00:01:39 -!- augur has quit (Remote host closed the connection).
00:02:19 -!- augur has joined.
00:03:59 <fizzie> Well, what do you think -- should I change the CAPTCHA back to something possible to see if the filter's coping? Or maybe at some later time when I'll be around to watch what happens.
00:04:21 <fizzie> (By the way, did you notice we had a likely spammer come by on the channel to ask for help in creating a new account?)
00:04:33 <ais523> fizzie: no, I wasn't here I don't think
00:04:41 <ais523> can you give a date so that I can logread it?
00:05:09 <fizzie> Today at 15:00..15:18 BST.
00:05:42 <myname> that guy who waited two whole minutes?
00:05:45 <fizzie> Logical today, as it's called where I come from.
00:05:54 <fizzie> myname: No, I think that was possibly someone legitimate.
00:06:55 <fizzie> This one wanted an account "to get information from this website", and left when I asked what sort of edits they wanted to make they needed an account for.
00:07:22 -!- augur has quit (Ping timeout: 272 seconds).
00:08:58 <shachaf> Hmm, I kind of want to create an account.
00:09:11 <shachaf> But apparently it's impossible.
00:09:58 <ais523> why is it impossible? we have a funge interpreter in the channel :-)
00:10:19 <shachaf> No, the CAPTCHA was changed.
00:10:23 <fizzie> ais523: I switched it back from the Befunge one to the "impossible" one when those two new spam accounts were made today.
00:10:43 <ais523> in what sense is it impossible? it doesn't give any information that would help you answer?
00:10:51 <fizzie> And it's a random answer.
00:11:07 <shachaf> So it's just a low probability.
00:11:08 <fizzie> I did let at least one person know the string when they messaged me and made sense.
00:12:48 -!- cemulate has joined.
00:14:02 -!- cemulate has quit (Client Quit).
00:14:29 <fizzie> I wonder if that particular visitor wanted an account too.
00:14:49 <izabera> $ inutility dd conv=swab,ucase obs=2 ibs=3 <<< abcdef
00:14:57 <izabera> 7 bytes (7.0B) copied, 0.000062s, 110.3K/s
00:15:19 <izabera> conv=swab now works!!!!!!!!!!!!!
00:15:31 <izabera> i know you're all excited for this
00:17:32 <ais523> izabera: I was going to say "that seems like a pretty eso use of dd"
00:17:38 <ais523> then realised there was a dd-based esolang already
00:19:08 <ais523> it is basically Bourne sh (not bash!) where the only program you're allowed to shell out to is dd
00:19:21 <ais523> the CLC-INTERCAL maintainer wrote an editor in it
00:20:46 -!- hppavilion[1] has joined.
00:21:35 <pikhq> ais523: Does that work in POSIX-compliant sh, or only the historical Bourne sh?
00:21:43 <pikhq> (they are not the same thing at all)
00:22:05 <ais523> http://dd-sh.intercal.org.uk/
00:22:13 <pikhq> Sadly, ddsh is not a very googlable string.
00:22:16 <ais523> looks like the language is defined wrt Bourne sh specifically
00:23:50 -!- hppavilion[2] has quit (Ping timeout: 272 seconds).
00:25:23 <pikhq> From what I can see, it looks like it's also valid POSIX sh.
00:25:58 <pikhq> Which makes sense as the differences between Bourne and POSIX sh are mostly in that there's a few unusual strings that in Bourne wouldn't do anything but in POSIX sh do something.
00:26:12 -!- hppavilion[1] has quit (Ping timeout: 244 seconds).
00:35:14 -!- MoALTz has quit (Ping timeout: 272 seconds).
00:44:08 -!- hppavilion[1] has joined.
00:49:26 -!- atrapado has quit (Quit: Leaving).
01:00:32 -!- augur has joined.
01:15:17 -!- hppavilion[1] has quit (Ping timeout: 244 seconds).
01:38:36 -!- augur has quit (Remote host closed the connection).
01:39:11 -!- augur has joined.
01:44:13 -!- augur has quit (Ping timeout: 252 seconds).
01:44:14 -!- tromp_ has joined.
01:47:27 -!- Phantom__Hoover has quit (Read error: Connection reset by peer).
01:55:29 -!- augur has joined.
02:02:05 -!- hppavilion[1] has joined.
02:02:49 -!- victo has joined.
02:03:12 -!- PinealGlandOptic has joined.
02:05:36 <HackEgo> Heck is where you end up if you don't believe in Gosh.
02:06:07 <hppavilion[1]> `learn Heck is where you end up if you don't believe in Gosh, or are darned for some other reason.
02:06:12 <HackEgo> Relearned 'heck': Heck is where you end up if you don't believe in Gosh, or are darned for some other reason.
02:06:49 -!- tromp_ has quit (Remote host closed the connection).
02:11:06 -!- victo has quit (Quit: Leaving).
02:11:28 -!- tromp_ has joined.
02:38:18 -!- jaboja has quit (Remote host closed the connection).
02:43:00 -!- tromp_ has quit (Remote host closed the connection).
02:56:03 -!- hppavilion[1] has quit (Ping timeout: 240 seconds).
02:56:35 -!- tromp_ has joined.
03:04:50 -!- ais523 has quit.
03:05:28 -!- ais523 has joined.
05:08:48 -!- ais523 has quit (Remote host closed the connection).
05:24:11 -!- Kaynato has quit (Ping timeout: 240 seconds).
05:34:57 -!- augur has quit (Remote host closed the connection).
05:37:48 <HackEgo> [wiki] [[Talk:FOSCode]] https://esolangs.org/w/index.php?diff=48925&oldid=48911 * Darkrifts * (+454) /* Answered question about loops */
05:38:01 <HackEgo> [wiki] [[Talk:FOSCode]] https://esolangs.org/w/index.php?diff=48926&oldid=48925 * Darkrifts * (+86)
05:40:51 <izabera> $ busybox dd conv=swab bs=3 status=none <<< 1234
05:40:53 <izabera> 213dd: can't swab 3 byte buffer
05:46:06 <HackEgo> [wiki] [[FOSCode]] https://esolangs.org/w/index.php?diff=48927&oldid=48910 * Darkrifts * (+147)
06:26:20 -!- augur has joined.
06:28:25 -!- PinealGlandOptic has quit (Quit: leaving).
06:33:55 <HackEgo> [wiki] [[FOSCode]] https://esolangs.org/w/index.php?diff=48928&oldid=48927 * Darkrifts * (+1313)
06:37:39 <HackEgo> [wiki] [[FOSCode]] M https://esolangs.org/w/index.php?diff=48929&oldid=48928 * Darkrifts * (+400) /* Documentation */
06:53:04 <HackEgo> [wiki] [[FOSCode]] M https://esolangs.org/w/index.php?diff=48930&oldid=48929 * Darkrifts * (+277)
07:20:24 -!- tromp_ has quit (Remote host closed the connection).
07:36:16 -!- hppavilion[1] has joined.
07:37:52 <Sgeo> So I've read that Pokemon TCG is cheap. But does it avoid pay-to-win? Or at least have a short ceiling?
07:40:09 <Sgeo> And is the online version better software than MTGO? (I'm assuming yes because everyone says MTGO is terrible)
07:40:38 <shachaf> mtgo is terrible but have you considered that all software is terrible
07:41:45 <pikhq> MTGO is a special kind of terrible.
07:41:50 <pikhq> But yes, I have considered that.
07:42:51 <Sgeo> I forget why I stopped playing Hearthstone. Might have been influenced by my Hearthstone-obsessed coworker stopping
07:46:07 <shachaf> maybe it was the pay-to-win aspect
07:53:11 -!- AnotherTest has joined.
07:54:57 <Sgeo> A lot of RNG in Pokemon TCG
07:55:26 <pikhq> Sounds like an accurate TCG version of the RPG, then.
07:55:50 * Sgeo doesn't care about the RPG >.>
07:55:57 <pikhq> Pokemon, I love you, but stop with some of the randomness.
07:56:58 <pikhq> 1/8192 odds of something are a real mean thing to put in a game.
07:57:30 <pikhq> (odds of a Pokemon being shiny, pre-gen VI)
07:57:48 <shachaf> Well, who cares about shiny?
07:57:50 <pikhq> Gen VI tweaked it! ... now it's 2/8192.
07:57:52 -!- ^^v has joined.
07:57:54 <pikhq> shachaf: Lots of people.
07:58:27 <pikhq> Where it's purely cosmetic? Yes.
07:59:03 <shachaf> Let them make their money from hats.
07:59:27 <pikhq> That's my stance on it. I've caught some shiny Pokemon, but I am *not* gonna spend any time trying to find 'em.
08:00:05 <pikhq> Keep in mind, the RPGs... don't have anything like a pay-money-to-get-things mechanic.
08:00:31 <pikhq> Aside from, well, buying the games.
08:02:31 <Sgeo> Hmm. This tutorial would be terrible for colorblind non-Pokemon players
08:02:40 <Sgeo> It's asking me to find the energy it asks for
08:02:45 <Sgeo> http://www.pokemon.com/us/pokemon-tcg/play-online/tutorial/
08:03:00 <Sgeo> Oh it's actually hilighting the correct answer
08:03:05 <Sgeo> I think this tutorial is for kids
08:03:25 <pikhq> That is the primary audience of nearly everything in the franchise.
08:12:31 -!- hppavilion[1] has quit (Ping timeout: 252 seconds).
08:12:51 -!- ^^v has quit (Ping timeout: 240 seconds).
08:21:16 -!- tromp_ has joined.
08:22:03 -!- AnotherTest has quit (Ping timeout: 264 seconds).
08:25:11 -!- LKoen has joined.
08:26:02 -!- tromp_ has quit (Ping timeout: 260 seconds).
