00:07:48 <zzo38> Is there a Magic the Puzzling that says something like "Your upkeep is beginning; place your triggered abilities on the stack." and your opponent also has some things that trigger during your upkeep too?
00:08:03 -!- jaboja64 has quit (Ping timeout: 248 seconds).
00:08:29 <shachaf> zzo38: I've never seen one.
00:08:40 <shachaf> But I think the only Magic: The Puzzlings that I've seen have been yours.
00:09:08 <zzo38> I have not myself written one like that (yet)
00:09:21 <zzo38> Do you know how to make up one?
00:09:36 <zzo38> (I mean one Magic: the Puzzling in general, it doesn't necessarily have to be like that)
00:09:37 <shachaf> All your Magic: The Puzzlings have seemed way too complicated to me.
00:09:40 <ais523> zzo38: might be interesting to port 3SAT into Magic
00:09:46 <shachaf> You need to know (or look up) so many cards.
00:09:50 <ais523> so we could set Magic: the Puzzlings that had nothing to do with Magic, just maths
00:10:04 <shachaf> Like the entirety of The Dark.
00:10:39 <zzo38> shachaf: That is just one of them though, others name cards explicitly and some do not have so many
00:11:08 <shachaf> What if the puzzle was presented as a picture that showed every card you need to know to solve it?
00:11:46 <zzo38> That would do yes, if you can print it out, although having text file would also help
00:12:03 <zzo38> Even so, some puzzles that do have pictures do not actually show all of the cards, or contain obsolete text
00:12:48 <shachaf> At least if it was a web page where you could mouseover a card name to see it.
00:13:23 <zzo38> I am using a format that is design to be not too difficult to parse by computer, so they could be converted into other formats
00:13:50 <shachaf> Oh, I did see at least one Magic: The Puzzling that wasn't yours.
00:14:13 <zzo38> My puzzles 5 and 6 and 7 do not have as many cards as the others
00:15:12 <shachaf> What do you call that ability on a card that lets you discard it from your hand?
00:16:03 <zzo38> Cycling you can discard it and also draw a card
00:16:16 <zzo38> Not all ability to discard from your hand may be cycling though
00:16:31 <shachaf> Yes, I was looking for that specific one.
00:18:37 -!- p34k has quit.
00:22:31 <zzo38> Perhaps look at puzzle.5 and/or puzzle.6 and/or puzzle.7 and see if you can understand it
00:23:39 <shachaf> You should include a full URL.
00:24:20 <zzo38> http://zzo38computer.org/textfile/miscellaneous/magic_card/puzzle.5
00:25:15 <zzo38> The others are in the same directory
00:45:09 <tswett> zzo38: why not use HTML?
00:46:05 <tswett> You could do funky stuff like...
00:47:24 <tswett> <div class="zone"><span class="zone-owner">Alice</span>'s <span class="zone-name">hand</span>: <ul class="zone-cardlist"><li><span class="card-name">Crumble</span></li></ul></div>
01:13:41 -!- lambda-11235 has joined.
01:27:23 <zzo38> I suppose that is another way yes, although I prefer to post as plain text. However conversion would be possible if wanted
01:30:45 <zzo38> Although that kind of HTML does make some sense. Not only can CSS (either included or user-defined) be used, but also allow to convert formats more easily by using the class names provided.
01:41:49 <zzo38> (But it does mean it is require to parse HTML in order to make the conversion)
02:12:17 -!- lynn has quit (Ping timeout: 260 seconds).
02:13:28 -!- ais523 has quit (Ping timeout: 252 seconds).
02:23:29 -!- oerjan has joined.
02:35:06 <oerjan> @ask fizzie <fizzie> Holy monologue, Batman. <-- you're not secretly oklopol, are you?
02:36:40 -!- ais523 has joined.
02:40:21 -!- bender| has joined.
02:43:11 -!- copumpkin has quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…).
02:43:49 <oerjan> <HackEgo> theory/To be theory is to be like a theorem, but inferior. <-- . o O ( physicists might disagree on the "inferior" )
02:44:09 -!- ais523 has quit (Ping timeout: 246 seconds).
02:44:34 <oerjan> ais523 has such exquisite timing
02:48:44 -!- boily has joined.
02:51:06 <lambdabot> KROC 130154Z 26004KT 10SM BKN250 11/M08 A3003 RMK AO2 SLP174 T01061078
02:51:09 <lambdabot> olsner asked 1d 2h 7m 37s ago: what does it mean to "expand int-e"?
02:51:31 <boily> olsner: hellolsner. to expand int-e's `wisdom entry
02:51:58 <boily> oerjan: we're in Rochester for the weekend. there is much tile shuffling.
02:52:44 <zzo38> Another idea of a Magic: the Gathering card is one which alters the stipulation of the puzzle. (What strangeness!!! I wonder if extra rules might be needed to support such thing?)
02:55:23 <oerjan> boily: now i'll just have to guess which one https://en.wikipedia.org/wiki/Rochester
02:55:29 <oerjan> (i suppose KROC is a hint)
02:55:38 <boily> KROC is a good hint.
02:56:34 <zzo38> Wikipedia reveals it right away
02:57:07 <zzo38> (Even the initial "K" indicates United States, so even if you do not look it up you might know that already)
02:57:26 <oerjan> although most of the places are in the US
02:57:53 <shachaf> are any of the places hard
02:58:43 <oerjan> shachaf: should i get the swatter
03:00:27 <shachaf> oerjan: only if you're stuck between KROC and a hard place hth
03:00:54 * oerjan hits shachaf with the saucepan ===\__/
03:01:06 <oerjan> i figure the swatter was too soft for this
03:03:01 <shachaf> i,i stuck between between a KROC and a DHAR EPLAC
03:03:21 <boily> I can't mapole shachaf; I'm travelling light.
03:08:08 <oerjan> <\oren\> if you have 4 cores, should you not tell it to have 4 threads? <-- that's what ghc recommends, although s/thread/capability/ because ghc has lightweight threads and calls the actual underlying OS ones something else
03:09:15 <boily> I thought the proper voodoo number is n+1 for gcc and make.
03:10:59 -!- ais523 has joined.
