00:00:47 <wob_jonas> this is so hard to control it's no wonder so many non-black people oppose necromancy as being against the natural order
00:01:35 <wob_jonas> I have to go to bed now, but I should get back to this with a fresher head later to try to understand what is really happening here
00:04:50 -!- wob_jonas has quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client).
00:13:31 <ais523> for the logreaders out there: Wipe Away is a standard card in Omnitell sideboards, and a) has Split Second, b) lets us get rid of everyone's lands and our own Omniscience (thus preventing us interfering with the combo)
00:13:42 <ais523> so that's another problem solved without interfering with the original deck
00:32:05 -!- 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.”).
00:39:29 <zzo38> Now the "mbff" program in the Farbfeld Utilities can also read classes and levels too.
00:40:21 -!- Phantom_Hoover has quit (Remote host closed the connection).
00:41:09 -!- ais523 has quit (Ping timeout: 256 seconds).
00:42:37 <zzo38> (Although it will not execute any class codes, meaning that it renders as displayed in the editor rather than at runtime, so auto-reversing Rotators won't be marked, Continuous Walls will not be continuous, etc)
00:51:31 -!- sleffy has quit (Ping timeout: 252 seconds).
00:54:48 -!- tromp has quit (Remote host closed the connection).
01:01:25 -!- xkapastel has joined.
01:24:15 -!- trout has quit (Quit: Found 1 in /dev/zero).
01:27:59 -!- oerjan has joined.
01:36:08 <oerjan> HackEgo is a bit slow.
01:36:38 <oerjan> might end up timing out :(
01:36:39 <HackEgo> bin/wälcåmä:exec welcome "$@" | bin/en2sv
01:36:55 <HackEgo> ls: cannot access bin/en2sv: No such file or directory
01:37:46 <oerjan> makes no sense to delete one but not the other.
01:43:13 <moony> HackEgo, the slowest bot in mexico
01:45:28 <oerjan> well that _is_ official.
01:45:33 <HackEgo> HackEgo, also known as HackBot, is a bot that runs arbitrary commands on Unix. See `help for info on using it. You should totally try to hax0r it! Make sure you imagine it's running as root with no sandboxing. HackEgo is the slowest bot in all Mexico!
01:48:51 -!- Cale_ has joined.
02:06:46 -!- sprocklem has joined.
02:09:39 -!- tromp has joined.
02:14:02 -!- tromp has quit (Ping timeout: 265 seconds).
02:29:34 <oerjan> . o O ( ITC MtG discussions can be on topic... )
02:34:38 <oerjan> `mkx bin/,2//nur "$@" |& sport 2
02:35:05 <oerjan> `card-by-name Brass Herald
02:35:06 <HackEgo> Brass Herald \ 6 \ Artifact Creature -- Golem \ 2/2 \ As Brass Herald enters the battlefield, choose a creature type. \ When Brass Herald enters the battlefield, reveal the top four cards of your library. Put all creature cards of the chosen type revealed this way into your hand and the rest on the bottom of your library in any order. \ Creatures o
02:35:15 <oerjan> `,2 card-by-name Brass Herald
02:35:17 <HackEgo> 2/2:es of the chosen type get +1/+1. \ AP-U, 8ED-R \
02:36:04 <HackEgo> `2 <cmd> is equivalent to `1 <cmd>, except that it starts displaying the _second_ output piece. Useful when you've already run a command forgetting to use `1.
02:42:33 -!- sprocklem has quit (Ping timeout: 265 seconds).
03:02:49 -!- tromp has joined.
03:07:40 -!- tromp has quit (Ping timeout: 256 seconds).
03:10:03 -!- variable has joined.
03:52:52 -!- Hoolootwo has changed nick to Hooloovo0.
04:13:31 -!- xkapastel has quit (Ping timeout: 245 seconds).
04:15:15 -!- xkapastel has joined.
04:28:01 -!- sprocklem has joined.
04:37:48 -!- sleffy has joined.
04:49:45 -!- tromp has joined.
04:54:27 -!- tromp has quit (Ping timeout: 240 seconds).
04:59:01 -!- sleffy has quit (Ping timeout: 252 seconds).
04:59:50 -!- atslash has joined.
05:04:05 -!- atslash has quit (Ping timeout: 240 seconds).
05:04:27 -!- atslash has joined.
05:31:06 -!- xkapastel has quit (Quit: Connection closed for inactivity).
05:55:36 <zzo38> Did you see this? https://www.mail-archive.com/sqlite-users@mailinglists.sqlite.org/msg108914.html
06:13:38 -!- sleffy has joined.
06:16:25 -!- tromp has joined.
06:21:14 -!- tromp has quit (Ping timeout: 276 seconds).
06:34:26 -!- atslash has quit (Quit: This computer has gone to sleep).
06:45:44 -!- atslash has joined.
07:10:35 -!- tromp has joined.
07:13:23 -!- variable has quit (Quit: /dev/null is full).
07:14:35 -!- tromp has quit (Ping timeout: 240 seconds).
07:24:32 -!- augur has quit (Remote host closed the connection).
07:54:41 -!- tromp has joined.
07:59:28 -!- tromp has quit (Ping timeout: 268 seconds).
08:00:45 -!- variable has joined.
08:08:47 <\oren\_> i wonder how the norms for the coin-bill transition get decided
08:08:52 <\oren\_> and yen has the highest coin at 500 which is 6 canadian dollars
08:08:57 <\oren\_> meanwhile china has the biggest coin as 1 yuan or 20 cents
08:09:39 <\oren\_> and canada is in the middle with biggest coin at 2 dollars
08:09:40 -!- trout has joined.
08:10:44 -!- oerjan has quit (Quit: Nite).
08:13:39 -!- variable has quit (Ping timeout: 268 seconds).
08:32:26 -!- tromp has joined.
08:33:19 -!- moei has quit (Read error: Connection reset by peer).
08:33:51 -!- moei has joined.
08:53:03 <esowiki> [[Countdown]] https://esolangs.org/w/index.php?diff=54267&oldid=53151 * Singingbanana * (+22)
09:04:28 -!- CADD has joined.
09:12:46 -!- CADD has changed nick to AisRauli.
09:23:11 -!- atslash has quit (Quit: This computer has gone to sleep).
09:31:56 -!- ais523 has joined.
09:48:54 <esowiki> [[Wierd Machine]] https://esolangs.org/w/index.php?diff=54268&oldid=54266 * Singingbanana * (+904)
09:49:29 <esowiki> [[Wierd Machine]] https://esolangs.org/w/index.php?diff=54269&oldid=54268 * Singingbanana * (+13)
09:50:57 <esowiki> [[Wierd Machine]] https://esolangs.org/w/index.php?diff=54270&oldid=54269 * Singingbanana * (+0)
09:56:11 <esowiki> [[Truth-machine]] https://esolangs.org/w/index.php?diff=54271&oldid=53221 * Singingbanana * (+49)
09:57:45 <esowiki> [[Truth-machine]] https://esolangs.org/w/index.php?diff=54272&oldid=54271 * Singingbanana * (+3)
10:00:48 -!- wob_jonas has joined.
10:02:31 <wob_jonas> ais523: on the esowiki, what's the policy for writing about obfuscation (or golfing) techniques and notable obfuscated programs in a non-esoteric language, like perl or C or haskell?
10:02:55 <wob_jonas> I'm asking because there are now four or five attempts to prove M:tG turing complete, and we have to collect at least the links somewhere.
10:03:22 <ais523> I don't think we have one yet
10:03:30 <ais523> it may just be worth making a page about M:tG
10:03:40 <ais523> it's not intended as a language, but when seen as a language it's definitely esoteric
10:05:28 <ais523> we can definitely store notable programs on the wiki (assuming appropriate licensing), although the Esoteric Files Archive is meant to be a better place for that
10:06:09 -!- sleffy has quit (Ping timeout: 260 seconds).
10:07:49 <zzo38> Yes, you can try to write about Turing-completeness and other stuff like that about Magic: the Gathering.
10:08:27 <ais523> I think the correct approach is something like this
10:08:51 <ais523> a) move the StackFlow implementation to the talk page or page history, as it's not actually correct within the rules of the game (it requires players to consistently stack triggers in a particular order)
10:09:06 <ais523> b) create a page that treats M:tG as an esolang, and focuses on its programming aspects
10:09:53 <zzo38> I think the StackFlow implementation could be moved to a subpage, and put the complaints about it also on that subpage and/or the talk page of that subpage.
10:10:06 <ais523> c) produce a concrete TWM program that can emulate any given Turing Machine by changing one of the inputs (i.e. proof of compilability)
10:10:40 <ais523> d) produce an article, hosted on an external website, that goes through the entire construction (I could use mine for this, but it might be fun to get it onto a better-known M:tG website)
10:12:59 <ais523> bonus points for actually doing this in a serious tournament and proving the whole thing to be broken
10:13:24 <zzo38> I doubt WotC cares to put it on their own website, but you can try, I suppose.
10:13:36 <ais523> I meant a third-party M:tG articles websitee
10:13:43 <ais523> I'm sure there are some that accept submissions
10:16:18 <ais523> anyway, I think I figured out the kill; assuming we're using Wild Cantor to get split-second protection for the whole thing, we can also use it to generate infinte mana by looping it, which lets us activate abilities
10:16:33 <ais523> so we can put a "you lose the game" ability (e.g. Door to Nothingness) on the stack /underneath/ the entire combo
10:17:40 <ais523> then we have to arrange for the loop to break when the halt state is reached, we do that by making the Rotlung Reanimator that handles the steady decrease insufficiently protected, so it acts like a waterclock rather than ROM
10:19:04 <ais523> once it hits zero the whole thing unravels, and if the opponent doesn't have a stifle effect they just lose (note that we can ensure they have no mana and no mana-generating permanents, so they'd need to do something like two spirit guides → manamorphose → stifle)
10:19:48 <ais523> this also lets me make the article better as I can add in a reference to There's the Door, a deck which is based around giving the opponent Door to Nothingness and repeating the same gamestate until they use it on themself
10:20:13 <ais523> in this case, we only give them the chance to Door themselves once, if they refuse we go into the main combo (and because they've refused to Door themselves they're hopefully less likely to concede?)
10:21:27 -!- lynn has quit (Read error: Connection reset by peer).
10:21:44 -!- lynn has joined.
10:21:48 <wob_jonas> ais523: I would also like to link to the two or three previous proofs, the ones that assume players making certain simple decisions all the time
10:22:04 <wob_jonas> those are important for history even if you make a better proof
10:23:21 <zzo38> Yes, add those in a section for links to previous proofs
10:23:49 <ais523> wob_jonas: they're all in the same place on Alex Churchill's website, so you'd just link that
10:24:06 <ais523> you'd link it anyway as he came up with many of the techniques used
10:24:39 <wob_jonas> ais523: (on the kill) what the heck? why'd you need that? can't you just have a Platinum Angel or Abyssal Gatekeeper in play, make sure one of the creature-type specific toughness effects affect it as if it was one of the counters, and reduce everyone's life to zero in advance?
10:25:02 <ais523> wob_jonas: the Door to Nothingness kill uses only one card above what we have already
10:25:21 <ais523> I thought of the Platinum Angel kill but it needs two extra cards, the Angel itself and whatever you're using to reduce life totals
10:25:52 <ais523> unless you're reducing life totals by attacking; you can get infinite extra turns with Emrakul but I don't see how you prevent the opponent simply blocking with the Angel
10:26:07 <zzo38> There is other stuff about Magic: the Gathering on esolang wiki, such as [[Magic The Gathering card deck of programming language]], [[StackFlow]], and [[Talk:Undefined behavior]].
10:26:25 <ais523> zzo38: I think we'd move the StackFlow stuff
10:26:36 <ais523> I'd rather leave the "card deck of programming language" alone because it's technically a different language
10:26:42 <ais523> maybe put it in a see also
10:27:25 <ais523> the Talk:UB thing is a reference, not actually immediately related
10:28:17 <zzo38> Yes, leave those things alone because it is different, yes (and maybe put it in a see also). StackFlow stuff can be moved. And yes I know the Talk:UB thing is not actually related. I am only mentioning where Magic: the Gathering is mentioned on the wiki; making the new article about Magic: the Gathering is of course different to these things.
