04:20:36 Good afternoon (from my timezone) esomark 04:48:27 Also it's spooky day for westerners now 04:50:08 I think making a brute-forcer should be easier than finding the Txy = yx expression myself 04:52:47 [[11Draft:Dimethylmercury]] 12https://esolanging.wiki/mediawiki/index.php?diff=948 * AnalysisOfVariance * (+109) * /* Candidate 5 */ 04:52:59 Ugh 04:53:10 My browser is allergic to \x03 04:53:21 It wants me to download the file instead of just letting me see it in the tab 05:10:20 I wonder if the IJ basis was already known to be Turing-complete 05:14:53 FloppaDiskDrive: see https://esolangs.org/w/index.php?title=Attached&action=history, it's apparently only Turing-complete if you can find out some place to dispose of the stuff you're not using 05:15:11 Oh ok 05:15:21 So that's what Blashyrkh meant by "moving the garbage beyond EOF" 05:15:52 actually now that i think about it why can't you just have the return result be a pair of the result and then the garbage 05:16:42 eh whatever i've never written a Crazy J program 05:17:38 I'm currently trying to study Blashyrkh's IJ brute-forcer and I think it's less about brute-forcing a specific expression and more about enumerating IJ expressions 05:17:53 Also wow is it fast 05:18:02 s/is it/it is 05:20:28 Output reminds me of binary counting 05:20:47 FloppaDiskDrive: yeah the way i did it for the little bit of SK brute-forcing i did was 05:20:53 1. keep a list of the combinators you want to find 05:21:02 2. look at every single expression of a certain length 05:21:16 Oh so it's like list all possible expressions then try them one by one 05:21:18 3. if any of those expressions match a combinator in the list, you've found it 05:21:32 yeah it's best to search for multiple combinators at once 05:21:56 problem is i'm doing it in a slower programming language on a slower computer so i can't brute force as much as Blashyrkh can :P 05:22:04 (or maybe he's made some optimizations i don't know of) 05:22:38 Does Blashyrkh have a SKI brute forcer or are you talking about the IJ one 05:23:22 i think he has a bunch of brute forcers for different combinator sets 05:23:30 there's SKI, BCKW, IJ, The Church i think 05:23:56 I guess that makes sense since he always seems to find tons of combinators in other bases 06:31:30 Helo yayimheretoo 06:31:43 floppadiskdrive: hi:) 06:32:08 [[11User:AnalysisOfVariance]] 12https://esolanging.wiki/mediawiki/index.php?diff=949 * AnalysisOfVariance * (+82) * /* drafts */ 06:35:21 yall working on anything interesting? 06:35:48 Check out [[Draft:1-866-025-4037]] and [[Draft:Dimethylmercury]] 06:38:22 saw 1-866-025-4037 06:38:34 dunno how interesting I find it personally but oh well 06:43:41 ill check out dimethylmercury 06:54:25 personally I think if it actually turns out to be Turing-complete it'll be interesting by accident 06:54:57 because both combinators it uses are HOT 06:55:05 (Hereditarily of Order Two) 06:55:43 1-866-025-4037 or dimethylmercury c4? 06:56:05 pkmnq: it could be interesting in that sense 06:56:15 I just think as a LANGUAGE it doesn't bring much new to the tale 06:56:21 but as a combinator thingy sure 06:56:54 yayimheretoo: i mean yeah the only difference between it, Crazy J, and Lazy K are the combinator sets 06:57:06 pkmnq: exactly 06:57:44 and yet Lazy K and Crazy J are entirely different experiences, and I expect 1-866-025-4037 would be too 06:57:58 Why the hell did me and esolangingbot get disconnected 06:58:11 who knows 06:58:13 lmao 06:59:59 hi 07:00:13 Finally 07:00:52 btw, do yall findd my draft interesting? 07:08:37 also floppadiskdrive i think the oncept for deriving cominators in dimethylmercury is pretty cool 07:36:10 the working backwards part (candidate 3) or the working forwards part (candidates 4 and 5) 07:37:03 backwards 07:52:45 Hello de67_ 07:52:50 hi 07:52:58 hi de67_! 