00:02:54 -!- arseniiv has quit (Ping timeout: 258 seconds).
00:09:51 <int-e> Okay, let's do this properly this time.
00:41:20 -!- _ has joined.
00:41:36 -!- _ has quit (Remote host closed the connection).
00:43:05 <esowiki> [[Nybblang]] M https://esolangs.org/w/index.php?diff=70096&oldid=70029 * PythonshellDebugwindow * (+127) /* Program import */
00:50:01 <esowiki> [[Orca]] M https://esolangs.org/w/index.php?diff=70097&oldid=70081 * Neauoire * (+177)
01:01:06 <int-e> (Dang it: I missed a corner case in my Ponder This approach.)
01:22:16 -!- sebbu has quit (Read error: Connection reset by peer).
01:25:43 -!- xelxebar has quit (Ping timeout: 240 seconds).
01:27:19 -!- xelxebar has joined.
01:28:29 <int-e> And it was so typical... the mistakes are never in the tricky part of the code, but in the places you consider obvious and easy :-/
01:36:57 <int-e> fungot: anything to lighten the mood?
01:36:57 <fungot> int-e: what really? i thought eval would bork when asked to input a character, or it's only in pity.
01:40:06 <esowiki> [[Eso2D]] M https://esolangs.org/w/index.php?diff=70098&oldid=70092 * PythonshellDebugwindow * (+410) /* Set the accumulator to 0 */ new extension. TC
01:40:25 <esowiki> [[Eso2D]] M https://esolangs.org/w/index.php?diff=70099&oldid=70098 * PythonshellDebugwindow * (+1) /* New commands */
02:07:57 -!- FreeFull has quit.
02:15:20 -!- tromp has quit (Remote host closed the connection).
02:20:57 -!- sebbu has joined.
02:28:05 -!- imode has joined.
02:31:25 -!- tromp has joined.
02:32:41 -!- tromp_ has joined.
02:33:55 -!- tromp__ has joined.
02:36:31 -!- tromp has quit (Ping timeout: 272 seconds).
02:37:47 -!- tromp_ has quit (Ping timeout: 272 seconds).
02:39:03 -!- tromp__ has quit (Ping timeout: 272 seconds).
03:10:21 -!- imode has quit (Remote host closed the connection).
03:10:54 -!- imode has joined.
03:33:03 -!- tromp has joined.
03:37:57 -!- tromp has quit (Ping timeout: 272 seconds).
04:03:39 <esowiki> [[AT]] M https://esolangs.org/w/index.php?diff=70100&oldid=70089 * Hakerh400 * (-5)
05:15:02 -!- imode has quit (Remote host closed the connection).
05:17:44 -!- imode has joined.
05:23:41 <zzo38> Apparently, someone used to sign their Usenet posts with a line about "evil umpire", but this was too early to be archived.
05:26:03 <zzo38> Also, I should probably implement scoring in bystand. There seems to be a lot of spam posted through Google to news.software.readers. (Maybe that is why some Usenet users do not want to receive messages posted through Google.)
05:27:12 -!- imode has quit (Ping timeout: 258 seconds).
05:29:18 -!- TellsTogo has joined.
05:35:43 <zzo38> (I do not wish to block all messages posted through Google, since some people who use Google have actual relevant stuff to write. However, I do not post using Google myself.)
06:17:42 <zzo38> Do you like Hollerith chording?
06:19:42 <kmc> zzo38: what is that?
06:20:54 <zzo38> It is my idea of how to enter text using a numeric keypad. There are other systems of entering text using a numeric keypad, and I think all of them are not very good.
06:21:49 <zzo38> Hollerith chording means if you push one button at a time then it makes numbers; if you push multiples at a time then it corresponds to multiple holes in one column of a computer card, and makes whatever letter that combination of holes represents.
06:22:44 <zzo38> (A separate button is needed for a space, although most systems using a numeric keypad already have a few other buttons anyways.)
06:23:47 <zzo38> What do you think is the best system for text entry using a numeric keypad?
06:41:08 -!- kritixilithos has joined.
06:45:26 -!- mniip has quit (Quit: This page is intentionally left blank.).
06:45:53 -!- mniip has joined.
07:23:05 -!- Cale has quit (Ping timeout: 240 seconds).
07:35:25 <TellsTogo> zzo38 Dotsies https://dotsies.org/
07:41:30 -!- Lord_of_Life_ has joined.
