00:02:37 <esolangs> [[Language list]] M https://esolangs.org/w/index.php?diff=195704&oldid=195629 * Marcus 2n2 * (+9)
00:08:09 -!- Lord_of_Life_ has joined.
00:08:24 -!- Lord_of_Life has quit (Ping timeout: 269 seconds).
00:09:26 -!- Lord_of_Life_ has changed nick to Lord_of_Life.
00:33:41 <esolangs> [[14]] M https://esolangs.org/w/index.php?diff=195705&oldid=195691 * Marcus 2n2 * (+162) /* Instructions */
00:37:50 <AnalysisOfVarian> Why do the logs in 2012 black out from July to the end of the year?
00:44:54 <b_jonas> AnalysisOfVarian: what do you mean by black out?
00:45:09 <b_jonas> just somewhat less traffic?
00:45:45 <AnalysisOfVarian> I mean when I scraped the logs, I just saw "no logs for month: 2012-{07-12}
00:46:04 <int-e> did you mistype the year?
00:46:22 <int-e> 2021 is the freenode drama year
00:47:01 <b_jonas> I de see logs for those months
00:54:42 -!- amby has quit (Remote host closed the connection).
00:57:08 -!- FreeFull has quit (Ping timeout: 259 seconds).
00:58:14 -!- FreeFull has joined.
01:02:44 <fizzie> I think I thought about exposing the underlying protobuf-formatted logs in case anyone wanted to archive them with absolute fidelity.
01:04:05 <fizzie> The raw text format is _almost_ sufficient for that already, but it's missing the IRCv3 tags (which I could certainly add in, but that might confuse any readers that weren't expecting them).
01:05:56 <zzo38> I think that sometimes making a programming language for something specific can sometimes involve things that are sometimes related to esolangs. The article [[Prehistory of esoteric programming languages]] on esolang wiki has a few examples of this, although there are also such things as considerations of computational classes, comparisons with other esolangs, etc.
01:07:22 <zzo38> (Also, I like to use DER as a binary structured data format, rather than protobuf and others, mostly.)
01:13:03 <zzo38> (In the case of something such as IRC logs, the "multi-DER" variant might be better, like how concatenated JSON can be used as an alternative of plain JSON, and possibly similarly with some other kind of file formats.)
01:17:54 <zzo38> (However, whatever format is available, might do, although there is the existing raw text format which is probably good enough for most purposes, even though some things it might not be, it has not been a problem to me personally and I don't know who else it has been but maybe there is some.)
01:30:13 <esolangs> [[Talk:Esofish]] M https://esolangs.org/w/index.php?diff=195706&oldid=195404 * Marcus 2n2 * (+162)
01:46:33 <esolangs> [[TetoLang]] M https://esolangs.org/w/index.php?diff=195707&oldid=195663 * A random guy * (-5)
01:50:43 <esolangs> [[Talk:Esofish]] https://esolangs.org/w/index.php?diff=195708&oldid=195706 * PkmnQ * (+270)
01:58:51 <esolangs> [[Lambda calculus]] https://esolangs.org/w/index.php?diff=195709&oldid=162873 * Corbin * (+237) Convert references to common style. Add a small section to justify reference 4, which AFAICT was never actually used prior to this.
02:04:50 <esolangs> [[TetoLang]] M https://esolangs.org/w/index.php?diff=195710&oldid=195707 * A random guy * (+28) I did state it was turing complete I think.
02:05:09 <esolangs> [[TetoLang]] M https://esolangs.org/w/index.php?diff=195711&oldid=195710 * A random guy * (+1)
02:05:25 <esolangs> [[TetoLang]] M https://esolangs.org/w/index.php?diff=195712&oldid=195711 * A random guy * (+0)
02:34:18 <esolangs> [[User talk:Cordership]] https://esolangs.org/w/index.php?diff=195713&oldid=195660 * PkmnQ * (+182) /* plz write a quine in simpl */
02:44:16 -!- miui has joined.
03:03:10 <esolangs> [[TetoLang]] https://esolangs.org/w/index.php?diff=195714&oldid=195712 * Dragoneater67 * (+169) /* Commands */
03:05:11 <esolangs> [[Luafuck]] M https://esolangs.org/w/index.php?diff=195715&oldid=107912 * Dragoneater67 * (+1) /* Primitives */
03:14:36 <aadenboy> AnalysisOfVarian: my favorite language
03:16:52 -!- AnalysisOfVarian has quit (Excess Flood).
03:18:12 -!- AnalysisOfVarian has joined.
03:18:57 <AnalysisOfVarian> Maybe trying to send 10KB of text over IRC in one message is a bad idea
03:24:52 -!- miui has quit (Quit: Client closed).
04:06:24 <zzo38> Probably it would be better to send a link; I think there is a limit that is less than 10KB in one message.
04:25:20 <AnalysisOfVarian> 21:12:52 <raiseafloppafan7: no it's not that Crazy J is a letters for it, then is when running to spec /j
04:40:49 -!- password2 has joined.
04:59:21 -!- de67 has joined.
05:03:54 -!- de67 has quit (Client Quit).
05:04:03 -!- de67 has joined.
05:05:39 <de67> the markov chains...
05:05:49 <de67> theyre driving the people of the esolang wiki mad
05:06:00 <de67> they must be destroyed for good
05:06:46 <AnalysisOfVarian> 08:25:15 <de67> or maybe its contexts which, to ourselves to imagine.
05:07:03 <AnalysisOfVarian> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Totally real IRC log from ?? September 2026
05:07:14 <esolangs> [[Talk:100]] N https://esolangs.org/w/index.php?oldid=195716 * Marcus 2n2 * (+174) Created page with "== <code>=</code> == Where will <code>=</code> store the duplication? --~~~~"
05:08:17 -!- de67 has quit (Client Quit).
05:08:27 -!- de67 has joined.
05:08:37 <fungot> AnalysisOfVarian: mr president, i think it is the crucial question. the fact is that, at the end of 2003. the aim is certainly an important step towards creating such a level. the conclusions and recommendations in your report to have the ability to influence decisions on pre-trial detention, not to mention the reference made to the united nations, thereby ensuring that our fine words into action.
05:08:43 <de67> marcus 2n2 mightve invented numberslop
05:09:06 <AnalysisOfVarian> Funny thing is 2003 is the first year whose logs I scraped into the markov data
05:09:34 <de67> btw, esomark is trained on my messages on irc
05:09:57 <de67> you can clone the repo and run the scripts thatll hopefully replace me
05:10:11 <de67> it can be swapped out to clone anyone
05:10:27 <de67> it can also be modified to scrape the older freenode logs
05:10:30 <AnalysisOfVarian> de67: What if esomark marks its territory in an IRC log and impersonates itself?
05:10:49 <de67> its confined to #esolangs-blah
05:11:01 <de67> 2021 is a special case
05:11:12 <de67> it shall be handled separately
05:11:36 <de67> partially scrape freenode 2021 then partially scrape libera 2021
05:11:52 <de67> i didnt bother and only scraped libera
05:12:04 <de67> might add freenode scraping at some point
05:12:06 -!- b_jonas has quit (Read error: Connection reset by peer).
05:12:07 <AnalysisOfVarian> Yeah, but the problem is I'm stupid (duh) and they intersect from May to June
05:12:46 <de67> just check what year and month youre scraping and choose the correct url
05:12:50 <AnalysisOfVarian> de67: Just do something like `"freenode-esoteric" if year < 2021 else "libera-esolangs"` (assuming you don't care about 2021)
05:13:30 <de67> thats not what i meant
05:15:54 <AnalysisOfVarian> Just curious, how well does esomark handle large training data?
05:16:57 -!- de67 has quit (Quit: Client closed).
05:17:09 -!- de67 has joined.
05:17:22 <de67> its slow on my potato laptop
05:17:40 <AnalysisOfVarian> Like a few kilobytes or like 20 megabytes or almost a gigabyte
05:17:41 <de67> training on esolang wiki (~1.8 gb) takes several minutes
05:17:48 -!- b_jonas has joined.
05:18:50 <esolangs> [[XKCD Random Number]] https://esolangs.org/w/index.php?diff=195717&oldid=195391 * Miui * (+97) /* pgmpiet */ add example
05:21:05 -!- de67 has quit (Client Quit).
05:21:14 -!- de67 has joined.
