonas has joined.
13:41:43 <wib_jonas> (18**3-1)/17 is 7**3; (22**3-1)/21 is divisible by 13**2
13:45:14 <wib_jonas> this type of statement of infinitely many squarefrees is the kind of statement that I'm not good at proving, I can just make a computer check small cases
13:45:48 <ais523> I'm not good at proving it either
13:47:23 <wib_jonas> I don't even know what that area of mathematics is called: I used to think it was called combinatorial number theory, but https://mathoverflow.net/questions/tagged/combinatorial-number-theory is so small while these problems are popular so that's apparently not the name for the area
13:47:38 <ais523> over half of the integers are square-free, apparently
13:48:03 <ais523> so it would be very weird if numbers formed of only 7s and 8s somehow suddenly stopped being square-free forever
13:49:09 <wib_jonas> I don't think that's enough of an argument, your pattern is too regular
13:49:33 <ais523> it isn't a proof, indeed
13:49:56 <ais523> it is a very strong conjecture when combined with the fact that the pattern doesn't show any particular inclination to stop being square-free…
13:50:02 <ais523> but still not a theorem
13:53:03 <int-e> Hmm so we can quickly establish bounds like "if q^2 | 10^p - 1 then q = 3 or q > 10^10"
13:54:02 <esolangs> [[BFFuck]] https://esolangs.org/w/index.php?diff=114051&oldid=113841 * None1 * (+77) /* Syntax */
13:54:18 <ais523> according to Wikpedia, there's a constant c such that for any n, there's always a square-free number in the range n…n+cn^(1/5)log n
13:54:48 <ais523> I don't think that solves the immediate problem, though, because you can't find a big consecutive range of integers formed out of 7s and 8s
13:55:03 <wib_jonas> ais523: https://mathoverflow.net/questions/291864/gaps-in-squarefree-numbers but I don't think that solves it either
13:57:30 <int-e> (if q^2 | 10^p - 1 then p | q(q-1), and p = q implies 10 = 1 (mod q), so only q = 3 works for that. So in all other cases, p | q-1 and we can enumerate and test all such p.)
14:00:20 <esolangs> [[Burgercamp]] M https://esolangs.org/w/index.php?diff=114052&oldid=113237 * None1 * (+0)
14:01:06 <esolangs> [[Category:Accumulator-based]] N https://esolangs.org/w/index.php?oldid=114053 * None1 * (+92) Created page with "Accumulator-based languages are based on accumulators, which are a finite number of numbers."
14:01:45 <ais523> ugh, someone is going to have to go and delete a bunch of categories again – I've been putting it off because I don't want to have to figure out which ones / fix all the resulting pages
14:02:04 <esolangs> [[BinaryLanguage]] M https://esolangs.org/w/index.php?diff=114054&oldid=114045 * None1 * (+45) It is not cell based
14:02:30 <esolangs> [[BinaryLanguage]] M https://esolangs.org/w/index.php?diff=114055&oldid=114054 * None1 * (+0)
14:04:13 <ais523> fwiw, this seems to be "register-based" which might actually be a good category, but it has been created at the wrong name
14:05:10 -!- Sgeo has joined.
14:15:35 <wib_jonas> [ q:<:10^21
14:15:35 <j-bot> wib_jonas: 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5
14:15:42 <wib_jonas> [ q:<:10x^21
14:15:42 <j-bot> wib_jonas: 3 3 3 37 43 239 1933 4649 10838689
14:16:09 <wib_jonas> [ q:<:10x^22
14:16:09 <j-bot> wib_jonas: 3 3 11 11 23 4093 8779 21649 513239
14:16:33 <int-e> (22 isn't prime)
14:16:39 <wib_jonas> yep, that's not prime length
14:17:11 <int-e> oh, neither is 21 of course.
14:17:48 <wib_jonas> but I think that goes against your bounds thing, doesn't it?\
14:18:03 <int-e> no
14:18:09 <int-e> my p and q are both prime
14:18:18 <wib_jonas> ah, you didn't say that. ok.