07:43:16 -!- Cale has joined.
07:43:34 -!- Lord_of_Life has quit (Ping timeout: 256 seconds).
07:43:34 -!- Lord_of_Life_ has changed nick to Lord_of_Life.
07:44:23 -!- kritixilithos has quit (Ping timeout: 240 seconds).
07:45:23 -!- kritixilithos has joined.
07:45:58 -!- tromp has joined.
07:52:40 -!- b_jonas has quit (Quit: leaving).
08:11:02 <Cale> `pbflist https://pbfcomics.com/comics/the-report/
08:11:04 <HackEso> pbflist https://pbfcomics.com/comics/the-report/: shachaf Sgeo quintopia ion b_jonas Cale kmc
09:04:02 -!- wib_jonas has joined.
09:04:58 -!- cpressey has joined.
09:27:57 -!- PkmnQ has joined.
09:29:07 -!- PkmnQ has quit (Remote host closed the connection).
09:36:36 <cpressey> Good morning. If a conventional OS has a Command-Line Interface (CLI) then it stands to reason that a Befunge-based OS would have a Command-Plane Interface (CPI).
09:40:30 <wib_jonas> cpressey: I'm not sure about that. a Befunge-based OS might use Befunge as a platform to run user-space programs, or for the kernel, neither implies that the shell has to be befunge too.
09:41:06 <wib_jonas> There are systems that run mostly machine code programs with system calls, but have a unix sh or dos shell or BASIC or Forth command line interface
09:41:41 <wib_jonas> On the other hand, some (but not all) BASIC-based interfaces are already command planes, in that they let you edit anywhere on a screen sized buffer of inputs and let you run any line there
09:42:38 <wib_jonas> By all rights that stupid system should be extinct, but it lives on in worksheet interfaces like that of Mathematica or SAGE, and the immediates window of VBA.
09:46:08 <wib_jonas> You could have a befunge command-plane shell on an OS that isn't based on befunge though, and befunge is particulary usable for this in fact,
09:46:43 <wib_jonas> because befunge already takes the program from the same command plane to where it can read data inputs and write outputs, just like how those BASIC shells work.
09:48:17 -!- wib_jonas has quit (Remote host closed the connection).
11:13:18 -!- laerling has quit (Quit: ZNC 1.7.3 - https://znc.in).
11:13:38 -!- laerling has joined.
11:26:23 -!- laerling has quit (Quit: ZNC 1.7.3 - https://znc.in).
11:26:43 -!- laerling has joined.
11:30:45 -!- laerling has quit (Client Quit).
11:31:35 -!- laerling_ has joined.
11:32:51 -!- b_jonas has joined.
11:42:21 -!- laerling_ has quit (Quit: ZNC 1.7.3 - https://znc.in).
11:42:40 -!- laerling_ has joined.
11:55:14 -!- laerling_ has quit (Quit: ZNC 1.7.3 - https://znc.in).
11:57:23 <int-e> tromp: Hmm, I tried to match your 74 bit equality test but I'm stuck at 79 for now. But I have a 73 bit *dis*equality test now.
11:57:55 <tromp> mine is 73 bit now
11:58:03 -!- laerling has joined.
11:58:31 <tromp> hint: it's just using scott_succ and scott_pred
11:59:01 <FireFly> is this the binary format of LC?
11:59:15 <tromp> a slightly tweaked scott_pred actually
11:59:27 <FireFly> or what's the bitcount for?
11:59:31 <int-e> tromp: But I also have 58 bit less than and less than or equal tests, and a 66 bit minimum.
11:59:39 <tromp> yes, size measured in blc encoding
12:08:04 <tromp> now embarking on a much more involved term looking for odd perfect numbers
12:13:12 <tromp> btw, modulo is <= 113 bits
12:13:49 <tromp> and div <= 126 bits
12:14:09 <b_jonas> tromp: could you look for Fermat primes 2**(2**k)+1 for 5<=k integer instead?
12:15:11 <tromp> that would be straightforward given our small primesieve
12:17:39 <b_jonas> though I guess odd perfect numbers would be a more interesting goal
12:19:02 <tromp> so far, smallest conjecture testing program is 213 bit Laver table one
12:19:31 <tromp> goldbach sits at 267 bits
12:20:09 <int-e> div is <= 105 bits.
12:20:13 <tromp> i expect oddperfect to come in between those
12:20:25 <tromp> int-e: cool, something new!