03:11:33 <ais523> bleh, kernel panics are so annoying
03:15:09 <oerjan> ais523: are these the kernel panics triggered by yesterday's spacetime disruption
03:15:38 <boily> there was a disruption?
03:16:18 <oerjan> around the time of the first mention of kernel panics
03:16:34 <ais523> it's triggered via plugging my laptop into my router
03:18:19 <Hoolootwo> sounds like some weird timing issue
03:18:32 <zzo38> What is the error message in the kernel panic message?
03:19:01 <ais523> zzo38: Attempted to kill init!
03:19:23 <ais523> but there were so many errors appearing beforehand that it's possible that something was just overwriting random memory
03:19:48 <boily> then don't kill init.
03:19:48 <ais523> also this thing has a manufacturer of 1, product of 2, serial number of 3, and a different serial number of 0123456789ABCDEF
03:20:01 <ais523> which makes me think that someone put dummy data in the device strings
03:23:52 <ais523> given the amount of error spam before that I have a suspicion that something was randomly corrupting kernel memory
03:24:19 <FreeFull> http://picat-lang.org/ I love this language
03:30:26 -!- boily has quit (Quit: RICE CHICKEN).
03:43:56 <oerjan> . o O ( they chose a less annoying waiting music today )
03:44:15 <oerjan> still room for improvement
04:04:53 <oerjan> that's the big remaining question
04:06:13 <izabera> in the press conference after match 3 he said something on the lines of "this is a defeat for lee sedol, not for mankind"
04:07:25 <\oren\> defiant forever! death to the undying!
04:09:49 <zzo38> It is a defeat for Lee Sedol for that tournament anyways; not forever. The problem I have is if they do not mention enough details of the computers in use
04:10:22 <oerjan> apparently they are copying the second game so far
04:14:16 <zzo38> A game with complete information (such as Go), I would expect theoretically a sufficiently powerful computer can win (unless a perfect strategy would force the other player to win or draw). Games such as Kriegspiel chess do not have complete information.
04:15:26 <zzo38> Someone on ChessVariants mentioned something very silly, which is to use the "Swizzelstick" rule: You must touch your nose while moving the pieces, and you must urinate over your opponent's pieces if you win, otherwise it doesn't count.
04:16:39 <zzo38> Anyways "theoretically a sufficiently powerful computer" doesn't necessarily mean that it is, and even if it isn't doesn't necessarily mean the computer program will lose either.
04:17:20 <zzo38> What is your opinion of these things please?
04:18:12 <izabera> are you asking serious opinions on the swizzlestick thing?
04:18:34 <zzo38> No, not on that, I mean in general
04:19:21 <zzo38> I mean about game vs computer programs in general, such as chess and Go and so on.
04:20:06 <izabera> we don't have sufficiently powerful computers that can analyze all the possible moves in go to a decent depth
04:20:34 <oerjan> i think generalized Go is something like EXPTIME-complete, so perfect play might be intractable.
04:22:14 <zzo38> Yes I know there is not sufficiently powerful computers, and probably would be too difficult to make up such a computer anyways (you can easily get into limits by laws of physics too I suppose).
04:22:36 <zzo38> Also by "generalized Go" what generalizations do you mean?
04:23:27 <izabera> possibly larger amount of players?
04:24:54 <zzo38> Another thing is consider what significant differences (if any) when the board size is odd than when it is even
04:26:16 <izabera> that's entirely unreasonable, obviously your chi can't flow on an even board
04:28:38 <zzo38> Whether or not your chi can flow on the board does not affect the rules of the game though, isn't it?
04:32:54 <ais523> izabera: B Nomic ran a "political go" for a while
04:33:06 <ais523> where players could become allies which meant their stones counted as the same colour for the purpose of capturing
04:33:19 <ais523> I'm not sure if there was a limit on alliances
04:33:36 -!- treaki_ has joined.
04:35:05 <zzo38> If you have only one colour of stones then you must play on all points in any order and then once it is full, you must remove all of the stones from the game.
04:35:26 <int-e> http://senseis.xmp.net/?OneColourGo is something different though
04:35:52 <int-e> so you've taken away the option to pass?
04:36:25 <ais523> clearly the optimal strategy is to play one stone then pass twice
04:36:55 -!- treaki__ has quit (Ping timeout: 250 seconds).
04:36:56 <ais523> and claim the entire board as territory
04:36:58 <zzo38> That is just something by the same name anyways
04:37:02 <zzo38> ais523: Yes of course
04:38:28 <zzo38> The "One Colour Go" mentioned on Sensei's Library is really you must use physically the same colours for both player just to confuse you as far as I understand it does not affect the rules of the game (although fouls might become more common, and you have to figure out how to deal with it; if you use a computer then the computer resolves it automatically)
04:39:51 <izabera> like non-flagging minesweeper
04:40:16 <ais523> I think speedruns play non-flagging
04:40:19 <ais523> because of the time it takes to click the button
04:41:15 <zzo38> izabera: Yes I suppose it is a similar thing
04:41:26 <zzo38> Although then you don't have to worry about fouls
04:41:47 <zzo38> It just mean, you cannot annotate the game during the play
04:43:14 <zzo38> Possibly if you play minesweeper with scratching cards it might be more difficult to mark where you believe are the bombs too, although maybe it depend what kind of pencil you use
04:45:57 <zzo38> What kind of pencil are you going to use?
04:51:35 -!- XorSwap has joined.
04:52:06 <\oren\> what if you could make or break alliances at any time?
04:55:48 <\oren\> then, you might cooperate with someone to capture some stones, and then they stab you in the back to get that territory
04:58:02 <zzo38> I sometimes get a kernel panic when switching the printer on or off
04:59:35 <\oren\> I have never gotten a kernel panic except for my stupid attempt at a kernel module
05:00:33 <\oren\> which, when loaded, accidentally overwrote something important
05:03:36 <\oren\> how would go work on a unbounded field?
05:06:05 <ais523> \oren\: oh you could definitely break alliances any time
05:06:13 <ais523> you might need a restriction on making them though
05:08:35 <\oren\> I have another idea but it will take some text to explain so I'll put it in my website after I figre it out.