07:53:08 hi yayimheretoo 07:53:21 have you seen my new draft and artppeice? 07:53:30 no actually! 07:53:33 im kinda stuck with the draft thing 07:53:42 can u link em both 07:54:22 ok 07:54:35 https://esolangs.org/wiki/Artppeice 07:54:45 dde67_ I could help with the draft if wanted 07:54:58 https://esolanging.wiki/wiki/1-866-025-4037 07:55:04 yayimheretoo: thank you! 07:55:12 oh wait wrong link 07:55:16 lmao 07:55:20 you type links..? 07:55:21 here: https://esolanging.wiki/wiki/Draft:1-866-025-4037 07:55:40 what are you stuck on? 07:55:49 yk 07:55:52 like, its a complete language, is anything missing? 07:55:55 everything pretty much 07:56:24 what are the lambdas for the two combinators now again? 07:56:38 M_2 x y = x y (x y) 07:56:43 \x. \y. x y (x y) and \x. \y. y I think 07:56:43 K' x y = y 07:56:55 huh 07:57:13 if I may suggest something that could make it interesting maybe? 07:57:18 sure 07:57:25 i was also thinking of the U K basis 07:57:33 i gotta think a bit gimme a sec but ye:) 07:57:38 yayimheretoo: ok 07:58:18 basically, a combinator kinda thing(but not really), lets name iiiiiiit 07:58:20 $ sure 07:58:58 it takes in a combinator that takes two inputs, and a value that could be anything 07:59:00 so like 07:59:30 $ K' (M2 M2 K') for example 07:59:50 and it produces a combinator X where 08:00:24 X x y = K' x (M2 M2 K') | K' y (M2 M2 K') 08:00:37 | would be some sort of concurrency / parallel execution thing 08:00:46 not completely formal but I find it interesting 08:00:52 hm 08:01:09 seth like 08:01:11 *smth 08:01:21 seth 08:01:22 i mean that sounds cooler as a basis for its own esolang than being randomly added to an existing one honestly 08:01:33 sure lmao 08:01:43 I got the idea cuz both combinators take two inputs 08:02:00 (x | y) A = (x A | y A) 08:02:17 but feel free to take it or not lmao 08:02:59 ill try UK basis first 08:03:03 it seems interesting 08:03:06 sure 08:03:26 your idea is quite interesting, but i agree with PkmnQ that it deserves its own esolang 08:03:31 say X takes two inputs, then X (x | y) = X x y 08:03:34 de67_: thanks! 08:03:38 sure ill develop it 08:03:46 then its just a question of what basis to use 08:10:43 U K seems to be even harder than M_2 K' 08:11:01 was that the goal? 08:11:36 idk 08:11:49 im just trying out combinatory logic 08:11:55 searching for interesting stuff 08:12:26 ah ok 08:12:27 cool:) 08:13:03 U U K eats everything after itself 08:13:19 so like... v? 08:13:21 wait no 08:13:32 so U U K x y z .... = U U K? 08:14:09 ye 08:14:28 yayimheretoo: that is unlambda's v, yes 08:14:38 ok cool 08:16:34 so, the idea I just gave to de67, ived developed it a bit 08:16:50 also, A * means A is normal/done executing 08:17:06 and A2 means A has 2 inputs 08:17:12 so 08:17:56 $ A2 C = B2, B2 x y = (x C | y C) 08:18:55 then 08:19:39 F (x * | y) = (F x | y) 08:19:54 F (x | y *) = (x | F y) 08:20:32 F (x * | y *) = (F x | F y) 08:20:44 F (x | y) = (x F | y F) 08:20:57 this OFC needs a turing incomplete but still loopable basis 08:21:33 such that the normal form is detectable 08:22:58 is that still interesting? 08:23:00 try M_2 K' 08:23:11 though its computational class is unknown 08:23:11 sure 08:23:21 i assume its not TC 08:23:40 isn't there smth about you need least one combinator with three inputs or whatever 08:23:43 i forgor 08:24:20 yayimheretoo: you need one with at least three to be universal (i.e. construct every combinator), but you don't need to be able to construct every combinator to be Turing-complete 08:24:40 pkmnq: oh yea thats true lmo 08:24:54 im not sure how to (dis)prove tcness for M_2 K' basis 08:25:44 K' is actually quite powerful 08:26:11 we had a looong conversation about K', me and corbbin 08:27:17 yayimheretoo: there's this set of closed lambda terms that's Turing-complete but can't even construct B 08:27:35 pkmnq: theres also crazy J. but ye 08:27:46 yeah but the one I mentioned was more of a surprise to me 08:27:50 although the proof seems to be rather inaccessible 08:28:13 it's some random paper from 1995 08:28:27 no wait 1988 i misremembered 08:28:32 i find it funny when u google "is IJ TC" gemini pulls from esolang 08:28:41 pkmnq: could u link it? 08:28:44 hm 08:28:51 isnt W just M*? 