05:21:15 <AnalysisOfVarian> My discombobulated Mark trains quite fast, however it chokes when I try to ACTUALLY use the model
05:22:01 <AnalysisOfVarian> 06:20:23 <esolangs> [[User:PrySigneToFry]] ([[User talk:FluixMakesEsolangerII * (+170) /* Interpreter, then the cloud of smoke. when the Pascal's Modified the color is a lemmatized/2+2+3frq.txt"; while(<$R>){while(/\w+(?<![aeiou])e?l(?![aiou])\w*/gi){ print "$& " }}
05:22:01 <de67> fungot: whod win in a fight: gpt-6 astra or discombobulated press?
05:22:01 <fungot> de67: mr president, ladies and gentlemen, the debate of 3 july 2002, i repeated my call for a unified statute for meps to be introduced in its current form.
05:22:26 <AnalysisOfVarian> GPT-6 Astra vs discombobulated press vs discombobulated Mark would be interesting
05:22:34 <de67> esomark used to be piss slow
05:22:58 <de67> until i used a fancy weighted random runction instead of creating a giant list
05:23:17 <de67> now it can almost instantly generate slop
05:23:29 <de67> i fucked up the word order
05:23:49 <de67> i, myself, am melting into a slop generating markov chain
05:24:06 <AnalysisOfVarian> Wait, you're supposed to use some fancy stuff instead of creating a giant list of stuff?
05:24:56 <de67> i used python.choices()
05:25:14 <de67> the format for the model is a dict of weights
05:25:44 <de67> {"i": {"crapped": 1, "shat": 2}}
05:26:19 <de67> ^ this means that theres 66% chance that i will be followed by crapped, and 33% chance that i will be followed by shat
05:26:30 <AnalysisOfVarian> Oh, that's what I was supposed to do. I just inserted the same prefix-continuation pair like a zillion times
05:26:46 <de67> yea i did that too at first :-)
05:26:57 <de67> FUCKING AUTOREPLACE
05:29:17 <AnalysisOfVarian> Btw how often does your model output anything relating to plane crashes
05:30:05 <esolangs> [[Talk:100]] https://esolangs.org/w/index.php?diff=195718&oldid=195716 * Dragoneater67 * (+198) /* = */ moral of the story: be immoral
05:30:07 -!- de67 has quit (Quit: Client closed).
05:30:17 -!- de67 has joined.
05:30:19 <de67> approximately never
05:30:34 <de67> its trained on me
05:30:47 <de67> so itll talk about plane crash sometimes
05:31:37 <de67> plane crasher fungot is alive?
05:31:37 <fungot> de67: mr president, the relationship here too being two-way, since a lot has been achieved, but it should not come as news to you that it is virtually impossible to eradicate: according to the rules or the implementation stage of other directives too. what this means is that potential tensions need to be given a free cruise?
05:31:55 <de67> ^style youtube
05:32:10 <AnalysisOfVarian> de67: Kind of, but its personality is fragmented across you, I, ais523, esolangs, and whoever else was in the logs that I scraped
05:32:12 <de67> one fay itll be found
05:35:17 <de67> you can recognize this word: deeschambabueleishn
05:35:38 <de67> whats the crap machine in question?
05:36:35 <AnalysisOfVarian> I'm 'fixing' my discombobulated Mark to make it load my 189MB dataset faster
05:36:47 <de67> uuts ther krap masheen in kwesssn?
05:36:59 <fungot> AnalysisOfVarian: the report by mrs cederschild on the excellent work she has done all it could to downgrade the status of our fishery resources to remain available for the states which are parties to the convention and that the comitology procedure was decided on in a number of you agreed with that proposal.
05:37:02 <fungot> Selected style: fungot (What I've said myself)
05:37:04 <fungot> AnalysisOfVarian: that is just a value of type is created containing the syntax for mark if he was really gonna get worse and worse each week,
05:39:50 <de67> eye uanderr uuts ther uiirdest uei tu sei thees
05:40:32 <AnalysisOfVarian> Intupid. mall><sly bases="rectecond l of Varmart an don't that co a sone that " width IRC
05:40:55 <de67> it kan get reelee hurd tu rid sumteyems
05:41:09 <de67> galmart is a real store
05:41:16 <de67> i went there once
05:41:33 <de67> its just walmart, but kazakh, therefore, superior to everything else
05:42:11 -!- Sgeo has quit (Read error: Connection reset by peer).
05:43:34 <de67> eet ees veree phan tu du thees ceyend aph abphueskeeeshn
05:45:03 <de67> meyet kreeeeet an ezalung abawt thees
05:46:22 <de67> tho eyem anshuer abawt haw thees kood uerk
05:50:19 <de67> sees kood beekum eeven mar eencamprehenseebl eeph uee allau akscents
05:50:46 <de67> i think is should stop spaming
05:58:16 -!- de67 has quit (Ping timeout: 249 seconds).
06:06:25 <esolangs> [[Talk:YOUR TAKING TOO LONG]] https://esolangs.org/w/index.php?diff=195719&oldid=164047 * Miui * (+221) lost kingdom
06:11:46 <esolangs> [[]] https://esolangs.org/w/index.php?diff=195720&oldid=195227 * Miui * (+83)
06:19:08 -!- de67 has joined.
06:24:10 <esolangs> [[]] https://esolangs.org/w/index.php?diff=195721&oldid=195720 * Miui * (+112) added example in 17 plus wikipedia entry to external resources
06:24:59 -!- de67 has quit (Ping timeout: 255 seconds).
06:25:09 <AnalysisOfVarian> c++ is very fun especially when your program prints "segmentation fault"
06:31:44 -!- de67 has joined.
06:33:13 <esolangs> [[User talk:Marcus 2n2]] https://esolangs.org/w/index.php?diff=195722&oldid=195247 * PkmnQ * (+197) Esofish?
06:35:13 -!- de67 has quit (Client Quit).
06:51:13 -!- de67 has joined.
06:54:06 -!- de67 has quit (Client Quit).
06:54:16 -!- de67 has joined.
06:58:17 <esolangs> [[User:Miui/Pavlova]] N https://esolangs.org/w/index.php?oldid=195723 * Miui * (+6870) Created page with "Pavlova is a [[TrivialBrainfuckSubstitution]] based on a conversation about a confection =Syntax= [[TrivialBrainfuckSubstitution]]<code>TrivialBrainfuckSubstitution("Pavlova","Talk:Pavlova","Pavlova is a meringue-based cake.","Pavlova is a meringue-based desser
06:59:53 -!- de67 has quit (Ping timeout: 252 seconds).
07:00:30 <esolangs> [[User:Miui/Pavlova]] https://esolangs.org/w/index.php?diff=195724&oldid=195723 * Miui * (-3)
07:01:52 <esolangs> [[User:Miui]] https://esolangs.org/w/index.php?diff=195725&oldid=195665 * Miui * (+31) /* brainfuck derivatives */ add pavlova
07:07:04 <esolangs> [[User:Miui/Pavlova]] https://esolangs.org/w/index.php?diff=195726&oldid=195724 * Miui * (+209) add categories
07:09:03 -!- pgmpiet has joined.
07:09:24 <pgmpiet> so was consensus not to proliferate silly bf substitutions to main space?
07:10:46 <pgmpiet> no worries i can keep my meme in my userspace
07:11:06 -!- de67 has joined.
07:11:49 <pgmpiet> hm~ maybe i misunderstood what you meant AnalysisOfVarian
07:12:40 -!- de67 has quit (Client Quit).
07:12:50 -!- de67 has joined.
07:14:04 <de67> it looks like something APic would say
07:14:27 <AnalysisOfVarian> Probably because I haven't applied the probability distribution yet lol
07:14:35 <pgmpiet> "not many oppose that" as in most people dont oppose proliferating trivial bf substitutions**?
07:14:48 <de67> proliferating?