12:20:42 <int-e> And \n m. n `div` (m+1) is even shorter.
12:25:21 <b_jonas> int-e: what representation of natural numbers are you using here for those bit counts?
12:27:43 <b_jonas> the original one, which encodes 4 as (\x.\y.x(x(x(x y)))) ?
12:29:33 <b_jonas> it's confusing that that's called Church encoding
12:29:41 -!- ArthurStrong has joined.
12:29:43 <b_jonas> but that's not easy to fix it now
12:31:00 <FireFly> what other encoding is referred to that way?
12:31:15 <FireFly> That's the only one I know of under that name
12:31:37 <int-e> tromp: let's make that 95
12:31:52 <int-e> (and 88 for n/(m+1))
12:34:05 -!- arseniiv has joined.
12:35:56 <b_jonas> FireFly: there's an encoding for algebraic types, where if you have a type with say three constructors A, B, C, then you encode a value A as (\a\b\c.a), the value (B x y) as (\a\b\c.bxy), and the value (C z) as (\a\b\c.cz)
12:36:30 <b_jonas> of which a special case is encoding False as (\f\t.f) and True as (\f\t.t)
12:36:38 <tromp> that's called the Scott encoding?
12:37:04 <b_jonas> is it? I think someone called it Church something, at least in the case of non-recursive data types
12:37:09 <b_jonas> let me look up the chat log
12:38:15 <b_jonas> hmm no, I'm probably just confused
12:38:19 <b_jonas> I don't know how all these encodings work
12:39:42 <FireFly> Ah, I guess the confusion might be due to there being a common subset of them?
12:40:12 <FireFly> So the Church booleans end up the same as the Scott encoding for a boolean type
12:40:14 <int-e> s/95/91/ (and 84 for n/(m+1)), hmm. But I think this may be it. mod will be worse.
12:41:30 <tromp> are you using any lists?
12:41:49 <tromp> mine are all list based:(
12:43:34 <int-e> I'm thinking of this as coroutines for the time being.
12:43:58 <tromp> that's way more awesome than lists
12:45:39 <int-e> trying the same with mod now, hmm. doesn't terminate... what did I do wrong :)
12:46:03 -!- kritixilithos has quit (Ping timeout: 240 seconds).
12:48:36 <b_jonas> int-e: try not to divide by zero
12:51:12 <int-e> b_jonas: that wasn't the issue ;)
13:01:48 <tromp> trust int-e to try with nonzero denominators:)
13:02:38 <tromp> did you also use coroutines for eq?
13:06:07 -!- kritixilithos has joined.
13:19:20 <esowiki> [[Eso2D]] M https://esolangs.org/w/index.php?diff=70101&oldid=70099 * PythonshellDebugwindow * (+18) /* Multi-cell extension (Turing-complete) */
13:22:18 <tromp> int-e: eq/mod/div added to AIT repo. feel free to make improvements
13:22:37 <int-e> Hmm, mod with this approach seems to come out at 115 bits.
13:23:16 <int-e> Which is okay I suppose. :P
13:24:20 <tromp> very competitive with mine!
13:24:31 <tromp> your div is way better though
13:25:21 <int-e> yeah it's 90 bits now :)
13:27:14 <tromp> would that be the first example of coroutines in the repo?
13:29:27 <int-e> pushed the div, so feel free to argue whether this is using coroutines or not.
13:33:10 <int-e> and mod is now 107 bits
13:34:02 <int-e> (Observation: n carries enough "recursive power" to avoid the full fixed point construction)
13:35:59 <esowiki> [[Eso2D]] M https://esolangs.org/w/index.php?diff=70102&oldid=70101 * PythonshellDebugwindow * (+160)
13:36:18 <esowiki> [[Eso2D]] M https://esolangs.org/w/index.php?diff=70103&oldid=70102 * PythonshellDebugwindow * (+0) /* Commands */
13:40:17 <tromp> https://github.com/tromp/AIT
13:40:59 <tromp> int-e: good point about using n as poor man's Y
13:41:38 <kritixilithos> so you're golfing functions in binary lambda calculus?
13:43:20 <kritixilithos> oh you're this guy https://tromp.github.io/cl/cl.html
13:45:17 <tromp> or rather, this guy https://github.com/tromp/
13:45:50 <tromp> int-e: that change also makes it so that mod n 0 no longer diverges:)
13:46:27 <tromp> now, we get mod n 0 = 1
13:46:43 <int-e> . o O ( very useful )
13:48:31 <tromp> it's really cool that you can define mod 0 m as m false, using only 2 bits for the inner argument
13:49:45 -!- sprocklem has quit (Ping timeout: 258 seconds).