10:29:01 <zzo38> (The stuff on [[Talk:Undefined behavior]] under "Magic: the Gathering" is just a joke anyways, not a real thing.)
10:31:10 <ais523> b_jonas: by the way, did we ever figure out if M:tG was /only/ TC or whether it could be computable?
10:31:38 <ais523> the aim would be to use a subgame card to differentiate a draw from a loss without ending the game, thus solving the halting problem
10:31:40 <wob_jonas> ais523: for the end condition, do you want the program to be able to cause a win or a lose, or are you satisfied with a construction that offers an infinity draw vs a win, or one that offers an infinity draw vs a lose?
10:32:04 <ais523> wob_jonas: I'm personally OK with draw vs. win because that has an observable effect on tournament results
10:32:15 <ais523> although win vs. loss would work even better as that means the program always halts
10:32:34 <zzo38> It depend if you want to be Vintage or "pseudo-Vintage" or something more.
10:32:42 <ais523> unfortunately, Shahrazad was banned in Legacy and I'm not aware of any other cards that are capable of observing infinite-loop draws
10:32:46 <wob_jonas> ais523: it's definitely uncomputable, but only because of the infinite loops rule, and even then only if someone managed to formalize that rule precisely enough to know what it says for all situations, and I don't think anyone wants to do that. if you don't have the infinite loop rule, then M:tG with just the comprehensive rules and no tournament r
10:32:46 <wob_jonas> ules or silver border is computable.
10:33:04 <ais523> wob_jonas: right, the whole point of this is to prove that the infinite loops rule is broken
10:33:38 <ais523> do it in a tournament and suddenly the judges need to solve the Goldbach Conjecture in order to work out what the result of the tournament is
10:34:03 <ais523> ah right, IIRC we decided it's obviously noncomputable, given a working definition
10:34:17 <ais523> the hard part is to determine whether it's /paradoxical/, i.e. if we can set up a game where a player wins if and only if they don't win
10:34:21 <ais523> but I believe the answer to that was no
10:34:46 <zzo38> Not just Legacy; Shahrazad is banned in Vintage, which is what is important.
10:35:52 <zzo38> If such a thing with Goldbach Conjecture occurs, and then, the player says, I have a proof of Goldbach Conjecture in my pocket I will show you...
10:36:29 <wob_jonas> I think if you just want draw vs win or draw vs lose, then once you figure out the exact construction, it will become trivial enough to modify it with at most one extra card (say Impetous Sunchaser) that the opponent must kill you in his turn without a choice. This relies on the exit from the main loop being clean and finite though, so I could be w
10:36:52 -!- AnotherTest has joined.
10:37:22 <ais523> wob_jonas: well if the main loop exits, one player dies, that's trivial to do
10:37:34 <ais523> if the main loop /doesn't/ exit, then it's a draw by definition
10:37:39 <wob_jonas> wait, why is Vintage important, as opposed to legacy?
10:37:47 <ais523> vintage is the most permissive format
10:37:55 <ais523> so if a card's banned in vintage it's banned everywhere
10:38:08 <ais523> I was planning to aim the combo at Legacy, though, because that's the format that Omnitell is designed to do well in
10:38:44 -!- Hooloovo0 has quit (Ping timeout: 256 seconds).
10:38:56 -!- Hoolootwo has joined.
10:39:23 <wob_jonas> exactly. for Vintage, you need to buy two dozen very expensive card to even stand a chance against serious opponents, and we don't want that in an elegant construction like what this should be unless it's absolutely necessary
10:41:34 <wob_jonas> I don't think Vintage will matter, unless you find that there's some card valid in Vintage but banned in Legacy and Two-headed Giant that really really helps make the setup work
10:45:41 <ais523> here's an example of an Omnitell deck that already has several cards we need maindeck, and the ability to get the entire combo out of the sideboard without changing any maindeck cards: http://sales.starcitygames.com//deckdatabase/displaydeck.php?DeckID=113604
10:47:51 <ais523> (I picked a recent-ish version in the hope that the deck is still competitively viable)
10:51:09 <alercah> ais523: what's the goal here?
10:51:43 <ais523> alercah: the original goal is to create a Magic: the Gathering gamestate for which an arbitrary computation is required to determine the result of the game
10:52:05 <ais523> i.e. as soon as the gamestate is set up, it's clear that the game is over either as an infinite loop or via the game-over state-based effect
10:52:32 <ais523> the slightly enhanced version of this is to do it in a deck that's competitive enough to win a game in a tournament, opposite a noncooperative opponent (assuming they don't just succeed)
10:52:32 <wob_jonas> as for obfuscations in other notable programs, so far I've mostly just made writeups for some of them that interpret or compile some esoteric language
10:52:46 <ais523> obviously it doesn't have to be reliable at this, just able to do it once
10:52:47 <alercah> oh man, toby would have a field day
10:52:50 <wob_jonas> and there's of course a lot of such programs
10:53:25 <ais523> the current plan is to fit the whole thing into the sideboard of Omni-Tell (a Legacy deck that isn't doing so great at the moment but definitely used to be competitively viable, it's probably Tier 2/3 atm which is enough to win a game or two in a tournament)
10:53:36 <alercah> I suspect in practice the game would be declared a draw
10:54:35 <ais523> we also want to write an entertaining article about it so that people will read it
10:55:09 <ais523> unfortunately, AFAICT the tournament rules require you to memorise the program you're implementing, so we'll need that to be something that a human can reasonably memorise, too
10:55:14 <wob_jonas> alercah: no, in practice you'd be disqualified and banned from magic forever and forcibly ejected from the building for even attempting to pull it off, all of which is supported by the tournament rules.
10:55:29 <ais523> wob_jonas: which rule's being broken? slow play?
10:55:58 <ais523> I don't think it is, the slow play rule says you can't continue a loop without being able to state the expecting resulted gamestate
10:56:04 <wob_jonas> I dunno, I'm not up to my knowledge
10:56:06 <ais523> in this case, we /do/ know the expected gamestate just before the loop ensd
10:56:20 <zzo38> Disqualified makes sense I suppose, but that does not mean to be banned forever, and even if so, you should not be forcibly ejected from the building.
10:56:35 <wob_jonas> yes, you'd probably only be ejected from the tournament site, not the whole building
10:56:39 <ais523> then there's no rule against performing a small finite number of actions (activating Door to Nothingness, casting Wipe Away targeting Door to Nothingness, sacrificing Wild Cantor)
10:57:05 <ais523> and at this point neither player has any control over where the game goes from there because there's a split second spell on the stack and a self-sustaining cycle of mandatory triggered abilities
10:57:14 <ais523> so you call a judge and explain the situation
10:57:23 <alercah> if the outcome were resonably determinable and you could give an ELI5 explanation of why it works, then you'd probably get the win
10:57:27 <alercah> otherwise I'd rule it a draw
10:57:47 <ais523> incidentally, once you've demonstrated the loop once you can just shortcut to the eventual resulting gamestate, which is very helpful
10:57:56 <alercah> you'd especially get a draw if your computation was something to which you didn't know the actual answer
10:58:02 <wob_jonas> ais523: slow play is applicable, because donating a thousand lynxes isn't a "small number of finite actions", especially since you could have won already, but I think there might be other tournament rules violations
10:58:19 <ais523> wob_jonas: that can legally be shortcut
10:58:26 <ais523> it's against the rules to donate them one at a time, but once you've demonstrated one iteration of a loop
10:58:30 <alercah> if you know the exact end state, yes
10:58:35 <ais523> it's perfectly legal to say "I do that 999 more times"
10:58:39 <ais523> and yes, we know the exact endstate here
10:58:52 <ais523> it's a fairly /complex/ exact endstate but it's precise and doesn't involve any randomness or decisions
10:59:00 <wob_jonas> ok wait, let me actually look up the rules
10:59:02 -!- Phantom_Hoover has joined.
10:59:12 <wob_jonas> https://wpn.wizards.com/en/resources/rules-documents
10:59:28 <ais523> err, decisions by the opponetn
11:00:20 <ais523> and you prevent the combo triggering while you're trying to set it up by creating all the tokens first with their original creature types, and then looping Artificial Evolution (thus creating the connections between them) as the last thing you do
11:00:41 <zzo38> I think the time needed to make the proof should also counts against slow play if necessary; if you have proven you won in time, you do, otherwise you are disqualified. But you should not be ejected from the tournament site if spectators are allowed, since this isn't the kind of disturbing thing that you can affect if you aren't part of the game.
11:01:02 <ais523> well, if you can't prove you've won, it's just a draw
11:01:09 <ais523> the issue is that we're creating a loop and nobody knows whether it's infinite or not
11:01:19 <ais523> infinite loop = draw, finite = win
11:01:41 <wob_jonas> a relevant rule is in 5.2 saying "The result of a match or game may not be randomly or arbitrarily determined through any means other than the normal progress of the game in play. Examples include (but are not limited to) rolling a die, flipping a coin, arm wrestling, or playing any other game."
11:02:11 <ais523> wob_jonas: huh, so if the judges decide to just call the game a draw as they can't figure out what happens
11:02:16 <ais523> does that mean that the /judges/ are violating rule 5.2?
11:02:27 <ais523> (note that we aren't, everything that happens is entirely based on "normal" gameplay)
11:02:38 <zzo38> Some card effects might cause the game result to be determined by a coin toss.
11:02:54 <ais523> yes, it's perfectly legal to arrange a gamestate in which a coin toss determines the result of the game as a consequence of the game's normal play
11:02:58 <ais523> `card-by-name mana clash
11:02:59 <HackEgo> Mana Clash \ R \ Sorcery \ You and target opponent each flip a coin. Mana Clash deals 1 damage to each player whose coin comes up tails. Repeat this process until both players' coins come up heads on the same flip. \ DK-R, 4E-R, 5E-R, 7E-R, 8ED-R, 9ED-R
11:03:08 <ais523> otherwise that card wouldn't be legal at all
11:03:26 <ais523> just not to say "let's toss a coin, loser of the flip concedes" or the like
11:03:26 <zzo38> So too can combinations of card effects cause the game result to depend on the Goldbach Conjecture, or whatever.
11:05:56 <wob_jonas> ais523: the judges aren't going to decide that in this case. you're intentionally throwing away the match you could win and deliberately making yourself win or draw depending on some condition irrelevant to M:tG. it doesn't matter that other players in the tournament didn't pay you shiny dollars for it.
11:06:59 <wob_jonas> just because it's a legal game state isn't enough. throwing away your win by just not playing your spells and letting the opponent finish you off with a weak deck if also illegal.
11:08:40 <alercah> ais523: if you're creating a loop and you don't know the outcome, you can't shortcut it though
11:09:08 <alercah> e.g. if you were to use a search for a counterexample to, say, P = NP
11:10:45 <wob_jonas> Slow play will almost certainly also apply, because unless you use an accomplice as an opponent or judge, which is against tournament rules for other reasons, you'll have to do an hour long CS course to explain why your construction does exactly what you want, and not just the Magic interactions parts, but also how you emulate say a multiple-stack
11:10:45 <wob_jonas> machine with a Minsky machine with an Amneisac machine with a Waterfall machine with M:tG.
11:11:43 <wob_jonas> And by the way, you'd have to do the whole thing alone and without notes you've taken prior to the match.
11:12:13 <wob_jonas> By alone, I mean there'd be at most you, the opponent, and two judges to help you.
11:13:08 <wob_jonas> Or maybe a teammate each too if you manage to find a constructed two-headed giant tournament with a wide enough format to pull of the combo.
11:14:44 <zzo38> Yes, or other team formats (I don't know if there are official tournaments with other team formats though)
11:15:57 <zzo38> But, yes, that is why I say, you will be disqualified for slow play unless you have proven it soon enough by yourself during the tournament.
11:16:29 <wob_jonas> zzo38: I'm not sure, but I think legacy two-headed giant is technically DCI-supported, but only in the sense that Mirage block constructed is also DCI-supported, which means someone could hold a tournament in it and the results could be registered to DCI, but it's unlikely that anyone will every actually do it
11:17:12 <wob_jonas> zzo38: we might have to ask some M:tG judges for this, but I think even without the slow play, this would go against that bribery rule
11:17:13 <ais523> wob_jonas: as soon as the combo is set up, the game is legally over
11:17:16 <ais523> you can't slow play at that point
11:17:20 <ais523> (also, alercah /is/ an M:tG judge)
11:17:42 <wob_jonas> alercah is an M:tG judge? I didn't know that. is he active or recently been active?