05:10:02 <zzo38> shachaf: Is 3.2.0-87-generic x86_64 a kernel version number?
05:11:10 <zzo38> I now published version 1 of the JavaScript Xlib package (called "remote-xlib") and also atarted to make another package to provide widgets for use with it
05:29:29 <\oren\> how does this sound so far http://www.orenwatson.be/ungriddedgo.htm
05:30:32 <\oren\> it's go on a board with no grid
05:38:06 <\oren\> Now I'm wondering what the correct radius of a liberty is
05:49:37 <ais523> wow at https://bugs.launchpad.net/ubuntu/+source/linux/+bug/882147 – someone is seriously arguing that it's OK if a kernel feature sometimes silently doesn't work, reporting success, because programs should be prepared for that situation
05:50:13 <ais523> their reasoning is that that's what the kernel does
05:50:21 -!- copumpkin has joined.
05:50:36 <ais523> someone else said that that's clearly a kernel bug as it doesn't match the docs, and they said the docs should be fixed
05:51:07 <ais523> then when someone else said that they fix the kernel to match the docs as they're closer to the intention, the original person suggested changing both (!)
05:51:20 <ais523> this may well be trolling at this point, but it's pretty creative
05:52:13 <\oren\> rrgh i hite whoever that guy it
05:53:08 <ais523> the interface in question is inotify, and the bug in question is tail -f not working
05:53:22 <ais523> it seems that tail already has an undocumented workaround, spelled "---disable-inotify"
05:54:17 <\oren\> see this kind of person is why we need Linus and his FUCK YOUs
05:54:24 <ais523> I like this naming convention
05:56:05 <ais523> haha, someone arguing with the user brought up the hypothetical of "what if select() has a bug" to which the response was "patch every application using select() with a check against the kernel version"
05:56:19 <ais523> or the alternative "give the kernels in question a database of select-using applications so that it can patch it"
05:56:27 <shachaf> The extra - is borrowed from CSS, I'm sure.
05:56:28 <ais523> pretty much everything up to that point was plausible stupidity
05:56:44 <ais523> that alternative, though, is clearly trolling
05:57:11 <ais523> I was parsing it as (---))disable-inotify
05:57:17 <ais523> * (---)disable-inotify
05:57:23 <ais523> but you're parsing it as (--)(-disable-inotify)?
05:58:10 <izabera> several programs have hidden options with 3 dashes
05:58:41 <izabera> `` factor ---debug 10 # first example that comes to mind
05:58:49 <shachaf> ais523: That's how I was trying to make sense of the triple -.
05:58:51 <HackEgo> factor: unrecognized option '---debug' \ Try `factor --help' for more information.
05:59:05 <shachaf> Wow, that factor option is really hidden.
05:59:13 <izabera> [using single-precision arithmetic] 10: 2 5
05:59:57 <ais523> does factor handle >64bit numbers (at all | in reasonable time) yet?
06:00:35 <shachaf> ais523: http://docs.factorcode.org/content/word-__gt__bignum,math.html hth
06:00:41 <izabera> so it works better if there are several smaller factors
06:01:12 <ais523> shachaf: factor(1), not factor the programming language
06:01:28 <shachaf> What sorts of programming languages can you factor?
06:01:40 <ais523> $ factor 1000000000000
06:01:42 <ais523> 1000000000000: 2 2 2 2 2 2 2 2 2 2 2 2 5 5 5 5 5 5 5 5 5 5 5 5
06:01:54 <ais523> shachaf: I assume that was intended as a joke, but it's arguably an interesting question
06:02:05 <shachaf> It was meant halfway in between.
06:02:06 <izabera> ais523: that's... < 64 bits
06:02:07 <ais523> C factors into the preprocessor and rest of the compiler, I guess?
06:02:14 <ais523> izabera: err, good point
06:02:27 <shachaf> Is there a sense in which you multiply programming languages?
06:02:29 <ais523> $ factor 1000000000000000000000000
06:02:30 <ais523> 1000000000000000000000000: 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5
06:02:38 <izabera> try with something less stupid...
06:03:01 <ais523> shachaf: if you interpret programming languages as being defined by their compilers, then multiplication could be compiler composition
06:03:02 <shachaf> izabera: that seems kind of rude, "no offense" or not
06:03:18 <shachaf> Can you factor regular languages?
06:03:21 <ais523> izabera: it worked for my purpose, which was seeing if factor ran quickly on a bignum with a lot of small factors
06:03:32 <izabera> $ time factor $(tr -dc 0-9 < /dev/urandom | head -c 40)
06:03:34 <izabera> 2352950823555604190942608370577849550714: 2 13 23 1291 4210103 723923547371247021321949091
06:03:36 <izabera> real: 0m0.034s, user: 0m0.000s, sys: 0m0.027s
06:04:13 <ais523> I guess that's one way to produce a random number…
06:04:15 <izabera> literally the only test that's more meaningless is to test how fast it can factor 2^n
06:04:35 <ais523> izabera: I would have done that but it's harder to spell
06:04:50 <ais523> I was just checking to see if it could parse the number /at all/
06:04:56 <shachaf> If a programming language is defined by a compiler, what's an interpreter? A special sort of compiler?
06:04:59 <ais523> and this was inspired by the mention of "single precision arithmetic"
06:05:23 <ais523> shachaf: compiles into user interactions, I guess
06:05:23 <izabera> it's backed by gmp so that's handled decently
06:05:32 <izabera> pollard rho doesn't really scale that well tho
06:06:40 <shachaf> ais523: Is it really compiling if it compiles a bit at a time?
06:08:22 <ais523> now you've got me trying to work out how to write a lazy compiler :-)
06:09:17 <shachaf> With a JIT compiler it's reasonable to define languages that compile into infinite programs.
06:09:42 <ais523> JITting sections would make sense, and probably be required
06:09:49 <ais523> how do you parse an infinite program, though?
06:09:55 <shachaf> The output is infinite, not the input.
06:10:03 <ais523> presumably it'd need to be specified as an AST rather than as a sequence of bytes
06:10:20 <shachaf> I'm thinking of compiling programs with higher-rank polymorphism by monomorphizing.