08:30:06 yayimheretoo: https://kilthub.cmu.edu/articles/journal_contribution/Combinators_hereditarily_of_order_two/6477068 <- it uses some weird notation but hopefully it's still understandable 08:30:06 also i just saw a cat 08:30:17 pkmnq: thanks! 08:30:22 de67_: great 08:30:27 i've been thinking of seeing if there's a simpler proof with how much more Turing-complete stuff is known to exist nowadays 08:31:12 its just ridiculous to me 08:31:24 how much turing complete stuff there is 08:31:29 ye 08:31:42 the more there iss 08:32:01 the more interested I get in those weird in between classes 08:32:10 like regex with capute groups or whtever 08:37:00 combinatorial explosion with combinatory logic is definitely *a* joke 08:38:39 uhhh 08:38:59 six seven 08:39:17 HE SAID THE THING! 08:40:15 Brute-force search with combinators is quite annoying 08:40:19 (in terms of time) 08:41:05 omgwtf 08:42:40 Wrong brute force target ;-; 08:43:26 A type of thing I love in life is a thing you can wield as a hammer to solve problems without having to think (e.g. quadratic formula and expression brute forcers) 08:44:32 hm 08:44:57 I basis in uninteresti g 08:45:02 I basis? 08:45:06 yes 08:45:08 I 08:45:13 Identity bird 08:45:37 I basis is just the equivalent of a constant language 08:46:04 Oh my god the expression for the Kestrel in E^K'M is at least 12 characters long 08:46:30 Combinatorial explosion with combinatory logic :D 08:46:49 ill try to make kestrel on ÊK'M 08:46:55 surely it cant be that hard 08:47:01 I'm using my brute forcer rn 08:47:12 It's currently searching length 12 has has been doing so for around a minute 08:47:15 ACTION disintegrates into ash and dust 08:47:16 Now it's at length 13 08:47:43 My search range ranges from lengths 10 to 20 since it didn't find an expression with length 9 or shorter 08:47:59 I think most of these permutations are illegal anyway but I'm too lazy to trim them 08:50:31 I think it's because I'm treating the entire program as a quaternary number (e k m `) which leads to lots of illegal configurations like ````` and k``m` which wastes a lot of time, while Blashyrkh's IJ enumerator uses a smarter approach 08:50:40 (and also because 4^13 is a pretty massive number) 08:52:09 hm 08:52:30 maybe a better approach would be to build a tree of combinators 08:52:48 I can see the vision 08:53:06 we can have 2 types of nodes 08:53:10 application node 08:53:12 That also prevents other illegal configurations (in Unlambda syntax) such as kkkk 08:53:15 and combinator node 08:53:27 application node cannot be the bottommost node 08:53:38 combinator node must not have children 08:54:30 bruteforcer also needs a decider 08:54:37 What decider 08:54:40 Like the target expression? 08:54:45 decider if it halts 08:54:48 Oh ok 08:54:55 we want to prune stuff that obviously doesnt halt 08:55:13 Maybe my brute forcer stumbled upon one of those patterns ;-; 08:55:29 Wait no my stupid evaluation logic is broken so it can't not halt HAHA 08:55:32 (fuck) 08:55:34 i have an idea 08:55:40 i need to test it first 08:56:12 Blashyrkh's IJ brute forcer has "counters" which I assume are for the applications 08:58:26 idk tho 08:58:33 I'll use the tree idea then 09:02:14 [[11Draft:Dimethylmercury]] 12https://esolanging.wiki/mediawiki/index.php?diff=950 * Dragoneater67 * (+265) * /* Candidate 5 */ add M_2 09:02:29 lemme add kestrel too 09:03:52 n about an hour or so 09:04:12 [[11Draft:Dimethylmercury]] 12https://esolanging.wiki/mediawiki/index.php?diff=951 * Dragoneater67 * (+32) * /* Candidate 5 */ add unlambda expression for M_2 09:04:30 what a nick 09:05:14 Woah 09:05:55 WOAH 09:05:57 NEW COMBINATOR 09:06:34 kestrel is (Ê(KK)(KK)(KK)M(KK))Kxy 09:06:44 dont ask how i managed to find it 09:07:26 Wtf 09:07:37 K = (Ê(K'K')(K'K')(K'K')M(K'K'))K' 09:08:21 M K and Ê are all useless combinators, so its surprising that they can actually do some stuff 09:08:31 You mean K'? 