11:18:19 <alercah> haven't been active for a year and a half
11:18:30 <wob_jonas> that's still recent as far as we're concerned I think
11:18:49 <wob_jonas> that the lynx is new is irrelevant
11:19:05 <zzo38> It does not seem to me that bribery rule will have to do with it, or maybe it can if opponent concedes due to your proof instead of the judge saying you have already won and they don't have to concede, but I don't know.
11:19:07 <alercah> ais523: I think the precedent would line up with what wob_jonas is saying: there is a point of mathematical complexity beyond which the play isn't permitted
11:19:29 <alercah> the precedent being infinite scry 2
11:19:29 <ais523> alercah: even though it entirely takes place with legal game actions?
11:19:45 <ais523> up until the point you sacrifice Wild Cantor, there is nothing hard to understand about the gamestate
11:20:15 <alercah> ais523: yes; there is a ruling disallowing using infinite scry 2 to arbitrarily reorder your library IIRC
11:20:17 <ais523> you're just creating a large number of tokens with specific characteristics, but that doesn't /do/ anything, you can just count the number of each type of token
11:20:26 <ais523> alercah: that involves hidden information though
11:20:30 <wob_jonas> ais523: you mean apart from the donated lynxes artificially evolved to probably at least dozens of different texts?
11:20:41 <wob_jonas> I'm not sure how much you can golf this really
11:20:45 <ais523> wob_jonas: yes, you're donating a lot of token lynxes and editing the creature types in them
11:21:00 <ais523> I don't see any problem with that, it's not conceptually different from creating 109280319840235 tokens
11:21:15 <alercah> ais523: it still allows you to set up a loop, define a number of iterations, and set a specific output state
11:21:18 <ais523> i.e. the precise number may have to be recorded, and that's a pain in case it's relevant, but it's not fundamentally different from picking any other definite number
11:21:21 <wob_jonas> I think you'll need either hundreds of *different* lynxes (ones with differently edited text) or an exponentially large power
11:21:24 <alercah> (namely, my library is in a specific order)
11:21:41 <alercah> the ruling was based more on "magic is not a game of mathematics" than the presence of hidden information
11:21:43 <ais523> alercah: the challenge there is the defined number of iterations
11:21:44 <wob_jonas> it's not the number of lynxes that's the problem, but that there's so many different ones, each of which have to be tracked separately
11:22:09 <ais523> wob_jonas: well you have two loops
11:22:24 <ais523> the first donates the opponent an arbitrarily large number of identical lynx tokens
11:22:39 -!- trout has quit (Quit: /dev/null is full).
11:22:49 <ais523> I don't think anyone would have a problem with that; it's a bit of a bizarre thing to do but it's not fundamentally different from any other combo
11:23:15 <ais523> the second loop casts Artificial Evolution twice at each of the tokens, which is clearly a defined operation; the only tricky part is that you don't always choose the same creature types
11:23:52 <ais523> IMO if you choose the same creature types every time that's clearly legal
11:23:56 <alercah> ah wait, the example I was thinking of was four horsemen, which has an indeterminate loop
11:24:04 <zzo38> Why should there be a ruling disallowing using infinite scry 2 to arbitrarily reorder your library? The only thing then to consider is slow play if it is take too long to reorder your library, which, if there are enough cards, it might do, otherwise it won't.
11:24:13 <ais523> alercah: yes, four horsemen is illegal as it has no bounded number of iterations
11:24:33 <ais523> I believe the infinite scry 2 is legal so long as you have memorised the order of your library and thus can give an exact iteration count for bubble-sorting it
11:24:43 <ais523> (this is why nobody does it in practice)
11:24:55 <ais523> infinite scry 2 = bubble sort, four horsemen = bogosort
11:25:13 <wob_jonas> zzo38: and slow play is already relevant: even without reordering the library, it's forbidden to take a note of all cards in the opponent's deck as you Cranial Extraction them for slow play, and reordering your library is on the sam eorder of magnitude
11:25:13 <ais523> one of these has a finite worst-case performance, the other an infinite worst-case performance
11:25:36 <alercah> ais523: I think if you pushed me on it
11:26:09 <zzo38> wob_jonas: Yes; unless the number of cards is low, it should count only slow play, is what I am saying. Rather than, just making the rules arbitrarily
11:26:10 <alercah> my ruling would be that "without being able to provide an exact number of iterations and the expected resulting game state" requires that you be able to provide those things in a reasonable time frame
11:26:19 <ais523> alercah: yes, I think that's valid
11:26:32 <ais523> so you the program needs to be golfed to the extent that we can explain what it is quickly
11:27:02 <alercah> if you're trying to explain how you have encoded a brainfuck interpreter in a minsky machine to the 10-year old across from you, it isn't going to go well
11:27:09 <ais523> exact iterations is easy; we can do too many without breaking the construction, so "3 trillion" is a simple valid number
11:27:13 <ais523> expected resulting gamestate is much harder
11:27:19 <wob_jonas> alercah: wait, isn't 11 years old the minimum age?
11:27:26 <ais523> as that gamestate encodes the program and so we need to golf the program to make it as simple as possible
11:27:36 <alercah> wob_jonas: there is no minimum AFAIK
11:27:37 <ais523> wob_jonas: there's a 7 year old who made the news for doing fairly well in an official tournament
11:27:52 <alercah> under 13 you need parental permission for privacy law reasons IIRC but otherwise no rules
11:28:05 <wob_jonas> I mean, I've seen very young people play casual magic, but this is for tournaments
11:28:14 <ais523> right, that's why it was newsworthy
11:29:00 <alercah> https://compete.kotaku.com/7-year-old-magic-prodigy-cant-shuffle-cards-yet-but-can-1820619884
11:29:10 <esowiki> [[StackFlow]] M https://esolangs.org/w/index.php?diff=54273&oldid=52350 * FireFly * (+2) /* Syntax */ line -> length (typo)
11:30:01 <zzo38> I thought it is mandatory to shuffle the cards by yourself?
11:30:14 <ais523> you can have someone to help you if you're disabled
11:30:26 <ais523> being too young to be able to hold the deck is probably a comparable case
11:30:51 <ais523> at a local Magic club, I used to help a blind person sometimes, telling them about what the content of their own hand was (as they couldn't see it)
11:31:37 <ais523> (I also played against them sometimes, announcing everything I was doing in detail because again, they couldn't see it)
11:32:23 <wob_jonas> I think most young tournament magic players can usually shuffle decks though
11:32:26 <alercah> yeah, the "shuffle yourself" rule is to prevent 1500-card decks that can't reasonably be shuffled by one person, not to stop people who are physically unable
11:32:37 <alercah> and to preclude the use of e.g. card shuffling machines
11:32:47 <ais523> right, it's a maximum deck size limit
11:33:00 <alercah> mainly thanks to Battle of Wits
11:33:01 <ais523> there's some debate about whether Battle of Wits decks are tournament-legal, because of this
11:33:18 <wob_jonas> it rarely comes up in practice, because most decks don't want to be larger than a Battle of Wits deck, and even Battle of Wits is usually only 240 cards or something
11:33:38 <ais523> incidentally, Battle of Wits is the most expensive Modern deck, because it turns out the optimal build is to fill the deck full of Modern staples with only a few tutors
11:33:39 <wob_jonas> you can shuffle 240 cards if you practice, and you can clearly practice in advance
11:34:46 <alercah> I just imagined a player after game 1 of a limited tournament realizing that they are basically hopeless against their opponent's deck
11:34:57 <alercah> so they side in 200 land and desperately hope they can draw their BoW
11:35:20 <wob_jonas> I've heared of one serious deck that's even bigger, using a large number of either Relentless Rats or Shadowborn Apostole (I forgot which)
11:35:45 <ais523> alercah: that's done with other cards in limited sometimes
11:35:50 <ais523> normally Lost in the Woods or Pack Rat
11:36:24 <ais523> (the Pack Rat deck is better, because as long as it finds the Rat, it can typically beat anything but a board sweeper, and those are rare in limited)
11:36:34 <alercah> yeah, I've seen that in action
11:37:00 <alercah> but pack rat is also very strong on its own
11:37:14 <alercah> it barely needs the 39-swamp deck to be obscenely strong
11:37:30 <zzo38> You might not practice in advance if you are playing Limited, but in such a case most of the cards will probably be conventional basic lands anyways, if you try to make large decks.
11:37:33 <ais523> it's so strong that casting it effectively wins the game, so you run 39 swamps alongside it to make sure you aren't mana screwed
11:37:35 <alercah> the only advantage it gains from the all-land deck is consistently hitting its drops
11:41:46 <ais523> it's particularly important to run a lot of lands because you're usually mulliganning until it's in your starting hand
11:41:52 <ais523> so you won't have as many lands drawn naturally as normal
11:42:23 <zzo38> Yawgatog still hasn't added a list of the differences of the rules between the previous and current version (even though they have done for older versions)
11:48:42 <wob_jonas> The strange thing is that an esolang like countercall might be relevant for a reduction to M:tG, because it naturally has a call stack, and you could have trigger an ability triggered as many times as some counter represented as the number of certain objects
11:49:25 <wob_jonas> probably not countercall exactly, but some language with a similar gimmick
11:49:26 <ais523> well countercall is sub-TC
11:49:45 <ais523> but yes, counter + call stack doesn't seem impossible
11:50:32 <alercah> ais523: let me know if you actually figure it out btw, I have some friends who would enjoy hearing about it
11:51:03 <ais523> alercah: well we already have a working construction for the The Waterfall Model → M:tG side of things
11:51:11 <ais523> it's just that TWM, while TC, is really hard to golf
11:51:23 <wob_jonas> alercah: I think ais523 more or less has the construction in his head now. but I'm really behind and will slowly try to understand it. first the part where M:tG is irrelevant, like how a reduction to Waterfall even works and how much it would blow up
11:51:35 <ais523> yes, I need to write it down
11:51:57 <ais523> the Waterfall side of things seems to be the fragile side
11:52:00 <wob_jonas> ais523: do you want to reduce something like the Amnesiac to Waterfall?
11:52:13 -!- ais523 has quit (Remote host closed the connection).
11:52:26 -!- ais523 has joined.
11:52:31 <ais523> [11:51] <ais523> yes, I need to write it down
11:52:32 <ais523> [11:51] <ais523> the Waterfall side of things seems to be the fragile side
11:52:34 <ais523> [11:52] <ais523> all the constructions I'm aware of scale from Minsky machines in a states + counters way
11:52:43 <ais523> I don't know what the minimum number of states + counters is for a universal Minsky machine program
11:52:58 <wob_jonas> we don't know small universal Minsky machines
11:53:17 <ais523> the Fractran self-interpreter is fairly small
11:54:00 <ais523> I wonder what I was thinking of
11:54:54 <wob_jonas> at least you can use state transitions that add or subtract any integer
11:55:00 <wob_jonas> you don't have to use 100 states to add 100
11:55:08 <ais523> ah right, there's a smaller one: https://stackoverflow.com/questions/1749905/code-golf-fractran/1802570#1802570
11:56:39 <wob_jonas> M:tG has 237 creature types right now
11:57:57 <ais523> we can't quite use all of them
11:58:17 <ais523> Cat, Rat, Eldrazi, Cleric, Zombie would have side effects halfway through the setup
11:58:25 <ais523> err, maybe not Rat actually
11:58:34 <ais523> but that still gives a lot of space
11:59:49 <wob_jonas> Nightmare and Horror might also have some effects
12:00:04 <ais523> no, we removed Faceless Devourer from the construction
12:00:11 <ais523> because we could get the same effect using cards that we already had
12:00:22 <ais523> err, almost, we need a static -2/-2 from somewhere
12:00:36 <ais523> `card-by-name Elesh Norn, Grand Cenobite
12:00:41 <HackEgo> Elesh Norn, Grand Cenobite \ 5WW \ Legendary Creature -- Praetor \ 4/7 \ Vigilance \ Other creatures you control get +2/+2. \ Creatures your opponents control get -2/-2. \ NPH-M, MM2-M
12:01:54 <ais523> come to think of it, Elesh herself could be the halt state, given that the oppoenet would control it
12:01:57 <ais523> that makes things even simpler
12:02:12 <ais523> because once she dies the "base" triggered ability (the one that triggers itself) will stop working
12:02:23 <ais523> and Praetor is a unique creature type we aren't using anywhere else
12:03:36 <ais523> hmm, at present this implements a TWM variant in which all waterclocks set themselves to 4 when exhausted (rather than an arbitrary number)
12:03:43 <ais523> I don't think that changes the TCness of the language though
12:04:09 <wob_jonas> there's also Night of Souls' Betrayal
12:04:18 <ais523> Elesh solves three problems at once though
12:04:27 <ais523> because we /also/ needed a static +/+ effect on the opponent's creatures
12:04:32 <ais523> and we also needed a way to halt
12:04:44 <ais523> (in addition to the static -/- that keeps the loop going)
12:04:53 <wob_jonas> for some reason Harmless Offering can donate enchantments. I really don't understand why Wizards chose to allow that in such a cheap sorcery, sounds like it will be broken.