13:50:48 <esowiki> [[Eso2D]] M https://esolangs.org/w/index.php?diff=70104&oldid=70103 * PythonshellDebugwindow * (+18) /* Commands */
13:51:05 <int-e> so basically what "kills" the coroutine approach for mod is that the counter isn't monotonic, so I need to deal with an extra argument... doing that in three places adds up to 20-ish bits.
13:53:01 <cpressey> dare I ask why the counter isn't monotonic
13:53:18 <tromp> mod m m < mod (m-1) m
13:53:18 <int-e> because it has to be reset when it reaches the modulus.
13:54:02 <int-e> in contrast to 'div', which is just, pass, pass, pass, bump counter, pass, pass...
13:55:17 <Taneb> Can you use mod x y = x - div x y to get any gain?
13:55:44 <int-e> Taneb: No, I can't do - and * in 25 bits.
13:56:25 <int-e> (- is the worse of the two... 56 bits is the best I have for that)
13:57:08 <int-e> But even \m\n. \f\x. m f (n f x) is already 33 bits.
13:57:50 <int-e> err. what am I doing?
13:58:39 <int-e> Multiplication *is* cheap, \m\n. \f. m (n f) = 19 bits. (These numbers change by about 10 when the thing is inlined.)
14:00:18 <tromp> i sometimes ponder if i should rename binary lambda calculus to "bitwise"
14:00:37 <tromp> but it would make googling way harder:(
14:00:59 <int-e> meh it's the "LC" part that's alluring to me
14:01:28 <tromp> yes, i like to mention that too. it's just quite a mouthful
14:02:22 <tromp> normally, programming languages names are single words
14:02:23 -!- kritixilithos has quit (Ping timeout: 240 seconds).
14:03:12 <fizzie> Just smoosh those words together, into "Bilaca" or something, Fortran-style.
14:07:52 -!- kritixilithos has joined.
14:20:58 -!- kspalaiologos has joined.
14:28:03 -!- kritixilithos has quit (Ping timeout: 240 seconds).
14:29:53 -!- kritixilithos has joined.
14:53:36 -!- ArthurStrong has quit (Quit: leaving).
14:57:06 <int-e> . o O ( BottLeneCk )
15:04:24 <tromp> ot in spanish BLanCa
15:07:58 -!- sprocklem has joined.
15:09:55 -!- cpressey has quit (Quit: A la prochaine.).
15:21:39 <tromp> regarding this month's Ponder This: the nice regular spacing of 20% unoccupied board squares is ruled out. so it's something irregular with higher unoccupied density towards the borders
15:23:11 <tromp> in particular, there can be no unoccupied square surrounded regularly by 4 others at knight's move distance
15:23:37 <arseniiv> tromp: you could portmanteau BLaCk and BLanCa into BLanCk which may be more googlable if not many people misspell “blank” (or write “blanck” intentionally)
15:23:42 <tromp> as its 4 directly adjacent squares would all need to be different
15:24:45 <arseniiv> (or for a z-grade linguistic pun, BLãCk)
15:33:21 -!- xkapastel has joined.
15:51:28 -!- sprocklem has quit (Ping timeout: 256 seconds).
16:06:50 -!- sprocklem has joined.
16:09:29 <int-e> tromp: We should probably discuss this in April :P
16:10:39 <int-e> (I know I talked about this a lot but I tried to keep it vague and avoid spoilers.)
16:16:03 -!- kritixilithos has quit (Ping timeout: 240 seconds).
16:17:21 -!- kritixilithos has joined.
16:30:41 <esowiki> [[Asm2bf]] https://esolangs.org/w/index.php?diff=70105&oldid=70095 * Palaiologos * (+644) new bfpp (based on Lua); r5 and r6 added to the register model
16:33:37 <tromp> oh, thought you talked about the ladders and snakes one
16:49:13 -!- laerling has quit (Quit: ZNC 1.7.3 - https://znc.in).
16:53:13 -!- laerling has joined.
16:55:40 -!- sprocklem has quit (Ping timeout: 258 seconds).
17:57:36 -!- FreeFull has joined.
17:59:56 -!- rain1 has joined.