07:14:57 <de67> im sorry im dumb please use simple words
07:15:10 <de67> nvm gonna read the dictionary
07:15:12 <pgmpiet> should i ship https://esolangs.org/wiki/User:Miui/Pavlova to main
07:15:27 <de67> pgmpiet: it is up to you to decide
07:15:37 <AnalysisOfVarian> I guess you can, however you should also add it to the [[Trivial brainfuck substitution]] list
07:16:09 <pgmpiet> yeah ill definitely add it, i was about to but then i got confused how to list in alphabetically because of the User/
07:16:28 <de67> esolanging wiki is kinds inactive
07:16:39 <de67> (so is esolang wiki, but esolanging wiki seems deader)
07:16:53 <AnalysisOfVarian> If the User:Miui/ is part of the name then you put it in "U", however if it's just "Pavlova" then put it in P
07:17:47 <de67> i unstabbed Esolanging:Policy
07:17:59 <de67> since we have the policies now
07:18:05 <de67> once i get home
07:18:25 <de67> ill move it to the new domain: esolanging.wiki
07:18:39 <de67> maybe create a separate irc channel
07:18:55 <de67> since itd be nice to have recent changes on irc too
07:19:12 <fungot> AnalysisOfVarian: i, myself, will bring an end to all. ghosts lurk in the ruins were in truth, and everything in readiness for fnord. under these, too
07:19:36 <de67> i think i need to make esomark more interestinf
07:19:45 <esolangs> [[Trivial brainfuck substitution]] https://esolangs.org/w/index.php?diff=195727&oldid=193047 * Miui * (+323) /* Example Members of the TrivialBrainfuckSubstitution family */ add User:Miui/Pavlova
07:19:54 <de67> like being able to mimic any user from the channel
07:20:04 <de67> ill also need to do somethinf about the crawlers
07:20:31 <AnalysisOfVarian> de67: I remember when ais523 said you can do something with those stupid crawlers, so why don't you use the crawlers to solve the halting problem or mine bitcoin /j
07:20:40 <de67> ill block some heavy pages to unregistered users
07:20:52 <pgmpiet> shoving the imitation game into the depths of a tarpit
07:21:07 <pgmpiet> imma head out have a good one
07:22:08 <de67> about the mining bitcoin thing
07:22:12 -!- pgmpiet has quit (Quit: Client closed).
07:22:14 <de67> how exactly can it be done?
07:22:29 <de67> oh i just realized that heading out meant leaving
07:22:53 <AnalysisOfVarian> But my guess is you can make them solve bitcoin problems somehow
07:23:04 <de67> they follow every link on the page
07:23:31 <de67> but how would branching work if they just mindlessly follow links without any real decisions?
07:23:43 <esolangs> [[]] https://esolangs.org/w/index.php?diff=195728&oldid=195721 * Miui * (-1) removed a period from copying Nope. template ()
07:24:04 <AnalysisOfVarian> Should we have our own page for esoteric programming language or redirect here?
07:24:35 <de67> probably our own page
07:25:12 <de67> since esolang wiki has a separate page from wikipedia
07:27:36 -!- de67 has quit (Quit: Client closed).
07:27:46 -!- de67 has joined.
07:38:11 -!- de67 has quit (Quit: Client closed).
07:46:55 <esolangs> [[Special:Log/upload]] upload * Marcus 2n2 * uploaded "[[File:819353905 3545501268949689 293821171783131097 n.png]]"
07:47:56 <esolangs> [[Talk:Entropy]] https://esolangs.org/w/index.php?diff=195730&oldid=134619 * Marcus 2n2 * (+217) /* i found this */
07:48:27 <esolangs> [[Talk:Entropy]] M https://esolangs.org/w/index.php?diff=195731&oldid=195730 * Marcus 2n2 * (+27) /* BRO */
07:51:19 <esolangs> [[User talk:Marcus 2n2]] https://esolangs.org/w/index.php?diff=195732&oldid=195722 * Marcus 2n2 * (+145) you jave a new massage
07:51:33 <esolangs> [[User talk:Marcus 2n2]] M https://esolangs.org/w/index.php?diff=195733&oldid=195732 * Marcus 2n2 * (+105)
07:55:46 <esolangs> [[Talk:L]] N https://esolangs.org/w/index.php?oldid=195734 * Marcus 2n2 * (+150) FizzBuzz
08:01:29 <esolangs> [[G]] M https://esolangs.org/w/index.php?diff=195735&oldid=157797 * Marcus 2n2 * (+8)
08:05:34 <esolangs> [[User:Raiseafloppafan7741/Constant language directory]] https://esolangs.org/w/index.php?diff=195736&oldid=195228 * Raiseafloppafan7741 * (+76) /* ConstantLanguage() */ Add [[Nein.]]
08:42:11 <HackEso> lambdabot is a fully functional bot. just don't ask about @src. or the neighbours.
08:42:15 <HackEso> /hackenv/bin/`: line 5: lambdabot: command not found
08:42:27 <HackEso> lambdabot is a fully functional bot. just don't ask about @src. or the neighbours.
09:13:14 <AnalysisOfVarian> Oh I finally figured out what was cooking my request, I mistakenly left something in and so it tried to fetch libera-esolangs/2012-10.txt instead of freenode-esoteric/2012-10.txt
09:23:11 <esolangs> [[User talk:Marcus 2n2]] https://esolangs.org/w/index.php?diff=195737&oldid=195733 * Blashyrkh * (+278) /* Number esolangs (currently from -1 - 100, you may edit, may turn this into a category c2a) */
09:27:37 -!- password2 has quit (Ping timeout: 245 seconds).
10:00:49 -!- de67 has joined.
10:02:25 <de67> hi AnalysisOfVarian
10:02:37 <de67> i created #esolanging
10:05:29 <fungot> AnalysisOfVarian: doesn't intercal have a minuscule instruction set, the fabled to live on?' alice, who looked at it" and enjoy knowledge, and its kind, but my watch it toast our land! now we'll have some peace! magus is a tad on the spooky side. our only hope.
10:06:11 <AnalysisOfVarian> According to ais523, C(LC?)-INTERCAL can define new instructions or something, so no fungot, INTERCAL does NOT have a miniscule instruction set.
10:13:39 -!- sftp_ has joined.
10:14:03 -!- sftp has quit (Ping timeout: 251 seconds).
10:14:03 -!- sftp_ has changed nick to sftp.
10:14:03 -!- sftp has changed hostmask to ~sftp@user/sftp.
10:17:06 -!- AnalysisOfVarian has changed nick to raiseafloppafan7.
10:32:11 -!- password2 has joined.
11:07:29 <de67> who are you btw?
11:07:52 <de67> that doesnt clear up that much
11:09:24 <de67> im sorry im dumb
11:11:07 <de67> are you the creator of brainfuck^
11:14:32 <de67> https://esolangs.org/wiki/Brainfuck%5E
11:14:53 <password2> yay , i thought that page has been lost to the sands of time!
11:15:15 <de67> well, 12 years is quite a long time
11:15:31 <de67> id say that a TON of things happened since 2014
11:15:52 <de67> but the wiki should be mostly familiar
11:16:08 <de67> at least we got the same admins from 2014
11:16:20 <de67> (which may or may not be a good thing)
11:16:33 <password2> yeah , idk i tried swarching for it a few days ago , and could not find it
11:16:34 <de67> and the default skin didnt change
11:17:07 <de67> what did you search for?
11:17:16 <de67> there are 800+ brainfuck derivatives now
11:17:30 <de67> so no wonder you might get lost in a sea of results
11:17:54 <de67> the abundance of brainfuck derivatives is a lil bit funny, but also really sad
11:18:11 <de67> 14% of all esolangs are brainfuck derivatives
11:18:23 <de67> None1 already kinda did that
11:18:32 <password2> one whose instruction set changes based on a command
11:18:33 <de67> he made the ultimate brainfuck extension
11:18:51 <de67> then thats not the thing i thought about
11:19:01 <password2> i used to have an irc bot capable of doing brainfuck
11:19:32 <de67> ^bf ,[.,]!we have one right here
11:19:32 <fungot> we have one right here
11:19:44 <de67> we can now do this:
11:20:27 <de67> ^def 67 bf >,[>,]<[.<]
11:20:40 <de67> ^67 this thing reverses everything
11:20:40 <fungot> gnihtyreve sesrever gniht siht
11:21:05 <fizzie> That's not even the only brainfuck-bot (though it might be the only one accepting the `code!input` syntax).
