00:02:14 -!- jaboja64 has quit (Remote host closed the connection).
00:06:33 <zzo38> With game such as chess and Go and so on it would be theoretically possible that a sufficiently powerful computer should solve the game and always win, but with some games such case is more unclear, and in the case of Snakes&Ladders, nobody can make any decision to have any kind of advantage (although a calculation may be possible to determine how much advantage the first player has)
00:08:41 <HackEgo> 1272) <prooftechnique> b_jonas: Most esolangs I've seen have more comprehensive docs than make
00:09:28 <oerjan> `` sed -i '1272s/^/<b_jonas> (make is an esoteric language) /' quotes
00:09:41 <HackEgo> 1272) <b_jonas> (make is an esoteric language) <prooftechnique> b_jonas: Most esolangs I've seen have more comprehensive docs than make
00:10:41 <int-e> Is make GNU make? I find info documentation quite useful and indeed comprehensive...
00:11:13 <oerjan> maybe he mispled comprehensible
00:11:34 <int-e> prooftechnique: any opinions?
00:11:38 <oerjan> hppavilion[1]: we're already doing that hth
00:11:46 <oerjan> you just happen to be inside
00:12:01 <oerjan> unfortunately, we didn't leave enough resources to recurse.
00:12:02 <myname> so, i have this data Foo = A | B | C deriving (Eq, Show) and i need to make something optional
00:12:38 <myname> right now i am soing data Bar = Optional Foo (Foo -> Bool) | Forced Foo
00:12:40 <hppavilion[1]> Nigh-Impossible Challenge: Computer that can play Nomic
00:12:48 <oerjan> hppavilion[1]: also your schemes seem to be escalating in a singularity
00:12:51 <myname> because the optional wouöd breal both eq and show
00:13:07 <int-e> oerjan: is there any recourse in that situation? (there must be a better way to use this pun...)
00:13:24 <myname> is there a way to not use that "Forced" constrictor wi
00:13:35 <myname> without breaking eq/show?
00:14:15 <int-e> hppavilion[1]: create a black hole before the world ends
00:14:21 <oerjan> hppavilion[1]: it's part of how that simulation saves resources hth
00:15:44 <int-e> hppavilion[1]: the last line reminded me of the "twitch installs arch linux" episode
00:18:31 <int-e> (what happened: some botnet took control, downloaded a Gentoo iso image, wrote it to /dev/sda and booted it... end of story)
00:19:09 <Riviera> same way docker wants to be installed
00:19:14 <int-e> and because even with the bots it was hard to type long commands, they put the commands into a pastebin, then executed them
00:20:32 <int-e> oh and they ran the job in the background... because the first attempts were interrupted by ^C :-) )
00:23:52 <int-e> hmm, stupid cloudflare. "Install and run [email protected]!" <-- I suspect this is SETI@home
00:24:58 <int-e> (source: http://twitchintheshell.com/ ... I have javascript disabled)
00:25:37 <oerjan> @tell prooftechnique <prooftechnique> If we do create one, I vote for måke as the name <-- only shachaf can make that one hth
00:27:07 <oerjan> shachaf: http://codu.org/logs/log/_esoteric/2016-03-11#172303prooftechnique
00:27:57 * shachaf prepares to be antiswatted
00:29:11 <shachaf> did i mention that i've been awake since 5
00:30:40 <vanila> http://www.muppetlabs.com/~breadbox/sketch/
00:30:43 <oerjan> shachaf: i recommend google translate
00:32:15 <Lymia> Can we make a page for GNU Make?
00:32:20 <Lymia> Or whatever the first version is
00:32:32 -!- APic has joined.
00:35:23 * oerjan antiswats shachaf. -----### a giant explosion occurs.
00:38:12 <int-e> oh I guess this was just an antimatter swatter
00:42:34 <olsner> what's the pun in måke again?
00:43:31 <olsner> @ask boily what does it mean to "expand int-e"?
00:44:04 <int-e> olsner: he meant the wisdom entry
00:44:37 <int-e> apparently boily's afraid of mixing languages.
00:45:40 <HackEgo> "Only sane man" boily is monetizing a broterhood scheme with the Guardian of Lachine, apparently involving cookie dealing. He's also a NaniDispenser, a Trigotillectomic Man Eating Chicken and a METARologist. He is seriously lacking in the f-word department. He is also a renowned Capitalist.
00:46:51 <int-e> Gregor Gregor Gregor
00:46:51 <olsner> bot seems to be dead/not-here
00:46:51 <int-e> it's here, but resting?
00:47:23 <int-e> fungot: what about you?
00:47:23 <fungot> int-e: os 9? that's, um, ' elucidating'? don't you find this amusing. who needs a desktop when you have 52 different fnord competing for the same thing.
00:48:45 -!- Sgeo has joined.
00:49:03 -!- Sgeo_ has quit (Ping timeout: 240 seconds).
00:50:59 <oerjan> <olsner> what's the pun in måke again? <-- you need to translate to hebrew hth
00:51:41 <int-e> nap time goodbye resting sleep mode activated...
00:54:41 <olsner> oerjan: oh, it's a norwegian/hebrew/#esoteric/make pun
00:55:54 <Lymia> fizzie, your chainlance repo is a mess. :(
01:04:13 -!- XorSwap has joined.
01:05:23 <fizzie> (Anything in particular?)
01:07:21 <Lymia> As I understand it
01:07:26 <Lymia> The web part needs Ruby?
01:07:43 <fizzie> Yes, but why would anyone be interested in the web part? It's horrible.
01:08:16 <Lymia> The visualization.
01:08:55 -!- Lilly_Goodman has joined.
01:09:40 <Lymia> Surely this can be rewritten into static HTML
01:09:58 <fizzie> I'm not sure what exactly you mean.
01:10:15 <Lymia> http://zem.fi/bfjoust/vis/
01:10:23 <Lymia> Statically rendered into .html files with no dependency on Ruby
01:10:48 <fizzie> Well, everything else than the json data files, sure.
01:11:12 <fizzie> But I wasn't expecing anyone else to want to take the website, so I didn't feel any reason not to try out this nanoc thing.
01:11:46 <Lymia> I need some kind of visualizer. :P
01:11:54 <Lymia> And there's a perfect good one right there.
01:12:42 <fizzie> Well. In theory, you can just take the "prerendered" one from the web, and replace the json files to change the contents.
01:12:54 <fizzie> I don't think there's anything dynamic baked in the generated HTML.
01:13:41 <fizzie> Also that's not "perfectly good", it's very incomplete (if a lot flashier) compared to the old egostats stuff.
01:14:25 <Lymia> What's it missing?