08:34:19 -!- hppavilion[1] has joined.
08:44:59 -!- AnotherTest has joined.
09:08:29 -!- augur has quit (Remote host closed the connection).
09:13:27 -!- ^^v has joined.
09:18:18 -!- ^^v has quit (Ping timeout: 276 seconds).
09:25:07 -!- hppavilion[1] has quit (Ping timeout: 252 seconds).
09:25:25 -!- hppavilion[1] has joined.
09:42:21 <hppavilion[1]> That's, like, the kind of thing I use as joke fodder
10:09:07 <hppavilion[1]> Are there any cthonic prayers besides Ph’nglui mglw’nafh Cthulhu R’lyeh wgah’nagl fhtagn?
10:21:23 -!- tromp_ has joined.
10:25:34 -!- tromp_ has quit (Ping timeout: 250 seconds).
10:44:01 <LKoen> hppavilion[1]: he obviously doesn't have the same background as you do. what's an overused joke to you might be a great serious innovation to him
10:44:33 <hppavilion[1]> LKoen: Yes, but bragging about humbility is absurd in any background
10:45:05 <LKoen> well, obviously he hasn't had the same training as you in the subtle art of logical reasoning
10:45:31 <LKoen> or maybe he's not quite sure of the meaning of humility but thought it sounded cool
10:57:55 -!- uberBear has joined.
11:02:52 -!- uberBear has quit (Quit: Leaving).
11:19:03 -!- hppavilion[2] has joined.
11:22:39 -!- hppavilion[1] has quit (Ping timeout: 264 seconds).
11:34:16 -!- boily has joined.
11:34:39 -!- hppavilion[2] has quit (Ping timeout: 264 seconds).
11:37:44 <fizzie> I think I'll change the esolangs.org CAPTCHA back to the Befunge one, to see if ais523's new filter is catching the spam. Now that I'm mostly around to monitor the situation.
11:57:43 -!- LKoen has quit (Remote host closed the connection).
12:06:34 -!- LKoen has joined.
12:09:19 -!- atrapado has joined.
12:26:59 <int-e> tromp: btw I convinced Isabelle that your CL fixed point combinator is a fixed point combinator, modulo a tiny encoding trick (I'm using a fresh constant instead of an arbitrary CL term for the equation `Tx = `x`Tx, and I have not proved the corresponding closure of rewriting under substitution)
12:30:02 -!- boily has quit (Quit: CONTEST CHICKEN).
12:38:04 -!- tromp_ has joined.
12:39:06 -!- MoALTz has joined.
12:42:56 -!- tromp_ has quit (Ping timeout: 272 seconds).
13:04:03 -!- AnotherTest has quit (Ping timeout: 264 seconds).
13:20:36 -!- AnotherTest has joined.
13:39:23 -!- Vorpal has joined.
13:39:23 -!- Vorpal has quit (Changing host).
13:39:23 -!- Vorpal has joined.
14:05:51 -!- tromp_ has joined.
14:20:35 -!- oerjan has joined.
14:24:59 <HackEgo> [wiki] [[Talk:Spoon]] https://esolangs.org/w/index.php?diff=48932&oldid=37343 * Erikkonstas * (+113) /* Interpreter? */ new section
14:29:19 -!- tromp_ has quit (Remote host closed the connection).
14:47:47 <tromp> int-e: i didnt realize that was a challenge. given how easy it is with a few equations
14:48:00 <oerjan> <int-e> hmmm, Category: low-level added ... where was that policy about new categories? <-- Esolang:Policy hth
14:49:22 -!- Guest7241 has joined.
14:50:58 <lambdabot> ENVA 191320Z 25011KT CAVOK 18/11 Q1015 NOSIG RMK WIND 670FT 29008KT
14:51:15 <lambdabot> EGLL 191320Z AUTO 16011KT 9999 NCD 32/14 Q1016
14:51:52 <fizzie> There's air conditioning at work.
14:51:58 <fizzie> I may never leave the building.
14:52:17 <lambdabot> ESSB 191320Z 30006KT 270V350 CAVOK 22/07 Q1017
14:54:15 * oerjan is trying out the automated hydra.
14:54:41 <oerjan> > 50000 heads so far. it's really an astoundingly bad strategy.
14:54:43 <lambdabot> parse error (possibly incorrect indentation or mismatched brackets)
14:55:24 <oerjan> http://www.madore.org/~david/math/hydra.xhtml using the automated play
14:55:43 <oerjan> manually, i've made it in less than 300.
14:55:50 -!- LKoen has quit (Remote host closed the connection).
14:56:47 <oerjan> it always chooses the leftmost head, which is about the worst thing you can do because it almost never prunes the oldest branch.
14:57:37 <oerjan> and also, when it does, there will be a _lot_ of room for new junk to appear.
14:58:19 <int-e> tromp: it's not much of a challenge... I'm dabbling. But it's not completely trivial to incorporate a rewriting strategy into a proof... and I didn't want to list all the intermediate steps.
14:58:29 <oerjan> the second oldest branch is getting small though, so maybe soon another cut.
14:59:29 <int-e> tromp: so the fun part was to get the machinery to the point where http://lpaste.net/7970676195693953024 works.
14:59:37 <oerjan> hm the second oldest branch exploded, but has no direness left.
15:02:28 -!- AnotherTest has quit (Ping timeout: 250 seconds).
15:03:04 <oerjan> i guess i'm really talking about the oldest branch of the oldest branch.
15:03:13 <tromp> the challenge will be in proving that no smaller combinator works
15:05:23 <\oren\> Just cut off the highest leftmost head
15:05:39 <\oren\> highest first, then pickleftmost at that level
15:06:52 -!- AnotherTest has joined.
15:07:30 <FireFly> But direness is important to avoid it exploding in complexity
15:08:10 <\oren\> but dire heads are always right of normal ones
15:08:34 -!- LKoen has joined.
15:09:18 <oerjan> my manual strategy is basically never to take a dire one unless everything on top of it is pruned first. i may reconsider that, though.
15:09:48 <oerjan> (since i haven't tested what happens when i don't)
15:10:19 <oerjan> not my best but better than my average.
15:10:31 <\oren\> highest level, leftmost on that level
15:11:02 * oerjan will try something like that when this automated eventually stops
15:11:52 <oerjan> note that afaict a lot of the strategy depends on the program having a limit to how much junk it will produce simultaneously.
15:12:17 <oerjan> otherwise it would obviously go on for >>> heat death
15:24:04 <oerjan> ah now the oldest branch has only the dire subbranch left
15:29:52 -!- tromp_ has joined.
15:31:35 -!- MoALTz has quit (Max SendQ exceeded).
15:34:11 -!- tromp_ has quit (Ping timeout: 240 seconds).
15:41:49 <oerjan> <fizzie> I'll be going to both of those places, and they seem hot. :/ <-- on the basis of a single sample, i can confirm that boston is hot hth
15:51:45 -!- centrinia has joined.
15:55:34 <quintopia> the mathematical type of hydras eh
15:58:23 <oerjan> @tell ais523 afaict your new filter says it depends on title but actually depends on entire article, is that intended?
16:10:58 <oerjan> <ais523> fizzie: for future reference, the spambots were creating pages and renaming them in order to hide the pages from the page creation log <-- hm that might explain why you got the google warning
16:12:56 <oerjan> makes more sense than google happening to catch it during the spam itself.
16:14:14 <oerjan> <shachaf> Taneb: What's the advantage of Real Fast Nora's Hair Salon 3: Shear Disaster Download over Binary Lambda Calculus? <-- readability, duh
16:23:42 -!- centrinia has quit (Quit: Leaving).
16:31:16 <oerjan> FireFly: after the spam flood, fizzie got an email from google's spam detection whatever saying they'd tagged the wiki for spam.
16:32:39 <oerjan> <\oren\> this upsets me <-- just learn haskell and submit a pull request to fix it hth
16:33:41 <oerjan> int-e: do commands in lambdabot have access to the uncorrected command name they were called with?
16:34:08 <oerjan> if so, it should be relatively simple.
16:34:38 <oerjan> except maybe what to do if the i is missing.
16:35:26 <oerjan> maybe it should just s/i/o/ to the whole command
16:37:28 -!- LKoen has quit (Remote host closed the connection).
16:38:31 <int-e> oerjan: I don't think so
16:38:33 -!- hppavilion[1] has joined.
16:39:06 <int-e> but I have not checked...
16:39:32 <int-e> ...it's quite possible actually that they get access to the whole IRC message and then they could parse it again themselves
16:39:55 <oerjan> well that would break in @@ stuff.
16:40:00 -!- LKoen has joined.
16:40:17 <lambdabot> echo; msg:IrcMessage {ircMsgServer = "freenode", ircMsgLBName = "lambdabot", ircMsgPrefix = "oerjan!~oerjan@hagbart.nvg.ntnu.no", ircMsgCommand = "PRIVMSG", ircMsgParams = ["#esoteric",":@echo"]} target:#esoteric rest:""
16:40:32 <lambdabot> "echo; msg:IrcMessage {ircMsgServer = \"freenode\", ircMsgLBName = \"lambdabot\", ircMsgPrefix = \"oerjan!~oerjan@hagbart.nvg.ntnu.no\", ircMsgCommand = \"PRIVMSG\", ircMsgParams = [\"#esoteric\",\":@@ @show @echo\"]} target:#esoteric rest:\"\""
16:41:05 <oerjan> well, might be close enough.
16:42:51 -!- hppavilion[1] has quit (Ping timeout: 240 seconds).
16:45:55 -!- gamemanj has joined.
16:52:49 -!- SopaXorzTaker has joined.
16:53:05 <SopaXorzTaker> What was the language where I could redefine literals called?
16:53:34 -!- Kaynato has joined.
17:07:04 <quintopia> a few missteps. haven't got the strategy yet.