18:00:36 <esowiki> [[Asm2bf]] M https://esolangs.org/w/index.php?diff=70106&oldid=70105 * Palaiologos * (-408) remove automated calls section
18:03:21 -!- TellsTogo has quit (Remote host closed the connection).
18:12:50 -!- LKoen has joined.
18:13:57 -!- sprocklem has joined.
18:21:27 -!- LKoen has quit (Remote host closed the connection).
18:21:52 -!- LKoen has joined.
18:43:14 -!- kritixilithos has quit (Quit: quit).
19:18:54 <int-e> tromp: I also talked about that one previously
19:21:36 -!- kspalaiologos has quit (Quit: Leaving).
19:24:57 <int-e> tromp: they tend to post these challenges a bit early (the March one was posted February 27th, and I discovered it on the 29th).
19:26:03 <int-e> They also tend to close them a bit late.
19:26:36 -!- LKoen has quit (Remote host closed the connection).
19:28:52 -!- imode has joined.
19:29:45 -!- LKoen has joined.
19:41:00 -!- Lord_of_Life_ has joined.
19:41:44 <tromp> i like that this one can be solved manually
19:44:23 -!- Lord_of_Life has quit (Ping timeout: 260 seconds).
19:44:27 -!- Lord_of_Life_ has changed nick to Lord_of_Life.
19:46:52 <int-e> I guess it's not inconceivable.
19:47:44 <int-e> (I didn't really try.)
20:43:43 -!- xelxebar has quit (Ping timeout: 240 seconds).
20:46:08 -!- xelxebar has joined.
22:00:40 -!- rain1 has quit (Quit: Lost terminal).
22:00:43 -!- LKoen has quit (Quit: “It’s only logical. First you learn to talk, then you learn to think. Too bad it’s not the other way round.”).
23:17:31 -!- FortyTwo has joined.
23:17:44 -!- FortyTwo has quit (Remote host closed the connection).
23:18:43 -!- Thousand-YearSto has joined.
23:18:57 -!- Thousand-YearSto has quit (Remote host closed the connection).
23:19:30 -!- TYS61 has joined.
23:19:34 -!- TYS61 has quit (Remote host closed the connection).
23:20:36 -!- MTGBusyBeaver has joined.
23:27:39 <MTGBusyBeaver> Hello, I have been working on doing the most possible damage in Magic the Gathering, without going infinite, and while we have gotten some big numbers, we would like to implement some sort of turing complete set of (forced) operations as the BusyBeaver function would grow even faster.
23:29:44 <MTGBusyBeaver> MTG has already been proved Turing complete, however all of the known setups actually allow for the creation of nondeterministic Turing machines, which could loop for an arbitrary amount of time before halting.
23:30:14 <longname> I wonder if YGO is turing complete.
23:31:30 <MTGBusyBeaver> We have been trying to get https://esolangs.org/wiki/The_Waterfall_Model to work but have run into a few problems
23:32:59 <longname> You do have an infinite hand and there's quite a lot of non-HOPT self bouncing effects
23:33:26 <b_jonas> MTGBusyBeaver: yes, ais523 was trying to get a simulation of that to work, and I think he did get it to work eventually, just didn't make a clear enough documentation of the details of that construction
23:35:04 <MTGBusyBeaver> There is a working construction, it just also allows for nondeterminism as nothing forces the encoded waterfall program to never have two registers zero simultaniously
23:36:51 <b_jonas> MTGBusyBeaver: sure, but you can choose the program
23:37:03 <b_jonas> so just choose a waterfall model program that never has two register zero simultaneously
23:37:19 <b_jonas> you can translate any program to twm that way
23:37:19 <int-e> b_jonas: that's not useful for the busy beaver function though
23:38:01 <b_jonas> take a busy beaver turing machine, translate it to TWM
23:38:03 <MTGBusyBeaver> yeah we can't allow for even the opportunity for nonderterminism
23:38:22 <b_jonas> translation is rather inefficient because TWM is like that, but it works
23:38:23 <int-e> b_jonas: so you're saying, take a TC language, translate it to TWM in a special way, and then do BB on top of that?
23:38:47 <b_jonas> and then translate the TWM program to an M:tG game plan
23:39:21 <int-e> Makes sense, I just missed the first indirection.