06:10:37 <shachaf> Particularly in the case of non-regular data types?
06:11:16 <shachaf> I was going to say that I don't like infinite programs. But of course the output of a compiler is also a program.
06:13:49 <ais523> hmm, is there some way in Haskell to pattern-match on a data structure only if it's already been forced (and thus its constructor is available)?
06:13:53 <ais523> some sort of "nonforcing match"?
06:14:19 <ais523> I was thinking about how you'd do optimizations on an infinite AST
06:15:08 <shachaf> Not in Report Haskell, but you can do it in GHC.
06:23:59 -!- llue has joined.
06:24:57 -!- lleu has quit (Read error: Connection reset by peer).
06:38:47 <shachaf> Well, you can certainly do it with false negatives.
06:39:06 <ais523> what are false negatives?
06:39:54 <shachaf> You might think something isn't evaluated when it is.
06:40:19 -!- lambda-11235 has quit (Quit: Bye).
06:41:07 <ais523> I didn't realise "false negative" was the phrase in that context
06:41:25 <ais523> less than 100% false negatives?
06:41:48 <shachaf> Maybe it's not the phrase.
06:43:29 <shachaf> The phrases "false positive" and "false negative" are scow.
06:56:14 <ais523> lee sedol in overtime, alphago with over an hour left
06:56:29 <ais523> one of them has been using time inefficiently, it's unclear which
07:00:12 -!- XorSwap has quit (Quit: Leaving).
07:11:27 <ais523> ooh, I think alphago got horizoned
07:11:36 <ais523> it's adding stones to a dead group
07:12:45 <ais523> as far as I can tell, that move it just played is clearly, objectively bad and it is obvious to a go player of pretty much nay skill level
07:15:27 <ais523> (and it's probably not just me missing something obvious as the 9p commentator agrees)
07:27:06 -!- augur has quit (Ping timeout: 276 seconds).
07:31:50 -!- augur has joined.
07:39:50 -!- adu has joined.
07:52:01 <ais523> the commentators don't seem to know about the horizon effect
07:52:21 <ais523> but it's a perfectly good description of the move at T11 (S11 on the commentator's board, which is misnumbered)
07:52:22 <oerjan> ais523: what's the horizon effect?
07:52:30 <shachaf> or that's what someone said in another channel?
07:52:35 <ais523> and it's inexplicable otherwise
07:52:37 <ais523> shachaf: still playing
07:52:46 <ais523> oerjan: basically if you see an impending large loss in your future
07:52:52 <oerjan> shachaf: no, they just discussed when it would resign
07:52:53 <ais523> you play a series of forcing moves to push it further back
07:53:01 <ais523> even if those moves are losing in their own right
07:53:02 <shachaf> oerjan: i don't have audio on hth
07:53:11 <ais523> so that it's out of the range that your evaluation function looks like
07:53:31 <ais523> it's a common issue with tree search AIs
07:53:53 <ais523> (it's believed to have caused Deep Blue to draw a won a game in one of its famous matches against Kasparov)
07:54:42 <ais523> the move at T11 forced Lee Sedol to play at S10; it was perfectly forcing
07:54:47 <ais523> it just obviously loses one stone for no benefit
07:57:03 <ais523> and Alphago was looking at a -16 loss around then, which is large but not necessarily gamelosing
07:57:12 <ais523> err, -14, I miscounted
08:04:45 <oerjan> hm now he says it's close
08:06:19 <ais523> alphago was commonly believed to be winning before it went haywire
08:06:25 <ais523> now it's probably losing
08:36:26 <oerjan> ais523: that A8 move looks like another case?
08:37:43 -!- haavard_ has joined.
08:38:54 -!- sewilton has quit (Ping timeout: 246 seconds).
08:38:54 -!- zemhill has quit (Ping timeout: 246 seconds).
08:38:54 -!- zemhill has joined.
08:38:54 -!- haavardp has quit (Read error: Connection reset by peer).
08:38:57 -!- sewilton_ has joined.
08:39:03 -!- erdic has quit (Ping timeout: 246 seconds).
08:39:24 -!- erdic has joined.
08:41:47 <oerjan> ais523: redmond seems to be recognizing the effect but he's just calling them forcing moves
08:43:07 <ais523> oerjan: he's realised that alphago is playing forcing moves but not why
08:43:21 <ais523> (it's a response to realising that your evaluation of the position was too optimistic)
08:53:09 -!- bender| has quit (Read error: Connection reset by peer).
09:03:42 <FreeFull> Seems AlphaGo isn't unbeatable
09:07:35 -!- AnotherTest has joined.
09:11:09 -!- oerjan has quit (Quit: Nite).
09:14:35 <izabera> cospiracy theory time: they lost on purpose
09:15:06 <izabera> lee is almost a national hero and they were humiliating him too much
09:44:57 -!- bender| has joined.
09:45:47 -!- lynn has joined.
09:57:55 <\oren\> izabera: isn't it good to humiliate another nation's hero?
09:58:21 <\oren\> alphago is supposedly a british computer
09:59:03 <Taneb> \oren\, do you want all the Brits to be driven out of South Korea
09:59:42 <Taneb> It's good if the other nation is your nation's enemy
10:00:00 <\oren\> well it adds to your natin's hero's prestige to have beaten someone else's
10:30:08 -!- Reece` has joined.
10:36:42 -!- adu has quit (Quit: adu).
11:03:00 <shachaf> zzo38: In puzzle.6, how is Artificial Evolution being used?
11:24:52 -!- Reece` has quit (Quit: Leaving).
11:31:08 -!- boily has joined.
11:33:58 -!- boily has quit (Client Quit).
11:42:15 -!- ais523 has quit.
11:47:59 <lifthrasiir> I'm late in the party and still I'm impressed of Lee's outcome
11:48:17 <lifthrasiir> probably the best thing to both Lee and Google
11:49:20 -!- J_Arcane has quit (Ping timeout: 244 seconds).
12:21:04 -!- AnotherTest has quit (Ping timeout: 260 seconds).