09:08:35 ye 09:08:39 K' 09:08:41 Also I wouldn't call a cancellative combinator useless 09:09:15 ye 09:09:43 but Ê definitely is quite useless in most cases 09:09:58 and again 09:10:05 yayimheretoo: hi 09:10:28 i made kestrel out of Ê K' and M 09:10:31 If we find S it's so over 09:10:41 FloppaDiskDrive: working on it! 09:11:02 [[11Draft:Dimethylmercury]] 12https://esolanging.wiki/mediawiki/index.php?diff=952 * AnalysisOfVariance * (-1) * /* Candidate 5 */ don't undersleep guys or this happens 09:11:15 I'm reworking my brute-forcer to be more like Blashyrkh's IJ one 09:11:25 It's crazy fast 09:11:39 and also smarter than the thing I'm doing rn 09:14:22 floppadiskdrive: do you know any combinator basis'es where, the halting rpolem is decidable, ,majority of cominators have two inputs? 09:15:05 yayimheretoo: whar 09:15:18 did you ingest any mold by any chance? 09:15:19 de67_ thats what I need for my programming language 09:15:32 oh the typos the typos 09:15:32 Not really 09:15:44 yayimheretoo: hm 09:15:44 What typos 09:16:00 try U, M and K' 09:16:06 a basis where its knowable wether the normal form will ever reached, AND the combinators are mostly 2 inputs 09:16:14 de67_ sure 09:16:21 also the typos are because of like 09:16:40 yayimheretoo: black mold black mold black mold black mold black mold black mold black mold black mold 09:17:12 1. some keys need to be pressed rlly hard for some reason, 2. sometimes my keyboard doubles characters for no reason, and 3. sometimes it delays the order of key presses for sum reason 09:17:25 and also theres certain words my brain cant wrap its head around idk why 09:17:51 yayimheretoo: oh i see 09:18:03 de67: lmao 09:18:20 also my spacebar is SO insensitive, I have to slam that thing 09:18:49 yayimheretoo: what came first slamming, or the insenstivity 09:19:06 de67_: insensitivty 09:19:18 idk why 09:19:28 but also this is a 12 yr old computer lmao 09:19:32 MacBook air 09:19:40 yayimheretoo: 12??? 09:19:44 holy 09:20:15 my laptop is also very old 09:20:16 [[11Draft:Dimethylmercury]] 12https://esolanging.wiki/mediawiki/index.php?diff=953 07M * AnalysisOfVariance * (+1) * /* Candidate 5 */ oops 09:20:37 well, it must be older than 12 12 is approximate 09:20:43 it came with ubuntu preinstalled, but by the time i got it it had windows 8.1 out of all oses 09:20:55 lmaooo 09:21:04 i had to wipe it since it was full of other crap apart from windows 8.1 09:21:11 the first thing i installed was debian 09:21:30 jeez lmao 09:21:59 anyway 09:22:03 back to combinators 09:22:07 yes 09:22:24 i think S x y z = (f z) (y z) 09:22:26 iirc 09:22:35 yes 09:22:41 (xz)(yz) actually 09:22:47 the f thing was a typo 09:22:51 lol 09:23:12 so im thinking if Ê K', and M can somehow express it 09:24:17 ``k`e`e`eek 09:24:23 FloppaDiskDrive: whats that 09:24:26 eek! 09:24:30 LMAO 09:24:36 oh sorry 09:24:38 i meant 09:24:41 owo wats dat? 09:24:49 I ran my brute forcer on lengths 10 to 20 and it gave me ``k`e`e`eek 09:24:55 hm 09:26:15 Wanna know what it is right now 09:26:29 also 09:26:36 Mxy = xx is wrong 09:26:41 its actually Mx = xx 09:26:41 Yea 09:26:43 I fixed it 09:27:00 Wanna know what ``k`e`e`eek is 09:28:24 Spoiler alert 09:28:53 ``k`e`e`eek in ÊK'M is just K' 09:29:05 K(EEEE)K = K 09:29:20 KEEEEK 09:29:28 thats just great lmao 09:29:47 1/1=1 09:29:51 1+1/1+1=1 09:29:54 1+1+1/1+1+1=1 09:29:59 1+1+1+1/1+1+1+1=1 09:30:03 yes? 09:32:51 bacc 09:32:59 Wtf so much of that is invalid 09:33:00 yay 