17:08:38 <quintopia> why doesnt that page give a proof? i've never seen a proof involving a hydra with dire heads.
17:11:31 <oerjan> i assume you have to define an ordinal for branches somehow.
17:16:48 <lynn> GolfScript famously lets you do that, too
17:17:32 <lynn> "3:2;" and then "2 2+" prints 6
17:18:36 <myname> quintopia: maybe there's a bachelor thesis for someone
17:20:13 <b_jonas> SopaXorzTaker: some dialects of FORTRAN I think
17:22:04 <oerjan> yay oldest branch finally got some pruning. which increased its direness, but anyway.
17:22:38 <int-e> oerjan: with the leftmost strategy?
17:22:39 <myname> i didn't get it in the first place
17:22:41 <oerjan> myname: surely it's a known result
17:22:52 <oerjan> int-e: the automatic one, yeah
17:23:04 <myname> i randonly clicked anywhere and made exactly one initial change
17:26:45 <int-e> nice to see some framework-free javascript code...
17:27:01 <int-e> (on the hydra page)
17:27:59 -!- SopaXorzTaker has quit (Remote host closed the connection).
17:28:05 -!- hppavilion[1] has joined.
17:28:32 <b_jonas> int-e: I think most of David's javascript toys (he has quite a few) are like that
17:29:13 <rntz> ... wrong channel
17:30:18 -!- tromp_ has joined.
17:30:33 <int-e> hmm, maybe I should make an svg version of http://int-e.eu/~bf3/squares/view.html :)\
17:32:10 <int-e> and maybe a version that uses the history less aggressively, hmm
17:33:48 <int-e> b_jonas: there's a "back" link at the top with explanations... it's about tiling rectangles with the minimum number of squares.
17:34:01 <int-e> or rather, browsing the tilings.
17:34:37 -!- tromp_ has quit (Ping timeout: 252 seconds).
17:36:06 -!- AnotherTest has quit (Ping timeout: 258 seconds).
17:36:21 <LKoen> quintopia: 876 heads cut off in my first try
17:36:30 -!- uberBear has joined.
17:36:31 -!- uberBear has quit (Max SendQ exceeded).
17:36:43 <LKoen> apparently a normal head segment should only be cut if it's a top depth
17:38:56 <b_jonas> int-e: wait, when you say "Ed Pegg provided results up to n,m ≤ 160!" then the exclamation mark doesn't denote a factorial, right?
17:39:24 <b_jonas> I guess it can't, because there are much smaller numbers than factorial(160) in the next lines
17:39:51 -!- 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.”).
17:42:21 <rntz> > help + unhelp^2
17:42:22 <lambdabot> Not in scope: ‘help’Not in scope: ‘unhelp’
17:42:37 <lambdabot> ‘h’ (imported from Debug.SimpleReflect),
17:42:38 <lambdabot> help <command>. Ask for help for <command>. Try 'list' for all commands
17:44:34 <oerjan> rntz: the single letters have been defined as a kind of symbols, but there's no real symbolic arithmetic.
17:45:11 <lambdabot> help <command>. Ask for help for <command>. Try 'list' for all commands
17:45:14 <b_jonas> > text "help" + text "unhelp" ^ 2
17:45:16 <lambdabot> No instance for (Num Doc) arising from a use of ‘+’
17:45:16 <lambdabot> In the expression: text "help" + text "unhelp" ^ 2
17:45:21 <b_jonas> > text "help" + text "unhelp"
17:45:22 <lambdabot> No instance for (Num Doc) arising from a use of ‘+’
17:45:22 <lambdabot> In the expression: text "help" + text "unhelp"
17:45:27 <b_jonas> > var "help" + var "unhelp"
17:45:33 <b_jonas> > var "help" + var "unhelp" ^ 2
17:45:50 <b_jonas> > var "help" + var "unhelp" ^ var "exponential"
17:46:00 <b_jonas> > var "help" + var "unhelp" ^^ var "exponential"
17:46:09 <b_jonas> > var "help" + var "unhelp" ** var "real exponential"
17:46:30 <oerjan> ^ doesn't work because it's not a method so cannot be defined sensibly for its right argument of Expr type
17:47:18 <oerjan> (it actually does arithmetic and comparisons to 0 or 1 that always fail, thus causing an infinite loop.)
17:47:33 <b_jonas> well, it does work in the sense of
17:47:50 <b_jonas> > var "one more than " ^ 9 + "one"
17:47:51 <lambdabot> Couldn't match expected type ‘Expr’ with actual type ‘[Char]’
17:47:51 <lambdabot> In the second argument of ‘(+)’, namely ‘"one"’
17:47:51 <lambdabot> In the expression: var "one more than " ^ 9 + "one"
17:48:14 <b_jonas> > (var "one more than ") ^ 9 + var "zero"
17:48:15 <lambdabot> one more than * one more than * (one more than * one more than ) * (one m...
17:48:53 <lambdabot> x * x * (x * x) * (x * x * (x * x)) * x
17:49:12 <fizzie> Interesting selection of parens there.
17:49:30 <lambdabot> * * ( * ) * ( * * ( * )) * ( * * ( * ) * ( * * ))
17:49:31 <FireFly> Indeed, I was expecting a foldl or so but I guess it's slightly more clever
17:49:39 <oerjan> fizzie: it treats * as left associative
17:49:45 <rntz> > let a = x * a in a
17:49:46 <lambdabot> x * (x * (x * (x * (x * (x * (x * (x * (x * (x * (x * (x * (x * (x * (x * (x...
17:50:03 -!- hppavilion[1] has quit (Ping timeout: 240 seconds).
17:50:06 <b_jonas> oerjan: which is good, because it is left associative
17:50:11 <oerjan> FireFly: it's a divide by 2 algorithm
17:50:26 <lambdabot> y * y * (y * y) * (y * y * (y * y)) * y
17:50:31 <oerjan> exponentiation by squaring
17:51:41 <fizzie> oerjan: Oh, so it's "really" just (((y * y) * (y * y)) * ((y * y) * (y * y))) * y?
17:52:08 <oerjan> and the identical subexpressions are shared.
17:52:22 <oerjan> they're just not printed that way.
17:53:15 <oerjan> huh it doesn't even do arithmetic when they're actual number literals
17:54:28 <oerjan> oh it's precisely textual comparison
17:55:35 <oerjan> > var "0.0" == 0 -- for completeness
18:00:11 <rntz> > let x = show x in take 10 . map ((+1) . length) . filter (/= "\"") $ group x
18:04:15 <shachaf> you figured it out too quickly
18:06:09 <lambdabot> <no location info>: not an expression: ‘--1’
18:06:13 <lambdabot> <hint>:1:3: parse error on input ‘-’
18:06:57 <oerjan> rntz: different type, actual Integers behave sensibly.
18:07:15 <oerjan> but a and z are of type Expr.
18:07:19 <rntz> oh, because ghc hard-codes defaults for literals
18:07:26 <rntz> as opposed to being like "which Num did you mean?"
18:07:39 <oerjan> rntz: no, that's not it
18:07:53 <oerjan> it does in fact depend on which Num you mean
18:07:59 <rntz> yes, that's what I mean
18:08:02 <oerjan> > (-1 :: Expr, -1 :: Integer)
18:08:25 <rntz> yes, but if you don't specify a type it assumes Integer unless you say otherwise
18:08:32 <rntz> *unless* it has to unify with something else that isn't integer
18:09:03 <oerjan> there's a warning you can enable for it
18:11:30 <oerjan> rntz: also it's not hardcoded either, you can change which types are defaulted to.
18:11:44 <oerjan> it's just not very common afaik
18:12:10 <shachaf> The default default is hardcoded.
18:12:24 <oerjan> default (Complex Double, Ratio Int) should be a hit
18:12:57 <oerjan> wait, add Integer to the end.
18:14:58 <oerjan> or wait, Ratio Int would probably never be chosen then.
18:15:04 <shachaf> rntz: You read _Stories of Your Life and Others_ by Chiang, right?
18:15:21 <shachaf> oerjan: Obviously CReal is the best choice.
18:15:59 <oerjan> no wait, Ratio Int can be chosen when RealFrac is needed
18:16:00 <shachaf> It's even compatible with the FFI.
18:16:16 <int-e> b_jonas: It's an ordinary exclamation mark.
18:16:19 <shachaf> Which C type does it correspond to?
18:16:27 * oerjan swats shachaf across the channel -----###
18:17:16 <gamemanj> is this channel one-dimensional or two?
18:17:25 <shachaf> in this channel, it's swat or be swatted
18:17:45 <HackEgo> #esoteric is the only channel that doesn't exist. After monqy left it became slightly off-centër. It's about 30 m (100 ft) across. oerjan seems to be making a lawn in the northern part, but it keeps getting dug up by free ranging moons. May contain crude drawings of nuts.
18:18:14 <oerjan> i think that implies at least 4.
18:18:48 <gamemanj> it implies there is time, and "north"
18:19:02 <gamemanj> So 3 dimensions, including time
18:19:27 <shachaf> oerjan: does #esoteric exist in minscowski space
18:19:37 <oerjan> i suppose it doesn't quite prove there's an east-west direction.
18:20:04 <gamemanj> I mean, you could imply it I guess based on that east/west would be the "default" for such things?
18:20:10 -!- hppavilion[1] has joined.
18:20:12 <oerjan> of course, since it doesn't exist, it has all the dimensions you want.
18:21:16 <int-e> b_jonas: the reason was that he had the program running for months.. which surprised me at the time
18:21:54 <oerjan> afk until the evening sun sets below the neighbors
18:22:31 <lambdabot> Local time for oerjan is Tue Jul 19 19:22:31 2016
18:23:54 <oerjan> it's particularly annoying now when it's right above the roof there
18:26:32 <quintopia> oerjan: re hydra: i suppose an optimal strategy must involve knowing what rules madore uses to "keep the screen from getting too cluttered". maybe strategies like getting the tree to take up as much horizontal space as possible at a height where it isn't too painful?