01:14:51 -!- hppavilion[1] has quit (Ping timeout: 264 seconds).
01:15:12 <Lymia> Just some analysis
01:15:17 <Lymia> Well, I suppose I can work on fixing that.
01:15:32 <fizzie> All the clustering stuff, at least.
01:15:50 <fizzie> Those used to be kind of interesting.
01:17:42 <oerjan> `bienvenido Lilly_Goodman
01:17:45 <HackEgo> Lilly_Goodman: ¡Bienvenido al centro internacional para el diseño y despliegue de lenguajes de programación esotéricos! Por desgracia, la mayoría de nosotros no hablamos español. Para obtener más información, echa un vistazo a nuestro wiki: http://esolangs.org/. (Para el otro tipo de esoterismo, prueba #esoteric en EFnet o DALnet.)
01:18:15 <Lymia> Maybe I'll work on it myself. :P
01:19:47 -!- hppavilion[1] has joined.
01:21:13 <oerjan> oh i had an idea the other day why so many people with canaima find their way here, and that maybe it's not really a mistake in their channel list after all
01:22:47 <oerjan> basically, if canaima presents channels in a long alphabetical menu, then it's quite likely that #esoteric is, entirely naturally, the last entry just before the #espana / #espanol ones...
01:23:34 <oerjan> so it could just people searching for the latter, and then seeing us just above the first search hit
01:28:26 <oerjan> Lilly_Goodman: i wonder how you found us
01:28:35 <oerjan> many people with canaima do, somehow
01:28:53 <oerjan> even if we are not a spanish chat
01:29:11 <oerjan> also, i'm waiting for the big Go game
01:30:42 <oerjan> even if i don't understand Go much either, but some here do
01:30:43 <zemhill> Lymia.nop: points -32.14, score 2.98, rank 47/47
01:31:10 <Lymia> !ztest parityscan (>)*9 ([(.)*-1]>>)*-1
01:31:10 <zemhill> Lymia.parityscan: points -34.00, score 2.25, rank 47/47
01:31:35 <Lymia> !ztest parityscan (>)*8 ([(.)*-1]>>)*-1
01:31:35 <zemhill> Lymia.parityscan: points -34.14, score 2.28, rank 47/47
01:32:02 <Lymia> In exchange for a super-fast poke
01:32:10 <Lymia> !ztest parityscan (>)*4 (+)*10 (>)*4 ([(.)*-1]>>)*-1
01:32:11 <zemhill> Lymia.parityscan: points -36.67, score 1.55, rank 47/47
01:32:24 <Lymia> !ztest nop (>)*4 (+)*10 (>)*4
01:32:24 <zemhill> Lymia.nop: points -33.90, score 2.33, rank 47/47
01:32:29 * oerjan wonders if the game has been mentioned outside nerd forums, he tried the two top norwegian newspaper web pages yesterday and it wasn't mentioned on their front pages
01:35:23 <oerjan> hppavilion[1]: as i said the other day, "the only way to win the Game is not to care" hth
01:35:46 <oerjan> which perhaps some nerds find unreasonably hard
01:35:52 -!- Lilly_Goodman has quit (Ping timeout: 260 seconds).
01:36:45 <oerjan> it's really a nerd sniping game, isn't it
01:37:30 <oerjan> nothing on front page, checking tech...
01:38:29 <zzo38> In Magic: the Gathering a lot more situations are possible, including the only way to win the game is to concede!!!
01:38:38 <oerjan> i expect slightly better than average from bbc
01:39:04 <hppavilion[1]> zzo38: Any ideas for GUTs for an esoverse off the top of your head?
01:40:45 <hppavilion[1]> zzo38: How about off the bottom? Heads work like a deque, remember
01:40:55 <oerjan> 'On Thursday, the Korea Times reported that locals had started calling AlphaGo "AI sabum" - or "master AI".'
01:48:54 -!- p34k has quit.
01:50:33 -!- Lilly_Goodman has joined.
01:52:50 <Sgeo> zzo38, ooh, details? (On conceding to win_
01:53:37 -!- vanila has quit (Quit: Leaving).
01:53:46 -!- XorSwap has quit (Ping timeout: 252 seconds).
01:54:22 <zzo38> Sgeo: http://zzo38computer.org/textfile/miscellaneous/magic_card/puzzle.5
01:55:58 <Sgeo> Oh, so you're on a team with someone else, and [I didn't actually read the cards] there's a situation where conceding prevents you from having to do something that hurts a teammate?
01:56:04 <Sgeo> `welcome Lilly_Goodman
01:56:10 <HackEgo> Lilly_Goodman: 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.)
01:56:55 <zzo38> Sgeo: Well, look for the text of the relevant cards (they aren't many), and see if it is of help to you
02:03:28 -!- XorSwap has joined.
02:14:07 <fizzie> oerjan: The "Finnish BBC" (Yle) had an article.
02:14:48 <fizzie> I think this has been copyedited to be broken, possibly while updating it.
02:15:15 <oerjan> Lilly_Goodman: this channel is sometimes silent
02:15:35 <oerjan> next game in 1:15 hours https://www.youtube.com/watch?v=qUAmTYHEyM8
02:15:45 <fizzie> There's a sentence that says (translated): "Alpha Go's advantage is that it never tires or gets worried, Hassabis."
02:15:57 <fizzie> I think it's supposed to be "-- Hassabis said/says".
02:16:20 <fizzie> Everyone keeps repeating the number-of-atoms-in-the-universe thing.
02:17:06 <oerjan> fizzie: humans are bad with exponential growth mkay
02:18:13 <fizzie> oerjan: Hassabis is the second, improved version of Hassa. (Compare the V.32, V.32bis modem standards.)
02:19:55 <zzo38> Even in the game of Pokemon card, a deck consisting of 1x Bulbasaur [Lv15] and 59 basic energy cards is guaranteed to win against a deck consisting of 1x Mewtwo [Lv53] and 59 basic energy cards.
02:20:51 <zzo38> (Even though the first deck in general is pretty bad, and in most cases the second deck is better.)
02:22:37 -!- Lilly_Goodman has quit (Ping timeout: 244 seconds).
02:23:10 <zzo38> (An alternative to Bulbasaur [Lv.15] would be Weedle [Lv.15] and that works too.)
02:30:05 -!- lynn has quit (Read error: Connection reset by peer).
02:35:02 -!- Lilly_Goodman has joined.
02:50:39 * hppavilion[1] calls an exterminator to eliminate our cricket problem
02:51:08 <oerjan> it's just so silent we can hear through the walls
02:55:07 <hppavilion[1]> oerjan: Why are there no functional-logic languages that can run on bare metal? xD
03:12:08 -!- mihow has quit (Quit: mihow).