12:25:53 <int-e> I'm glad we got to see that Alphago still displays classical MC tree search weaknesses (a horizon effect, essentially, leading to ludicruous moves when the AI is losing)
12:26:42 -!- jaboja has joined.
12:30:24 <izabera> is it something like "there's still a chance for humans" or...?
12:30:50 <int-e> No. I was wondering whether the Alphago team had somehow solved this problem... now I know they didn't.
12:31:30 <\oren\> to find out if they solved it I guess
12:31:34 <int-e> Sorry, did you see the "we got to see" part of what I wrote?
12:32:31 <int-e> Anyway, I'm interested in Go AI progress ... as an observer. So yeah, what oren said.
12:32:37 <izabera> doesn't look like good news
12:33:18 <int-e> I like knowing where we stand.
12:34:32 <int-e> If it hadn't done this while behind that would also have made me happy. But it takes a really strong player (or some artificial setup; I'm sure the Alphago team itself has also experimented with positions where Alphago would've been behind) to get Alphago into that position.
12:34:35 -!- idris-bot has joined.
12:35:52 -!- Melvar` has changed nick to Melvar.
12:35:53 <int-e> izabera: The thing that annoys me about your question, btw, is that I initially wrote "I'm glad to see" and realized that this isn't quite what I meant. So I rephrased that part... and then you start asking why I'm glad :-(
12:38:21 <int-e> And now I'm sorry that I made you apologize. Oh well, communication is hard.
12:40:23 <int-e> And of course I'm also glad to see the human win, but that's on a much more emotional level, not really connected to my scientific curiosity.
13:20:46 <b_jonas> heh heh, humans… you think you're better because you make something you think are the right moves when you're already losing
13:21:49 <b_jonas> If you think those moves are somehow better or more elegant or something, you should have put that in the rules so we get more points for a match played elegantly, or something, rather than complaining that the bots don't do something you didn't ask for.
13:22:53 <b_jonas> Also, congrats to both players.
13:24:10 <int-e> b_jonas: you're not serious about the "humans" remark, right
13:25:08 <Lymia> Too bad I don't know Go well. ;d
13:30:46 <int-e> b_jonas: The thing is, it's a situation where there are no "right moves" anymore; every move is losing. In such a situation, humans try to make the game more complicated (increase the variance of the likely outcomes), as far as it's possible without losing points; they bank on opponent's mistakes (which tend to be merely human)... Alphago apparently has not learned to do that; it assumes an...
13:30:52 <int-e> ...almost perfect opponent, where this kind of strategy would just prolong the inevitable.
13:32:44 <int-e> I'm sure it's a solvable problem (and one that a commercial Go program will have to solve; people will want to take handicaps against it, and then the computer will be behind for most of the game... There are some known approaches too, "dynamic komi" being perhaps the most promising)
13:33:51 <b_jonas> int-e: maybe, but how much of that works in practice against such a strong player as Li Sedol?
13:34:32 <int-e> Ultimately one needs some measure of how hard an overplay (technically a mistake, but one that is hard to refute, only incurs a small loss if refuted and is profitable otherwise) is to refute.
13:35:16 <int-e> b_jonas: Well, the endgame is hard... it's easy to lose half a point here, half a point there, when moves still have sizes of 5 points an more.
13:40:00 <int-e> and from what I've read the game was still quite close 20-30 moves before Alphago resigned; around that point it started to play forcing moves that lose points.
13:56:24 -!- jaboja has quit (Remote host closed the connection).
14:02:58 -!- J_Arcane has joined.
14:04:51 -!- earendel has quit (Ping timeout: 248 seconds).
14:53:00 <fizzie> Bernard Madoff wants to give me 100 million euros.
14:53:04 <fizzie> He regrets his crimes and wants my assistance in giving this money away to charity.
14:53:09 <fizzie> He's "rightly convinced that somebody of your statue [sic] will not stoop so low to derail in this kind of social and humanitarian task --".
14:53:13 <fizzie> I wasn't even aware I had a statue.
14:53:26 -!- Lord_of_Life has quit (Excess Flood).
14:53:52 <int-e> well you have a coin
14:53:59 <int-e> why wouldn't you have a statue as well?
14:54:14 <fizzie> That's true. And statues usually don't stoop, if they're well made.
14:55:37 -!- Lord_of_Life has joined.
14:56:37 -!- nys has joined.
14:57:10 -!- p34k has joined.
14:57:56 <int-e> but I guess you could put a statue on track to derail a train
15:20:39 <tswett> \oren\: I read someone's idea for ungridded Go once. In that version, two stones have to actually touch in order to be connected. A liberty is any spot where you could put a connected stone.
15:21:45 <b_jonas> tswett: heh. someone did mention go on grids other than a plain rectangular grid, but I haven't heared of an ungridded one yet.
15:21:51 <tswett> http://senseis.xmp.net/?GoOnABoardWithoutLines
15:22:00 <b_jonas> in fact, wait, I might even have a link
15:23:12 <int-e> there's the whole http://senseis.xmp.net/?Variants page... some links are broken though
15:23:39 <int-e> in particular the one on http://senseis.xmp.net/?GeneralGraphGo :-(
15:26:13 -!- lambda-11235 has joined.
15:36:58 -!- carado has quit (Ping timeout: 268 seconds).
15:38:15 -!- nys has quit (Ping timeout: 244 seconds).
15:41:12 <tswett> So, looks like people like Bridge Go: http://senseis.xmp.net/?BridgeGo
15:46:58 <Taneb> So, I'm at my parents' place for the Easter break
15:47:05 <Taneb> Forgot one or two things
15:47:11 <Taneb> Such as my phone charger
15:47:28 <Taneb> And the doohickey to get my desktop to connect to wi-fi
15:49:14 <b_jonas> Borrow a phone charger from them
15:50:10 -!- shikhin has changed nick to Their.
15:50:23 -!- Their has changed nick to shikhin.
15:54:14 -!- carado has joined.
15:54:54 <Taneb> b_jonas: that solves one of the issues
15:59:39 <b_jonas> wait wait. Easter break? But Easter is more than two weeks away, isn't it?