18:27:32 -!- hppavilion[2] has joined.
18:30:31 -!- hppavilion[1] has quit (Ping timeout: 240 seconds).
18:31:47 <b_jonas> quintopia: for an optimal strategy, you might want to decide what your goal is at first. is your goal to kill the hydra fast because you're Hercules, or to kill the hydra without using automation so that king Eurystheus can't find an excuse, or to keep the hydra alive for as long as possible to find an apparent counterexample to the hydra theorem, etc?
18:35:09 <\oren\> is the symbol for the integer gird (between 3 and 4) in Unicode
18:37:33 <Vorpal> b_jonas, this sounds interesting. How much chat log do I need to read to find out what the question is?
18:39:23 <Vorpal> Deewiant, With regards to the calendar in DATE, it seems my code explicitly skips year 0. But you could aruge that astronomical date numbering (which includes year 0) should be used instead
18:39:29 <lambdabot> http://www.madore.org/~david/math/hydra.xhtml
18:40:24 <shachaf> I was trying to remember who Madore was.
18:40:26 <b_jonas> Vorpal: I think it concerns David Madore's game http://www.madore.org/~david/math/hydra.xhtml which is related to the Hydra theorem which is one of those often mentioned few examples of easily mentionable theorems provable in ZFC but known to be not provable in Peano Arithmetic,
18:40:36 <b_jonas> and is related to ordinals somehow.
18:40:38 <shachaf> But now I remember. Unlambda.
18:40:55 <b_jonas> shachaf: and more. I think I even added a page on esowiki listing him or something.
18:41:08 <Vorpal> Deewiant, I assume it is unspecified?
18:41:54 <b_jonas> shachaf: the recent interesting developments in this area about ordinals is https://johncarlosbaez.wordpress.com/2016/07/07/large-countable-ordinals-part-3/#comment-81550 , comments on John Baez's blog where Baez wrote three entries on topics very similar to what David had written about earlier,
18:42:46 <b_jonas> I'm still trying to untangle what happened in the comments there, because there are too few links, so it's not obvious which of David's articles are relevant
18:43:00 <int-e> chopping merrily away at a speed of 200 heads per second...
18:43:08 <Vorpal> Deewiant, I was also looking at handling negative julian day numbers, and that doesn't seem to work too well either. When it is less than about year -7000
18:43:29 <Deewiant> Vorpal: Mh. Given that it only says "Gregorian calendar is assumed for calendar dates" it seems unspecified, yes.
18:44:00 <Vorpal> DATE is an rcfunge one?
18:44:17 <int-e> Hmm I guess leftmost is not the most stupid thing one can do.
18:44:35 <b_jonas> but at least http://www.madore.org/~david/weblog/d.2008-03-16.1534.ordinaux-et-hydres.html (about the hydra), http://www.madore.org/~david/weblog/d.2015-11-16.2337.html , http://www.madore.org/~david/weblog/d.2013-01-16.2104.grands-nombres.html , http://www.madore.org/~david/weblog/d.2011-09-18.1939.nombres-ordinaux-intro.html , http://www.madore.org/~david/weblog/d.2011-10-02.1946.html are relevant
18:44:58 <b_jonas> um, was that truncated? http://www.madore.org/~david/weblog/d.2013-01-16.2104.grands-nombres.html , http://www.madore.org/~david/weblog/d.2011-09-18.1939.nombres-ordinaux-intro.html , http://www.madore.org/~david/weblog/d.2011-10-02.1946.html are relevant
18:45:31 <shachaf> Now I feel like b_jonas has posted links to madore.org before.
18:45:32 <Vorpal> Deewiant, Yeah I hope I remember to fix it to use the same calendar in both directions before you get around to testing that though
18:45:38 <shachaf> But they were in French so I couldn't read them.
18:45:49 <b_jonas> and wait, I forgot the ordinal browser (whcih is linked from Baez's post already)
18:45:52 -!- ^v has quit (Ping timeout: 258 seconds).
18:46:11 <b_jonas> no wait, http://www.madore.org/~david/weblog/d.2011-10-02.1946.html is the ordinal browser
18:46:28 <b_jonas> shachaf: http://esolangs.org/wiki/David_Madore
18:46:52 -!- lambdabot has quit (Remote host closed the connection).
18:47:09 <b_jonas> shachaf: I'm a big fan of David Madore's blog really. I've been reading it for a very long time, but mind you, I haven't seen like half of it because that blog is OLD,
18:47:34 <b_jonas> and for a while it's on the top of one of my external links page http://math.bme.hu/~ambrus/sc/grn on my homepage
18:47:55 <oerjan> quintopia: "getting the tree to take up as much horizontal space as possible at a height where it isn't too painful" sounds like about what i've been trying.
18:47:58 <Vorpal> Anyone know if there is any property based testing framework (quickcheck style that is) for C?
18:48:12 <Vorpal> I know there is one for C++ called rapidcheck, but for C I don't know of any
18:49:03 <shachaf> not sure about this disclaimer
18:49:13 -!- ^v has joined.
18:49:42 <b_jonas> And there are also at least two old articles on this topic that are probably completely superceded and that I probably didn't link
18:50:13 <oerjan> <\oren\> is the symbol for the integer gird (between 3 and 4) in Unicode <-- yes, but good luck writing down its codepoint number hth
18:50:17 <b_jonas> He also has several useful maths articles on the blog that aren't connected to ordinals and infinity
18:50:34 <Vorpal> b_jonas, is the hydra problem solvable?
18:51:00 <Vorpal> shachaf, the one you linked
18:51:06 <Vorpal> <lambdabot> http://www.madore.org/~david/math/hydra.xhtml
18:51:17 <shachaf> The hydra always dies eventually, if that's what you mean.
18:51:22 <Vorpal> shachaf, defeating it? According to the title
18:51:41 <shachaf> It says at the bottom of the page.
18:51:42 <b_jonas> Vorpal: solvable in what sense?
18:51:43 <Vorpal> I only fooled around a bit with it, and it just seemed to keep growing
18:52:04 <b_jonas> but it is a game that always ends in finite number of steps, no matter what you do
18:52:07 <shachaf> @google goodstein's theorem
18:52:19 <b_jonas> shachaf: ah, that's what it's called
18:52:37 <shachaf> It's not exactly the hydra but it's similar.
18:52:59 <b_jonas> rumour is, this is connected to how a descending sequence from naturals to ordinals is always eventually constant
18:53:08 <b_jonas> shachaf: but what's this about dire hydras?
18:53:22 <b_jonas> how are those different from non-dire hydras?
18:53:43 <b_jonas> (and are there werehydras, celestial and infernal hydras, giant hydras, half-elemental hydras, etc?)
18:53:53 <int-e> oerjan: have you seen the automatic strategy finish a hydra?
18:54:10 <oerjan> <int-e> Hmm I guess leftmost is not the most stupid thing one can do. <-- it isn't?
18:54:19 <b_jonas> http://www.madore.org/~david/weblog/d.2008-03-27.1537.html is the actual entry for the hydra game!
18:54:26 <b_jonas> I've been looking for that for a while
18:55:23 -!- lambdabot has joined.
18:55:36 <shachaf> b_jonas: But it's in French, and therefore meaningless.
18:56:16 <oerjan> <shachaf> But they were in French so I couldn't read them. <-- rings a bell.
18:56:24 <int-e> oerjan: it just seems that by strategically picking dire branches deep down the tree when the treep is deep should make things worse.
18:56:50 <int-e> oerjan: but I see that this would prevent those from being duplicated later. It's a tough question.
18:56:51 <shachaf> oerjan: It was something about Hebrew grammar, maybe?
18:57:04 <oerjan> as the sun sets, the (other) neighbors start up their subwoofer :(
18:57:49 <int-e> Anyway... The hydra currently has: 29 segments, 9 heads, and depth 7. Hercules has cut 175347 heads so far. ... still looking quite healthy.
19:03:17 -!- AnotherTest has joined.
19:04:45 <oerjan> <int-e> oerjan: have you seen the automatic strategy finish a hydra? <-- not YET hth (it's currently paused at step 56888 although it _did_ make progress on the oldest dire branch around 56000.)
19:06:02 <oerjan> <b_jonas> how are those different from non-dire hydras? <-- i assume the latter don't have the dire segments from the game?
19:07:39 <oerjan> and also, without the dire segments the proof looks easy: just give a branch an ordinal equal to { sum(omega^ordinal b) | b child branch }
19:08:25 <oerjan> ("commutative" sum, i guess)
19:10:13 <oerjan> int-e: btw i haven't experimented much with trying to cut dire branches eagerly, so i'm not sure if that would actually be worse.
19:10:42 <oerjan> shachaf: no, it was about not reading madore blog links hth
19:11:31 <oerjan> (24 segments, 11 heads and depth 5 here btw)
19:13:26 <oerjan> i don't think those numbers are that important. i think the "ordinal" size of the oldest branch is what really determines progress, the rest is just temporary junk that _will_ disappear as long as there's enough of it.
19:13:36 <quintopia> cutting dire branches eagerly is p good, except that it ends up making tall tall green branches, which take forever to clear
19:14:55 <oerjan> i do suppose using more height means there's more room for junk, which would slow it down.
19:15:59 <oerjan> assuming madore's algorithm is about actual fitting into space somehow, rather than just total number of segments or the like.
19:16:36 <oerjan> hm and i should be going
19:17:05 <quintopia> if a bettor can bet, why can't a debtor debt?
19:17:28 -!- oerjan has quit (Quit: Later, sorry).