03:12:21 <Lilly_Goodman> I speak Spanish and English as I do not understand because I feel lonely and I do not have friends
03:17:23 -!- Lilly_Goodman has quit (Ping timeout: 248 seconds).
03:23:29 -!- Lilly_Goodman has joined.
03:25:13 <Lilly_Goodman> I speak Spanish and English as I do not understand because I feel lonely and I do not have friends
03:33:05 -!- XorSwap has quit (Quit: Leaving).
03:37:09 <oerjan> i left the youtube channel open so i'd hear when it started but now they've got this annoying waiting music
03:38:44 <oerjan> which ruins the whole point
03:39:03 <oerjan> because i have to turn it off
03:44:38 <oerjan> Lilly_Goodman: https://www.youtube.com/watch?v=qUAmTYHEyM8 the go match
03:46:25 <oerjan> Lilly_Goodman: most of us here are geeks, when we talk we talk about geeky stuff
03:49:43 <oerjan> most of us here do not speak spanish
03:54:06 <hppavilion[1]> Lilly_Goodman: Would you happen to be Venezuelan? If so, you've got the wrong channel
03:56:19 <Lilly_Goodman> hppavilion[1]: In Venezuela speak spanish, no Engish
03:57:48 <hppavilion[1]> Lilly_Goodman: Este es un canal de programación informática para surrealistas, básicamente. Si estabas buscando algún otro tipo de elementos esotéricos, recomiendo probar #esoteric en DALnet.
03:58:27 <oerjan> hppavilion[1]: we have a spanish translation already, you know
03:58:40 <HackEgo> ¡Bienvenido al centro internacional para el diseño y despliegue de lenguajes de programación esotéricos! Por desgracia, la mayoría de nosotros no hablamos español. Para obtener más información, echa un vistazo a nuestro wiki: http://esolangs.org/. (Para el otro tipo de esoterismo, prueba #esoteric en EFnet o DALnet.)
03:58:43 <hppavilion[1]> oerjan: I know, but I couldn't remember the command xD
04:00:12 <coppro> I've got to hand it to them on one point
04:00:15 <coppro> they've been prompt every time
04:00:48 <Lilly_Goodman> hppavilion[1]: y nadie habla español en esta pagina?
04:02:00 <Lilly_Goodman> hppavilion[1]: y si yo hablo español y ustedes lo traducen a ingles, pueden?
04:02:15 <Lilly_Goodman> hppavilion[1]: lo que pasa es q estoy sola y sin amigos y nadie me quiere :-(
04:03:27 <hppavilion[1]> Parece que usted puede tener una depresión. Simplemente no han encontrado su gente, sin embargo, probablemente
04:05:11 <Lilly_Goodman> hppavilion[1]: no quiero ayuda de depresion solo quiero saber si puede ser mi amigo?
04:06:07 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: bemvindo: not found
04:06:15 <Lilly_Goodman> hppavilion[1]: lo q pasa es que me expulsaron de el #canaimasocial
04:06:44 <Lilly_Goodman> hppavilion[1]: lo q pasa es que me expulsaron de el #canaima-social por decir perdon, que le parece eso?
04:07:08 <HackEgo> エソテリックプログラミング言語のディザインとデプロイメントの国際な場所へようこそ！詳しく、ウィキを見て： http://esolangs.org/wiki/Main_Page。（他のエソテリック、irc.dal.netの#esotericへ）
04:14:25 <prooftechnique> Canaima parece tranquilo. Usted debe tratar de senderismo. Usted puede hacer un montón de amigos durante una excursión.
04:30:07 <oerjan> fines de semana son a menudo tranquilo aquí
04:34:33 -!- treaki__ has joined.
04:38:29 -!- treaki_ has quit (Ping timeout: 244 seconds).
04:43:35 -!- oerjan has quit (Quit: Later).
04:49:54 -!- hppavilion[1] has quit (Quit: Leaving).
04:59:49 -!- Lilly_Goodman has quit (Ping timeout: 260 seconds).
05:02:00 <\oren\> I don't know how to evaluate who's winning
05:07:20 <\oren\> zzo38: can you tell who's winning in a go match?
05:07:27 <pikhq> \oren\: That's a Hard Problem.
05:08:30 <\oren\> yah, that's why i asked a human instead of one of our bots
05:08:43 <izabera> one of the commenters at the alphago challenge is a 9 dan professional player and the best he could do is to roughly guesstimate the score
05:11:21 <zzo38> \oren\: I don't know?
05:11:40 <zzo38> You could calculate the score from the current position but if the game is not finished it won't tell you who is winning.
05:12:29 <pikhq> Evaluating the strength of a position in Go is not just hard for computers, it's hard in general.
05:16:58 <\oren\> well, actually, that's good
05:20:15 <\oren\> but somewhat annoying for the spectator
05:21:26 <izabera> how do you tell who's winning in chess?
05:29:06 <\oren\> I think yoyu count how many pieces they have left and which kinds
05:29:46 <zzo38> Yes you can do that in chess at least, which provides some more easily estimate I suppose
05:32:05 -!- oerjan has joined.
05:37:21 <shachaf> <shachaf> Why did it take the computer a minute and a half at the beginning of the game to respond to what seems like a very obvious opening move that you could precompute the response to?
05:38:57 <izabera> they ran /bin/alphago --suspense=7
05:39:35 <oerjan> shachaf: iiuc the computer doesn't have any opening book
06:04:59 <oerjan> `learn iiuc i understand iiuc correctly
06:05:03 <oerjan> `learn iiuc i understand iiuc correctly.
06:05:06 <HackEgo> Learned 'iiuc': iiuc i understand iiuc correctly
06:05:08 <HackEgo> Relearned 'iiuc': iiuc i understand iiuc correctly.
06:09:16 <HackEgo> Testing can only do so much.
06:09:29 <shachaf> `learn Testing can do so much
06:09:34 <HackEgo> Relearned 'testing': Testing can do so much
06:10:03 <shachaf> Hmm, maybe it should cat lastfiles too.
06:13:35 <HackEgo> sed -i "$1" "$(lastfiles)"
06:13:43 <int-e> `cat bin/lastfiles
06:13:44 <HackEgo> hg log -l 1 --template "{files}\n"
06:14:00 <int-e> sorcery and witchcraft
06:23:16 -!- hppavilion[1] has joined.
06:23:22 <int-e> how does alphago manage time anyway? is it always taking 1:30 minutes (decreasing slowly) per move or is it "taking time to think" sometimes, playing obvious moves more quickly?