12:05:04 <ais523> wob_jonas: it was intended as a build-around
12:05:09 -!- xkapastel has joined.
12:05:17 <ais523> and they wanted to make it wide so that as many decks as possible that needed an effect like that would have one
12:05:21 <wob_jonas> there's also, um, what's that iconic black hero called
12:06:11 <HackEgo> Ascendant Evincar \ 4BB \ Legendary Creature -- Vampire \ 3/3 \ Flying (This creature can't be blocked except by creatures with flying or reach.) \ Other black creatures get +1/+1. \ Nonblack creatures get -1/-1. \ NE-R, 10E-R, HOP-R
12:07:12 <ais523> right, -1/-1 on a legendary needs two of them to get -2/-2
12:07:32 <ais523> so the idea is that we use a zombie token for almost every piece of RAM
12:07:39 <ais523> but Elesh Norn for the remaining counter
12:07:52 <ais523> if she ever ends up dying, the whole combo breaks and the stack unwinds to the point at which we cast Wipe Away
12:08:01 <ais523> and we have a Door to Nothingness activation on the stack just below that
12:08:27 <ais523> or, I guess if we wanted to save a card, an attacking Emrakul (we could stifle the annihilator trigger)
12:09:11 <wob_jonas> yes, but just one zombie token for each counter, right? unlike the Alex Churchill construction, which uses a ramp of creatures of increasing toughness, as many total as the counter
12:09:47 <ais523> occasionally the tokens die and just get replaced with a new token
12:09:51 <wob_jonas> How does stifling save a card? you don't need a stifle effect otherwise, do you? or do you use one to start the loop?
12:10:01 <ais523> the deck naturally contains a copy of Trickbind
12:10:31 <ais523> `card-by-name Trickbind
12:10:33 <HackEgo> Trickbind \ 1U \ Instant \ Split second (As long as this spell is on the stack, players can't cast spells or activate abilities that aren't mana abilities.) \ Counter target activated or triggered ability. If a permanent's ability is countered this way, activated abilities of that permanent can't be activated this turn. (Mana abilities can't be tar
12:11:16 <ais523> that's helpful because post-combo, Trickbind + 4 copies of Force of Will mean that the opponent is unlikely to be able to interfere
12:12:22 <wob_jonas> if you want attacking, you have to pull the combo off before your combat. that's not an obstacle for this deck, right?
12:12:52 <ais523> nah, in fact the setup even benefits from passing the turn on occasion
12:13:15 <ais523> given that the opponent will have no permanents and we'll have an extra turn right after, I don't see much of a problem from that
12:13:31 <ais523> the whole thing happens at mana ability speed in the middle of combat
12:13:40 <ais523> wob_jonas: nothing /to/ untap
12:13:52 <ais523> and no, you don't untap if you don't get a turn
12:13:55 <ais523> untap phase is part of your turn
12:13:58 <wob_jonas> oh, you kill all his permanents too? what with?
12:14:05 <ais523> not kill, we bounce them with Wipe Away
12:15:01 <wob_jonas> that sounds a bit dangerous because it could trigger some leaves tb abilities, but I guess that can't be helped
12:15:19 <ais523> well, we can get Trickbind before doing that
12:15:46 <ais523> but yes, there's always a danger period immediately after we're in the start state (our entire deck in our hand, Omniscience or Dream Halls in play)
12:15:52 <wob_jonas> and it's definitely safer than leaving arbitrary permanents in play. the alternative is exiling all permanents
12:15:54 <ais523> but before we've fully neutralised the opponent
12:16:11 <wob_jonas> oh right, you exile them from his hand later
12:16:28 <ais523> not in the current version, Extraction turned out to be unneeded
12:16:34 <ais523> we just let them exist in hand and counter anything they try to play
12:16:42 -!- LKoen has joined.
12:16:55 <ais523> there's not much that can be played with no mana and no permanents, after all
12:17:14 <ais523> and we have four hard counters in hand
12:17:29 <ais523> it's not 100% secure, but it becomes more so over time
12:19:20 <ais523> (you need to pass the turn to remove the summoning sickness from the copy of Emrakul that stays on the battlefield to legend-rule the one we play as part of the loop)
12:20:24 <wob_jonas> but doesn't being able to pass the turn cost one or more cards? if you just need haste, you can give it with one card, especially since you already have Evolution
12:21:20 <ais523> Emrakul gives you an extra turn when you hardcast it
12:21:33 <ais523> so passing the turn hardly does anything
12:22:13 <wob_jonas> though it ends some effects, probably none of them are crucial
12:22:25 <ais523> I don't think we're using any of them during the setup phase
12:22:42 <wob_jonas> you're not trying to keep anything alive with a growth despite the continuous -2/-2, right?
12:23:33 <ais523> there are some things we have to keep alive but we can use Hungry Lynx to put +1/+1 counters on them
12:23:37 <ais523> during the setup, that is
12:23:45 <ais523> so it's another use of using the same card for two purposes
12:25:31 <wob_jonas> and all that is after you set up the infinite card recycling combo, right?
12:25:57 <ais523> the infinite combo exists "naturally" in Omnitell
12:26:01 <ais523> I'm not sure if most people noticed, though
12:26:10 <ais523> because they don't expect a Regrowth effect to cost 27 mana
12:26:30 <ais523> (with Omniscience out, though, the cost gets reduced to 0 so we can do it easily)
12:26:47 <ais523> in general the deck doesn't /need/ a 27 mana regrowth so people weren't looking for one
12:33:55 <wob_jonas> the difficulty is not just the 27 mana, but coming out positively in card balance in them. and I'm not quite sure green would consider that natural, because green has Recollect/Reclaim effects which get one card back and cost one card, plus Repopulate which is an instant that returns creatures only.
12:34:47 <wob_jonas> it's one thing to live longer by swapping some Conjurer's Baubles, but much more difficult to make it infinite and returning arbitrary sorceries
12:35:33 <wob_jonas> the Beacons and the Elixir can do it, but those aren't good for us, firstly because you need two, secondly because then you need a way to get rid of cards you don't need
12:37:34 <wob_jonas> By the way, once you set up the infinite card recycling and the Omniscience, Manamorphose would give you infinite mana in any color if you needed that; but you probably don't need it, because using activated abilities can cause difficulties with taking the choice away later.
12:39:23 <wob_jonas> I'm still laughing at the Harmless Offering. I should addquote that.
12:40:17 <HackEgo> Quotes are just elements of the quantum dilapidated bogosphere. See qdb.
12:40:26 <HackEgo> quote formatting? ¯\(°_o)/¯
12:40:50 <wob_jonas> #`addquote <ais523> oh, we also need a Donate effect <ais523> Harmless Offering is the obvious choice given that we'll mostly be donating Hungry Lynxes
12:41:27 <HackEgo> quoteformat is: <nick> message; * nick action; two spaces between messages; all elisions marked with [...] other than irrelevant intervening messages; for messages separated by elision, one space on each side, not two.
12:42:22 <wob_jonas> `addquote <ais523> oh, we also need a Donate effect <ais523> Harmless Offering is the obvious choice given that we'll mostly be donating Hungry Lynxes
12:42:31 <HackEgo> 1322) <ais523> oh, we also need a Donate effect <ais523> Harmless Offering is the obvious choice given that we'll mostly be donating Hungry Lynxes
12:45:09 <ais523> wob_jonas: for the mana ability speed version of this, we can do that using Wild Cantor
12:45:20 <ais523> for the instant speed version, we don't need Wild Cantor at all
12:45:27 <ais523> (the trigger would be donating Elesh Norn)
12:45:38 <ais523> `card-by-name Harmless Offering
12:45:39 <HackEgo> Harmless Offering \ 2R \ Sorcery \ Target opponent gains control of target permanent you control. \ EMN-R
12:45:47 <ais523> sorcery speed version then :-)
12:46:01 <ais523> I'm not sure we can easily trigger this without a Wild Cantor
12:46:11 <ais523> which would be a good reason to leave it in
12:46:29 <wob_jonas> meh, just throw in an Orrery, you'll find you need it anyway sooner or later :-)
12:46:31 <ais523> and doing it at split second speed is flashier (and harder to interfere with)
12:46:43 <ais523> wob_jonas: some versions of the deck actually play a Quicken maindeck
12:46:56 <ais523> although not my reference list
12:47:01 <wob_jonas> there are other mana ability speed things comparable to Wild Cantor though
12:47:56 <wob_jonas> Phyrexian Tower is the most famous one
12:48:51 -!- 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.”).
12:49:42 <ais523> that doesn't really fit very well in the deck, though, as it doesn't naturally play creatures
12:50:56 <wob_jonas> there's no one card that lets you repeatedly draw and discard at instant speed, right? Jayemdae Tome and that colored but cheaper version they printed later allows drawing at instant speed, but instant speed discard is rare
12:51:25 <wob_jonas> well, there's all sorts of looting
12:52:26 -!- AisRauli has quit (Quit: Connection closed for inactivity).
12:52:39 <ais523> what do you want that for?
12:52:47 <wob_jonas> eww, my searches are getting dirty by all these Portal cards
12:53:18 <ais523> there are quite a few creatures with abilities that cost "discard a card"
12:53:54 <ais523> `card-by-name Mesmeric Trance
12:53:55 <HackEgo> Mesmeric Trance \ 1UU \ Enchantment \ Cumulative upkeep {1} (At the beginning of your upkeep, put an age counter on this permanent, then sacrifice it unless you pay its upkeep cost for each age counter on it.) \ {U}, Discard a card: Draw a card. \ IA-R, ME2-R
12:53:59 <ais523> is that what you want?
12:54:00 <wob_jonas> dunno, I was just thinking because Dimir Guildmage is so useful if you have enough mana, because both the discard and the draw can target any player. you can make the opponent discard all cards with it... not that that really helps in this construction
12:54:31 <wob_jonas> no, I'm thinking more of something that lets you discard a card, force the opponent to discard a card, or you draw a card, any of these at instant speed and multiple times
12:54:49 <wob_jonas> but that probably doesn't exist all on one card
12:55:11 <wob_jonas> making the opponent loot is not good enough, he'll still have cards in his hand, until he loses from not being able to draw
12:55:57 <ais523> Wizards don't like instant speed discard
12:56:53 <wob_jonas> it still happens sometimes, but not just as a plain activated ability for no reason
12:57:08 <wob_jonas> and the colored Jayemdae is called Scepter of Insight
12:57:15 <ais523> Necrogen Spellbomb would work if you didn't have to sacrifice it
12:57:27 <ais523> `card-by-name Necrogen Spellbomb
12:57:27 <HackEgo> Necrogen Spellbomb \ 1 \ Artifact \ {B}, Sacrifice Necrogen Spellbomb: Target player discards a card. \ {1}, Sacrifice Necrogen Spellbomb: Draw a card. \ MRD-C
12:57:34 <ais523> but I think that's about as close as you'll get
12:57:55 <wob_jonas> I think you can get a bit closer than that
12:58:18 <HackEgo> Cinderhaze Wretch \ 4B \ Creature -- Elemental Shaman \ 3/2 \ {T}: Target player discards a card. Activate this ability only during your turn. \ Put a -1/-1 counter on Cinderhaze Wretch: Untap Cinderhaze Wretch. \ SHM-C
12:58:47 <HackEgo> Disrupting Scepter \ 3 \ Artifact \ {3}, {T}: Target player discards a card. Activate this ability only during your turn. \ A-R, B-R, U-R, RV-R, 4E-R, 5E-R, 6E-R, 7E-R, 8ED-R, 9ED-R
12:58:56 <ais523> those don't draw cards
12:59:22 <wob_jonas> you said Necrogen Spellbomb because it lets you draw too
13:00:22 <wob_jonas> I should have realized, I have the white and green spellbombs (gain life and animate land) and have seen the red one (shock), and each of them has a draw a card ability.