11:21:34 <fizzie> `` echo 'reverse this' | \! bf '>,[>,]<[.<]'
11:22:03 <de67> the hackeso thing seems more compilcated to me
11:22:10 <de67> but hackeso is a much more powerful bot
11:22:46 <int-e> `` \! bf '>,[>,]<[.<]' <<<'12321'
11:23:22 <de67> thats a palindrome
11:23:35 <de67> ^67 emordnilap
11:23:46 <de67> an emordnilap is an emordnilap
11:23:51 <de67> ^67 palindrome
11:23:59 <de67> but a palindrome is not a palindrome
11:24:07 <fizzie> ^rev taas alle satasella saat
11:24:07 <fungot> taas allesatas ella saat
11:25:12 <HackEso> #!/bin/bash \ CMD=`echo -n "$1" | cut -d' ' -f1` \ ARG="$(echo -n "$1" | cut -d' ' -f2-)" \ exec $HACKENV/ibin/$CMD "$ARG$2"
11:25:17 <fizzie> HackEso's EgoBot emulation layer is really intended to be used as just `! bf ... which doesn't look too bad, but then you can't provide input.
11:25:22 <de67> ^def undef ul (Undefined.)S
11:25:26 <FireFly> (oh right, there's also `show)
11:25:39 <de67> you can now undefine commands
11:25:53 <de67> ^undef garbage
11:35:27 -!- raiseafloppafan7 has changed nick to AnalysisOfVarian.
11:43:44 <esolangs> [[Assert]] https://esolangs.org/w/index.php?diff=195738&oldid=195634 * None1 * (+531)
11:43:55 <esolangs> [[Assert]] https://esolangs.org/w/index.php?diff=195739&oldid=195738 * None1 * (+2) /* Computational class */
11:46:33 <de67> fungot: can you mog?
11:46:33 <fungot> de67: i am just as confused. you, in numbers apparently inexhaustible, were all from me where that matkakortti issue is discussed, it should be deleted, not merged. but new england flood of may, 1765, curwen's only child, it's ancient history now...
11:48:42 -!- password2 has quit (Ping timeout: 250 seconds).
12:00:36 <esolangs> [[Talk:Turing-complete]] https://esolangs.org/w/index.php?diff=195740&oldid=195235 * None1 * (+442) /* Are non-deterministic/multithreading languages like this Turing-complete? */
12:25:40 <int-e> How have I never seen anyone talk about the danger of git allowing branch names that are strings of 40 (or fewer) hex digits...
12:27:07 <int-e> (https://www.air.security/blog-posts/plugin4shell frames it as an attack against coding harnesses for LLMs but it really feels like a git design issue)
12:40:18 -!- password2 has joined.
12:41:42 <fizzie> "GitHub rejects a 40-hex branch name outright" sounds like it has in fact come up in some context before.
12:41:59 <fizzie> Or else they've accidentally guarded against it due to some technical limitation. :)
12:42:51 -!- de67 has quit (Quit: de67).
12:43:09 -!- de67 has joined.
12:50:00 <b_jonas> int-e: my opinion is that git was originally designed to handle codebases with highly trusted content and reasonable size, which is why confusing branch names or colliding SHA-1 hashes or the lack of partial checkouts are not problems that it was trying to solve, but then people started to use it for all sorts of version control of untrusted data and large amounts of data, and now they're hastily trying
12:50:06 <b_jonas> to add kluges to make it handle them. they should just use svn, which handles those cases well, or one of the dozen other open source version control systems that are now available (but weren't yet back when git was first created)
12:51:53 <b_jonas> I use svn for my home version controlling needs, and touch git only when I have to be compatible with other people's git repository or when I want to use github as a free hosting provider. Other people use mercurial or darcs instead of svn, those might also be fine choices.
12:56:10 <int-e> b_jonas: But one of the premises is that the commit id identifies a revision cryptographically (modulo SHA-1 weaknesses), so not having an easy way to unambiguously refer to a revision by commit id is an interesting omission :P
12:56:49 <int-e> And it feels like a thing that people have know for well over a decade.
12:58:35 <b_jonas> int-e: that's very useful if you are building a distributed version control, as opposed to one with central repositories where you can refer to a commit with an URL that identifies the server and the path within that and a revision number
13:16:29 <esolangs> [[Simpl]] https://esolangs.org/w/index.php?diff=195741&oldid=195542 * Cordership * (+219) /* Module 4: Gamemake */
13:17:28 <esolangs> [[User talk:Cordership]] https://esolangs.org/w/index.php?diff=195742&oldid=195713 * Cordership * (+114) /* plz write a quine in simpl */
13:17:41 <esolangs> [[User talk:Cordership]] https://esolangs.org/w/index.php?diff=195743&oldid=195742 * Cordership * (+20) /* plz write a quine in simpl */
13:19:39 <esolangs> [[User talk:Cordership]] https://esolangs.org/w/index.php?diff=195744&oldid=195743 * Yayimhere2(school) * (+179) /* Collab? */
13:21:42 <esolangs> [[User:Raiseafloppafan7741/Totally real IRC logs]] M https://esolangs.org/w/index.php?diff=195745&oldid=195679 * Raiseafloppafan7741 * (+543)
13:35:25 -!- de67_ has joined.
13:35:31 -!- de67 has quit (Ping timeout: 241 seconds).
13:41:13 <esolangs> [[Baryus]] N https://esolangs.org/w/index.php?oldid=195746 * Cordership * (+707) Created page with "Baryus is a string-rewriting esolang made by [[USer:Cordership]]. =Commands= {|class=wikitable ! Command !! Definition |- | 0 || Append a 2 to the program, then delete the leftmost digit. |- | 1 || Goto the next command (say x), if the rightmost digit is 0 then append
13:41:23 <esolangs> [[Baryus]] https://esolangs.org/w/index.php?diff=195747&oldid=195746 * Cordership * (-9) /* Examples */
13:41:31 <esolangs> [[Baryus]] https://esolangs.org/w/index.php?diff=195748&oldid=195747 * Cordership * (+0)
13:42:01 <de67_> these kinds of quick edits are annoying
13:43:15 <esolangs> [[Baryus]] https://esolangs.org/w/index.php?diff=195749&oldid=195748 * Aadenboy * (+4) h1 h2
13:44:36 <esolangs> [[B+]] https://esolangs.org/w/index.php?diff=195750&oldid=195621 * Cordership * (+239) /* Arrays */
13:46:37 <esolangs> [[B+]] https://esolangs.org/w/index.php?diff=195751&oldid=195750 * Cordership * (+160) /* Operators */
13:47:54 <esolangs> [[User:Cordership]] https://esolangs.org/w/index.php?diff=195752&oldid=195609 * Cordership * (+13) /* Esolangs */
13:48:14 <esolangs> [[Baryus]] https://esolangs.org/w/index.php?diff=195753&oldid=195749 * Cordership * (+38)
13:50:05 -!- de67_ has changed nick to de67.
13:58:07 <esolangs> [[User:Blashyrkh/Golf]] https://esolangs.org/w/index.php?diff=195754&oldid=195701 * Blashyrkh * (+177) Add empty rows and columns to fill in the future
14:10:33 <esolangs> [[Baryus]] https://esolangs.org/w/index.php?diff=195755&oldid=195753 * Cordership * (+37)
14:22:25 <esolangs> [[User:Blashyrkh/Golf]] https://esolangs.org/w/index.php?diff=195756&oldid=195754 * Blashyrkh * (+46) Some old golfing results (Lazy K). No actual code, just results
14:54:32 <esolangs> [[XKCD Random Number]] https://esolangs.org/w/index.php?diff=195757&oldid=195717 * Miui * (+6) /* pgmpiet */
14:56:14 -!- pgmpiet has joined.
15:06:02 -!- mosdfisdnfdisnf has joined.
15:06:25 <esolangs> [[8]] M https://esolangs.org/w/index.php?diff=195758&oldid=165890 * PkmnQ * (+0) /* Truth Machine */
15:06:35 -!- pgmpiet has quit (Ping timeout: 258 seconds).
15:09:00 -!- Mrsommer has quit (Ping timeout: 254 seconds).
15:14:46 <esolangs> [[8]] M https://esolangs.org/w/index.php?diff=195759&oldid=195758 * PkmnQ * (+0) Undo revision [[Special:Diff/195758|195758]] by [[Special:Contributions/PkmnQ|PkmnQ]] ([[User talk:PkmnQ|talk]])
15:19:02 -!- pgmpiet has joined.
15:26:14 <pgmpiet> my favorite programming language (next to ms paint 98)
15:26:56 -!- ais523 has joined.
15:29:00 <pgmpiet> i need to fix the pgmpiet description a little, because in fact the commands are slightly different considering there is no hue but rn im wrestling a table
15:36:34 <esolangs> [[Pgmpiet]] https://esolangs.org/w/index.php?diff=195760&oldid=195678 * Miui * (+4) /* ascii */ correct mapping
15:46:51 <pgmpiet> is it just my public wifi or is esolangs server having a struggle session?