06:24:51 <int-e> anyway, after an exciting fight starting in the top left corner, it looks like alphago stands quite a bit better already. too bad.
06:25:05 -!- Sprocklem has quit (Quit: brb).
06:25:27 <hppavilion[1]> oerjan: I had some idea earlier I'm now trying to remember xD
06:28:56 -!- Sprocklem has joined.
06:29:20 -!- Sprocklem has changed nick to Guest55048.
06:29:56 <hppavilion[1]> A language based on nothing but arrows and brackets (asciicized)
06:30:45 -!- Guest55048 has quit (Changing host).
06:30:45 -!- Guest55048 has joined.
06:31:57 -!- Guest55048 has changed nick to Sprocklem.
06:59:51 -!- hppavilion[1] has quit (Ping timeout: 264 seconds).
07:22:35 <int-e> well, last ditch effort, a fight he cannot hope to win
07:22:52 <int-e> that's admirably fighting spirit though... and a great show for spectators
07:23:16 <int-e> of course, if that group at the bottom lives, then that would turn the game around
07:23:27 <int-e> I just don't see it.
07:24:43 <pikhq> I admit I'm really weak at Go strategy, but I agree.
07:27:59 -!- Sprocklem_ has joined.
07:28:23 -!- Sprocklem_ has changed nick to Guest91351.
07:28:34 -!- Sprocklem has quit (Ping timeout: 260 seconds).
07:30:59 -!- lambda-11235 has quit (Quit: Bye).
07:37:23 -!- lleu has quit (Quit: That's what she said).
07:46:24 <oerjan> that was a fast move by alphago
07:57:37 <int-e> too many kos on board... I guess we get to find out whether alphago has inherited the monte carlo bot's weaknesses in that area
08:08:20 <HackEgo> 1260) <mauris> MAHJONG TILE AUTUMN / HIRAGANA LETTER YA / SNOWMAN WITHOUT SNOW
08:14:34 <HackEgo> [U+1F028 MAHJONG TILE AUTUMN] [U+3084 HIRAGANA LETTER YA] [U+26C4 SNOWMAN WITHOUT SNOW]
08:18:45 <HackEgo> [U+002D HYPHEN-MINUS] [U+2013 EN DASH]
08:41:08 -!- oerjan has quit (Quit: Nite).
08:41:13 -!- yorick__ has joined.
08:44:01 -!- yorick has quit (Ping timeout: 240 seconds).
08:51:33 -!- yorick has joined.
08:53:18 -!- yorick__ has quit (Ping timeout: 248 seconds).
09:12:34 -!- AnotherTest has joined.
09:45:06 <fizzie> int-e: It definitely played some "forced" moves very quickly.
09:45:19 <fizzie> Well, "very" -- not as quickly as a human, but noticeably faster than others.
09:46:02 <fizzie> I'm sure someone's already plotted the timing.
09:47:06 <fizzie> http://i.imgur.com/22zZDSZ.png
09:47:21 <fizzie> (Thanks, r/dataisbeautiful.)
09:47:45 <fizzie> That's for the first match, but it's clearly different from the human.
09:49:49 <fizzie> The intro-to-AI course tournament I keep mentioning every now and then allocated 1 hour of (single-thread) CPU time to each participant, and watching how they manage time was kind of interesting.
09:50:51 <fizzie> For each bot, my result page plots included a graph of turn/cumulative time used, with one line for each game they played, and colored according to win/tie/loss.
09:50:59 <fizzie> The results were nicely distinctive.
09:52:25 <fizzie> http://users.ics.aalto.fi/htkallas/ai2012/img/cpu_ajeeb.png <- there's one example, where the bot clearly has three separate strategies depending on how much time it has left.
09:53:18 <fizzie> Or I guess the first turning point could also depend on how many moves it has made -- the games are so grouped initially, it's hard to tell.
09:55:37 <fizzie> But there's an obvious "start going fast when you have only 100 seconds left" mode, and also a faster-than-midgame opening mode, which might last either for the first 300 seconds or 40 moves. And those changes are so abrupt, it's clearly a programmed-in feature.
09:55:56 <fizzie> (Whereas the general sloping rightwards is probably just a consequence of positions getting simpler.)
09:56:16 <fizzie> http://users.ics.aalto.fi/htkallas/ai2012/details.xhtml has same plots for all the 2012 participants.
09:57:59 <fizzie> (Also I haven't worked for the ICS department for the last three years, but they don't seem to be in a hurry to clean my personal web directory.)
10:00:26 <fizzie> http://users.ics.aalto.fi/htkallas/ai2012/img/cpu_justice.png and this one on the other hand clearly has some sort of a gradually decreasing time-per-turn thing.
10:00:55 <fizzie> (The occasional flat horizontal lines are probably because the game in question has some situations where there's only one legal move.)
10:01:33 <fizzie> Holy monologue, Batman.
10:10:11 -!- bender| has joined.
11:00:34 -!- bender| has quit (Quit: lel wait a sec).
11:32:12 -!- augur has quit (Read error: Connection reset by peer).
11:32:19 -!- augur has joined.
11:36:26 -!- bender| has joined.
11:50:45 <Lymia> fizzie, can you paste your test/reference.out
11:51:03 <Lymia> In the chainlance repo.
11:51:11 <Lymia> I'm pretty sure something's wrong with my nodejs
11:51:16 <Lymia> Because I'm getting lines like ">>><<<<<<<<<<XXXXXXXX XXXXXXXXXXXXXXXXXXXXX NaN"
11:55:02 <izabera> `` tac -r -s $'.\|\n' le/rn
11:55:16 <HackEgo> /hackenv/bin/tac: line 2: -r: No such file or directory \ /hackenv/bin/tac: line 2: -s: No such file or directory \ /hackenv/bin/tac: line 2: .\| \ : No such file or directory \ sep="/"; [[ "$0" == *//* ]] && sep="//"; [[ "$1" == ?*"$sep"* ]] || exit 1; key="$(echo "${1%%$sep*}" | lowercase)"; value="${1#*$sep}"; [ -e "wisdom/$key" ] && verb="Re
11:55:36 <HackEgo> /hackenv/bin/tac: line 2: --version: No such file or directory
11:55:54 <izabera> it's the one i wrote haha <.<
11:56:02 <izabera> `` /bin/tac -r -s $'.\|\n' le/rn
11:56:03 <HackEgo> /hackenv/bin/`: line 4: /bin/tac: No such file or directory
11:56:11 <HackEgo> tac is /hackenv/bin/tac \ tac is /usr/bin/tac
11:56:15 <izabera> `` /usr/bin/tac -r -s $'.\|\n' le/rn
11:56:17 <HackEgo> \ "»yek$« brev$" ohce && ")"yek$/modsiw" p-ohce($" > "eulav$" ohce ;"denraeL"=brev || "denraeleR"=brev && ] "yek$/modsiw" e- [ ;"}pes$*#1{$"=eulav ;")esacrewol | "}*pes$%%1{$" ohce($"=yek ;1 tixe || ]] *"pes$"*? == "1$" [[ ;"//"=pes && ]] *//* == "0$" [[ ;"/"=pes
12:46:34 <int-e> fizzie: interesting tournament, thanks; what game is that?