19:20:17 <\oren\> quintopia: because english is a crappy language that hopefully will die soon
19:21:30 <HackEgo> [wiki] [[FOSCode]] https://esolangs.org/w/index.php?diff=48933&oldid=48930 * Darkrifts * (+111) /* Arguments */
19:21:51 <HackEgo> [wiki] [[FOSCode]] https://esolangs.org/w/index.php?diff=48934&oldid=48933 * Darkrifts * (+12) /* Arguments */
19:21:53 <\oren\> I bet the programming languages and libraries with english naming schemes will outlive the spoken english language
19:22:14 <Vorpal> \oren\, oh? What language do you think will replace English?
19:22:43 <\oren\> Vorpal: whatever English evolves into.
19:22:57 <Vorpal> \oren\, and is there any indication that is ongoing currently?
19:23:42 <Vorpal> I highly doubt English in one way or another will go away any time soon. Going to take several generations.
19:24:57 <\oren\> The fact I can't really understand certain American English dialects, indicates that the spoken language is splitting
19:25:25 <Vorpal> Which sort of English do you speak then?
19:25:53 <shachaf> I think that if anything the language is unsplitting.
19:26:49 <\oren\> History seems to indicate that written languages can remain the same even after noone speaks their spoken version.
19:27:13 <Vorpal> I'm not a native speaker and I managed to understand most variants of English I came into contact with. Some are harder than others yes. I find Indian English especially difficult, and I have to deal with that (over video conference systems) at work.
19:27:29 <int-e> The hydra currently has: 34 segments, 10 heads, and depth 9. Hercules has cut 500325 heads so far. ... still going.
19:28:47 <shachaf> isn't it easily going to go to ackermann numbers
19:29:19 <shachaf> ackermann, ackermann, does whatever an acker cann
19:29:41 -!- augur has joined.
19:29:43 <int-e> shachaf: I have not done any anlysis... for all I know this won't finish before the heat death of the universe
19:31:05 <int-e> ah the rightmost subtree changed.
19:32:07 <int-e> http://int-e.eu/~bf3/tmp/hydra.png "proof"
19:33:01 <int-e> (maybe oerjan can predict how long this will still last ;)
19:33:45 <int-e> re: "proof" - the only thing I changed in the html file is the interval timer (from 750 to 5 milliseconds)
19:35:10 <int-e> hppavilion[2]: what's up with the nick...
19:35:22 -!- hppavilion[2] has changed nick to hppavilion[1].
19:35:28 -!- jaboja has joined.
19:38:32 <HackEgo> [wiki] [[FOSCode]] https://esolangs.org/w/index.php?diff=48935&oldid=48934 * Darkrifts * (+1062) /* Documentation */
19:41:08 <HackEgo> [wiki] [[FOSCode]] https://esolangs.org/w/index.php?diff=48936&oldid=48935 * Darkrifts * (+159)
19:44:15 <HackEgo> olist 1045: shachaf oerjan Sgeo FireFly boily nortti b_jonas
19:44:52 <int-e> shachaf: actually since the tree size is more or less bounded (thanks to a "magic" probability), this implementation can only be exponential in that target tree size...
19:44:52 <shachaf> HackEgo: stop being so slow
19:45:28 <int-e> I wonder whether HackEgo being slow is HackEgo's fault or CoC's fault (they may be putting a lot of VMs on the same machine...)
19:45:53 <int-e> ah... I was waiting for this... touch: cannot touch ‘a’: Read-only file system
19:46:20 <int-e> (previously: [3193390.301619] Aborting journal on device dm-0-8.)
19:47:14 <Vorpal> int-e, that sounds really really bad?
19:47:51 <int-e> Vorpal: that's my usual experience with CaC (oops, sorry, bad typo/thinko there)
19:48:48 <int-e> (Call of Cthulhu really doesn't deserve to be confused with Cloud at Cost.)
19:51:27 -!- wob_jonas has joined.
19:52:00 <int-e> let's see if it still boots :P
19:52:25 -!- augur has quit (Quit: Leaving...).
19:52:30 <wob_jonas> oerjan: but, um, what's a "dire segment" then?
19:52:57 <int-e> it's explained on http://www.madore.org/~david/math/hydra.xhtml
19:58:38 <HackEgo> [wiki] [[FOSCode]] https://esolangs.org/w/index.php?diff=48937&oldid=48936 * Darkrifts * (+301) /* Added section on errors */
20:04:00 <int-e> nice, it booted but the web server seems to have suffered... such fun...
20:07:35 <int-e> ... http://tunes.org/~nef/logs/esoteric/?C=M;O=D
20:12:49 <Vorpal> int-e, Cloud at Cost? What is that
20:13:32 <pikhq> Rather cheap VM host.
20:13:42 <Vorpal> One time fee? How on earth does that work
20:14:17 -!- augur has joined.
20:14:26 <Vorpal> pikhq, yeah especially given what I pay for a not-quite-lowest-end linode, that price as a one-time fee is insane
20:14:36 <Vorpal> Mind you, my linode is *rock solid*
20:14:40 <int-e> shachaf: well, they just don't buy any new hardware is what I think.
20:14:44 <shachaf> I should probably switch away from Linode.
20:14:55 <int-e> also, they have NO abuse management whatsoever
20:15:01 <Vorpal> shachaf, what is the issue with linode?
20:15:04 <wob_jonas> Is it one of those where you have a monthly cap of network use, if you transfer more data than that cap any month then you pay an expensive overuse fee, and there's no way to just ask in advance that you don't want to pay any fees so if you get close to the cap they should automatically shut off the network access or slow it down to very slow?
20:15:08 <shachaf> They have all these security issues, for one.
20:15:14 <int-e> no people involved (just a few developers) = cheap.
20:15:17 <Vorpal> shachaf, oh? I have not heard about that
20:15:34 <int-e> you get what you pay for, or perhaps a bit less ;)
20:15:35 <shachaf> I often hear negative things about them.
20:15:41 <Vorpal> int-e, still the electricity, IP block assignment and all other running costs?
20:16:01 <shachaf> "just a few developers" doesn't sound very cheap
20:16:03 <int-e> Vorpal: as long as they find new customers, they will pay for that
20:16:16 <Vorpal> int-e, that sounds like a pyramid scheme
20:16:19 <wob_jonas> int-e: oh, so it's a pyramid scheme
20:16:21 <int-e> Vorpal: and from I read their main business is a telco one.
20:16:31 <shachaf> I think it's just a Ponzi scheme.
20:16:44 -!- tromp_ has joined.
20:16:45 <pikhq> Yeah. I bought one a while back, but I have *not* been impressed.
20:16:46 <Vorpal> well okay, pyramid, ponzi, something like that
20:17:04 <pikhq> Thank goodness it's not an ongoing cost?
20:17:09 <int-e> pikhq: I have... the QoS is soooo bad!
20:17:16 <Vorpal> shachaf, what better option is there to linode then? Same price class, same performance class
20:17:22 <shachaf> Do you get money for your referrals' referrals or something?
20:17:35 <int-e> The ongoing cost for this channel is that HackEgo is becoming increasingly slower.
20:17:36 <Vorpal> shachaf, I looked at one that seemed really nice, that fizzie used. Forgot the name of it. Until I realized the price was in euros
20:17:49 <int-e> (AFAIK it's on one of the CaC VMs)
20:18:04 <shachaf> Or maybe AWS/Google Cloud/Microsoft Azure.
20:18:08 <shachaf> I don't remember how the pricing compares.
20:18:10 <Vorpal> shachaf, I heard shit things about DO
20:18:17 <pikhq> I wonder if there's a Raspberry Pi colo. That might be an upgrade. :P
20:18:22 <Vorpal> AWS or such is probably pretty good though
20:18:26 <FireFly> I'm on prgmr, they're pretty nice though I dunno how the pricing compares to other services
20:18:34 <shachaf> The reason I don't use Google Cloud is the free trial.
20:18:39 <Vorpal> pikhq, I heard of such a thing in Sweden. No space for external HDD. Also they sold out iirc
20:19:21 <Vorpal> Well prgmr has quite a spartan site don't they
20:20:06 <Vorpal> FireFly, seems slightly cheaper than linode from what I can see.
20:20:22 <shachaf> Vorpal: They give you $300 to spend on Google Cloud over 60 days.
20:20:24 <int-e> NICE. could not open mime types config file /etc/mime.types
20:20:31 <Vorpal> Ah, linode started switching from Xen to KVM recently
20:20:36 <shachaf> Vorpal: I don't want to squander it, so I haven't even started the trial.
20:21:00 <shachaf> Vorpal: Which is a shame, because I want to use the plain old virtual machine service, which I don't really need a trial for.
20:21:10 <shachaf> What I'd want a trial for is all the fancy Google-specific services.
20:21:12 -!- tromp_ has quit (Ping timeout: 260 seconds).
20:21:17 <shachaf> So really they should let me activate the free trial whenever I want.
20:21:30 <shachaf> In the end I'm not using it at all.
20:21:39 <shachaf> AWS and Azure have similar deals. :-(
20:21:50 <int-e> shachaf: where's the business sense in that? you'd be willing to pay them then...
20:21:59 <Vorpal> shachaf, anyway, so far I found linode to be really solid. And I guess you have as well?
20:22:02 <pikhq> Though, AWS's free trial is a year long and covers less.
20:22:19 <shachaf> int-e: I'm willing to pay them now, for the boring service that everyone provides.
20:22:33 <int-e> shachaf: I'm suggesting that letting you start the free trial when you want would be bad for business.
20:22:54 <int-e> Or perhaps I'm speculating. Or joking. I haven't decided.
20:22:57 <shachaf> Well, setting it up such that I don't even sign up is also bad for business.
20:23:05 <shachaf> Or maybe I'm just unusual.