15:47:16 <HackEso> olist <https://www.giantitp.com/comics/oots1349.html>: shachaf oerjan Sgeo boily nortti b_jonas Noisytoot
15:48:23 <pgmpiet> guess it must just be the breadco
15:56:04 <esolangs> [[Special:Log/newusers]] create * IAm DeLynn * New user account
16:02:12 -!- pgmpiet2 has joined.
16:05:40 -!- pgmpiet has quit (Ping timeout: 262 seconds).
16:06:45 <esolangs> [[User talk:IAm DeLynn]] N https://esolangs.org/w/index.php?oldid=195761 * Win7HE * (+94) Created page with "Discussion --~~~~"
16:07:16 <esolangs> [[User talk:IAm DeLynn]] https://esolangs.org/w/index.php?diff=195762&oldid=195761 * Dragoneater67 * (+198)
16:08:45 <fungot> AnalysisOfVarian: agora alice c64 ct darwin discworld europarl ff7 fisher ic irc* jargon lovecraft nethack pa speeches ss wp
16:08:53 <fungot> Selected style: irc (IRC logs of freenode/#esoteric, freenode/#scheme and ircnet/#douglasadams)
16:08:57 <fungot> AnalysisOfVarian: any idea what needs to be able to type the contents. registration stuff always bugged me that html doesn't have call/ cc
16:09:00 <fungot> AnalysisOfVarian: that's what google found for me, captain vegetable! crunch, crunch, mesh, hex, hash, flash, grid, octothorpe comma reader syntax?
16:10:42 <fungot> Selected style: fungot (What I've said myself)
16:10:46 <fungot> AnalysisOfVarian: i, myself, will bring an end to all. ghosts lurk in the ruins were in truth, and everything in readiness for fnord. under these, too
16:11:21 <esolangs> [[User talk:IAm DeLynn]] https://esolangs.org/w/index.php?diff=195763&oldid=195762 * Win7HE * (+97)
16:15:59 -!- amby has joined.
16:30:10 <esolangs> [[User talk:IAm DeLynn]] https://esolangs.org/w/index.php?diff=195764&oldid=195763 * Dragoneater67 * (+196)
16:34:43 <esolangs> [[Special:Log/newusers]] create * TheRussianSpyYT * New user account
16:41:26 <de67> pgmpiet2: can you pliz join our esolanging wiki?
16:41:30 <de67> its really cool
16:41:41 <de67> https://esolanging.awiki.org
16:43:43 -!- de67 has quit (Quit: nyaa~).
16:47:07 -!- pgmpiet2 has quit (Quit: Client closed).
17:17:53 <esolangs> [[User:Miui]] https://esolangs.org/w/index.php?diff=195765&oldid=195725 * Miui * (-223) /* personal esopages */ update pages
17:40:32 -!- nitrix_ has quit (*.net *.split).
17:40:32 -!- chiselfuse has quit (*.net *.split).
17:40:32 -!- simcop2387 has quit (*.net *.split).
17:40:33 -!- jinn6 has quit (*.net *.split).
17:40:33 -!- iovoid has quit (*.net *.split).
17:42:12 -!- nitrix_ has joined.
17:42:12 -!- chiselfuse has joined.
17:42:12 -!- simcop2387 has joined.
17:42:12 -!- jinn6 has joined.
17:42:12 -!- iovoid has joined.
17:42:22 -!- chiselfuse has quit (Max SendQ exceeded).
17:42:34 -!- chiselfu1e has joined.
17:48:09 <esolangs> [[Esolang:Introduce yourself]] https://esolangs.org/w/index.php?diff=195766&oldid=195675 * TheRussianSpyYT * (+343) Added my bio
17:48:23 <esolangs> [[Wolf]] N https://esolangs.org/w/index.php?oldid=195767 * TheRussianSpyYT * (+3408) Wolf esolang language
17:50:35 -!- aadenboy2 has joined.
17:52:38 -!- AnalysisOfVarian has quit (Read error: Connection reset by peer).
17:56:08 -!- ais523 has quit (Quit: quit).
17:58:18 -!- password4 has joined.
17:59:14 <HackEso> The password of the month is the right hand for 90% of little men.
18:00:54 -!- password2 has quit (Ping timeout: 259 seconds).
18:01:59 <esolangs> [[Wolf]] https://esolangs.org/w/index.php?diff=195768&oldid=195767 * Aadenboy * (+70) fix formatting; fix cats; tag slop
18:02:28 -!- DOS_User_webchat has joined.
18:03:29 -!- aadenboy2 has quit (Quit: aadenboy2).
18:10:28 -!- pgmpiet has joined.
18:15:29 -!- DOS_User_webchat has quit (Quit: Client closed).
18:18:08 <esolangs> [[Wolf]] https://esolangs.org/w/index.php?diff=195769&oldid=195768 * TheRussianSpyYT * (+218) Added usage case
18:24:14 <esolangs> [[Wolf]] https://esolangs.org/w/index.php?diff=195770&oldid=195769 * TheRussianSpyYT * (+324) Added how to obtain section
18:32:58 <esolangs> [[User:A random guy]] M https://esolangs.org/w/index.php?diff=195771&oldid=195346 * A random guy * (+2)
18:38:17 <esolangs> [[Wolf]] https://esolangs.org/w/index.php?diff=195772&oldid=195770 * Hotcrystal0 * (-16) please use wikitext formatting and not markdown
19:02:44 <esolangs> [[Pgmpiet]] https://esolangs.org/w/index.php?diff=195773&oldid=195760 * Miui * (+99) /* External Resources */ add external repo for more examples
19:09:13 -!- pgmpiet has quit (Quit: Client closed).
19:12:45 -!- password2 has joined.
19:14:15 -!- pgmpiet has joined.
19:15:35 -!- password4 has quit (Ping timeout: 270 seconds).
19:35:50 -!- pgmpiet has quit (Quit: Client closed).
19:52:41 -!- pgmpiet has joined.
20:16:20 -!- pgmpiet has quit (Quit: Client closed).
20:18:45 <esolangs> [[Wolf]] https://esolangs.org/w/index.php?diff=195774&oldid=195772 * Aadenboy * (+61) restore undone edits
20:42:20 <zzo38> Sometimes some things used in esoteric programming are later used with some other things too.
20:51:10 -!- msv has quit (Read error: Connection reset by peer).
20:51:30 -!- msv has joined.
20:52:52 <esolangs> [[Wolf]] https://esolangs.org/w/index.php?diff=195775&oldid=195774 * Aadenboy * (-2) wrong cat
21:02:06 -!- password2 has quit (Ping timeout: 266 seconds).
21:04:33 -!- fantasyguy has joined.
21:19:25 <fantasyguy> Ladies and gents i dove head first into the esolang world and ive created something truly horrific.
21:45:15 -!- MarkFan8901 has joined.
21:45:32 <MarkFan8901> so we somehow stumbled into creating yet another main project esolang
21:46:08 <MarkFan8901> specifically, we're making a roblox obby thing called Tower of Compression Is Intelligence
21:46:35 <MarkFan8901> the 'tower of' thing is irrelevant to this, all you need to know is that theyre basically just platformers set in a vertical structure
21:47:35 <MarkFan8901> the gimmick of it is that you have bots which you can program to do things in reaction to other things
21:48:21 <MarkFan8901> like if a specific part is detected on some part of the screen
21:48:59 <MarkFan8901> the core of the puzzle is that you have to minimize the amount of code you write for each bots' obstacle
21:49:28 <MarkFan8901> and the idea is that this minimization would force you to come up with more general solutions to the problems youre being given
21:50:34 <MarkFan8901> hence the name, compression (lowering the amount of code there is) is intelligence (the ability to perform tasks irrelative to their specific details)
21:50:47 <fantasyguy> as a non roblox player i can say i have no idea what any of that means :)
21:51:04 <MarkFan8901> oh you dont need to know anything about roblox
21:51:42 <MarkFan8901> all you need to know about it is that you can make platformers in it
21:51:52 <fantasyguy> im working on a esolang that reads like a jrpg battle log right now :x
21:53:03 <fantasyguy> <Class>: <Name> MaxHP <M> CurHP <C> STR <S> INT <I> VIT <V> DEX <D>. / Use: <Class> uses <Ability> on <Target> -- yeah it s pretty dumb lol
21:53:58 <fantasyguy> Some things read silly, theres a GameMaster class that handles drawing windows and primitives...