12:47:38 <int-e> "World Journal of Research and Review", yay, sounds legitimate
12:50:54 <int-e> A few days ago one of these spams promised a "blind peer review" within 4 days; I found that refreshingly honest.
12:51:57 <myname> "that one really looks like a paper"
12:52:13 <int-e> myname: no looking, just feel it!
12:52:36 <myname> it's only a matter of time before you can publish scigen stuff in nature
12:59:50 <izabera> are they playing 5 games anyway even if lee already lost 3?
13:14:55 <int-e> izabera: and lee can still earn a bonus if he wins ($20k or so?)
13:20:39 <izabera> yeah i've just read it on wikipedia
13:20:54 <izabera> + 150k if he plays all 5 games
13:29:12 -!- bender| has quit (Quit: [reboot]).
13:49:28 -!- jaboja has joined.
13:58:29 <b_jonas> I realize that Moody has lost an eye, a leg, and much of his nose and doesn't have them back. However, given the above I think that's probably because he lost them to Dark Magic, being an Auror who fought Death Eaters.
13:58:59 <HackEgo> theory/To be theory is to be like a theorem, but inferior.
14:07:16 -!- jaboja64 has joined.
14:10:43 -!- jaboja has quit (Ping timeout: 248 seconds).
14:12:01 -!- jaboja64 has quit (Ping timeout: 244 seconds).
14:16:14 -!- p34k has joined.
14:18:54 -!- lynn has joined.
14:27:24 -!- lynn has quit (Ping timeout: 246 seconds).
14:28:37 -!- bender| has joined.
14:35:47 -!- lynn has joined.
14:41:03 -!- lynn has quit (Ping timeout: 246 seconds).
14:43:15 -!- lynn has joined.
14:57:09 -!- lynn has quit (Ping timeout: 246 seconds).
15:08:49 -!- J_Arcane has quit (Ping timeout: 244 seconds).
15:29:28 -!- Yurume has quit (Ping timeout: 252 seconds).
15:29:29 -!- newsham has quit (Ping timeout: 252 seconds).
15:30:05 -!- idris-bot has quit (Ping timeout: 252 seconds).
15:30:43 -!- Yurume has joined.
15:44:44 -!- J_Arcane has joined.
15:52:24 -!- lambda-11235 has joined.
15:53:35 <Lymia> I am now crunching a 9000x9000 BF Joust hill.
15:53:45 <Lymia> I hope my laptop doesn't overheat.
16:00:57 -!- jaboja64 has joined.
16:08:09 <Lymia> 3.7G./run/evaluation_tmp
16:08:23 <Lymia> I didn't think wrenchlance would use so much disk space.
16:10:26 -!- mysanthrop has joined.
16:11:50 -!- jameseb- has joined.
16:12:18 -!- fizzie has quit (Ping timeout: 246 seconds).
16:12:19 -!- myname has quit (Ping timeout: 246 seconds).
16:12:19 -!- sewilton has quit (Ping timeout: 246 seconds).
16:12:19 -!- jameseb has quit (Ping timeout: 246 seconds).
16:12:19 -!- cnr has quit (Ping timeout: 246 seconds).
16:12:19 -!- ^v^v has quit (Ping timeout: 246 seconds).
16:12:19 -!- fungot has quit (Ping timeout: 246 seconds).
16:12:22 -!- sewilton has joined.
16:12:24 -!- fizzie` has joined.
16:12:32 -!- sewilton has quit (Changing host).
16:12:32 -!- sewilton has joined.
16:13:11 -!- ^v has joined.
16:13:27 -!- cnr has joined.
16:13:27 -!- cnr has quit (Changing host).
16:13:27 -!- cnr has joined.
16:26:22 -!- newsham has joined.
16:40:33 -!- Melvar has quit (Ping timeout: 240 seconds).
17:01:05 -!- Melvar has joined.
17:05:58 <\oren\> I want to write a new code editor
17:06:33 <\oren\> it will use advanced Unicode characters to have a cool interface
17:09:06 <\oren\> like using ▁▂▃▄▅▆▇█ to make a scrollbar showing where you are in the file
17:14:03 -!- jaboja64 has quit (Ping timeout: 240 seconds).
17:19:29 -!- jaboja64 has joined.
17:22:18 <Lymia> Python's multiprocessing library is a little arbitary sometimes. ^.^;
17:22:45 <Taneb> In terms of syntax, semantics, API?
17:24:29 <mysanthrop> \oren\: i thought about something similar, too, but i am too comfortable using vim
17:24:38 -!- mysanthrop has changed nick to myname.
17:30:36 <Lymia> However it distributes work.
17:30:41 <Lymia> It's letting cores go inactive for minutes at a time.
17:31:05 -!- lambdabot has quit (Remote host closed the connection).
17:32:23 <int-e> prevents overheating :P
17:33:48 <Lymia> ~800/9000 done. \o/
17:34:01 <int-e> that doesn't sound so bad
17:34:16 <int-e> 100 minutes so far?
17:34:31 <Lymia> Around 1 and a half hour, yeah.
17:34:37 <Lymia> And it'll get faster near the end.
17:34:42 <Lymia> Since there's less matchups left per program
17:35:23 <Lymia> Memory usage's wtf though.
17:37:42 -!- lambdabot has joined.
17:40:23 -!- bender| has quit (Ping timeout: 268 seconds).
17:41:29 -!- lleu has joined.
17:46:18 <fizzie`> Lymia: reference.out: http://sprunge.us/ZdCL
17:47:00 <Lymia> I fixed it already
17:47:17 <Lymia> https://github.com/Lymia/chainlance/commit/0316061c60ffbe577131805ea92199eac9e488b7
17:48:09 <fizzie`> Oh, I didn't even know/notice it had something like that.
17:48:32 <fizzie`> Guess I've never had it trip.
17:48:36 -!- fizzie` has changed nick to fizzie.
17:49:21 <fizzie> int-e: The game's called "Hierarchy" (Hierarkia); it was made by a Finnish chess I-don't-know-which-title, who originally wanted to make it an actual board game, but that never went anywhere; he then donated it for the course.