09:33:08 Or my profiling is shit 09:35:23 It says that all expressions of length 10 are invalid somehow 09:35:28 1048576 expressions 09:35:46 And for length 9 it says 98.7022% are invalid (out of 262144) 09:35:56 So 258742 expressions out of 262144 are invalid 09:36:08 just build a tree!! 09:36:13 Good point 09:36:37 I assume the way to do it is to do breadth-first ordering 09:37:30 ye 09:38:12 you better make it brute force S combinator 09:38:32 bye imma take an esolang break 09:46:06 cya 09:52:23 I think I finally understand Blashyrkh's IJ brute-forcer 10:03:40 Using ternary instead of quaternary made it a lot better 10:03:50 (my brute force machine) 10:09:41 ok 10:09:42 uh 10:09:56 my Kestrel for M_2 K' basis is so wrong 10:10:01 its completely wrong 10:10:04 what was i thinking 10:10:16 ugh 10:10:23 i may need to use a different basis 10:12:34 hm 10:12:40 psi looks interesting 10:12:44 but also very powerful 10:12:55 Tree brute forcer it is 10:13:34 Maybe you could make the next crazy j but with psi 10:14:17 Psichosis sounds like a great name 10:14:33 and PsiK 10:14:48 Psichosis is a very cool name 10:15:01 ok 10:15:08 heres an amazing basis: 10:15:19 Also I think psi can give I* very easily if you have I as well 10:15:22 PsiK'Ê 10:15:31 That's fucking genius 10:17:11 [[11Draft:1-866-025-4037]] 12https://esolanging.wiki/mediawiki/index.php?diff=954 * Dragoneater67 * (-90) * /* Overview */ remove broken derivations 10:20:53 [[11Draft:Psichosis]] 12https://esolanging.wiki/mediawiki/index.php?diff=955 07N * Dragoneater67 * (+502) * Created page with "{{WIP}} '''Psichosis''' is an [[esoteric programming language]] made by [[User:Dragoneater67]] in [[:Category:2026|2026]]. 10:26:44 I see that for any expression with N combinators, there can be at most N - 1 applications 10:26:50 (I think) 10:27:43 So there is a ternary number (the combinators) and another thing determining where the applications are 10:27:56 Actually wait not even "at most" it's "exactly" 10:30:07 psi is probably universal 10:30:19 it may not be that interesting tbh 11:03:20 I wonder why the table of combinators doesn't have a combinator that is applicative, cancellative, and duplicative all at once 11:03:50 I guess for example a combinator @ x y z w = yz(wz) 11:04:02 (basically S but with 4 arguments and it discards the first one) 11:05:09 Wait no it's not exactly S 11:07:51 Or # x y z w t u v i j k = t(ji(ij)xjx(kk)y)(yxz)ij(wz)v 11:08:21 I wonder if this is complete 11:08:25 Or if you have the I# basis 11:09:08 %xyz = xy(xx) 11:09:34 Combinators so weird de67's connection died 11:10:12 Helo again 11:10:21 Remember # x y z w t u v i j k = t(ji(ij)xjx(kk)y)(yxz)ij(wz)v 11:10:22 hi 11:10:28 uh 11:11:02 I put some letters and parentheses in a blender and it gave me that, is it good? 11:11:10 FloppaDiskDrive: lemme see 11:11:22 # x y z w t u v i j k = t(ji(ij)xjx(kk)y)(yxz)ij(wz)v 11:12:19 i i vented a new combinator: 🐱f g x y (f x y) (f y x) (g x y) (g y x) 11:12:39 I think 😺 could actually be useful 11:13:27 also i give up on psichosis 11:14:03 [[11Special:Log/delete]] * Dragoneater67 deleted page [[11Draft:Psichosis]] (content was: "{{WIP}} '''Psichosis''' is an [[esoteric programming language]] made by [[User:Dragoneater67]] in [[:Category:2026|2026]]. It is based on [[combinatory logic]]. 11:14:26 alr, back to M_2 K' 11:14:32 i need to figure out kestrel 11:14:50 Do you already have a brute forcer for M_2K'? 11:15:29 not yet 11:18:31 hi 11:18:38 im onl aptop now 11:18:49 Hi 11:19:05 I found another enumerator by Blashyrkh 11:19:22 enumerator? 11:19:46 It just lists all expressions in the IJX basis between two lengths 11:19:50 https://github.com/rs-blashyrkh/between-IJ-and-SK/blob/master/enumerate.c 11:20:29 ok 11:22:19 E^MK' is somehow at least as powerful as M_2K' 11:22:58 I propose two new combinators 6 and 7 11:23:12