20:24:02 <Vorpal> shachaf, I thought googles cloud only provided rather unusual environments? Like just python "web apps". Or is that google app engine? which is different??
20:24:13 <wob_jonas> shachaf: what you're saying sort of makes sense. I mean, it doesn't matter whether it's the boring services or the interesting ones, but if you start the free trial but only use the *cheap* services and so you don't spend all $300 before the timeout, then that might be a waste in some sense
20:24:31 <shachaf> Vorpal: You might be thinking of an older version of App Engine.
20:24:46 <shachaf> Vorpal: I think they still have that but they also have a lot more services.
20:24:51 <Vorpal> shachaf, can you set up your own mail server there? Or do you need to use their SMTP sever?
20:25:03 <Vorpal> because that was one thing for me. I want to run my own mail
20:25:18 <lambdabot> https://cloud.google.com/compute/docs/tutorials/sending-mail/
20:25:40 <shachaf> "Google Compute Engine does not allow outbound connections on ports 25, 465, and 587."
20:28:26 <Vorpal> shachaf, so not good for me then
20:28:44 <Vorpal> Since I run my own mail
20:29:04 <Vorpal> I specifically wanted to get away from google seeing it all through gmaik
20:29:47 <fizzie> 26.7 degrees inside. :/
20:30:09 <gamemanj> Vorpal: well, I tried running an email server on a home connection, and...:
20:30:18 <gamemanj> GMail accepts my mail but puts it in the spam folder
20:30:27 <fizzie> Vorpal: So far, I've used prgmr.com, tilaa.com and digitalocean.com (in that order) for my VPSen.
20:30:32 <gamemanj> I think Yahoo does the same as GMail IIRC
20:30:35 <wob_jonas> Vorpal: couldn't you use someone else (not gmail, but not self-ran) as a mail provider?
20:30:48 <Vorpal> gamemanj, right, which is why I have a proper VPS
20:31:33 <lambdabot> http://docs.aws.amazon.com/ses/latest/DeveloperGuide/send-email-smtp.html
20:31:33 <lambdabot> Title: Using the Amazon SES SMTP Interface to Send Email - Amazon Simple Email Service
20:31:39 <Vorpal> wob_jonas, I want to be in control of my own mail. That was the point. Then I might as well pay for a VPS and run other stuff on it too. Since all free email providers obviously need to have some other way to make me a profit
20:31:46 <Vorpal> (i.e. profiling for ads)
20:31:50 <shachaf> Vorpal: What, you don't trust Gmail?
20:32:24 <Vorpal> shachaf, oh I trust they will deliver my mail and provide a solid service. Certainly. But it is free. How do they make a profit from it do you think?
20:32:31 <wob_jonas> Vorpal: some ISPs give you a mail server as a bonus if you pay for net access, isn't that an option for you?
20:32:52 <Vorpal> wob_jonas, sure, but those services are shittier than shitty
20:33:08 <Vorpal> I think it is like 20 MB limit, POP only, no IMAP
20:33:11 <shachaf> Vorpal: (I used to work at Google on the team that runs Gmail.)
20:33:31 <Vorpal> shachaf, okay. Then you probably know how they make money from it, and can't comment on it
20:33:41 <Vorpal> But I assume it is from targeted ads
20:34:02 <Vorpal> And even with ad blocking, that profile they build of you can be used elsewhere
20:34:53 <shachaf> I don't think I can say anything that isn't public. But it's not a great business mystery how it works.
20:35:05 <wob_jonas> shachaf: I don't completely trust GMail either. It's not really the ads, but that they're a big company that can decide at any time that they don't need me a user and ban my gmail account for any reason, for they can afford some false positives when they're trying to ban actually malicious accounts, and if they do so, I won't be able to do anything
20:35:05 <wob_jonas> to get that email address back. I need a *stable* email address, one that is likely to work even ten years from now,
20:35:42 <shachaf> wob_jonas: Good reasoning.
20:35:45 <fizzie> GMail gives me that nasty red lock icon for emails that I send from my own mail server, I really should figure that out.
20:36:06 <shachaf> fizzie: Do you want me to write you an extension that disables the nasty red lock?
20:36:07 <wob_jonas> not necessarily at the same server, but the same address should work. ambrus@math.bme.hu is that stable address, since if that server goes down, I know who I have to bribe or threaten to fix the server, or to point the domain elsewhere if they really don't want to run the server,
20:36:10 <fizzie> Because I thought I set up opportunistic TLS on it, yet it's all "zem.fi did not encrypt this message".
20:36:25 <fizzie> shachaf: I'd rather just have it speak TLS to Google servers.
20:36:26 <Vorpal> fizzie, I believe that is due to lack of SSL
20:36:40 <shachaf> wob_jonas: whoa whoa whoa, is that your name?
20:36:57 <wob_jonas> and also think that even without me doing anything, it's very likely that mail on that domain will be served for a long time in some way, due to other more important people having their mail there.
20:37:10 <Vorpal> fizzie, set up so your server prefers sending and receiving to/from other servers with SSL when possible (opportunistic encryption)
20:37:23 <fizzie> Vorpal: <fizzie> Because I thought I set up opportunistic TLS on it, yet it's all "zem.fi did not encrypt this message".
20:37:23 <Vorpal> Oh yeah you said that below
20:37:35 <Vorpal> fizzie, and I just found that line :P
20:37:51 <fizzie> Possibly I just misconfigured it. Maybe it lacks root CAs or something.
20:38:03 <shachaf> fizzie: just use end-to-end encryption hth
20:38:12 <fizzie> I'm sure it's something simple, I just never remember to look into it.
20:38:53 <fizzie> My university gave me a "forever" alumnus email forwarding address.
20:39:03 <Vorpal> fizzie, something like this should work, replace your certificate paths of course http://termbin.com/ets2
20:39:16 <shachaf> I should switch my email address to a custom domain name instead of gmail.com
20:39:21 <shachaf> But I haven't decided which domain name.
20:39:23 <wob_jonas> So anyway, I do use gmail, but I give a non-gmail email address anywhere I want a future-stable address.
20:39:25 <Vorpal> I think there needs to be a few lines in master.cf as well
20:39:57 <HackEgo> [wiki] [[FOSCode]] https://esolangs.org/w/index.php?diff=48938&oldid=48937 * Darkrifts * (+164) /* Misc */
20:40:13 <lambdabot> Local time for fizzie is Tue Jul 19 20:40:13 2016
20:40:43 <fizzie> Vorpal: Oh, by the way -- I changed the esolangs.org captcha to be Befunge instead of brainfuck.
20:41:18 <Vorpal> fizzie, http://termbin.com/ujog is from master.cf. The only relevant line I could see
20:41:26 <HackEgo> Vorpal writes software for boring machines. Really big ones.
20:42:09 <Vorpal> Hm though that line appears to be about client to server
20:42:17 <Vorpal> So may not be relevant
20:42:27 <fizzie> Vorpal: Here's a description I wrote for someone else: It shows you a string of the form 9NNNNNNNNN>\#+:#*9-#\_$.@ (where each N is a random base-9 digit) and asks you to provide the number a Befunge interpreter would output when given that.
20:42:39 <wob_jonas> Most of the mail to my main gmail address is public mailing lists by the way.
20:43:10 <shachaf> FireFly gave me a Befunge puzzle.
20:44:24 <Vorpal> fizzie, any recent panorams btw?
20:44:28 <shachaf> Not very. But it was slightly harder than you implied.
20:45:14 <wob_jonas> fizzie: whoa, I can't read Befunge, I have no idea what those things mean...
20:45:33 <wob_jonas> I think # is a trampoline, but what the heck are those backslashes?
20:46:13 <Vorpal> wob_jonas, swap top items on stack
20:46:25 <Vorpal> been so long since I wrote a befunge interpreter
20:46:39 <wob_jonas> I was wondering if it's a mirror, but that wouldn't make sense there
20:46:56 <HackEgo> [wiki] [[FOSCode]] https://esolangs.org/w/index.php?diff=48939&oldid=48938 * Darkrifts * (+94) /* Errors */
20:47:39 <Vorpal> wob_jonas, befunge doesn't have those. It does have handle bars though for turning. [ and ]
20:47:55 <fizzie> Funge-98 has those. Befunge-93 doesn't.
20:48:03 <fizzie> wob_jonas: Backslash is a swap, yes.
20:48:34 <shachaf> did you invent any good maths lately
20:48:41 <fizzie> The >\#+:#*9-#\_$.@ part is essentially right swap jump plus dup jump mul 9 sub jump swap if drop print exit.
20:48:51 <wob_jonas> and guessing from the > and the trampolines there's probably a loop there that runs both left and right, oscillating
20:48:58 <fizzie> Where "if" goes right if 0, left otherwise.
20:49:00 <wob_jonas> but I don't know which part turns back
20:49:32 <FireFly> shachaf: I haven't mathsed at all recently, to be honest
20:49:36 <Vorpal> fizzie, wait a second... Isn't that program just going to ouput 0 and exit in the end? I see no output instruction prior to the _$.@
20:49:40 <fizzie> >..._ is the traditional oneliner loop, with judicious use of #s when you need to do different things in different directions.
20:49:51 <fizzie> Vorpal: Why would it output 0?
20:49:54 <shachaf> FireFly: what did you invent recently
20:50:39 <shachaf> did you invent any cute cat pictures maybe
20:50:47 <Vorpal> fizzie, Hm... Where is the output decimal except right before the @?
20:50:53 <FireFly> I didn't; that's more zgrep's department
20:51:07 <fizzie> Vorpal: Nowhere else, but the top of stack is not 0 at the .
20:51:11 <Vorpal> Oh wait I see now. Because it checks if there is a swapped \
20:51:18 <shachaf> zgrep: please provide kitten pics twh
20:51:29 <zgrep> shachaf: zgrep.org/cats.html
20:51:53 <zgrep> hint: click space, also try the arrow keys if you want to
20:51:53 <fizzie> (It's just a base-9-to-base-10 converter, essentially.)