17:49:47 <fizzie> int-e: We used it to make it a bit less probable people just submitted code they copy-pasted from someone's chess program.
17:50:12 <fizzie> It's also got a little bit higher branching factor, but it's no Go.
17:50:27 -!- jaboja64 has quit (Ping timeout: 248 seconds).
17:51:07 <fizzie> The rule page has 404d.
17:54:03 <Lymia> My laptop's CPU doesn't have the greatest clock speed, so.
17:55:05 <Lymia> Not the greatest single-threaded preformance.
17:58:19 <fizzie> Huh, what the: someone's made a copy of the page.
17:59:25 <fizzie> int-e: http://www.cs.hut.fi/~framling/hierarkia/ here you go. (The Scheme stuff was no longer in use when I was running the thing.)
17:59:32 <Lymia> https://github.com/Lymia/chainlance/blob/master/wrenchlance-stub.c
17:59:41 <Lymia> This is what I'm doing right now, and it seems to work pretty well.
18:01:55 -!- idris-bot has joined.
18:05:06 <fizzie> Made by this guy: http://www.chessgames.com/player/jyrki_parkkinen.html
18:24:56 <Lymia> I'm turning down the split size for the parmap next time.
18:25:03 <Lymia> So it doesn't end up putting 3 cores idle
18:29:50 <tswett> The things I do for neural nets.
18:30:16 <tswett> Apparently CUDA doesn't actually support Debian or something. Ubuntu, but not Debian.
18:30:28 <tswett> I am, of course, running Debian.
18:30:41 <Lymia> why is there nothing running
18:31:21 -!- ais523 has joined.
18:31:27 <\oren\> if you have 4 cores, should you not tell it to have 4 threads?
18:31:47 <tswett> So... this is a little tricky to do. CUDA wants packages that don't exist on Debian.
18:31:54 <ais523> hmm, threads, unlike processes, use shared-memory concurrency
18:32:05 <\oren\> or is that just my instinct
18:32:30 <ais523> if a lot of the sharing is actually used, it might be inefficient to run the threads on different processors due to all the synchronization that has to take place between the processors
18:32:51 <ais523> OTOH if the threads are mostly independent then you could run 4 with no problems, and might or might not benefit from more depending on the properties of the scheduler
18:32:52 <Lymia> I did tell it to use 4 threads.
18:32:56 <\oren\> i mean, the number of cores is the limit of actual concurrency that you can do, right?
18:33:32 <Lymia> ais523, I have literally 0 dependence between cores.
18:33:51 <\oren\> then why not use an actual fork instead of threads
18:34:16 <ais523> \oren\: you can get more concurrency if the threads are using completely different parts of the computer
18:34:37 <ais523> e.g. if one thread is using the memory bus, it can reasonably act in pseudo-parallel with another thread on the same CPU that's just doing arithmetic
18:35:37 <Lymia> Why would I rather deal with distributing work between multiple independent processes when it's already hard enough when that's done with shared memory and threads.
18:35:42 <Lymia> What benefit do I derive from that.
18:36:18 <\oren\> maybe the processes will actually run on separeate cpus?
18:37:04 <Lymia> Perhaps you should stop being so presumptuous about what the problem is. :/
18:37:31 <\oren\> well, i thought the problem was that on;ly one cpu core was bing used?
18:37:31 <ais523> I don't know what the problem is, but I think my advice is reasonable so far
18:37:48 <ais523> if the threads are completely independent, then the correct number to use will be comparable to the number of CPUs, possibly a little more
18:38:20 <Lymia> Because your advice is thoroughly off base? How many wrong things did you assume. Let's see.
18:39:09 <Lymia> First of all, Python already uses multiple processes because it has that dumb global interpreter lock.
18:39:41 <Lymia> Second of all, even if I wasn't, the threads would only be spinning off external processes anyway then doing a little parsing on the output. I'm only using its multiprocessing library at all because it splits the workload for me.
18:40:14 <Lymia> Third of all, the problem isn't even that, for some reason, I have 4 threads that decided to run on one core. That doesn't even make much sense on Linux, where threads are much like processes to the scheduler anyway in the first place.
18:40:15 <fizzie> tswett: This must be about something like a more modern version of CUDA? Because there are CUDA packages right there in the Debian repository.
18:40:34 <Lymia> It's that for some reason multiprocessing's work distribution is not running and letting threads in the thread pool sit idle for minutes at a time.
18:40:36 <fizzie> I've got nvidia-cuda-dev 6.5.19-3 installed.
18:40:53 -!- jaboja64 has joined.
18:42:46 <tswett> fizzie: uhhh, you have? Lemme see that.
18:43:04 <Lymia> The ideal solution would really be to wait asynchronously in one thread in the Python code.
18:43:12 <fizzie> https://packages.debian.org/jessie/nvidia-cuda-dev
18:43:17 <fizzie> It's non-free, of course.
18:43:21 <Lymia> But there's no library code for that, and I'd need to write it out on my on.
18:43:30 <fizzie> (And that's 6.0.37-5 because jessie.)
18:44:36 <tswett> fizzie: do you have nvidia-alternative installed?
18:44:50 <fizzie> (Oh, I think "nvidia-cuda-toolkit" is maybe the more topper-level package.)
18:45:17 <tswett> What version? I'm trying to install 340.96-1 and it says it can't find it.
18:46:02 <fizzie> I've got 352.79-1, but this is also something like stretch or sid, I forget what exactly.
18:46:39 <fizzie> stretch has a higher-priority pin, so technically that, I guess.
18:49:44 <tswett> Let me see if doing random things with aptitude will fix the problem.
18:49:50 <tswett> Without also creating even worse problems.
18:50:51 <tswett> It seems like aptitude keeps saying "I can't install that thing you asked me to install, but I can uninstall a whole bunch of other stuff instead."
18:53:55 -!- MoALTz has quit (Ping timeout: 248 seconds).
18:54:00 <\oren\> Sometimes if you uninstall everything (for a given value of "everything") and reinstall, it fixes the dependency tree
18:54:48 <tswett> Hmm, it looks like it actually did manage to install it.
18:55:58 -!- MoALTz has joined.
18:56:42 <\oren\> aptitude is annoyingly intolerant of badly written packages
18:58:29 <\oren\> ideally, it shouldn't be possible for a package with wrong dependencies to screw up other, correctly made packages, but that doesn't seem to be the case
19:02:39 <\oren\> What shortcut should be "save"?
19:02:47 <\oren\> I'm thinking control-S
19:02:58 <\oren\> but nano uses control-O
19:05:06 <tswett> C-Q and C-S are occasionally used for XON and XOFF, right?