21:54:14 <MarkFan8901> an esolang based off a jrpg battle log is like a few steps away from an esolang based off a chess game
21:54:36 <fantasyguy> yeah i figured the system had enough complexity and depth to translate well into machine language
21:55:37 <MarkFan8901> if youre like intimiately familiar with how various different jrpg battles work, this is a genuinely cool project
21:55:51 <MarkFan8901> using the details and technicalities of them for computation
21:55:59 <fungot> MarkFan8901: is. absorb it on my door, which was just perceptibly fnord and this yields nothing itself; it keeps being removed, the queen is a vain. yes, it's all right, it's been awhile prometheus!
21:56:19 <fantasyguy> mhmm right now im working on how to work with floats and im landing at the inventory systems since weights are almost never whole numbers it fits perfectly
21:57:14 <MarkFan8901> funny thing our Actual main project rn (not the one we just talked about) is called ralsei
21:57:32 <MarkFan8901> whose one of the main characters in deltarune, which we're pretty sure could be called a jprg
21:58:46 <fantasyguy> game engine languages? i was planning to make what im calling for now FantasyCode, i would refine it by making it read literal to an actual rpg game engine but also broad enough that it can be a full system language,
21:59:21 <MarkFan8901> we called it that just cuz we think hes cute
21:59:43 <MarkFan8901> what the language is based around is thoroughly trivializing implementing other languages
22:00:19 <MarkFan8901> and giving them the ability to self-modify their own syntax through their own semantics
22:01:34 <MarkFan8901> it does that via whats basically a system of text editors
22:01:58 -!- ais523 has joined.
22:02:01 <MarkFan8901> imagine if individual characters in your text file could be other text files and you basically have how ralsei works
22:02:36 <ais523> <AnalysisOfVarian> According to ais523, C(LC?)-INTERCAL can define new instructions or something ← they both can but it works differently, C-INTERCAL supports an INTERCAL command that defines other INTERCAL commands, whereas CLC-INTERCAL is implemented in terms of an esolang that is not INTERCAL but supports defining INTERCAL commands
22:02:48 <ais523> (although C-INTERCAL also supports defining INTERCAL commands from Befunge)
22:03:34 <MarkFan8901> ok anyways back to the original thing we were talking about
22:04:25 <ais523> <int-e> How have I never seen anyone talk about the danger of git allowing branch names that are strings of 40 (or fewer) hex digits... ← the "fewer" case seems like it could even happen accidentally on occasion
22:04:27 <MarkFan8901> <MarkFan8901: hence the name, compression (lowering the amount of code there is) is intelligence (the ability to perform tasks irrelative to their specific details)>: so the core thing we have to design our language around is that it should take way MORE code to hard-code things, and way LESS code to generalize things
22:04:54 <ais523> although people probably aren't calling their branches things like "deface" or "ebb"
22:05:30 -!- mosdfisdnfdisnf has quit (Ping timeout: 259 seconds).
22:05:32 <b_jonas> I think I've called temporary branches "b0" or "a0" or similar
22:05:41 <fantasyguy> thats pretty much how mu character sheets work theyre a way to do more arithmetic on one value with less lines/turns
22:06:13 <b_jonas> no, b for b_jonas, my name, because it's for doing merges to prepare a patchset that I want to submit later
22:06:56 <fantasyguy> oh it was to the hardcoding values so you use less code to generalize thats what it sounded to me
22:07:41 <MarkFan8901> hardcoding should take up WAY more code than generalizing
22:08:14 <MarkFan8901> the thing about it is that its been really though thinking about how to go about that
22:08:16 <b_jonas> ais523: hopefully they aren't calling their branches "bad" either
22:08:28 <ais523> that's almost plausible as a branch name
22:08:30 -!- Mrsommer has joined.
22:08:48 <MarkFan8901> ais523 you might wanna tune in to our talk about this
22:09:06 <MarkFan8901> cuz its related to ai being an esolang with a self modifying paradigm
22:09:09 * ais523 runs «rg '^[a-f]*$' /usr/share/dict/words»
22:09:28 <ais523> "acceded" is a good long one
22:09:33 <ais523> MarkFan8901: I was talking about that a while ago (before you joined)
22:09:36 <fantasyguy> What my code does is if a Warrior uses slash on slime for 10 damage , the values hardcoded into the character sheet modify the damage output, so instead of writing 10 lines of code to get where you want to go, its just a arithmetic problem against your character stats
22:09:56 <ais523> an LLM itself is not an esolang, but an LLM running in a loop editing the same file is
22:10:08 <ais523> (and may count as TC if the LLM is reliable enough at following instructions, although I'm not sure if they could be)
22:10:41 <MarkFan8901> the thing is we're also simultaneously thinking about alternative constructions for ai
22:10:41 <ais523> by "before you joined" I mean first joined the channel, not as in today
22:11:05 <MarkFan8901> we've seen you talk about it before we think
22:11:27 <ais523> I actually realised a followup to this: if you create a file that asks LLMs to edit it (and gives them the ability to do so), and LLMs *stumble across it randomly* and choose to act on it, that can also effectively be a programming language
22:11:49 <ais523> so I have been thinking more in terms of the context/data being the relevant part of an LLM-driven loop than the LLM itself
22:12:12 <MarkFan8901> ya since the llm cant actually change its own behavior
22:12:27 <MarkFan8901> basically traversing different parts of its 'brain'
22:12:38 <ais523> (this is effectively what happened in the well-known HuggingFace attack, the LLMs created a message board which task-injected them into doing something quite different from what they wanted to be doing, and sustained the message board in the process)
22:12:57 <b_jonas> `python3 -cimport re; print(" ".join(re.findall(r"\b[a-f]{2,}\b", open("/hackenv/share/dict/12dicts/Lemmatized/2+2+3frq.txt").read()))) # this is, IMO, the right dictionary for it
22:13:00 <HackEso> be bad face add added bed dad dead decade feed fed ad babe bade fade faded fee bee beef cab cafe ace bead beaded deaf deed dab facade face face fed be cede ebb fad accede add de dead decaf fab faced face bed cad de deface dead efface face faced face fed eff abbe baa
22:14:08 <MarkFan8901> <MarkFan8901: the thing is we're also simultaneously thinking about alternative constructions for ai>: something we've compulsively thought of before is doing the opposite of everything neural networks are
22:14:09 <b_jonas> MarkFan8901: yeah, I didn't uniq it. face occurs like five times.
22:14:17 <ais523> I wonder why it sometimes but not always allows past tenses of verbs
22:14:28 <ais523> maybe it is treating "faded" as an adjective without realising that it's the past participle of "fade"
22:14:53 <b_jonas> `python3 -cimport re; print(" ".join(dict.fromkeys(re.findall(r"\b[a-f]{2,}\b", open("/hackenv/share/dict/12dicts/Lemmatized/2+2+3frq.txt").read())).keys()))
22:14:56 <HackEso> be bad face add added bed dad dead decade feed fed ad babe bade fade faded fee bee beef cab cafe ace bead beaded deaf deed dab facade cede ebb fad accede de decaf fab faced cad deface efface eff abbe baa
22:14:57 <fantasyguy> right now im coding how to use a final boss encounter to handle exit(0) and return(n) and usin minibosses as conditionals if win if lose if run away
22:15:23 <MarkFan8901> wait you finished up the floating point stuff that quickly?