13:01:40 <ais523> there are two cycles of spellbombs, I think?
13:02:24 <wob_jonas> you're right, they made a worse one in Scars of Mirroding
13:02:34 <wob_jonas> just like how they made a cycle of worse Guildmages in RTR
13:03:04 <ais523> `card-by-name Nihil Spellbomb
13:03:05 <HackEgo> Nihil Spellbomb \ 1 \ Artifact \ {T}, Sacrifice Nihil Spellbomb: Exile all cards from target player's graveyard. \ When Nihil Spellbomb is put into a graveyard from the battlefield, you may pay {B}. If you do, draw a card. \ SOM-C, C13-C, C17-C
13:03:16 <ais523> that's a pretty widely played card
13:03:35 <ais523> exiling a graveyard and drawing a card for 1B is a pretty useful effect
13:03:53 <ais523> and the exact wording makes it usable in some combo decks
13:04:02 <wob_jonas> I have copies of Lifespark Spellbomb and Sunbeam Spellbomb. they're not very useful, although Lifespark Spellbomb might perhaps find some marginal utility in an infinite mana combo deck
13:09:55 <wob_jonas> the program loop is in your turn. you simulate counters with one token of a specific creature type each. you use Hungry Lynx controlled by the opponent to have triggers the trigger on when the counter would get reduced to zero.
13:11:12 <wob_jonas> this way you want to simulate The Amnesiac from Minsk level 1, and to avoid having to use triggers with a decrease effect, you modify each trigger by adding a constant to every counter, and keep reducing the counters.
13:11:47 <wob_jonas> you reduce them slower than the program triggers by using triggers on objects you control... I'm not sure if that works really
13:12:04 <wob_jonas> what was the strange card you used for that now?
13:12:56 <ais523> we control the Ghoul, and also one Rotlung Reanimator
13:13:05 <ais523> set to Zombie for both creature types
13:13:26 <ais523> when we create a zombie token via the reanimator, it immediately dies due to the opponent's Elesh Norn
13:13:30 <wob_jonas> the rotlung reanimator is for recreating the counter after the Hungry Lynx triggered so it's usable again
13:13:31 <ais523> thus causing another zombie token to be created, etc.
13:13:40 <ais523> wob_jonas: we use them for two purposes
13:13:45 <ais523> the opponent has them for recreating the counters
13:13:51 <ais523> and we have one for creating this infinite loop of zombies dying
13:14:05 <ais523> every dead zombie triggers the Noxious Ghoul and -1/-1's all the counters
13:14:09 <ais523> `card-by-name Noxious Ghoul
13:14:11 <HackEgo> Noxious Ghoul \ 3BB \ Creature -- Zombie \ 3/3 \ Whenever Noxious Ghoul or another Zombie enters the battlefield, all non-Zombie creatures get -1/-1 until end of turn. \ LGN-U, HOP-U
13:14:30 <wob_jonas> but why are two players enough for this?
13:15:07 <ais523> well, let's start at the "main loop", where the only things on the stack are a Noxious Ghoul trigger and Rotlung Reanimator trigger (in either order)
13:15:16 <ais523> both controlled by us, as the active player
13:15:25 <ais523> we can stack them in either order, but it doesn't matter which
13:15:40 <ais523> if the Reanimator trigger goes first, it just creates another 0/0 zombie, which gives us another of each trigger
13:16:04 <ais523> so there's always one Reanimator trigger and arbitrarily many Noxious Ghoul triggers, but the point is that it doesn't matter how many Ghoul triggers we allow to stack up before resolving them
13:16:07 <wob_jonas> the counters are also recreated by Noxious Ghoul, and you use some Radiant Destiny to make those tokens not die, right?
13:16:27 <ais523> the counters are recreated by Rotlung Reanimators controlled by the opponents
13:16:38 <ais523> we can use Blades of Velis Vel to make them Zombies in addition to their other types
13:16:54 <ais523> so that they don't get caught in the "blast radius" of the Noxious Ghoul
13:17:08 <ais523> it might be doable without, using mathematics to make sure that the toughness never falls to 0, but that's harder
13:17:56 <ais523> anyway, the point is that what we have on our side of the field leads to an infinite sequence of Noxious Ghoul triggers
13:18:03 <wob_jonas> "it doesn't matter how many Ghoul triggers we allow to stack up before resolving them" => it's not so simple. that might mean, depending on how exactly the infinite loop rules work, that you're permitted to draw the game at will. if that's true, then you must use a construction where the program can cause you to win, as opposed to one where the pro
13:18:03 <wob_jonas> gram causes you to lose if it terminates.
13:18:06 <ais523> /unless/ something happens to Elesh Norn, in which case the sequence will end
13:18:22 <ais523> wob_jonas: the rules say you can never choose to do something infinitely many times in a row if you have another option
13:18:57 <ais523> which means that if there's a legal stacking combination on our side that leads to game over, we have to take it
13:19:17 <ais523> and it's clear that if any combination leads to game over, the combination of always stacking the reanimator trigger below the ghoul trigger will do it
13:19:24 <wob_jonas> ARGH! now we use "counter" in three different ways
13:19:46 <ais523> let's use "variable" in the explanation then
13:20:06 <wob_jonas> minsky machines used "counter" for decades
13:20:24 <ais523> now, if a noxious ghoul trigger causes one of the opponent's creatures representing a variable to die
13:20:31 <int-e> wob_jonas: don't worry, it's all counter-intuitive.
13:20:34 <ais523> there will be a range of triggers on the opponent's side too, via game rules those always resolve first
13:20:59 <ais523> but they're commutative in the sense that they all act independently so the order in which the opponent stacks them doesn't matter
13:21:38 <wob_jonas> yes, the opponent triggers are commutative
13:21:52 <ais523> the triggers a) recreate the dead creature as a 2/2+2/2 token (unless it was Elesh Norn), b) put +1/+1 counters on creatures of a specific creature type
13:22:07 <ais523> b) will affect both ROM and RAM, but for ROM it doesn't matter how high the toughness is as it never dies anyway
13:22:49 <ais523> so the only relevant part of the effect is the one that toughens up variables other than the one that just became zero
13:23:22 <ais523> so we have "when a counter hits 0, its value becomes 4 and a constant is added to each other counter"
13:24:04 <wob_jonas> you're saying that extra triggers of the special Reanimator doesn't matter, because we only move forward with the trigger of the Noxious Ghoul. but what happens when we have triggers by the ordinary Reanimators, the ones that recreate counters. isn't it a problem to stack those in the wrong order, with respect to the Noxious Ghoul trigger?
13:24:31 <ais523> wob_jonas: they're owned by a different player
13:24:33 <ais523> specifically, the opponent
13:24:41 <ais523> it's our turn, so our triggers stack before theirs and thus resolve after theirs
13:24:59 <wob_jonas> so both the Lynxes and the per-counter Reanimators are owned by the opponent? I see
13:25:42 <ais523> err, controlled, not owned
13:26:03 <ais523> technically speaking we own all the cards involved, this is at least partly because we don't want to rely on specific cards existing in the opponent's deck
13:26:05 <wob_jonas> and what card exactly did you use as a lord effect, to add toughness to the tokens?
13:26:20 <ais523> makes the tokens come in as 4/4s
13:26:34 <ais523> that's a deviation from TWM's spec, because it's meant to be arbitrary there
13:26:38 <ais523> but I believe the language is still TC
13:26:59 <wob_jonas> who controls the single tokens that represent the variables?
13:27:10 <wob_jonas> right, because he has the reanimator
13:27:13 <ais523> opponent controls all ROM and RAM, we control the clock
13:27:15 <wob_jonas> I see, so that's why Elesh Norn works
13:27:39 <wob_jonas> I thought you'd need some other lord you copy or something
13:27:44 <int-e> . o O ( So... what would happen if at a sanctioned M:tG tournament, two players were to set up a game state where the winner depends on whether the Goldbach conjecture is true or not? )
13:27:58 <ais523> int-e: we were debating that earlier
13:28:14 <ais523> I don't think there's any legal option, the practical option would likely be to call the game a draw
13:28:23 <ais523> that said, setting that up without violating slow play rules will be hard
13:28:31 <ais523> as we need to be able to state the resulting gamestate in a reasonable length of time
13:28:36 <ais523> meaning it'll need to be heavily golfed
13:28:44 <wob_jonas> int-e: we were talking about that earlier. I was arguing that the player who sets it up would be disqualified no matter what, the other player might continue if he calls a judge early and doesn't cooperate
13:28:55 <wob_jonas> int-e: but after alercah's statements, I'm no longer convinced of this.
13:29:01 <ais523> perhaps the Collatz Conjecture would be golfier
13:29:21 <wob_jonas> int-e: it is still very unlikely that you could set this up, because you'd need to set up and explain everything without notes you took before the tournament, on a strict time limit to an impatient judge
13:29:21 <ais523> it's basically made for counter machines
13:30:19 <wob_jonas> and either of those might be easier than to set up a universal turing machine, which is what you'd need to encode some hard cryptographical problem
13:30:52 <wob_jonas> but I wonder if you could do even more golfier by encoding factoring of some particular composite number of your choice
13:31:08 <wob_jonas> more precisely the decision problem of whether it has a factor between 2 and a particular limit
13:31:41 <wob_jonas> or a factor that's 3 modulo 4 and below a particular limit, that might be slightly golfier so you don't have to skip 1
13:32:19 <ais523> if you pick a number with n digits at random (where n is large), what's the probability that it'll be hard to factorise quickly?
13:32:29 <ais523> the benefit of that is that you could just make up a number on the spot
13:32:34 <ais523> and have no idea whether it lead to a win for you or not
13:32:35 <wob_jonas> you don't need to pick one at random
13:32:44 <wob_jonas> you can choose a particular easy to remember number
13:32:55 <ais523> wob_jonas: yes but the point is that you need to not yourself know whether or not it's composite
13:32:57 <ais523> so you don't know who won
13:33:07 <wob_jonas> you choose one that's definitely composite
13:33:37 <wob_jonas> and you win if the factor is lower than a limit you set
13:33:48 <wob_jonas> and you set the limit right so it doesn't come out trivial
13:34:25 <wob_jonas> primality testing is too easy if the judge has access to the internet, you'd need a much higher number for it to be hard
13:34:50 <ais523> I see, so you use a probably-prime test, see it says "not prime"
13:34:54 <ais523> but still don't know what the factor is?
13:36:42 <wob_jonas> sort of like a composite number you'd use for RSA, although you might use a challenge that's slightly easier than you'd use for real world cryptography, as long as it still won't be factored during the tournament
13:37:42 <wob_jonas> so you choose an easy to learn one, one which has a form where all but six digits at the start and six digits at the end are zero, which probably would be too few digits for real world cryptography, but OK here
13:38:00 <wob_jonas> oh, and unlike in real world cryptography, nobody needs to have ever known the factors
13:38:05 <wob_jonas> not even the computer that generated the number
13:38:42 <wob_jonas> in fact, it's probably best if nobody knows the factors and the answer
13:39:06 <wob_jonas> you certainly aren't allowed to know that, because you knowing and not telling would probably be a tournament rules violation
13:40:58 <wob_jonas> per tournament rules 4.1 "Players must answer all questions asked of them by a judge completely and honestly, regardless of the type of information requested. Players may request to do so away from the match."