19:05:28 <\oren\> int-e: that's a good idea, since it's one keypress
19:07:32 <ais523> \oren\: Ctrl-S is vulnerable to potentially really major terminal misparses
19:07:36 <\oren\> I'm planning on having a "hint bar" like nano
19:08:00 <ais523> so if you want to use it as a keybinding in a terminal-based program, you'd better be really confident in your ability to configure the terminal correctly
19:08:11 <ais523> nano mocks you for pressing Ctrl-S for this reason (in cases where it's parsed correctly)
19:08:19 <\oren\> are the F keys generally safe?
19:09:04 <ais523> F1-F5 are vulnerable to ambiguities
19:09:45 <ais523> (this is because F1-F5 aren't sent to the application at all on an actual VT100, so people had to invent codes for them; some of the invented codes are used for other keys that actually did exist on a VT100 but not on a PC keyboard, but in an inconsistent way)
19:10:37 <ais523> the problem here being that the question when things like xterm were originally ported to PC-like keyboard layouts was not "how do I tell the program what keys on a PC keyboard are being pressed", but "how do I allow the user to use this program, which was designed for a VT100, on a PC keyboard"
19:11:18 <ais523> so for example, in practice there are two codes you're likely to see for F1, one is unambiguously F1, the other is ambiguous with Num Lock
19:12:52 <tswett> aptitude's suggestions can seem kinda cheeky sometimes.
19:13:10 <tswett> "Hey aptitude, I want you to install this package, version 6.5." "All right, how about if I don't install it instead?"
19:13:26 <tswett> "No, you need to actually install it." "All right, how about if I install version 6.0 instead?"
19:13:50 <tswett> "No, it needs to be version 6.5." "Oh, why didn't you say so?"
19:22:02 -!- Treio has joined.
19:23:03 <zzo38> tswett: I have had problem like that but found out ways to resolve it by manually specifying version of packages in different orders and then it won't say there is a conflict
19:23:45 -!- ais523 has quit (Remote host closed the connection).
19:23:55 -!- ais523 has joined.
19:40:16 -!- digitalc1ld has changed nick to digitalcold.
19:44:11 <tswett> Time to reboot and see if X still works!
19:44:40 -!- Treio has quit (Quit: Leaving).
19:46:16 -!- jaboja64 has quit (Ping timeout: 244 seconds).
19:46:47 <Lymia> It's reading 347M of data every round
19:46:57 <Lymia> I can't imagine that being good for efficiency
19:49:12 <Lymia> But on the other hand
19:49:19 <Lymia> I really don't want to give a shot at concurrency in c
19:51:21 <tswett> Do they make SATA power extension cables?
19:51:24 <\oren\> I really need to finish my string and array libraries before I can implement this
19:51:47 <ais523> Lymia: I was trying to work out if it'd be possible to run this sort of thing on a GPU
19:51:51 <ais523> but I think the answer is "not efficiently"
19:51:55 <tswett> Looks like they certainly do.
19:52:07 <ais523> I can't see any way to make BF Joust data-parallel across warps
19:52:19 <ais523> which is needed or you're only using about 1/32 of the GPU's power
19:53:25 <tswett> And the answer is no, X does not still work.
19:53:32 <\oren\> tswett: do you have a giant computer case or something?
19:54:33 <tswett> I want to hook up five SATA devices: four hard drives and one optical drive.
19:54:49 <tswett> My PSU, as far as I can see, has two SATA power cables, one with two connectors and one with three connectors.
19:55:41 <tswett> But the optical drive is too far away from the hard drives for a single SATA power cable to supply both drives.
19:56:32 <\oren\> uh, would it just be stupid to put ahard disk in one of the optical drive bays?
19:56:47 <\oren\> assuming you have more than one
19:57:10 <tswett> I've got three or four optical drive bays. But those are bigger.
19:57:56 <tswett> And presumably I could get some sort of adapter so I can fit a hard drive in an optical bay, but that sounds harder than just getting an extension cable.
19:57:57 -!- jaboja64 has joined.
19:58:07 <\oren\> oh, are you afraid of leaving something not securely screwed into a bay?
20:01:08 * \oren\ had an extra psu jammedinto an optical bay on his old computer
20:03:11 <tswett> Don't you have your computer mounted in gimbals?
20:03:59 <\oren\> heh. this was the computer that I left open with a cardboard manifold connecting it to a big room fan
20:04:25 <\oren\> thus pressurizing the case, sort of
20:05:12 <\oren\> hmm, what of you actually kept the air pressure higher inside a computer?
20:06:36 -!- lambda-11235 has quit (Quit: Bye).
20:08:48 <\oren\> yeah, i think theoretically, it woudl increase the rate at which temperature exchages with the air
20:09:15 <Lymia> ais523, I was thinking about running multiple instructions at onec.
20:09:31 <Lymia> There are many conditions in BF Joust when the two opposing programs are "non-interacting", so to speak.
20:09:45 <tswett> All right, let me try to get X working again.
20:09:46 <Lymia> For example, when they're on different cells entirely, or the cell they're both on is 128
20:10:15 <tswett> This has definitely happened to me before.
20:10:16 <ais523> actually one thing you could to is check the number of cycles before a program goes past the fifth cell
20:10:20 <Lymia> It'd be a torquelance style compilation with serious optimization involved.
20:10:33 <ais523> and just start at that location
20:11:24 -!- variable has joined.
20:13:27 <\oren\> you could also detect when both programs are doing [+] (with various numbers of +'s) on the same cell, and make a lookup table of the outcomes?
20:14:25 <\oren\> maybe there's too many possibilities....
20:17:06 <Lymia> The approach I was thinking would be splitting every part of the program into two "segments", kinda.
20:17:40 <Lymia> "How many cycles I can run in one/few instructions in a row"
20:17:51 <Lymia> "Execute x cycles, which may be the previous value, or less than such"
20:19:59 <ais523> come to think of it, at any moment both programs are in an inner loop
20:20:10 <ais523> you could probably do some sort of optimization on it
20:20:18 <ais523> via calculating the outcome in advance
20:20:35 <ais523> that would help a lot because [-] loops are very common
20:29:06 <tswett> All right, the nouveau module is not loaded. So I just need to make it so that the other module is loaded? And what module would that be, exactly?
20:32:54 <tswett> Aha, I need to install nvidia-driver.
20:34:29 <tswett> There we go, now it's working.
20:34:52 <tswett> Now let's see if I can train a neural net of size SIXTEEN FUCKING HUNDRED.
20:34:58 <tswett> (Hint: PROBABLY FUCKING NOT.)