16:01:12 <Taneb> b_jonas: I get a month off
16:01:47 <int-e> b_jonas: minus the "more than"? (unless you insist on waiting until Monday)
16:02:56 <b_jonas> int-e: um yes, sorry, exactly two weeks away
16:15:49 -!- bender| has quit (Ping timeout: 268 seconds).
16:20:26 -!- Reece` has joined.
16:25:30 -!- jaboja has joined.
16:32:33 <izabera> `` seq 10 | head ---presume-input-pipe -c -5
16:32:46 <HackEgo> 1 \ 2 \ 3 \ 4 \ 5 \ 6 \ 7 \ 8
16:34:07 <izabera> `` seq 10 | head ---presume-input-pipe ---disable-inotify -1
16:34:09 <HackEgo> head: unrecognized option '---disable-inotify' \ Try `head --help' for more information.
16:34:24 <izabera> `` seq 10 | tail ---presume-input-pipe ---disable-inotify -1
16:34:25 <HackEgo> tail: option used in invalid context -- 1
16:34:32 <izabera> `` seq 10 | tail ---presume-input-pipe ---disable-inotify -f
16:34:35 <HackEgo> 1 \ 2 \ 3 \ 4 \ 5 \ 6 \ 7 \ 8 \ 9 \ 10
16:35:17 <izabera> `` touch x; rm ---presume-input-tty xc
16:35:19 <izabera> `` touch x; rm ---presume-input-tty x
16:35:22 <HackEgo> rm: cannot remove `xc': No such file or directory
16:37:57 <b_jonas> fungot, which level are you stuck on?
16:54:17 -!- lambda-11235 has quit (Ping timeout: 260 seconds).
17:02:43 -!- lambda-11235 has joined.
17:11:23 -!- vanila has joined.
17:11:35 <vanila> https://gist.github.com/rui314/3b66fe949b5ee3b21120
17:13:02 <vanila> https://raw.githubusercontent.com/shinh/bflisp/master/bflisp.bf
17:14:28 -!- fungot has joined.
17:23:35 <izabera> how large a subset of c does it compile?
17:24:41 <vanila> im sure you can find the answer by reading the source
17:31:37 -!- jaboja has quit (Ping timeout: 240 seconds).
17:35:04 -!- gremlins has joined.
17:35:36 -!- Reece` has quit (Ping timeout: 246 seconds).
17:45:55 -!- jaboja has joined.
17:49:16 -!- earendel has joined.
18:00:19 -!- sebbu2 has joined.
18:00:39 -!- sebbu has quit (Ping timeout: 268 seconds).
18:00:50 -!- sebbu2 has changed nick to sebbu.
18:01:53 -!- Sgeo_ has joined.
18:04:03 -!- Sgeo has quit (Ping timeout: 240 seconds).
18:04:16 <lifthrasiir> https://twitter.com/senokay/status/709077093192077312 I'm not really good at bc.
18:43:44 -!- heroux has quit (Ping timeout: 244 seconds).
18:47:38 -!- Reece` has joined.
18:48:45 -!- gremlins has quit (Ping timeout: 246 seconds).
18:49:00 -!- Reece` has quit (Read error: Connection reset by peer).
18:59:51 -!- heroux has joined.
19:06:22 -!- XorSwap has joined.
19:23:35 <b_jonas> fungot, do you have values that are static and some that aren't?
19:23:35 <fungot> b_jonas: another t-shirt worthy phrase from rodgerthegreat x-d a language based on actual music. where the wiki files go
19:28:47 -!- AnotherTest has joined.
19:36:22 <olsner> fungot: what does your t-shirt say?
19:36:22 <fungot> olsner: should i do that?
19:37:05 <fungot> int-e: no. you cannot solve the halting problem... and a political career ruined by the fact that we're whining about ook! on irc is that it's overloaded as the extension for scheme files written in elisp
19:37:24 <HackEgo> fungot is our beloved channel mascot and voice of reason.
19:37:47 <int-e> I like the idea of a political career being ruined by the fact that we're whining about ook! on irc.
19:38:46 <olsner> maybe any one of us could've had a political career if it wasn't for this channel
19:39:19 <int-e> You may be right. So thanks #esoteric for being here!
19:39:39 <fungot> shachaf: where do you want the number of
19:40:32 <shachaf> fungot: what are your views on taxes
19:41:04 <fungot> int-e: when your projects are there, o fnord. eihn siit silloin oo mit fnord."
19:41:25 <int-e> fungot: but I suspect that's going too far
19:41:25 <fungot> int-e: it's just a fnord sexpr to html converter... but right now i'm preoccupied with the oracle of bacon fat and bacon strips? yum. fnord saa fnord fnord verkkotunnuksia.', oh yes.
19:41:40 <fungot> Available: agora alice c64 ct darwin discworld enron europarl ff7 fisher fungot homestuck ic irc* iwcs jargon lovecraft nethack oots pa qwantz sms speeches ss wp youtube
19:42:10 <fungot> shachaf: strange layout. care about one thing. why isn't it just a different way no. i keep thinking of mold for some reason, i decided to write a mandelbrot for some esolang
19:42:11 <b_jonas> fungot: ok, so you don't want to talk about healthcare. in that case, what are your views about the government interfering with the education system?
19:42:11 <fungot> b_jonas: oh i must've seen her before then. all the standard stuff that pc-select does i have a non-vector version at my homepage is fnord lg n)
19:44:36 <int-e> fungot: may the fnord be with you...
19:44:37 <fungot> int-e: in poland, comics are free to implement it? seat of their pants?) those few hours the basement computer class was open for all. i cannot take this imperative madness, i have written an infix-to-prefix converter but it isnt what he though
19:45:56 <shachaf> fungot: who do you think should be the next us supreme court justice twh
19:45:56 <fungot> shachaf: if you ask me
19:46:02 <fungot> shachaf: yep. black lung."
20:11:09 -!- variable has joined.
20:16:11 -!- Vorpal has quit (Ping timeout: 244 seconds).
20:21:18 -!- XorSwap has quit (Quit: Leaving).
20:22:32 -!- Vorpal has joined.
20:25:46 -!- variable has quit (Quit: 1 found in /dev/zero).