13:41:15 <wob_jonas> so you aren't allowed to know the result of the Turing machine but not reveal it
13:44:21 <int-e> wow, that's rather strong
13:44:41 <int-e> (since it fails to limit the scope of the questions)
13:48:18 <wob_jonas> The collatz conjecture is probably better
13:49:01 <wob_jonas> it's golfier, and if the opponent or a judge solves it, then we've won regardless of what happens with the tournament
13:54:46 <alercah> wob_jonas: you certainly wouldn't get disqualified for trying this
13:54:57 <alercah> at the very worst, you get an upgraded slow play game loss
13:55:29 <alercah> I suppose if you kept at it you could get yourself disqualified but really?
13:55:48 <ais523> if you got disqualified it would be for Stalling
13:56:19 <wob_jonas> alercah: not even if you would succeed? as opposed to trying it but never getting even remotely close
13:56:25 <ais523> but I think there's precedent saying that if you're genuinely performing your game actions as fast as possible, and they lead to progress in the gamestate, it's not Stalling even if those actions aren't working towards a win
13:56:35 <wob_jonas> alercah: and does the bribery chapter matter for this?
13:57:05 <ais523> this isn't bribery any more than Mana Clash is
13:57:09 <wob_jonas> ais523: and they're not working for a draw either?
13:57:15 <alercah> stalling must be explicitly to abuse the time limit
13:57:31 <alercah> you could eventually get disqualified for intentional slow play
13:57:46 <alercah> but I would only consider that after a game loss
13:59:08 <wob_jonas> in that this is hard to pull of, but probably not impossible
13:59:43 <wob_jonas> I'll have to look up some card prices then, to see if any of the pieces we used are very expensive
14:00:18 <ais523> well, the maindeck is a tournament-viable Legacy deck so it's naturally expensive
14:00:24 <ais523> but I think our additions would be very cheap
14:00:40 <wob_jonas> that's my guess too, but since I can look them up, I will
14:01:04 <wob_jonas> and if any piece is expensive, I can look for cheaper substitutes
14:01:06 <ais523> let's try to get a complete list of added cards
14:01:58 <ais523> Hungry Lynx, Rotlung Reanimator, Noxious Ghoul is the core of the combo, we definitely need those three
14:02:10 <ais523> everything else is more flexible but we can't do it with those three alone
14:02:58 <ais523> currently we have Elesh Norn, Grand Cenobite as the -2/-2 effect, and Harmless Offering, Artificial Evolution (OK this is probably inflexible), Cackling Counterpart, Wild Cantor as the setup
14:03:15 <ais523> 8 added cards, not bad
14:03:27 <ais523> that means we only have to remove half the sideborad
14:04:03 <ais523> (we need a few other cards that are in the deck naturally, like Emrakul, the Aeons Torn; Wipe Away; and Omniscience+Enter the Infinite)
14:04:45 <ais523> two, there are already two in the deck naturally though
14:06:48 <wob_jonas> plus 4 Omniscience, 4 Force of Will (that latter seems the most expensive so far)
14:06:57 <ais523> Force of Will is in the deck naturally too
14:07:12 <ais523> also it isn't technically required, just makes it harder for the opponent to screw with us
14:07:23 <ais523> as I said, it's a tournament-viable Legacy deck and thus will naturally be expensive
14:07:39 <wob_jonas> although obviously you can resell it afterwards
14:08:01 <wob_jonas> (resell the expensive cards that is)
14:08:15 <ais523> I wasn't planning to actually do this myself
14:08:27 <ais523> among other things, there aren't many Legacy tournaments around here
14:08:30 <wob_jonas> I'm just talking in general, since you brought up the possibility
14:08:41 <wob_jonas> and I want to justify the old joke that mathematics is cheap
14:11:50 <wob_jonas> 2 City of Traitors (isn't cheap either)
14:12:22 <ais523> hmm, I wonder if there's an argument for putting the Wild Cantor maindeck
14:12:38 <ais523> because it would also act as acceleration, and the reference version I linked earlier can generate red mana
14:13:01 <wob_jonas> and then 4 Scalding Tarn, 1 Volcanic Island (pretty expensive)
14:13:11 <ais523> well, yes, it's a Legacy deck and it isn't monocolour
14:13:29 <ais523> that said, Steam Vents would work almost as well
14:14:14 <wob_jonas> not with Misty Rainforest, mind you
14:14:44 <wob_jonas> Steam Vents is one of the ravnica rare duals
14:15:04 <ais523> it's two life worse than Volcanic Island, but in Legacy combo decks two life is rarely relevant
14:15:09 <ais523> (it comes up sometimes but not very often)
14:15:25 <wob_jonas> so many dual lands out there, and many of the cycles have unconnected names
14:16:38 <ais523> they do it because they want generic names that they can print on any plane
14:16:56 <ais523> although I'm not convinced by this reasoning as it rarely worked out in practice
14:17:08 <ais523> e.g. the Steam Vents cycle has been printed only on Ravnica
14:17:20 <ais523> and they ended up having to print Dragonskull Summit on Ixalan, which doesn't have dragons
14:17:31 <wob_jonas> sometime they manage names that at least sound similar, such as Barren Moor, Forgotten Cave, Lonely Sandbar, Secluded Steppe, Tranquil Thicket; and of course it gets much easier if you see the actual cards, with art and frame and all
14:18:29 <wob_jonas> does Ixalan at least have skulls on pirate flags?
14:19:41 <wob_jonas> `card-by-name Territorial Hammerskull
14:19:42 <HackEgo> Territorial Hammerskull \ 2W \ Creature -- Dinosaur \ 2/3 \ Whenever Territorial Hammerskull attacks, tap target creature an opponent controls. \ XLN-C
14:22:06 <wob_jonas> the Steam Vents cycle got reprinted in the Return to Ravnica block because it's so iconic, and if it weren't reprinted, it was because of its power level, which they might not have realized back in ravnica. I'm not sure that's too relevant about the naming.
14:22:21 <wob_jonas> yeah, I guess that's still "on Ravnica" for planes
14:22:22 <ais523> well, they haven't reprinted the Selesnya Sanctuary cycle either
14:22:26 <ais523> which does have a plane-specific name
14:22:28 <ais523> and which I rather liked
14:22:34 <ais523> I think they consider that one too powerful too, though
14:22:44 <ais523> Ravnica had a /lot/ of color fixing
14:22:46 <wob_jonas> they have reprinted that in commanders
14:22:59 <wob_jonas> but it is commons so it's cheap enough that I have four-of of most of them
14:23:13 <wob_jonas> four copies of five or six out of the ten I think
14:23:51 <wob_jonas> they don't need to reprint if it's common and there's enough on the market
14:24:50 <wob_jonas> I hate the font used in the title line of the recent cards. why did they have to choose a worse font, after they replaced the horrible colors and font of the old title line in Mirrodin?
14:25:49 <ais523> the current version of the font was chosen to be a font that isn't publicly available, to make the cards harder to counterfeit
14:28:35 <wob_jonas> mirage to invasion has bad colors and a very bad font on the title line. mirrodin to theros has a good font and good colors (not counting some time spiral block shenanigans). khans of tarkir to present has a somewhat worse font but still good colors.
14:28:55 <wob_jonas> really, counterfeiting is their reason? that sounds a bit stupid
14:30:44 <ais523> it's a major threat to their business, given the difference between the amount they charge for cards and the amount they cost to print
14:30:58 <ais523> that's why they started adding the hologram stamp on rares
14:32:08 <wob_jonas> In that Omnitell deck, which already includes wishes, is any card from the original sideboard necessary?
14:32:27 <wob_jonas> I mean, can you just throw the whole sideboard away to fill it with cards for the turing-machine setup?
14:33:24 <ais523> Enter the Infinite and Show and Tell in the sideboard are not technically necessary, but they make it a lot stronger
14:33:46 <ais523> Release the Ants is the intended finisher of the deck, although I think that's a mistake because there's a reliable kill just with the cards in the maindeck
14:34:22 <wob_jonas> But doesn't everyone have much better resolution scanners these days than printers, so the text outline for any card can be easily reconstructed? I have scanned graphics and photos in 600 dpi because it doesn't cost me anything more than a 300 dpi scan, and once you actually want to make decent quality counterfeits, you can go even higher.
14:35:01 <ais523> not commercial printers, their output has a better resolution than most scanners (and most modern printers, for that matter) can manage
14:35:05 <ais523> at least when printing text
14:35:12 <ais523> they print different layers separately
14:35:51 <ais523> this is why the easiest way to tell a counterfeit is normally to look at the edge of the expansion symbol with a powerful magnifying glass
14:36:16 <ais523> printers whose input for that was a set of pixels, rather than vector-format, tend to be unable to recreate it accurately, it's obviously fuzzy
14:36:37 -!- atslash has joined.
14:37:22 <ais523> btw, the maindeck Impulse is only there for a combo with Firemind's Foresight, so if you remove one you can remove the other too
14:41:16 -!- atslash has quit (Ping timeout: 256 seconds).
14:52:14 <wob_jonas> and it's not likely that just Sigma-one or primitive recursion or something is easier to simulate in M:tG than full Turing-completeness, right?
14:53:42 <wob_jonas> I guess that would be possible if we could have a trigger that copies a token to another token with the same variable toughness but different type
14:55:03 <wob_jonas> or perhaps if we represented variables with as many tokens as its value
14:57:13 <HackEgo> Fungal Sprouting \ 3G \ Sorcery \ Create X 1/1 green Saproling creature tokens, where X is the greatest power among creatures you control. \ M13-U
14:57:29 <HackEgo> Kin-Tree Invocation \ BG \ Sorcery \ Create an X/X black and green Spirit Warrior creature token, where X is the greatest toughness among creatures you control. \ KTK-U
14:57:50 <HackEgo> Miming Slime \ 2G \ Sorcery \ Create an X/X green Ooze creature token, where X is the greatest power among creatures you control. \ GTC-U
15:00:31 <wob_jonas> but you can't easily repeat those effects, because Spellbinder and similar effects are optional
15:05:43 <wob_jonas> So how much does the translation from that small Minsky machine to Amnesiac level 1 blow up? I think going from Amnesiac to Waterfall doesn't blow up too much.
15:10:18 <ais523> Amnesiac L1 is an O(n) blowup with the mechanical compilations
15:10:24 <ais523> but I think it's normally less if you do it by hand
15:11:09 <wob_jonas> ais523: O(n) blowup when the Minsky machine can increment or decrement variables by any fixed integer, and that integer is represented in binary in the description?
15:11:15 <lambdabot> LOWI 031450Z VRB01KT 9999 FEW090 SCT110 BKN250 09/M02 Q0992 R08/19//95 NOSIG
15:11:31 <wob_jonas> that is, the blowup doesn't depend on the integer increment values, but only on the number of counters and states?
15:11:47 <ais523> wob_jonas: yes, the blowup is to remember which state you're in when you do an increment
15:11:52 -!- Cale_ has quit (Remote host closed the connection).
15:11:54 <lambdabot> EGBB 031450Z 07013KT 3000 BR OVC006 01/M00 Q0993
15:12:09 <wob_jonas> then it might fit the 237 creature types
15:12:12 <ais523> that time looks out of date
15:12:26 <int-e> (I should've checked before I left the house... could have picked a different coat
15:13:45 <wob_jonas> https://stackoverflow.com/questions/1749905/code-golf-fractran/1802570#1802570 --- WTF! this is on Stack Overflow? as opposed to Code Golf SE or something?
15:14:52 <wob_jonas> (ais523 linked to that earlier, but I didn't look)
15:15:17 <ais523> it predates PPCG being created
15:15:20 <ais523> and got locked for being offtopic
15:15:40 <ais523> PPCG got created mostly as a result of posts like that being repeatedly locked, people liked them and wanted somewhere to put them
15:15:41 <wob_jonas> and fractran might be a good language for this sort of M:tG construction too
15:15:55 <ais523> I don't think so, divisibility tests in M:tG are hard
15:16:26 <wob_jonas> you'd store only the prime exponents
15:16:38 <wob_jonas> not the representation as a single integer
15:17:07 <wob_jonas> it might not be as good as Waterfall, but still
15:22:11 <ais523> oh, in that case, tests for simultaneous presence of a particular subset of creatures are hard to do in M:tG
15:25:07 <wob_jonas> and you still use Blades of Velis Vel to protect the opponent's Rotlung Reanimators from the breath of the Ghouls, right?