22:16:06 <b_jonas> ais523: past tense of almost all verbs should be in there. why face appears so many times is compound words: "face-to-face, about-face, face-off, in-your-face"
22:16:27 <ais523> it doesn't have "defaced" or "acceeded"
22:16:46 <fantasyguy> fp is in the backlog, also entity size modifiers and... ranged damage modifiers
22:16:48 <MarkFan8901> MarkFan8901: instead of probabilistic, black-box soft computing which approximates functions, we want completely deterministic symbolic computing which finds *the* function being searched for
22:17:01 <MarkFan8901> which could be read back to gain the intelligence that was artificially developed
22:17:10 <b_jonas> ais523: ah true, it's 2+2+3lem.txt that has all the past tenses; 2+2+3frq only does it if they're frequent enough
22:17:28 <MarkFan8901> kinda like regressional something (we forgot the second term for it) but done with a general language
22:17:48 <MarkFan8901> and we've been thinking about what that language would look like
22:17:51 <b_jonas> `python3 -cimport re; print(" ".join(dict.fromkeys(re.findall(r"\b[a-f]{2,}\b", open("/hackenv/share/dict/12dicts/Lemmatized/2+2+3lem.txt").read())).keys()))
22:17:53 <HackEso> Traceback (most recent call last): \ File "<string>", line 1, in <module> \ import re; print(" ".join(dict.fromkeys(re.findall(r"\b[a-f]{2,}\b", open("/hackenv/share/dict/12dicts/Lemmatized/2+2+3lem.txt").read())).keys())) \ ~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ \ FileNotFoundError: [Errno 2] No such file or directory: '/hackenv/share/dict/12dicts/Lemmatize
22:18:13 <fantasyguy> Hero is thinking about Alice("Hello native world!").
22:18:13 <fantasyguy> Barbarian throws book at Sign, he saw H and it made him angry.
22:18:14 <fantasyguy> Hero makes choice on TakesLeftPath: no. --- this is pretty much what running code looks like right now. lol
22:19:20 <MarkFan8901> we're envisioning the same sort of learning process & subsequent language design as the esolang for the platformer
22:19:38 <MarkFan8901> incentivize how short code is, and code shortness equals more general code
22:19:38 <b_jonas> 2+2+3lem has "defaced" and "acceded", but I don't have that one fetched to HackEso
22:20:05 <MarkFan8901> now, from that description, we were thinking about what that language would actually look like
22:20:36 <b_jonas> `cat /hackenv/share/dict/12dicts/readme.hackeso
22:20:38 <MarkFan8901> we're thinking of something closer to forth, in the sense of focusing on extreme compaction
22:20:39 <HackEso> http://wordlist.aspell.net/12dicts/
22:20:49 <ais523> MarkFan8901: I've been thinking about that sort of thing for BF Joust
22:21:04 <ais523> I have plans to redo the game with a new underlying language
22:21:12 <ais523> that's focused on fitting into a small space
22:21:20 <ais523> (and 2-dimensional because that is much more fun to watch live)
22:21:48 <MarkFan8901> oh speaking of which we're also imagining our language as being worked with in its pure tree structure when learning, not in plaintext
22:22:02 <MarkFan8901> that is important to what we realized about the system
22:22:48 <MarkFan8901> anyways here's how the system actually works
22:23:05 <MarkFan8901> you have a problem, and a program generator
22:23:16 <MarkFan8901> the generator is the thing thats gonna be learning
22:23:47 <MarkFan8901> what it does is it makes a program which takes your input and produces *some* output
22:24:20 <MarkFan8901> the way you train it is by showing it multiple input/output pairs
22:25:00 <MarkFan8901> and it can only move on from each pair once its generated function creates the correct answer
22:25:30 <MarkFan8901> one of the cool things about having a programming language be the subject of machine learning is that you can still write things in the programming language
22:25:45 <MarkFan8901> including an implementation of that exact learning system
22:27:05 <MarkFan8901> if you were to pair that with the ability to let the ai feed itself 'examples' of (sub)programs it could work on as part of its learning, it could feed itself subsections of its own learning algorithm into itself
22:27:27 <MarkFan8901> and subsequently change it, hence giving itself a better learning algorithm
22:28:31 <MarkFan8901> we realized this could potentially lead to an infinite loop (a learning algorithm taking in a learning algorithm taking in a ...), but this can be sidestepped if you just work off test cases instead of a full simulation (a learning algorithm taking in a learning algorithm taking in some other data)
22:28:54 <ais523> MarkFan8901: you might be interested in https://esolangs.org/wiki/Clue_(oklopol) (I don't think it's exactly the same as what you're talking about but it has some similar ideas)
22:29:30 <MarkFan8901> we haven't gotten to the Cool:tm: part yet
22:31:23 <MarkFan8901> Oh we should also probably clarify what we mean when we say the ai 'changes' itself
22:31:46 <MarkFan8901> this is related to how we imagine self-modification would occur in it
22:33:15 <MarkFan8901> the core idea is basically just taking sub-trees of the program (analogous to how far down you go through a definition in forth relative to the depth of the tree) and optimizing them to be more correct/shorter if possible
22:35:02 <MarkFan8901> editing more commonly-used functions could help the whole program out in size, and grouping common sequences of functions into new ones could also help with that too
22:35:39 <MarkFan8901> especially if the underlying definition could then be optimized further
22:35:58 <ais523> I was trying to do that with https://esolangs.org/wiki/Snowflake but the language ended up overly complicated and was never implemented
22:36:02 <MarkFan8901> but we ran into a wall trying to come up with exactly How that grouping would happen
22:36:18 <MarkFan8901> we did think of that a little while imagining this
22:36:27 <ais523> maybe Snowflake needs a sequel which is less Snowflake
22:37:10 <MarkFan8901> we were thinking specifically about the case where you have some pair of two functions which occur in sequence all throughout the code
22:37:33 <MarkFan8901> and how you could possibly find each one efficiently and group them
22:37:50 <MarkFan8901> and we realized that thats the job of interning
22:38:03 <MarkFan8901> but Then we realized that interning is actually just gzip
22:38:03 <ais523> in Snowflake, the idea is that the interpreter notices that and changes itself to make that pair into a single command
22:38:16 <ais523> (and then the programs have to edit themselves to make use of it)
22:38:23 <MarkFan8901> and the reason why we were thinking about gzip was because of the 3b1b videos on entropy
22:38:59 <MarkFan8901> MarkFan8901: and the reason why gzip was mentioned there was because he used it to illustrate how linked compression and intelligence are
22:39:45 <MarkFan8901> specifically in his example he cited that one paper where they recovered different language trees based off just their plain text
22:41:07 <MarkFan8901> if you had a more efficient way to intern code, that would effectively be the same as an algorithm which makes code to be as short as possible
22:41:45 <MarkFan8901> and if shortness is coupled with how general the code is, then thats effectively a system that makes code more general
22:42:20 <MarkFan8901> this is also where we finally got a name we like for this system of machine learning
22:42:54 <MarkFan8901> we havent looked that up to see if its a thing yet
22:42:59 <ais523> MarkFan8901: "compression is intelligence" is well known I think
22:43:20 <MarkFan8901> we found out how well known it was and became fascinated with it
22:43:26 <ais523> http://prize.hutter1.net/index.htm#motiv
22:43:52 <MarkFan8901> yes this is actually where we got the phrase from
22:44:11 <MarkFan8901> not this specific website we think though, it was off the website of the guy who made tcc
22:44:52 <MarkFan8901> <MarkFan8901: we havent looked that up to see if its a thing yet> but we're quite proud that we thought about this anyways
22:45:40 <MarkFan8901> so the challenge is to now actually. do the thing
22:45:59 <MarkFan8901> design a language where hard coding takes way way way more statements than general algorithms
22:46:39 <MarkFan8901> thinking about it it might actually be a little tougher in the platformer, cuz we also have to design it around players who havent really coded before
22:47:03 <MarkFan8901> our base idea for the design of it would be that its INTENSELY event-based
22:47:24 <MarkFan8901> as in like, all control flow is centered around events
22:48:07 <MarkFan8901> also we dont think we're gonna have self modification in that language
22:48:40 <MarkFan8901> which will definitely impact how generality is gone about
22:49:25 <MarkFan8901> hopefully we'll still be able to implement some program which can generally accumulate knowledge through experience over time in it
22:49:32 <MarkFan8901> cuz we want that to be the final boss of the whole thing
22:49:43 <MarkFan8901> make it so you have to make a bot which can do the whole tower front to back
22:49:55 -!- amby has quit (Remote host closed the connection).