20:28:05 -!- Vorpal has quit (Ping timeout: 244 seconds).
20:33:13 -!- Vorpal has joined.
20:40:03 -!- jaboja has quit (Ping timeout: 246 seconds).
20:50:50 -!- Phantom_Hoover has joined.
21:03:30 <zzo38> While the bug I have reported in SQLite has been fixed in the code repository much earlier than when 3.11.0 was released, it seem that the change is not merged into the released version until 3.12.0 though. Do you know why?
21:03:54 <shachaf> zzo38: Did you see my question above?
21:04:35 <zzo38> No, can you repeat please?
21:04:55 <shachaf> <shachaf> zzo38: In puzzle.6, how is Artificial Evolution being used?
21:05:15 <zzo38> O, well, it is on the stack.
21:05:24 <shachaf> What choices are being made?
21:05:35 <zzo38> The choices aren't made until it resolves.
21:05:46 <zzo38> However, the target is Airdrop Condor.
21:05:52 -!- hppavilion[1] has joined.
21:06:23 <zzo38> You have to guarantee you can win; it is the default stipulation.
21:06:35 <zzo38> (See the Codex if this part is unclear)
21:06:43 <shachaf> What's next? Andlanda, GA?
21:08:03 <hppavilion[1]> A board game where one side is the Axis, the other side is the Allies. The Axis creates encryption using the mechanics of the game, the Allies have to crack it
21:08:22 <lynn> I went there last summer, too. Epcot was neat
21:08:29 <shachaf> Don't they have to crack each other's encryptions?
21:08:58 <hppavilion[1]> Best part is, you don't /know/ when the allies crack it, you have to calculate and make a new encryption when they do. But there's something that gives you an aversion to making a new code, so you have to be sure it's cracked
21:09:38 <hppavilion[1]> I think it'd have to be digitized for optimal effect
21:09:58 <shachaf> hppavilion[1]: But the Enigma cipher was symmetric.
21:12:45 <hppavilion[1]> shachaf: And the cypher you use can be symmetric, or it can be asymmetric. Part of the game is making your own cypter
21:12:47 <vanila> I like the idea fro the game
21:13:22 <hppavilion[1]> vanila: Yay! Of course, knowing me, I'll get bored of it in 3 days and never think about it again :/
21:14:52 <hppavilion[1]> vanila: I'm still thinking about how to balance cyphers though; my first thought is to make it unbounded
21:15:09 <hppavilion[1]> But if I do that, there's nothing to prevent someone from implementing PGP at game startup
21:15:35 <vanila> maybe you could limite the compuutation people can do somehow
21:16:21 <hppavilion[1]> vanila: I was thinking that I make the encryption some sort of pseudo-ASM and require (a) you unlock commands and (b) every block costs you
21:17:08 <hppavilion[1]> You would instead use an ASM with access to JSON data
21:17:35 <hppavilion[1]> vanila: My favourite part of the game is that you're making actual encryption in it :)
21:17:54 <hppavilion[1]> Most games would start with a caearian cypher and slowly escalate into something hideous
21:18:52 <hppavilion[1]> vanila: I'm thinking that, in the game, you manufacture "objects" that you can "protect"
21:19:04 <hppavilion[1]> The more protected an object, the harder it is to steal
21:19:25 <hppavilion[1]> vanila: Not quite; this is about protecting the keys, actually
21:20:03 <hppavilion[1]> The best strategy is to have a "codebook" object (which is really just some JSON or an algorithm) and put as much protection as you can into it, because once the allies get their hands on both a device and a codebook, you're screwed
21:21:21 <hppavilion[1]> vanila: The key catalog (the thing that tells units how to encrypt/decrypt a message for each day) (which is not strictly necessary, but is advised) is an object
21:21:49 <hppavilion[1]> When you "protect" it, it gets harder for your enemies to obtain a copy of it
21:22:42 <hppavilion[1]> (You don't need a key catalog, but if you don't use one you're pretty much just using a fixed encryption that'll get cracked the moment they get a copy)
21:23:19 <hppavilion[1]> vanila: There will /also/ be tendency for error that occurs, with frequency proportional (or otherwise related) to code complexity
21:24:15 <hppavilion[1]> The tendency for error especially comes in any time you call random in your units. They use a very lopsided randomization algorithm.
21:24:30 <hppavilion[1]> So it's advised you keep all randomization at the toplevel
21:35:59 <hppavilion[1]> vanila: One of the best strategies in the game is to send the occasional junk transmission xD
21:41:25 <hppavilion[1]> vanila: I just found that Scotch quality = 1/malts
21:49:05 -!- MoALTz_ has joined.
21:52:35 -!- MoALTz has quit (Ping timeout: 248 seconds).
21:54:18 -!- MoALTz_ has changed nick to MoALTz.
22:01:08 <HackEgo> [wiki] [[Libra]] N https://esolangs.org/w/index.php?oldid=46589 * Erinius * (+862) Created page with "'''Libra''' is a language that provides no real features, so the programmer must rely entirely upon libraries to do anything. These standard libraries use extremely esoteric t..."
22:04:12 <HackEgo> [wiki] [[Libra]] https://esolangs.org/w/index.php?diff=46590&oldid=46589 * Erinius * (+111)
22:06:56 -!- AnotherTest has quit (Quit: ZNC - http://znc.in).
22:16:21 <HackEgo> invincibility frames? ¯\(°_o)/¯
22:18:52 <b_jonas> vanila: dunno. I think we once speculated on a language where if control falls out of too many braces at the time without a sentence break;ing the fall, you die, unless you cast feather fall before
22:19:37 -!- Melvar has quit (Ping timeout: 252 seconds).
22:19:37 -!- newsham has quit (Ping timeout: 252 seconds).
22:19:39 -!- Sgeo__ has joined.
22:20:05 -!- Melvar has joined.
22:20:10 -!- hppavilion[1] has quit (Ping timeout: 252 seconds).
22:20:33 -!- hppavilion[1] has joined.
22:20:43 -!- Sgeo_ has quit (Ping timeout: 252 seconds).