20:52:44 <wob_jonas> fizzie: yes, I guess it's something like that, but does it also take a 9's complement for converting, or is that only for checking the 9 at the bottom?
20:52:50 <shachaf> zgrep: Seems like a nonobvious UI.
20:53:02 <shachaf> zgrep: You should add a spot of text to it that says how to use it.
20:53:20 <zgrep> shachaf: But then there will be a spot of text that says how to use it.
20:53:32 <shachaf> Sure, but it can disappear after you press space.
20:53:40 <HackEgo> [wiki] [[FOSCode]] https://esolangs.org/w/index.php?diff=48940&oldid=48939 * Darkrifts * (+428) /* Added section on creating for-like constructs */
20:53:54 <FireFly> shachaf: I did find a silly bug yesterday, does that count as an invention
20:54:10 <FireFly> Actually I had already found the same bug elsewere before
20:54:33 <shachaf> was the bug that a program crashed on argc==0
20:54:40 <zgrep> shachaf: But that takes effort... I'm spending my effort on doing everything else that I'm doing right now.
20:55:19 <shachaf> zgrep: cats are so good though
20:55:28 <zgrep> shachaf: cats are better than instructions
20:55:28 <FireFly> No, the bug is that all innocents on Foresight items in Disgaea PC are level 1
20:56:14 <FireFly> Turns out it's because the item description field in one of the data files exceeds the width that the field is supposed to be
20:56:36 <FireFly> so it overwrote the next column with (string-terminating) 0, which isn't even a legal value for that column
20:57:18 <Vorpal> FireFly, I concur with wob_jonas here, you need to give some context?
20:57:19 <FireFly> The PSP port of the game had a similar bug, but even worse because the bytes that went in the column were part of the text string.. which caused even weirder behaviour
20:57:32 <shachaf> zgrep: what if it showed more than one cat at a time
20:57:37 <Vorpal> FireFly, what is this game
20:57:49 <lambdabot> https://en.wikipedia.org/wiki/Disgaea
20:58:01 <zgrep> shachaf: Feel free to make thy own ui. It just grabs a URL from zgrep.org/cats . :P
20:58:12 <Vorpal> ah, tactical RPG. It makes very slightly more sense now
20:59:02 <FireFly> https://twitter.com/FireyFly/status/710549067089252352 the "Rank" column is supposed to only take on values in [1..42]
20:59:18 <FireFly> (This is the older of the bugs)
21:00:44 <FireFly> I'm not sure if it'd be interesting to explain how it misbehaves since it essentially just boils down to numbers being off-the-charts
21:02:09 <FireFly> Anyway, you can see how the item description is cut off and the fields after it seem to have fishy values
21:02:35 -!- izabera has quit (Ping timeout: 244 seconds).
21:02:45 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: assemble: not found
21:05:52 <wob_jonas> did Primal Clay change at some point, and if so, was it a functional change?
21:08:07 <fizzie> I heard there's a cat-collecting easter egg in N.
21:10:29 <gamemanj> ...is that supposed to be an attempt to addict me to a game involving a ninja
21:10:38 <fizzie> Where I come from, uppercase single letters are always Android releases if at all possible given the context.
21:10:49 <fizzie> http://www.androidcentral.com/android-70-nougat-developer-preview-5-actually-comes-awesome-easter-egg
21:11:41 <wob_jonas> The change I'm looking for was at Shadowmoor.
21:12:16 <wob_jonas> But the bulletin totally doesn't explain anything.
21:12:22 <gamemanj> the food makes people hungrier for the releases...
21:12:33 <fizzie> Because the name isn't made public until pretty late.
21:12:39 <wob_jonas> Is it no longer an effect that changes the copiable values?
21:12:56 <gamemanj> What happens when they reach "Z"?
21:13:01 <gamemanj> They'll run out of foodstuffs.
21:13:21 <fizzie> Android Zucchini, very tasty.
21:13:55 <fizzie> Maybe it'll wrap. Did Ubuntu already run out?
21:14:01 <wob_jonas> As in, what happens now if you Cytoshape from Primal Clay as the template?
21:14:18 <fizzie> Apparently they're at Yakkety Yak now.
21:14:47 <fizzie> Or will be, for 16.10.
21:14:49 <fizzie> So they still have Z for the 17.04 release, but then they're out.
21:15:45 <wob_jonas> Maybe they'll use letters other than ascii ones.
21:15:56 <wob_jonas> Or they'll follow with a left square bracket.
21:16:48 <FireFly> gamemanj: the N game is good too, though
21:17:31 <Vorpal> <wob_jonas> "Yakkety"? Seriously? <-- is that a real word??
21:17:52 <fizzie> Not according to WordNet. But I'm sure it's colloquial.
21:18:28 <fizzie> I've missed all these recent ones. Utopic Unicorn, Vivid Vervet, Wily Werewolf, Xenial Xerus.
21:19:17 <Vorpal> fizzie, well since I go on LTS I miss most of them
21:19:26 <wob_jonas> How big is the Toy Story franchise? How long till Debian runs out of release codenames?
21:24:15 * gamemanj prepares a thwomp in front of firefly, and waits anxiously with a button...
21:24:34 -!- hppavilion[1] has quit (Ping timeout: 258 seconds).
21:24:50 <FireFly> oerjan usually sticks to a good ol' swatter
21:25:10 <shachaf> Did you play "Spider and Web"?
21:25:11 <gamemanj> if you're going to turn someone into a flattened mess, you have to do it with class!
21:25:13 <shachaf> That's a pretty good game.
21:26:04 <\oren\> I prefer to use an electric racket to kill bugs
21:26:50 <\oren\> and you don't have to wait till they land on something
21:27:21 <fizzie> wob_jonas: The Pixar wiki "Toy Story Characters" category has 58 pages, though some are groups, and some aren't really that suitable for code names.
21:27:32 * gamemanj gives the thwomp button to FireFly
21:27:35 <fizzie> wob_jonas: Thought I'd certainly install Debian Woman on Pizza Planet P.A. system.
21:27:47 * gamemanj then puts a sign in front of the Thwomp, saying "firefly food here"
21:28:13 <fizzie> (And that category is for the first film alone, I think.)
21:28:25 * FireFly buzzes toward the supposed firefly food
21:29:06 <gamemanj> * The thwomp activates on it's own because someone wasn't keeping the button deactivated, and gamemanj is thwomped. FireFly manages to escape.
21:29:32 -!- izabera has joined.
21:29:53 <HackEgo> [wiki] [[FOSCode]] https://esolangs.org/w/index.php?diff=48941&oldid=48940 * Darkrifts * (+220) Added information on using Calcutape
21:30:00 <gamemanj> (the intention was that FireFly would press the button and thus doom me...)
21:30:05 <shachaf> FireFly: did you play that game
21:30:16 <wob_jonas> fizzie: so when they get close to exhausting that, they can start collecting money to pay for some new supplementary novels or something (new novels are cheaper than feature films)
21:30:21 <Vorpal> FireFly, what game is that
21:30:30 <FireFly> Old flash platformer game thing
21:30:55 <Vorpal> Oh, sounds vaguely familiar now
21:31:01 <Vorpal> FireFly, isn't there an N+?
21:31:14 <FireFly> I've only played the original
21:31:30 <FireFly> http://www.freewarefiles.com/screenshot/N.jpg looks approximately so
21:31:31 <Vorpal> Think there is a remake with more levels and such
21:31:41 <Vorpal> Remember seeing something about it maybe a year ago or so
21:31:55 <Vorpal> Like that but better looking
21:32:10 <Vorpal> FireFly, if you liked it, that might be worth checking out
21:32:26 <shachaf> FireFly: No, Spider and Web.
21:32:35 <gamemanj> (as everybody knows, N++ is like N but with classes and templates.)
21:32:56 <Vorpal> FireFly, http://store.steampowered.com/app/230270/
21:32:57 <FireFly> shachaf: nope, hadn't heard of it before but it sounds a bit interesting
21:33:05 <Vorpal> May be available on GoG or similar, I have not checked
21:33:42 <lambdabot> http://eblong.com/zarf/zweb/tangle/
21:36:33 <wob_jonas> by the way, efnet/#mtgrules tells me that the Primal Clay ability still modifies the copiable values of the type, subtype, ability, p/t
21:40:16 -!- Cale has quit (Ping timeout: 250 seconds).
21:42:43 -!- Cale has joined.
21:43:38 <int-e> there, this way (prefer the deepest head) the Hydra dies much faster... http://int-e.eu/~bf3/tmp/hydra.xhtml
21:46:22 <shachaf> Another way to make it die faster is to make the auto-clicked 7.5 times faster.
21:49:08 <int-e> and http://int-e.eu/~bf3/tmp/hydra2.xhtml is faster still.
21:49:33 <int-e> shachaf: did you pull that number out of your hat at random?
21:49:45 <wob_jonas> int-e: what's the difference? does it cheat?
21:50:04 <int-e> wob_jonas: no, it just uses a different strategy to select the heads to cut
21:50:14 <shachaf> int-e: No, I compared "autoInterval = window.setInterval(autoPlay,750);" with "autoInterval = window.setInterval(autoPlay,100);"
21:50:53 <int-e> shachaf: okay. it seemed a bit unlikely indeed :)
21:51:28 <wob_jonas> int-e: one sort of cheating strategy to defeat the hydra faster is to start by clicking on "Restart" until you get a short enough starting hydra
21:52:15 <int-e> meanwhile... The hydra currently has: 28 segments, 8 heads, and depth 7. Hercules has cut 1328805 heads so far.