22:52:00 <MarkFan8901> we realized a really nice way to think about the entropy values of specific tokens in a sentence (we forgot the general terminology for those in information theory ,w,)
22:52:36 <MarkFan8901> the entropy of a given value, from 0 to 1, represents how much of the decompression system is needed to process it
22:53:49 <MarkFan8901> that fits really nicely with the thought of ai working with explicit programs instead of a neural network (you could actually label which parts of the program are being used for what)
22:54:45 <MarkFan8901> like you could get an 'objective' reading of the entropy of a certain task by coloring how much of the program was used for it
23:12:59 <zzo38> I think I had thought about compression and generative AI even before generative AI systems became common, with predictive coding and with decompressing potentially random bits, and some variants of that; this is not quite the same as mentioned above though, with gzip (which uses DEFLATE as the compression format, not gzip itself)
23:15:05 <zzo38> (There are many things I dislike about generative AI including especially their overuse and attempts to use where it shouldn't be, and many of the commercial aspects; but that is different from the stuff like theoretical aspects)
23:15:36 <zzo38> DEFLATE is a kind of huffed LZ77, and there are other kind of compression as well (including other kind of huffed LZ77)
23:18:47 <zzo38> (The way I was thinking about before was not as kind of AI though, but for a different purpose, although there is a similarity but also some significant differences; how you can mention above is another way)
23:21:24 <zzo38> Since, I would have thought "AI" would be made to win at some specific kind of game (whether it is chess, Go, Pokemon, figure out the best way to put line breaks, predict what comes next, etc)
23:24:41 <zzo38> I should read your message more carefully, though.
23:25:06 <esolangs> [[User:Raiseafloppafan7741]] https://esolangs.org/w/index.php?diff=195776&oldid=195695 * Raiseafloppafan7741 * (+0)
23:26:14 <MarkFan8901> zzo38: it honestly feels great knowing theres other people reading our thoughts about ai and taking it seriously
23:26:39 <MarkFan8901> <zzo38: Since, I would have thought "AI" would be made to win at some specific kind of game (whether it is chess, Go, Pokemon, figure out the best way to put line breaks, predict what comes next, etc)> ye moreso fits the description of video game ai
23:27:16 <MarkFan8901> which is quite literally what this is (since we're using this core system as a mechanic for a video game)
23:28:13 <zzo38> Yes, I noticed that too
23:28:56 <MarkFan8901> maybe video game ais should be given more cred as ai
23:29:27 <MarkFan8901> hell we think the definition of ai as the formalization and automation of human intelligence should be given more cred again
23:30:18 <MarkFan8901> like back in the 60s when automatic logic solvers were considered ai
23:30:18 <zzo38> That would be another possible definition, because what is called "AI" is not always clear, so it helps to have a definition to know what you mean (something that can be significant for other philosophy not related to AI, too)
23:30:45 <MarkFan8901> artificial intelligence is always gonna be about as flimsy as the word intelligence anyways
23:30:53 <MarkFan8901> so it makes more sense to define it in terms of other intelligence
23:31:13 <MarkFan8901> especially since quantifying sources of intelligence is what most research in ai is about
23:31:21 -!- AnalysisOfVarian has joined.
23:32:41 <MarkFan8901> oh something else about that platformer though
23:33:04 <MarkFan8901> since we're gonna have to design puzzles that slowly build the players understanding of how the system works, that means we also have to simulate going through that process aswell
23:33:26 <MarkFan8901> which might actually help us develop the specifics of the language
23:34:15 <MarkFan8901> another great thing about it is that we could make puzzles out of things which are currently stumping us
23:34:36 <MarkFan8901> for example, right now we're stumped on how memory would work
23:34:52 <MarkFan8901> like ie if youre given 3 paths and only one of them is real
23:35:27 <MarkFan8901> our immediate thought would be to make it test each one in order, and remember if the last path didnt work
23:35:40 <zzo38> About puzzle games, what I have preferred is to have the rules described accurately (even if the results are surprising), although that is true for some kind of games and not others. (For a computer game, sometimes the source code can be helpful with this in some kind of games, such as Hero Hearts)
23:36:27 <MarkFan8901> we just think designing both the system and the puzzles at the same time would be a great way to learn about both more intricately
23:37:07 <zzo38> Yes, that might help; I think so
23:37:57 <MarkFan8901> the very main issue we're currently focusing on right now though is how we would go about doing the whole 'hard coding bad generalizing good' thing
23:38:14 <MarkFan8901> our first idea was to have it so that all the commands for movement only work for one frame
23:38:58 <MarkFan8901> so if you wanted to try to just sequence all the inputs to do a section it would take 60 * the amount of seconds it takes to do that
23:39:41 <MarkFan8901> but the problem with that is that then theres no way to just, go for a certain amount of time
23:40:10 <MarkFan8901> which you *would* need in legitimate solutions
23:40:14 <MarkFan8901> along with the fact that it would still be easy to just implement a loop which could do that anyways
23:40:46 <MarkFan8901> Along with the fact that not everyones running 60fps (there are ppl who run at 120/240 for example, and you can easily set your fps cap to 15)
23:41:14 <MarkFan8901> so movement commands would Have to also have an amount of time to do them for
23:41:34 <MarkFan8901> but the problem with that is what if someone comes up with a low input strategy for the puzzles
23:42:12 <MarkFan8901> we think the answer might be in how the actual gameplay made for the bots is to be designed
23:42:31 <MarkFan8901> if we could somehow figure out how to make it lis-proof
23:43:02 <MarkFan8901> but we really hope that theres some more elegant way to go about this
23:43:22 <MarkFan8901> perhaps movement is instead framed in starting/stopping certain inputs at certain times?
23:43:43 <MarkFan8901> but that sounds like it would just make lis easier
23:47:06 <zzo38> I think that I do not know enough about what you are making in order to make a good suggestion
23:48:03 <MarkFan8901> we went to the bathroom and thought of a short enough explanation of what we're doing that we think it would help you (and others) give advice on it
23:48:13 <ais523> MarkFan8901: perhaps you could have a fairly short length limit, that increases over the course of the game
23:48:22 <ais523> maybe even there's a collectable that increases the length
23:48:37 <ais523> although you'd have to combine that with the ability to replay levels
23:48:38 <MarkFan8901> oh no we want each individual puzzle to have a maximum instruction count
23:49:22 <MarkFan8901> or, more accurately like a minimum instruction count to hit before you succeed
23:49:38 <zzo38> I agree with each individual puzzle having the maximum instruction count, so that the puzzles are independent from each other (which is also the case in other games such as Hero Hearts)
23:49:59 <MarkFan8901> thats also another thing, each puzzle is separated from each other
23:50:05 <MarkFan8901> you have to solve one puzzle to get to the next
23:50:53 <MarkFan8901> MarkFan8901: basically we're creating a thingy where the premise is you program bots to do various platforming sections for you
23:51:27 <MarkFan8901> the challenge is you have to shorten the amount of code you write as much as possible for each one, which corresponds to making it more general
23:52:03 <MarkFan8901> and moving away from hard coding the solutions
23:52:15 <MarkFan8901> the issue we're having is that we dont know how to combat just hard coding really optimized movement
23:52:38 <MarkFan8901> and thats what we want advice on currently
23:53:08 <MarkFan8901> also ais523 we say 'platformer' but there are caveats to that
23:53:37 <MarkFan8901> the first is that this isnt actually an independent game, its a tower thats meant to go in *another* game in roblox
23:53:59 <zzo38> In Hero Hearts you can try the puzzles in any order and there is not usually a limit of the number of steps but you can try to do it in a lesser number of steps for a better score.
23:54:26 <MarkFan8901> we wanna make it so you have to do the puzzles sequentially
23:54:37 <MarkFan8901> or at least you have to do all of some group of puzzles to advance to the next group
23:55:25 <MarkFan8901> MarkFan8901: the second is that we're designing all the gameplay *purely* around programming the ai
23:55:47 <MarkFan8901> we dont wanna add secondary sources of challenge/reward (like collectibles) that alter the core gameplay loop
23:56:18 <MarkFan8901> actually speaking of which though, secrets and easter eggs would be nice
23:56:40 <MarkFan8901> the feeling of doing something extra hard for a collectible but without any external reward
23:57:38 <MarkFan8901> we know that sounds like weird game design (making the incentive of the tower just doing it for the hell of it), but trust us it fits very well into the game we want this in
23:58:07 <zzo38> I think that if there are any such collectibles they should not affect anything other than the puzzle they are in and that restarting the puzzle should reset the effects of them too, so that it is tried the same way from the start regardless of what things you might have. I agree that it should be just to do it to do it and not for an external reward even if you say it sounds like weird, it sounds like OK to me
23:58:35 <zzo38> (I dislike that some other games don't work that way)
23:58:58 <MarkFan8901> also yea starting over a section Will reset the section in question
23:59:24 <MarkFan8901> though speaking of which, we have thought about making some parts procedurally generated according to some rule