22:26:53 <b_jonas> `learn walk without return or you'll attract bugs
22:26:58 <HackEgo> Learned 'walk': walk without return or you'll attract bugs
22:27:08 <b_jonas> no wait, that makes no sense
22:27:13 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: unlearn: not found
22:29:49 <HackEgo> rm: cannot remove `wisdom/': Is a directory \ Forget what?
22:31:24 <shachaf> Is the set of frenemies clopen?
22:31:26 <b_jonas> `learn Frenemy is the relationship between Kirby and king Dee Dee
22:31:33 <HackEgo> Learned 'frenemy': Frenemy is the relationship between Kirby and king Dee Dee
22:32:23 <shachaf> That seems like some inside joke no one else in the channel would know about.
22:32:32 <shachaf> But maybe it's just my knowledge that is lacking.
22:32:41 <pikhq> Just your knowledge.
22:33:10 <shachaf> Oh, there's a King Dedede.
22:33:27 <b_jonas> Dedede? Is that his real name?
22:39:02 <HackEgo> cat "$(find evil -type f | shuf -n1)" | tr '[:lower:]' '[:upper:]'
22:41:51 <hppavilion[1]> vanila: Another one of my projects is actually making Kitten
22:51:56 -!- hppavilion[1] has quit (Ping timeout: 250 seconds).
23:00:13 <HackEgo> Yoda object-verb dialogue adopts.
23:01:21 -!- hppavilion[1] has joined.
23:06:09 -!- oerjan has joined.
23:16:08 -!- newsham has joined.
23:16:59 <hppavilion[1]> vanila: Yet another project of mine is making a good shell for windows xD
23:17:16 -!- p34k has quit.
23:17:16 <vanila> how about making a good shell for linux
23:17:43 <izabera> cygwin -> shells -> pick one
23:18:01 <hppavilion[1]> myname: I've tried getting Ubuntu running on this laptop, but it wouldn't load
23:18:19 <izabera> cygwin -> shells -> choose one -> download source -> compile with msvc
23:18:32 <hppavilion[1]> izabera: I mean a shell for windows that I designed on my own
23:18:43 <myname> the problem isn't the shell
23:18:55 <myname> download the gnu utils and you can kinda work with it
23:18:59 <hppavilion[1]> myname: It's a fun project I'm doing because I want to
23:19:13 <izabera> what's your standard for a good shell?
23:19:14 <HackEgo> [wiki] [[Gbagbo]] N https://esolangs.org/w/index.php?oldid=46591 * Qpliu * (+3758) Created page with "The Gbagbo programming language expresses calculations on bags. It is named after the former Ivory Coast President Laurent Gbagbo. ==Lexical structure== Single character to..."
23:19:27 <HackEgo> [wiki] [[Language list]] https://esolangs.org/w/index.php?diff=46592&oldid=46582 * Qpliu * (+13) /* G */
23:19:56 <hppavilion[1]> izabera: Designed not because we need a shell, but because it's something that would be useful
23:19:58 <izabera> that's not enough of a description
23:23:01 <hppavilion[1]> the @ is a redirection that calls its lhs then its rhs until the lhs fails
23:23:42 <hppavilion[1]> for binds the name the first argument suggests to each successive item of the second argument each time it is called, failing when it runs out
23:23:50 <hppavilion[1]> And cat and >> behave exactly as you expect them to
23:23:53 <myname> so... it's like cat $foo > $bar where $bar is 1 and $foo is this weird "everything but the first" var
23:24:40 <izabera> for x in "${@:2}"; do cat "$x" >> "$1"; done
23:24:52 <HackEgo> mountaintop boatmaker? ¯\(°_o)/¯
23:25:20 <hppavilion[1]> myname: merge takes a destination filename then a list of filenames and appends each of the filenames in the list to the destination in order
23:26:01 <myname> that is exactly what i said
23:26:21 <hppavilion[1]> myname: I wasn't sure whether what you said would have the behavior I expected
23:26:23 <izabera> cat "${@:2}" >> "$1" you could also write it like this in bash, although it's not entirely equivalent because the file is only opened once
23:26:28 <myname> there is some $@ or the like which is "every argument but the first" and would do exactly what you wnt
23:26:43 <myname> izabera: hard disadvantage
23:27:32 <hppavilion[1]> myname: I'm not doing this for practicality, I'm doing it to improve my knowledge of PX
23:27:48 <hppavilion[1]> (PX is like UX, but the users are programmers themselves)
23:28:06 <hppavilion[1]> myname: Also, this way I have a shell that I understand 100% perfectly
23:30:32 <hppavilion[1]> myname: What commands are necessary for a good shell I wonder...
23:30:54 <myname> the commands don't make the shell
23:31:22 -!- lynn has quit (Ping timeout: 250 seconds).
23:32:05 <hppavilion[1]> The primary ones I have so far are echo, cat, cd, df (delete file), chk (evaluate its argument and fail if it is false), chk.f (fail if the flag its argument talks about did not occur in the flags), and for (the for mentioned above)
23:32:32 <hppavilion[1]> myname: Because... um... I didn't know how else to read a file?
23:32:45 <izabera> launch the external cat command?
23:33:13 <myname> the actually needed builtins are only a few
23:33:32 <hppavilion[1]> Those are builtins, and I have a basic mv (with -d to delete the source) and merge, which was explained above
23:33:40 <izabera> commands that do something meaningful to the shell's internal state need to be builtins
23:33:45 <hppavilion[1]> Then there are a few calculator languages launched from shell commands
23:33:58 <hppavilion[1]> But those are more external programs than commands
23:34:04 <myname> i am not sure if an echo executable would make sense
23:34:20 <HackEgo> echo is a shell builtin \ echo is /bin/echo
23:34:56 <izabera> i'm surprised by how fast it replied
23:37:17 <hppavilion[1]> ? unary redirector (well, not really redirection but some kind of modifier) just checks if the command it modifies succeeded
23:38:20 <myname> what should the check do?
23:38:34 <myname> it sounds like: if true then true else false
23:43:10 <myname> if common linux shells don't have it, there is probably a reason
23:44:30 <izabera> common linux shells have a lot of idiosyncrasies
23:46:04 <izabera> https://github.com/izabera/shellnotes i've been keeping a list