14:26:21 <int-e> wib_jonas: I mean if p isn't prime, the p | q(q-1) part still works. In your case, 22 | 110.
14:26:47 <int-e> But splitting that into p = q or p | q-1 doesn't work.
14:31:47 <esolangs> [[Catshark]] M https://esolangs.org/w/index.php?diff=114056&oldid=114036 * D * (+30)
14:33:53 <ais523> anyway, I think I've found a way to solve the underlying problem without this (turns out that my problem can be solved by finding a number for which the smallest repeated prime factor is greater than a given number, which can trivially be done by taking the factorial, adding 1, and then repeating 7 that many times)
14:34:00 <ais523> * the factorial of the given number
14:34:05 <esolangs> [[Deadfish]] M https://esolangs.org/w/index.php?diff=114057&oldid=113571 * D * (+91) /* Variants of deadfish */
14:34:25 <ais523> e.g. if I want to ensure a number has no repeated prime factor that's 5 or lower, I can just use (5!+1) (i.e. 121) copies of 7
14:35:54 <ais523> or, hmm, does that work?
14:40:35 <ais523> is it the case that all repunits with prime numbers of digits are necessarily coprime with all repunits in the same base with smaller numbers of digits?
14:44:14 <ais523> oh, here's a proof that works: for any number coprime with the base, it's a factor of *some* repunit: so you can multiply all the primes you don't want to see twice together, and find a repunit that's a multiple of all of them: then adding one more digit produces a repunit that's coprime to all of them, and thus none of them can appear twiec
14:44:49 <ais523> (or indeed, even once)
14:48:29 <int-e> the other thing also works... you take a multiple of the lcm of phi(p) of all the primes you want to avoid and add 1.
14:49:11 <int-e> so in particular p! + 1 where p is the largest prime of interest will work
14:51:33 <ais523> anyway, gtg for now
14:51:40 -!- ais523 has quit (Quit: quit).
15:01:21 -!- SGautam has quit (Quit: Connection closed for inactivity).
15:04:50 -!- wib_jonas has quit (Quit: Client closed).
15:05:09 -!- wib_jonas has joined.
15:29:04 <esolangs> [[BinaryLanguage]] M https://esolangs.org/w/index.php?diff=114058&oldid=114055 * PythonshellDebugwindow * (-6) /* External Resources */ category
15:29:22 -!- Thelie has quit (Read error: Connection reset by peer).
15:30:15 <esolangs> [[Burgercamp]] M https://esolangs.org/w/index.php?diff=114059&oldid=114052 * PythonshellDebugwindow * (+25) Categories
15:43:19 -!- rodgort has quit (Quit: Leaving).
15:46:27 -!- rodgort has joined.
16:04:47 -!- cheater has joined.
16:05:06 -!- cheater has left (Closing Window).
16:24:23 -!- razetime has joined.
16:59:07 -!- rodgort has quit (Ping timeout: 245 seconds).
17:08:00 -!- razetime has quit (Ping timeout: 245 seconds).
17:28:28 -!- rodgort has joined.
17:38:10 -!- wib_jonas has quit (Quit: Client closed).
17:40:00 -!- Thelie has joined.
17:49:10 -!- SGautam has joined.
18:08:15 -!- Wryl has quit.
18:27:37 -!- tromp has joined.
18:52:26 -!- Thelie has quit (Remote host closed the connection).
19:58:35 -!- SGautam has quit (Quit: Connection closed for inactivity).
20:24:00 -!- VzxPLnHqr has joined.
20:32:35 <VzxPLnHqr> hi esolang people! is anyone here familiar with Amelia? (https://esolangs.org/wiki/Amelia)
20:32:42 <VzxPLnHqr> I am wondering if it is turing complete
20:40:10 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
20:47:49 -!- tromp has joined.
21:06:31 -!- bgs has quit (Remote host closed the connection).
21:54:09 -!- Lord_of_Life has quit (Quit: Laa shay'a waqi'un moutlaq bale kouloun moumkine).