21:52:29 <int-e> (that's the leftmost strategy still going at 200 chops per second)
21:53:29 <\oren\> Argh! I've forgotten English grammar!
21:53:56 <wob_jonas> \oren\: are you sure you have ever known it?
21:54:14 <\oren\> I just legit wrote "This a circular dependency isn't?"
21:55:20 <\oren\> it should have been "isn't this a circular dependency"
21:57:14 <\oren\> but it didn't even look wrong until after I sent the email
21:57:40 <HackEgo> [wiki] [[FOSCode]] https://esolangs.org/w/index.php?diff=48942&oldid=48941 * Darkrifts * (+245) /* Misc */
21:57:53 <gamemanj> package "isn't this a circular dependency?" depends on "circdep"
21:58:08 <gamemanj> package "circdep" depends on "isn't this a circular dependency?"
21:59:02 <HackEgo> You might expect a reference to recursion here, but to make it interesting you'll actuallSTACK OVERFLOW
21:59:15 * \oren\ mutters to himself "english is never SOV, english is never SOV"
21:59:45 <pikhq> Except when it is.
22:01:00 <int-e> shachaf: I hope I didn't do anything too embarrassing in that Javascript code :)
22:02:14 <shachaf> You can send me a CL if you want to.
22:02:26 -!- FreeFull has quit (Quit: reboot).
22:03:19 <int-e> computational logic, combinatory logic... I'm confused.
22:03:28 <shachaf> pikhq: Is there a general-purpose word for a CL?
22:03:46 <myname> computational linguistics
22:03:55 <HackEgo> [wiki] [[FOSCode]] https://esolangs.org/w/index.php?diff=48943&oldid=48942 * Darkrifts * (+235) /* Documentation */
22:04:49 <int-e> nice number: The hydra currently has: 0 segments, 0 heads, and depth 0. Hercules has cut 555 heads so far.
22:05:53 <shachaf> int-e: It means "Change List". In this case a CL contains code to be reviewed before committing.
22:06:01 <shachaf> But it can also mean code that has already been reviewed and committed.
22:09:06 <pikhq> shachaf: It matches somewhat to the git concept of "commit".
22:09:28 <pikhq> Particularly as used in Linux kernel development.
22:09:39 <shachaf> But the concept of code review doesn't really exist in git in the same way.
22:10:23 <shachaf> I didn't appreciate code review until I used Google's system.
22:10:25 <pikhq> Though to be fair, that particular bit isn't part of P4's CLs, just google3's.
22:10:27 <shachaf> Now I think it's essential.
22:10:43 <shachaf> pikhq: whoa whoa whoa, you mean google[redacted]'s
22:11:00 <pikhq> Nah, that much is public knowledge.
22:12:03 <wob_jonas> wait, both of you worked at google?
22:12:09 <pikhq> It already was. I saw the presentation that was approved by legal and given to a bunch of students at RIT.
22:12:37 -!- atrapado has quit (Quit: Leaving).
22:13:00 <wob_jonas> how many people here are known to have worked there?
22:13:23 <shachaf> Do you count interns and contractors?
22:13:33 <shachaf> How about people who are no longer in this channel?
22:17:20 <shachaf> hydraz: What do you make of int-e's strategy?
22:17:28 <wob_jonas> do we know whether any of you worked on the super-secret military or medicine industry research projects that are so secret that you're not even allowed to say you're working on them?
22:17:42 <HackEgo> hydraz: Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: <http://esolangs.org/>. (For the other kind of esoterica, try #esoteric on EFnet or DALnet.)
22:17:57 -!- tromp_ has joined.
22:17:58 <shachaf> wob_jonas: i can neither confirm nor deny htdnh
22:18:39 <shachaf> oerjan: Hmm, I don't think I've seen "htdnh" in here before.
22:19:01 <shachaf> oerjan: Though it might be the most accurate of all the acronyms in that family.
22:19:38 <wob_jonas> strangely as a change I'm now working (not at google) on a project where the project itself is quite non-secret, so much that you can probably even find out what project it is with some clever internet searches
22:19:47 <wob_jonas> this is the first such project for a while
22:19:53 <wob_jonas> all the previous ones were much more secret
22:20:11 <wob_jonas> of course the details of this one are secret too, but not the project goals themselves
22:21:39 <wob_jonas> shachaf: just because I'm allowed to tell doesn't mean I want to tell
22:22:06 <shachaf> That sounds like a secret.
22:22:20 <myname> http://www.nerdcore.de/2016/07/19/live-action-futurama-trailer/ wat
22:22:27 -!- tromp_ has quit (Ping timeout: 258 seconds).
22:25:44 <hydraz> The best way to kill a hydra is to flatten that tree into a list then kill it using foldr.
22:26:28 <wob_jonas> hydraz: yes, but that's not how Hercules kills the hydra
22:27:06 <hydraz> He traverses it depth-first parallely, I know.
22:27:17 <wob_jonas> hydraz: this time, Hercules wants a solution that king Euthingy accepts
22:27:45 <shachaf> The Indian variant of the legend is much worse.
22:28:06 <shachaf> that's why it's called hydrabad
22:28:27 <hydraz> I assume this Euthingy fella doesn't fork enough to see a parallel solution as acceptable?
22:30:38 <wob_jonas> Does it have invincibility frames?
22:30:49 <hydraz> Yes and I don't think so
22:38:12 -!- Guest7241 has quit (Quit: This computer has gone to sleep).
22:46:47 -!- AnotherTest has quit (Quit: ZNC - http://znc.in).
22:49:18 <wob_jonas> so http://www.madore.org/~david/math/hydra0.xhtml is the ordinary hydra, and http://www.madore.org/~david/math/hydra.xhtml is the dire hydra
22:50:31 <int-e> the ordinary hydra looks so tame now...
22:51:11 <int-e> (well, one reason is that its depth doesn't increase)
22:52:28 * gamemanj gets out a head-removal device, and aims it in izabera's general direction
22:53:13 <gamemanj> You will pay for your crimes against the ** As gamemanj did not know, the device was aimed backwards. **
22:53:16 <wob_jonas> int-e: the gcd dragon is much tamer. That dragon is parametrized by two relative prime positive integers m and n. It starts with exactly one head. In each step, you can use an acid sword to cut off m heads (but only if the dragon has at least m head), or an ordinary sword to cut off one head, but in that case n+1 heads immediately grow (even if you
22:54:33 <gamemanj> Well, on the other hand, at least it can never sustain >m heads.
22:56:40 <int-e> wob_jonas: ah, but the gcd dragon can survive an unskilled attack forever... the hydra can't.
22:57:03 <int-e> (assuming it has at least two heads initially)
23:10:19 <\oren\> When you cut a dire neck, that permanently decreases the number of possible consecutive dire necks.
23:11:08 -!- gamemanj has quit (Ping timeout: 258 seconds).
23:12:03 <\oren\> similarly, the regualr hydra without dire necks can never increase in height
23:12:14 <\oren\> this limits its growth
23:14:04 -!- boily has joined.
23:14:50 <\oren\> Also, on the regualr hydra, the maximum number of heads attached to any node on the second highest level, can never be increased
23:16:39 <\oren\> so perhaps a good strategy would be to seek to ratchet down these invariants
23:17:33 <\oren\> attack the dire heads. attack heads with the most brothers first. attack the highest heads.
23:18:12 <wob_jonas> \oren\: I thought a good strategy would be to just look for a ready-made proof in articles referenced by http://www.madore.org/~david/weblog/d.2008-03-27.1537.html or http://www.madore.org/~david/weblog/d.2008-03-16.1534.ordinaux-et-hydres.html
23:19:34 <int-e> The leftmost hydra has survived 1.5 million chops... I stopped there.
23:19:54 <\oren\> yeah you need to attack highest heads
23:20:23 <int-e> yeah, I automated that (see above) and it works pretty well
23:24:36 <\oren\> the question is which rules should be evaluated first?
23:27:57 <\oren\> dire heads > highest heads > heads with most siblings?
23:28:16 <\oren\> highest heads > dire heads > heads with most siblings?
23:29:54 <int-e> my best attempt weighs dire edges by 2 and normal edges by 1 and takes the (leftmost) deepest one according to that measure... but mainly it's just easy to implement and slightly better than the version without weights.
23:31:43 <wob_jonas> int-e: just try forking the staet, deleting each head, playing the game from there recursively, see which one leads to the fastest completion of the game, and delete that head
23:32:03 <int-e> it's possible that the optimal strategy would start with creating a very wide subtree (which in Madore's implementation causes growth to become very much restricted) and then prune the rest before trimming that subtree
23:32:35 <int-e> wob_jonas: how about you implement that and come back here with the results?
23:33:08 * int-e needs to sleep... and anyway it's approaching the point where this is no longer fun.
23:33:30 <wob_jonas> hmm, are we on browsers where the javascript random function isn't cryptographically secure? what if you break that and roll until you believe the hydra will roll favorably for you
23:34:00 <int-e> wob_jonas: the code comes with its own random number generator... I guess you could peek at its state
23:34:05 <\oren\> the number of sibling dire heads can only indrease when you kill O-o=o heads
23:35:48 -!- Phantom_Hoover has joined.
23:38:45 <\oren\> which imples that dire heads first might be the primary rule
23:40:39 -!- wob_jonas has quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client).
23:43:31 <\oren\> actually, you might want to kill the LOWEST dire heads first
23:43:43 <\oren\> they have the most potential
23:45:48 -!- FreeFull has joined.
23:49:07 -!- deltab_ has joined.
23:50:11 <\oren\> yeah, something along the lines of computing the heads that are most dangerous and attakcing them
23:52:00 <Sgeo> I assumed the ping was for olist?
23:54:24 -!- adu has joined.
23:58:28 <\oren\> hmm yeas... "seek to cut the lowest dire neck segment in the tree"