20:35:28 -!- variable has quit (Quit: 1 found in /dev/zero).
20:37:52 <tswett> But I can do one of size 1300. Wow.
20:39:16 <tswett> I can even do one of size 1500. I wasn't expecting that.
21:07:22 <HackEgo> [wiki] [[Talk:Minebit]] M https://esolangs.org/w/index.php?diff=46588&oldid=46256 * Erinius * (+34)
21:21:53 <tswett> Heheh, my computer's graphics aren't working too hot.
21:24:43 <\oren\> wait, why are you running X at the same time as this?
21:25:51 <tswett> That's a good question.
21:26:58 <Hoolootwo> hmm okay, probably better than any of mine
21:26:59 -!- AnotherTest has quit (Quit: ZNC - http://znc.in).
21:27:28 <Hoolootwo> last time I tried to do anything on a GPU, my cpus were faster by about an order of magnitude :/
21:28:55 <tswett> Let's see. Released in 2010 at $79. 40 nanometer fab. 1 GB memory. 269 gigaflops.
21:29:45 <tswett> Lemme compare that to something a little more modern.
21:30:48 <tswett> The GTX 950. Released in 2015 at $159. 28 nanometer fab. 4 GB memory. 1573 gigaflops.
21:30:58 <tswett> That's a lot floppier.
21:35:50 <Hoolootwo> very floppy, though not as much as some of my 8 inch wangs :P
21:35:53 <Hoolootwo> http://siarchives.si.edu/wp3/wp-content/uploads/2010/08/newwang.jpg
21:36:33 -!- Reece` has joined.
21:41:43 <Lymia> That's, like, tiny
21:45:11 <int-e> double density... expensive stuff
21:47:55 <Lymia> Are floppy disks good for anything nowadays
21:48:52 <int-e> I'd expect there's legacy hardware in the industry that still uses them
21:50:07 <Hoolootwo> I have been using them a bit to get a 286 onto the internet, but for practical purposes no
21:52:01 -!- MoALTz has quit (Ping timeout: 240 seconds).
21:53:34 -!- ais523 has quit (Ping timeout: 248 seconds).
21:55:39 -!- MoALTz has joined.
22:05:48 -!- hydraz has changed nick to thetypo.
22:05:53 -!- nortti has changed nick to norrti.
22:06:04 -!- norrti has changed nick to nortti.
22:06:19 -!- thetypo has changed nick to hidraz.
22:07:03 -!- hidraz has changed nick to hydraz.
22:08:39 <b_jonas> When playing causal constructed Magic: the Gathering with friends, if I offer the people sweets like bonbons or assorted wafers or jelly beans, is it usually considered bad form to offer a selection with more than four copies of any one type of bonbon or wafer or jelly bean?
22:11:11 -!- ais523 has joined.
22:11:11 <int-e> wouldn't those be counters...
22:11:39 <b_jonas> int-e: no, they'd make the cards dirty.
22:11:54 <shachaf> Player counters, of course.
22:12:02 <b_jonas> shachaf: heh, that might work
22:12:06 <shachaf> Hopefully you don't offer poison counters, though.
22:12:19 -!- hydraz has changed nick to thetypo.
22:12:24 -!- thetypo has changed nick to hydraz.
22:12:30 <b_jonas> shachaf: hmm, how many other named player counters are there even?
22:15:51 <b_jonas> ais523: about that machine that only accesses a limited set of stacks, I think it's actually possible to do symbol table translation too efficiently, only you have to do it on the whole file and all symbols together, rather than incrementally
22:16:05 <shachaf> b_jonas: I guess I don't know of any others
22:16:13 <ais523> b_jonas: yes, it's definitely a different syle of programming
22:16:32 <ais523> fwiw programs written in this style might work more efficiently even on regular CPUs
22:16:37 <b_jonas> ais523: however, you brought up this machine because it would get rid of all aliasing problems, but you don't really need just stacks for that
22:16:50 <ais523> b_jonas: not just aliasing, cacheing too
22:16:57 <ais523> it was originally invented to get rid of cache stalls
22:17:08 <ais523> also to satisfy hppavilion[1]-like thinking of "let's be different for the sake of it"
22:18:40 <b_jonas> ais523: I see. I like the model of a pointer machine (with a fixed finite size of registers) that accesses trees with *immutable* nodes and can create new nodes, decide whether a node is null, and access the children of nodes, and its core is finite state.
22:19:24 <ais523> how is this garbage-collected?
22:19:57 <b_jonas> ais523: presumably anything that's not accessible from the root anymore would be collected.
22:20:09 <ais523> no, I mean hardware-wise
22:21:02 <b_jonas> ais523: I don't know, it's not really a specific hardware thing, just a theoretical model. hardware implementations typically have more complicated rules to be optimized, which are useful for real world programming, such as built-in fixed integer arithmetic.
22:45:30 -!- ais523 has quit (Read error: Connection reset by peer).
22:45:39 -!- ais523 has joined.
22:47:39 <b_jonas> ais523: I found out about two cards in older sets that seem to be in the wrong color if you apply the modern color pie. There's an instant for {W} and an instant for {G} that untaps a target permanent, and the latter even has more options than that.
22:48:31 <b_jonas> These days that's exclusively blue, with Dream's Grip doing this.
22:49:26 <b_jonas> White can untap creatures, green can untap lands or specific creatures that have a self-untap ability, and then there's Magewright Stone.
23:05:52 -!- Reece` has quit (Quit: Leaving).
23:07:22 -!- idris-bot has quit (Ping timeout: 260 seconds).
23:07:30 -!- Melvar` has joined.
23:08:19 -!- Melvar has quit (Ping timeout: 248 seconds).
23:19:00 <zzo38> I know that with standard cards a player might have poison counters and experience counters; with a custom card I made up it is possible for a player to have any kind of counter.
23:24:16 <zzo38> (I don't know of others but you can tell me if there is; maybe I forgot some?)
23:26:30 -!- lynn has joined.
23:39:15 <zzo38> I had to shorten the kitchen sink picture posted here yesterday because it was previously too long and it should be maximum 16x16 like the standard X cursor font is
23:39:25 <zzo38> (I did fix it though.)
23:45:37 -!- asie has joined.
23:45:45 <asie> oh, font design?
23:46:25 <zzo38> Yes some people on here have designed other fonts
23:47:07 <zzo38> Although I also did a supplementary X cursor font and a new character encoding too (which requires new font to go with it though)
23:51:30 <ais523> asie: the 1×5 font is pretty eso
23:51:43 <ais523> it uses subpixel antialiasing; it's actually readable if you get really close to the screen