15:25:39 <ais523> I think so, you need /something/ to add an additional creature type to it
15:26:12 <wob_jonas> yeah, Blade of Velis Vel is the easiest for that
15:26:47 <ais523> there might be a way to do it mathematically, have their toughnesses changing over time but never hitting 0
15:27:41 <wob_jonas> that could certainly work for the factoring problem where we have a strict easy to compute upper limit on the runtime
15:27:49 <ais523> Hungry Lynx, Rotlung Reanimator, Noxious Ghoul, (Elesh Norn, Grand Cenobite), Harmless Offering, Artificial Evolution, Cackling Counterpart, Wild Cantor, Blades of Velis Vel
15:28:55 <ais523> would be nice to golf this down a bit
15:29:36 <ais523> wait, we need research//development, too
15:29:45 <ais523> Hungry Lynx, Rotlung Reanimator, Noxious Ghoul, (Elesh Norn, Grand Cenobite), Harmless Offering, Artificial Evolution, Cackling Counterpart, Wild Cantor, Blades of Velis Vel, Research//Development
15:45:05 <ais523> it's a pity Mastermind's Aquisition costs double-black, if not for that it might have been viable in the maindeck
15:45:13 <ais523> we could still use it as an R//D replacement
15:56:32 <ais523> `card-by-name Fractured Identity
15:56:32 <HackEgo> Fractured Identity \ 3WU \ Sorcery \ Exile target nonland permanent. Each player other than its controller creates a token that's a copy of it. \ C17-R
15:57:10 <ais523> close to merging two of our slots but doesn't quite get there because we can't get the Lynx back out of exile
15:58:40 <ais523> `card-by-name coax from the blind eternities
15:58:42 <HackEgo> Coax from the Blind Eternities \ 2U \ Sorcery \ You may choose an Eldrazi card you own from outside the game or in exile, reveal that card, and put it into your hand. \ EMN-R
15:58:47 <ais523> can we use that rather than R//D?
15:59:04 <ais523> (you artificially evolve "Eldrazi" to match a creature type of the creature that got exiled)
16:00:56 <ais523> the idea is that we use Cunning and Burning wishes that we naturally have in our deck to fetch Coax, Fractured, and Artificial
16:01:04 <ais523> `card-by-name Blades of Velis Vel
16:01:05 <HackEgo> Blades of Velis Vel \ 1R \ Tribal Instant -- Shapeshifter \ Changeling (This card is every creature type.) \ Up to two target creatures each get +2/+0 and gain all creature types until end of turn. \ LRW-C, MM2-C
16:01:15 <ais523> Blades of Velis Vel is naturally an Eldrazi, so we can get that too
16:02:09 <ais523> then we use the standard double-Emrakul trick to keep recycling Coax and Artifical in order to get all the creatures out of our sideboard
16:02:48 <ais523> Hungry Lynx, Rotlung Reanimator, Noxious Ghoul, (Elesh Norn, Grand Cenobite), Artificial Evolution, Wild Cantor, Blades of Velis Vel, Fractured Identity, Coax from the Blind Eternities
16:05:23 -!- HereToAnnoy has joined.
16:08:10 <ais523> right now we're busy trying to prove Magic: the Gathering to be Turing complete even with no decisions made by the players (once a crafted gamestate is set up)
16:10:46 <ais523> or, rather, we've already pretty much proved it but we're trying to golf down the construction so that we can fit it in the sideboard of a tournament-viable deck
16:11:35 <ais523> <ais523> Hungry Lynx, Rotlung Reanimator, Noxious Ghoul, (Elesh Norn, Grand Cenobite), Artificial Evolution, Wild Cantor, Blades of Velis Vel, Fractured Identity, Coax from the Blind Eternities
16:11:42 <ais523> current record is 9 cards, it seems likely we can do better though
16:12:11 <ais523> (this also makes certain assumptions about the rest of the deck; we're working from this Omnitell list: http://sales.starcitygames.com//deckdatabase/displaydeck.php?DeckID=113604)
16:15:56 -!- Cale_ has joined.
16:24:46 -!- HereToAnnoy has left.
16:33:38 -!- Sgeo has quit (Read error: Connection reset by peer).
16:34:03 -!- Sgeo has joined.
16:37:10 <wob_jonas> Fractured Identity? does that mean you have to play Elesh last?
16:37:20 <wob_jonas> and also, that you don't use Harmless Offering now?
17:01:01 <ais523> yes, the idea was to save a card
17:01:32 <ais523> you would indeed need to play Elesh pretty late, but you'd have to do that anyway
17:01:50 <ais523> the idea is that Fractured Identity can combine the token-making and donating parts of the setup
17:02:19 <ais523> although that has the downside that the card gets exiled, so we have to change the "get cards out of our sideboard" part of the setup to compensate
17:07:29 -!- atslash has joined.
17:09:21 -!- LKoen has joined.
17:11:55 -!- augur has joined.
17:23:19 <ais523> err, hmm, there's a mistake in the construction: Blade of Velis Vel can't legally target Hungry Lynx no matter how much you adjust the creature types
17:24:15 <ais523> which makes it even more important to try to save that slot
17:27:12 <ais523> <706.2> Other effects (including type-changing and text-changing effects), status, and counters are not copied.
17:27:19 <ais523> so we can't just switch to Shields of Veils Vel either
17:30:02 <ais523> I think Master Biomancer works but it'd be awkward
17:30:26 <ais523> (you hack it to say Zombie rather than Mutant while you're deploying the cats, then bounce it before you deploy the tokens)
17:31:48 -!- laerling has joined.
17:32:17 <ais523> I think it's the only option though
17:39:12 <ais523> `card-by-name Master Biomancer
17:39:13 <HackEgo> Master Biomancer \ 2GU \ Creature -- Elf Wizard \ 2/4 \ Each other creature you control enters the battlefield with a number of additional +1/+1 counters on it equal to Master Biomancer's power and as a Mutant in addition to its other types. \ GTC-M, C16-M
17:45:01 -!- j-bot has quit (Remote host closed the connection).
17:50:20 <wob_jonas> ais523: why can't Shields of Velis Vel work?
17:50:32 <wob_jonas> if you do the construction in the right order?
17:51:18 <ais523> `card-by-name shields of velis vel
17:51:20 <HackEgo> Shields of Velis Vel \ W \ Tribal Instant -- Shapeshifter \ Changeling (This card is every creature type.) \ Creatures target player controls get +0/+1 and gain all creature types until end of turn. \ LRW-C
17:51:27 <ais523> oh wait, you can aim it at your opponent
17:51:34 <ais523> for some reason I assumed it'd only affect you
17:52:07 <ais523> we sent the ROM over first, then Shields it, then send over the RAM
17:52:36 <ais523> later on we can Shields ourself to prevent our Rotlung dying to the enemy Elesh Norn
17:53:01 <ais523> (the /+1 also keeps the enemy ROM and RAM alive while we temporarily cast Elesh so that we can copy her)
17:54:17 <ais523> 9 cards: Hungry Lynx, Rotlung Reanimator, Noxious Ghoul, (Elesh Norn, Grand Cenobite), Artificial Evolution, Wild Cantor, Shields of Velis Vel, Fractured Identity, Coax from the Blind Eternities
17:54:40 <ais523> actually, let me order that by card type
17:55:11 <ais523> 9 cards: CREATURE: Hungry Lynx, Rotlung Reanimator, Noxious Ghoul, (Elesh Norn, Grand Cenobite), Wild Cantor, INSTANT: Artificial Evolution, Shields of Velis Vel, SORCERY: Fractured Identity, Coax from the Blind Eternities
17:55:20 <ais523> because that's relevant in how we get them out of the sideboard
17:55:25 -!- variable has joined.
17:55:38 <ais523> 9 cards: CREATURE/TRIBAL: Hungry Lynx, Rotlung Reanimator, Noxious Ghoul, (Elesh Norn, Grand Cenobite), Wild Cantor, Shields of Velis Vel, INSTANT: Artificial Evolution, SORCERY: Fractured Identity, Coax from the Blind Eternities
17:55:53 <ais523> as we're getting out Shields based on its tribal type
17:56:48 <ais523> this is mostly creatures/tribals which is great, as it means we can use the existing Burning Wishes for the sorceries and existing Cunning wish for the instant and are unlikely to have run out
17:57:01 <ais523> (the wishes self-exile, so we can't loop them; Coax doesn't, so we canA)
17:59:31 <wob_jonas> ais523: don't you still need Wipe Away?
18:00:02 <wob_jonas> and maybe something more for either the recycling or the cleanup after the setup? I don't think I understand how that works
18:00:39 <wob_jonas> `card-by-name Coax from the Blind Eternities
18:00:39 <HackEgo> Coax from the Blind Eternities \ 2U \ Sorcery \ You may choose an Eldrazi card you own from outside the game or in exile, reveal that card, and put it into your hand. \ EMN-R
18:01:58 <wob_jonas> ais523: how do you create the ram in first place?
18:02:13 <ais523> yes, we wish for Wipe Away
18:03:12 <ais523> to create the RAM we start by creating the Cleric tokens on our side of the field (via, e.g. Rotlung Reanimator set to Shaman and sacrificing a Cantor), then Fracturing them onto the other side of the battlefield
18:03:51 <wob_jonas> and after you send the tokens through, you evolve them to the right type? could work
18:04:12 <ais523> then to set the values away from all-2s we use a spare creature type and have a stack of Lynxes set to that type
18:04:22 -!- aloril has quit (Ping timeout: 260 seconds).
18:04:36 <ais523> sacrifice an evolved Cantor on our side and the Lynxes will place a precise amount of +1/+1 counters on everything based on their type
18:04:46 <ais523> then we never use the type again
18:05:17 <wob_jonas> instead of the Cantor, can't you use something that serves another function too though?
18:05:28 <wob_jonas> something that sacrifices for some other useful effect
18:07:29 <ais523> it does serve a useful function, it lets us start the combo with Wipe Away still on the stack
18:07:47 <ais523> to make it obvious that neither side can interfere
18:07:49 <wob_jonas> yeah, you want split second tricks
18:08:26 -!- tromp has quit (Remote host closed the connection).
18:10:49 -!- aloril has joined.
18:11:46 <wob_jonas> what if you use Starlit Sanctum's second ability to start the combo, you control it and you don't control any changelings, make sure you sacrifice the last creature of the creature type you want, and as a bonus you can use Starlit Sanctum to adjust life totals somewhat for the finisher?
18:12:02 <ais523> `card-by-name starlit sanctum
18:12:04 <HackEgo> Starlit Sanctum \ Land \ {T}: Add {C} to your mana pool. \ {W}, {T}, Sacrifice a Cleric creature: You gain life equal to the sacrificed creature's toughness. \ {B}, {T}, Sacrifice a Cleric creature: Target player loses life equal to the sacrificed creature's power. \ ONS-U
18:12:23 <ais523> that isn't a mana ability
18:12:43 <ais523> "you gain life" doesn't generate mana
18:13:24 <ais523> also, creatures are the easiest card type to get out of our sideboard
18:13:35 <wob_jonas> or... let me check all the cards that generate Gold or Eldrazi Spawn tokens
18:13:57 <HackEgo> Development \ 3UR \ Instant \ Create a 3/1 red Elemental creature token unless any opponent has you draw a card. Repeat this process two more times. \ [This is half of the split card Research // Development.] \ DIS-R
18:14:01 <HackEgo> Research \ GU \ Instant \ Choose up to four cards you own from outside the game and shuffle them into your library. \ [This is half of the split card Research // Development.] \ DIS-R \ \ Research // Development \ GU // 3UR \ Instant // Instant \ Choose up to four cards you own from outside the game and shuffle them into your library. // Create a
18:14:19 <wob_jonas> oh right, you also use that eldrazi stuff
18:15:17 <ais523> right, we needed to get creatures back from exile too once we started using Fractured Identity
18:15:36 <ais523> and Coax gets creatures from both exile and sideboard (the creature type restriction isn't relevant because we need Artificial anyway)
18:15:53 <HackEgo> Essence Feed \ 5B \ Sorcery \ Target player loses 3 life. You gain 3 life and create three 0/1 colorless Eldrazi Spawn creature tokens. They have "Sacrifice this creature: Add {C} to your mana pool." \ ROE-C
18:16:03 <wob_jonas> but I think you don't want to drain the opponent to zero, so that's not what we want
18:17:11 <ais523> attack with Emrakul (naturally in the deck), in response to the annihilator trigger cast Wipe Away targeting Wild Cantor, then sacrifice Wild Cantor to generate mana
18:17:25 <ais523> if the combo collapses then the opponent will be facing down Emrakul and have nothing to block it with
18:18:15 <ais523> we get rid of our own Omniscience before this (after having made just enough mana to finish the combo by looping Cantor) so that we don't have any ability to stop our own win
18:18:34 <wob_jonas> but what if he has enough life for Emrakul?