21:55:30 -!- Lord_of_Life has joined.
22:23:54 <int-e> VzxPLnHqr: Not familiar but the fibonacci example indicates that you have flexible flow control, addition (so you can make arbitrarily large numbers. well, abstractly; the Perl implementation puts limitations on the numbers), so you can make a Minsky machine. You don't need the indirect addressing for that, not most of the operations
22:25:17 <VzxPLnHqr> int-e, thanks! How do you think something like amelia would stack up (pun intended?) against a language like push-forth for genetic programming?
22:27:20 <int-e> I imagine random programs in Amelia aren't interesting... and the fact that "genes" are referred to by numbers makes composing programs harder. So... no clue, but I suspect it wouldn't work great.
22:27:59 <int-e> (No clue - I've never done any genetic programming.)
22:28:45 <int-e> Though I've read about it and thought about it abstractly.
22:29:37 <VzxPLnHqr> Thanks. Yeah, I came to the same tentative conclusion regarding Amelia programs.
22:38:05 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
22:48:15 <esolangs> [[BinaryLanguage]] M https://esolangs.org/w/index.php?diff=114060&oldid=114058 * None1 * (-32)
22:48:52 <esolangs> [[Category:Accumulator-based]] M https://esolangs.org/w/index.php?diff=114061&oldid=114053 * None1 * (-92) Blanked the page
23:13:10 <VzxPLnHqr> int-e: most of my experience with genetic programming has also been abstract, and here is something that has puzzled me:
23:15:12 <VzxPLnHqr> say the target program is something like a cryptographic hash function, sha256 for example, coming up with a fitness function to reward evolution towards that algorithm is not straight forward. I wonder if it is even possible.
23:16:37 <VzxPLnHqr> by "target program" what I mean here is that the genetic algorithm is supposed to create its own implementation of sha256 but using only a stream of valid inputs and outputs
23:21:26 <VzxPLnHqr> a typical desirable property of cryptographic hash functions is precisely to destroy (or, rather, hide beyond recovery) any link between input and output. Yet a genetic programming algorithm is supposed to reconstruct such links and represent them in the form of a program.
23:23:11 <VzxPLnHqr> so how to design a fitness function which will (eventually) recover such a program is not at all obvious to me
23:24:07 <int-e> Yeah I'm pretty sure you won't learn a full hash function. Maybe if you break it apart, down to a single round. Even then... it'll be mostly brute force; the "genetic" part won't come into play I think.
23:28:29 <int-e> I have no story for when genetic programming (an optimization technique) is really appropriate... I suspect *programming* is, in most cases, outside of its scope. It may work in some niches where you *can* incrmentally tune a program (so "genetics" work because you can replace fragments by other fragments that behave similar in some sense, maybe)
23:28:38 <VzxPLnHqr> exactly my thought as well. I find it philosophically amusing though and wonder whether such a fitness function could even exist.
23:29:49 <VzxPLnHqr> int-e, yeah. If you have to specify/encode into the fitness function the entire algorithm you are wanting it to "learn," it defeats the purpose entirely :-)
23:29:57 <int-e> That's a big reason why I've never done genetic programming. (Except for maybe some straightforward "recombination" of parts of solutions to certain puzzles whose solutions are sequences of moves.)
23:30:43 <int-e> (Which I think technically checks the "genetic programming" checkmark, but really feels too mundane for such a fancy name.)
23:31:21 <VzxPLnHqr> I agree, and have had similar experience there.
23:32:06 <int-e> There's also the undeniable fact that genetic programming has *not* taken the world by storm.
23:32:36 <VzxPLnHqr> that too!
23:38:45 -!- Melvar has quit (Ping timeout: 246 seconds).
23:41:29 -!- imode has joined.
23:44:01 -!- Lord_of_Life has quit (Ping timeout: 260 seconds).
23:44:24 -!- Lord_of_Life has joined.
23:51:25 -!- Melvar has joined.
←2023-08-20 2023-08-21 2023-08-22→ ↑2023 ↑all