23:39:30 <MTGBusyBeaver> But the MTG game plan with the same cards could make a nondeterministic TWM program and beat the BB bound
23:39:59 <b_jonas> MTGBusyBeaver: the game with the same cards could also just not enter any TWM loop, you might just never play the right combos
23:40:07 <b_jonas> setting up the combo is nontrivial, it's easy to mess it up actually
23:40:09 <int-e> b_jonas: I imagined you'd filter TWM programs for those that are "good" in the sense that they never have two zero registers as they run.
23:40:33 <b_jonas> int-e: you can't really "filter" them after the fact, that's uncomputable and at least as halt as telling if a turing machine halts
23:40:47 <int-e> b_jonas: hence my initial response (and objection)
23:41:31 <MTGBusyBeaver> but we get to pick our draws and plays, we wont "mess up" the combo, and if there is a line that does more damage we will take it
23:42:07 <MTGBusyBeaver> and if that amount of damage is unbounded, the deck is disqualified.
23:42:19 <b_jonas> MTGBusyBeaver: sure, but nothing in the deck forces you to do that specific combo. if you can go off with the deck, you can just win instantly rather than try to set up a combo
23:43:12 <b_jonas> MTGBusyBeaver: but then the TWM indeterminacy doesn't matter either, you could also set up a TWM program that deals infinite damage, or just deal infinite damage without going through the hoops
23:44:05 <b_jonas> longname: infinite hand like https://www.xkcd.com/1099/ ?
23:44:18 <MTGBusyBeaver> The idea is that the busy beaver machine would incidentially output tokens that could only attack after the TM halts
23:46:18 <b_jonas> https://www.xkcd.com/2275/ a Giant Spider? aaargh
23:47:23 <int-e> b_jonas: Have I complained already that covid-19 is a totally awful name? "corona virus disease 2019".
23:47:40 <b_jonas> int-e: yes, it should have a four-digit year name in there
23:48:03 <int-e> But the virus name got some actual symptoms, SARS-CoV-2.
23:48:37 <longname> You can only have 60 cards in your deck, max
23:48:44 <int-e> b_jonas: No, my complaint is about descriptiveness.
23:48:51 <longname> So I guess you could only have 60 cards in hand total.
23:49:27 <longname> Unless you took cards from your opponents hands too actually, so total 120
23:49:51 <int-e> "Standard decks must contain a minimum of sixty cards. There is no maximum deck size; however, one must be able to shuffle one's deck without assistance."
23:50:22 <longname> I'm talkin' YGO, Beaver's talking MTG
23:52:04 <b_jonas> in M:tG, you can momentarily put tokens into your deck or your hand, though they disappear the next time state-based actions are processed, and in some crazy cases they actually matter in your deck and make it hard to figure out how to follow the rules
23:52:25 <int-e> b_jonas: "is just too catchy" -- is that a play on its virility?
23:53:11 <longname> Ah, hm. Then yeah I guess maybe YGO is impossible to make turing incomplete without some change in rules just 'cus of a lack of memory.
23:53:25 <longname> *impossible to make turing complete.
23:54:08 <b_jonas> longname: really? that's surprising
23:56:08 <longname> You can summon shitloads of tokens (actually infinitely in theory with some mayakura the destructor + utlimate offering + phantom skyblaster + gy looping shenanigans) but they can only appear on the field
23:57:24 <MTGBusyBeaver> The useful operations we have: 2xArcbond (when this creature takes damage it deals that much damage to each other creature) allow us to set up a 'clock' that locks us out of interaction once the TM has been put into motion until one of the targeted creatures dies, artificial evolution lets us change the creature types a card refers to, rotlung
23:57:24 <MTGBusyBeaver> reanimator (whenever a creature of type X dies create a 2/2 Y token) allows us to remake/change things Dralnu's Crusade (all X's are also Ys) lets us have things with multiple types.
23:58:34 <b_jonas> I mean, if it were impossible to make it Turing-complete because it lacks crazy building blocks like (Rotlung Reanimator or Xathrid Necromancer or Hungry Lynx) and Artificial Evolution
23:58:55 <MTGBusyBeaver> I'm thinking about trying some sort of cyclic tag system using Coat of arms (each creature gets +x+x for each creature that shares a type with it) but I dont think that is enough
23:58:55 <b_jonas> that wouldn't surprise me. but if you can actually prove that YGO is not TC because of lack of memory, that _would_ surprise me.
23:59:36 <b_jonas> it wouldn't surprise me for a newer game like Hearthstone that may be specifically designed to avoid such crazy combos, but Yu Gi Oh is old