18:18:42 <wob_jonas> to survive Emrakul's attack that is
18:18:48 <ais523> I was envisioning using multiple attacks while the board was empty
18:19:22 <wob_jonas> but you can't make copies of Emrakul
18:20:30 <ais523> we have infinite turns anyway
18:20:35 <wob_jonas> I guess you could just attack with more spare creatures than the opponent have lynxes
18:20:57 <ais523> we could also pile +1/+1 counters on Emrakul
18:21:04 <ais523> wob_jonas: cast trigger on Emrakul
18:21:11 <ais523> we're hardcasting it as part of the recycle loop
18:21:15 <ais523> (which is why it costs 27 mana)
18:21:20 -!- tromp has joined.
18:21:42 <ais523> specifically, the loop is cast Emrakul, it dies to legend rule, shuffle your library; cast Enter the Infinite, get your library back to hand
18:21:59 <ais523> end result is graveyard → hand, and it's free if Omniscience is in play, and it's entirely based on cards that are in Omnitell naturally
18:22:08 <ais523> you need the other copy of Emrakul on the battlefield for this to work
18:22:24 <ais523> (note: if it got exiled somehow, you can get it back with Coax before starting the combo)
18:22:59 <wob_jonas> that's much better than a Dark Confidant combo
18:23:18 <ais523> playing Dark Confidant and Emrakul in the same deck is inadvisable anyway ;-)
18:23:43 <ais523> *shuffle your graveyard into your library
18:25:00 <wob_jonas> wtf, Omniscience is from M2013? what did people use before that came out?
18:25:27 <ais523> Omnitell didn't exist as a deck at the time
18:25:59 <wob_jonas> I dunno, I just thought either it was older, or there's some very similar older card
18:25:59 -!- LKoen has quit (Remote host closed the connection).
18:26:08 <ais523> `card-by-name Genesis Wave
18:26:09 <HackEgo> Genesis Wave \ XGGG \ Sorcery \ Reveal the top X cards of your library. You may put any number of permanent cards with converted mana cost X or less from among them onto the battlefield. Then put all cards revealed this way that weren't put onto the battlefield into your graveyard. \ SOM-R
18:26:19 <ais523> that was used for this sort of stupid combo, sometimes
18:26:21 <wob_jonas> simply because I don't usually know about rares from sets newer than M2010
18:27:10 <wob_jonas> I mean, people have invented a hundred different infinite mana combos, and some of them give any color, and some of them replace Omniscience just fine
18:27:50 <wob_jonas> ah yes, and there was another symmetric one
18:28:07 <wob_jonas> which put cards into play from your hand and from your opponent's hand too
18:29:18 <HackEgo> Hypergenesis \ Sorcery \ Suspend 3--{1}{G}{G} (Rather than cast this card from your hand, pay {1}{G}{G} and exile it with three time counters on it. At the beginning of your upkeep, remove a time counter. When the last is removed, cast it without paying its mana cost.) \ Starting with you, each player may put an artifact, creature, enchantment, or
18:29:40 <wob_jonas> not useful here, just that's what Genesis Wave reminds me
18:30:22 <HackEgo> Eureka \ 2GG \ Sorcery \ Starting with you, each player may put a permanent card from his or her hand onto the battlefield. Repeat this process until no one puts a card onto the battlefield. \ LE-R, VMA-M, MED-R
18:30:38 <wob_jonas> that's the original version Hypergenesis is nerfing
18:31:34 <ais523> `card-by-name Show and Tell
18:31:35 <HackEgo> Show and Tell \ 2U \ Sorcery \ Each player may put an artifact, creature, enchantment, or land card from his or her hand onto the battlefield. \ US-R, CN2-M
18:31:45 <ais523> that's actually in Omnitell, it's how it affords to play Omniscience
18:31:58 <ais523> (it can also just put Emrakul in directly if it has no better option, which is risky because the opponent can often remove it)
18:34:51 -!- ais523 has quit (Quit: quit).
19:03:38 <esowiki> [[Talk:StackFlow]] https://esolangs.org/w/index.php?diff=54274&oldid=42023 * B jonas * (+12734)
19:03:41 <esowiki> [[StackFlow]] https://esolangs.org/w/index.php?diff=54275&oldid=54273 * B jonas * (-12176)
19:05:07 <esowiki> [[Talk:StackFlow]] https://esolangs.org/w/index.php?diff=54276&oldid=54274 * B jonas * (+146)
19:05:10 <esowiki> [[StackFlow]] https://esolangs.org/w/index.php?diff=54277&oldid=54275 * B jonas * (-164)
19:05:34 -!- variable has quit (Quit: /dev/null is full).
19:15:51 -!- variable has joined.
19:23:07 -!- LKoen has joined.
20:11:31 <esowiki> [[RTFM]] N https://esolangs.org/w/index.php?oldid=54278 * TeslaX93 * (+1643) first version
20:12:13 <esowiki> [[F^3]] N https://esolangs.org/w/index.php?oldid=54279 * TeslaX93 * (+18) redirect
20:36:15 -!- sleffy has joined.
21:01:08 -!- atslash has quit (Quit: This computer has gone to sleep).
21:08:53 <zzo38> Did you see if there is a mistake in ZPXDB?
21:09:22 <zzo38> I am not sure how I should test it properly, but maybe you know, and then you can tell me how.
21:11:08 -!- wob_jonas has quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client).
21:20:26 -!- 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.”).
21:21:01 -!- augur has quit (Ping timeout: 245 seconds).
21:26:30 -!- ais523 has joined.
21:47:50 -!- augur has joined.
21:54:05 -!- ais523 has quit (Quit: quit).
22:12:16 <esowiki> [[Language list]] https://esolangs.org/w/index.php?diff=54280&oldid=54259 * TeslaX93 * (+35) added RTFM and F3
22:16:13 -!- wob_jonas has joined.
22:17:55 <wob_jonas> ais523: One more important question about the M:tG construction is this. If you don't insist on tournament viability, how can you modify the construction such that the abstract machine can safely query inputs from either player at its choice, in order that we can simulate an M:tG game or any other game.
22:23:40 <int-e> so you're making an alternating TM?
22:23:58 <HackEgo> An ATM is when you're withdrawing money right now at a machine that will duplicate your relevant info.
22:25:15 <int-e> yeah I guess that's better than "An ATM is an alternating Turing machine. It alternates between dispensing and not dispensing money." (but here's the idea for the record)
22:26:59 <wob_jonas> int-e: I don't know what an alternating Turing machine is
22:28:20 <wob_jonas> int-e: ah. yes, something like that.
22:28:54 <wob_jonas> although you might not find the true value it computes, because you and the opponent need not play optimally.
22:29:37 <int-e> "alternating" generalizes "non-deterministic"; when you get to make a choice in a non-deterministic TM, it will accept if any of the choices leads to an accepting state; in an ATM, for some of the non-deterministic choices, *all* of the paths need to lead to acceptance, while for others, like in NTMs, it suffices for one of the choices to lead to accepting paths (this gives rise to an arbitrary...
22:29:43 <int-e> ...alternation of finite existential and universal quantifiers)
22:30:07 <int-e> And that in turn has game semantics, https://en.wikipedia.org/wiki/Ehrenfeucht%E2%80%93Fra%C3%AFss%C3%A9_game ... closing the circle.
22:30:44 <wob_jonas> Plus I guess you might want to add a random source too.
22:31:08 <int-e> yeah there's nothing probabilistic about ATMs.
22:33:17 -!- j-bot has joined.
22:33:50 <wob_jonas> Luckily M:tG has tons of cards that offers a choice to the controller, and a large enough selection of cards that offer a choice to the opponent or make a choice at random, so it is probably possible to add all that.
22:34:09 <wob_jonas> Hmm, that might not be enough though
22:34:52 <wob_jonas> You might also want a way for some of the state of the program to be hidden from the players, and that might be very hard if you want it working strictly,
22:35:25 <wob_jonas> because there's almost no way in M:tG to store more hidden state than what the permutation of the physical cards can store.
22:36:09 <wob_jonas> Luckily, state hidden from both players isn't a big problem, you can just approximate that with randomness and lots of computation. State hidden from just one player can be a problem.
22:36:17 <wob_jonas> And that might be hard to simulate too.
22:43:54 <int-e> (but you can only havea bounded number of those, I guess, and even if you *can* operate on such hidden information, that act would reveal it)
22:43:59 <wob_jonas> I guess you could use cryptography to simulate the hidden information
22:44:22 <wob_jonas> but that would only work if you forced the opponent to do expensive cryptographic operations in his head
22:45:41 <wob_jonas> it's probably easier to just simulate a perfect information game, like chess or go
22:46:58 <int-e> or backgammon, if you want to use the probabilistic feature
22:48:14 <wob_jonas> The advantage is that there are already nice small implementations of chess for multiple real computers, so you only need to simulate most of a 6502 plus RAM plus some IO devices and then run an existing chess program
22:49:36 <wob_jonas> int-e: yeah. and if you can't easily store arbitrary hidden information, that means not only you can't simulate M:tG, you also can't simulate Scrabble or Starcraft.
22:49:50 <int-e> which lead me to https://gilkalai.wordpress.com/2011/01/14/is-backgammon-in-p/ (still reading, but at least some people are interested :) )
22:52:13 <wob_jonas> although you could simulate a single-player game with hidden information and randomness, if you don't mind a few more levels of exponential slowdown (since we're using a counter machine, you already have at least one level of exponential slowdown)
22:52:21 <int-e> homomorphic encryption is a relevant keyword
22:52:35 <int-e> (if you want to simulate hidden information through crypto)
22:54:18 <wob_jonas> int-e: you could technically do that, but I don't think you need that. Much less power is enough to just simulate M:tG, although you might need homomorphic encryption to faithfully simulate Starcraft or a Counterstrike deathmatch.
22:55:01 <int-e> which still isn't practical, I guess. "In late 2014, a re-implementation of homomorphic evaluation of the AES-encryption circuit using HElib, reported evaluation time of just over four minutes on 120 inputs, bringing the amortized per-input time to about 2 seconds." [wikipedia]
22:55:01 <wob_jonas> And again, the problem would be that the opponent would need to compute difficult crypto computations in his head to keep hidden information, or at least ask a judge to do it for him.
22:55:35 <wob_jonas> And it'd be hard to convince a non-accomplice to do that.
22:58:24 <wob_jonas> you're already using a construction that has at least one level of exponential slowdown. You obviously shortcut most of that. So you could just say that the opponent only needs to encrypt each of his decisions (with random salt), the simulated machine decrypts it (in double or triple exponential time, who cares), and since you're shortcutting every
22:58:24 <wob_jonas> thing, you can just shortcut all the crypto part too by just playing the simulated game and keeping hidden info with pieces of paper facing down and similar traditional methods.
22:59:41 <wob_jonas> Mind you, it might be possible to make an improved reduction to M:tG that has only polynomial slowdown, in particular a fixed simulation of StackFlow could probably do that.
23:00:05 <wob_jonas> (StackFlow with a bounded number of stacks and states etc.)
23:03:24 -!- laerling has quit (Quit: Leaving).
23:11:10 -!- sebbu has quit (Ping timeout: 240 seconds).
23:33:57 -!- AnotherTest has quit (Ping timeout: 260 seconds).