←2018-12-09 2018-12-10 2018-12-11→ ↑2018 ↑all
00:00:02 -!- danieljabailey has quit (Quit: ZNC 1.6.6+deb1ubuntu0.1 - http://znc.in).
00:00:21 -!- danieljabailey has joined.
00:15:23 -!- tromp has quit (Remote host closed the connection).
00:16:03 -!- tromp has joined.
00:17:30 -!- MDude has quit (Ping timeout: 250 seconds).
00:34:01 -!- imode has joined.
00:39:01 <izabera> i have a thing which has several possible "states"
00:39:26 <izabera> these states are finite and not even that many
00:39:34 <izabera> 66bits are enough to encode all of them
00:39:35 <izabera> but
00:39:51 <izabera> state 1 feels less random than state 138924533525526
00:40:34 <izabera> how do i quantify this randomness?
00:42:58 <ais523> one method is by seeing how likely each state is
00:43:12 <izabera> they're all equally likely
00:43:29 <ais523> then I don't see any reason why state 1 would be less random than state 138924533525526
00:43:42 <ais523> the only reason we perceive 1 as being less random is that's a more commonly used number in general
00:43:42 <izabera> state 0 is special and we like it a lot for some reason
00:44:25 <izabera> an easy way to measure this thing would be to take the distance of each state from state 0
00:44:27 <izabera> but
00:44:46 <izabera> there's some states that are as far as possible from state 0, which are not very random at all
00:45:43 <izabera> i.e. you can describe those states in a very compact way
00:45:52 <ais523> here's the start of a list of all integers, ordered by how frequently they appear in the Google Books corpus: 1 2 3 4 5 6 10 8 7 9 20 12 15 11 14 0 30 13 16 18 25 17 19 24 100 21 22 50 23 40 28 26 27 29 31 60
00:46:09 <ais523> err, I'm not sure if negative integers are included, the - might be parsed as a punctuation mark
00:46:19 <ais523> so perhaps it's the start of a list of all /nonnegative/ integers with that ordering
00:46:19 <izabera> ok...
00:46:45 <izabera> ais523: do you know how to solve a rubik's cube?
00:46:50 <ais523> hmm, I guess if you're treating state 0 as special, then the specialness of other states would be based on how they related to state 0
00:46:59 <ais523> izabera: I know one algorithm for it, and the general principles behind many others
00:47:10 <izabera> yes that's correct, the specialness is in relation to state 0
00:47:31 <ais523> if this is about a Rubik's Cube, I recommend reading up on the Thistlethwaite algorithm, it seems relevant here
00:47:40 <izabera> i know that
00:47:54 <ais523> in particular, it categorises states by what /sort/ of moves are required to go between them and state 0
00:47:57 <izabera> i know
00:48:19 <izabera> but still i don't think it's a very good measure
00:48:45 <ais523> hmm, this list of integers ordered by frequency cuts off at frequency 40, containing such integers as 19111919 and 16841721
00:48:59 <ais523> there are 245164 integers that appear at least 40 times in the corpus
00:49:19 <ais523> oh, 83000000000 too
00:49:34 <izabera> so there's this state that's called a superflip
00:49:54 <ais523> meanwhile, 395000000000000 has a full 46 occurrences!
00:49:57 <izabera> basically all the corners are solved, and all the edges are in the right position but flipped
00:49:57 <ais523> this is like art in its own way
00:50:12 <ais523> and by some metrics, it's the furthest you can get from the starting state
00:50:29 <izabera> yes it takes 20 moves
00:50:38 <izabera> it's the first case that was proved to require 20 moves
00:50:46 <izabera> 20 moves is the diameter of the graph
00:51:01 <izabera> can't go farther than that from the solved state
00:51:13 <izabera> but it's absolutely trivial and totally not random
00:51:38 <ais523> there are two metrics, in one of them the furthest-known position is apparently unique (as in, the only known position), in the other, there are lots of ties
00:51:54 <ais523> but both of them have either superflip or a variant of it as one of the furthest-known positions
00:51:55 <izabera> it's not unique
00:52:00 <izabera> there's a lot of ties
00:54:19 <ais523> if it's not unique in half-turn metric, then it must be quarter-turn metric which has the believed-unique maximal-moves pattern
00:54:25 <ais523> (which IIRC is superflip with some of the centres swaped)
00:54:26 <ais523> *swapped
00:55:58 <izabera> https://ruwix.com/pics/rubiks-cube/patterns/six-spots.svg things like this have 0 solved pieces, 0 pieces with the right orientation, and yet are absolutely trivial and they don't feel random
00:56:12 <izabera> thistlethwaite would vomit by only looking at it
00:57:17 <ais523> you might want to categorise by symmetries, in that case
00:57:36 <ais523> patterns like that are easy because their symmetries make it easier to spot a solution
00:57:55 <ais523> besides, isn't that pattern solvable with only slice moves? that's the exact sort of rule that thistlethwaite-like algorithms are good at picking up on
00:58:30 <izabera> what
00:58:36 <izabera> that's like
00:58:38 <ais523> the first step in the Thistlethwaite algorithm is to orient the corners
00:58:47 <izabera> do you even know what you're talking about
00:58:53 <ais523> and the corners in that pattern are clearly correctly oriented, if all in the wrong position
00:59:19 <izabera> none of them is correctly oriented, and the first step in th is not to orient corners
00:59:51 <ais523> izabera: the aim in Thistlethwaite, as written, is to get into a position from which you can ban moves on two opposite faces
00:59:55 <izabera> corner orientation is in <R U>
01:00:03 <izabera> i know
01:00:05 <izabera> you don't
01:00:20 <ais523> err, ban quarter-moves on two opposite faces
01:01:01 <ais523> I'm having trouble remembering what the restrictions are for each of the four phases
01:01:03 <izabera> you go from <R L U D F B> to <R L U D F2 B2> to <R2 L2 U D F2 B2> to <R2 L2 U2 D2 F2 B2> to solved
01:01:18 <ais523> it's the second step that locks the orientation of the corners, right
01:01:27 <ais523> because whatever face of the corner is on U/D is stuck there
01:02:14 <ais523> anyway, you can imagine a Thistlethwaite-like algorithm which uses different move sets
01:02:35 <ais523> e.g. one in which one of the sets allows the three slice moves (both quarter-turn and half-turn), but not any individual face moves
01:03:06 <ais523> the pattern you linked belongs (I think) to that set, which is a notable property of it, even if it doesn't neatly fall into any of the "normal" Thistlethwaite sets
01:15:23 <izabera> ok i'll correct myself, the six-dots case only has 8 flipped edges
01:16:09 <izabera> and depending on how you look at it, the first step in th is 4 moves and the second one is 4 moves and the last 2 are skipped
01:16:49 <izabera> well depending on how you look at it, you get to distribute the same 8 moves over the 4 steps
01:17:03 <ais523> anyway, even if I have the details wrong (and I probably do), six-spot is the sort of pattern that you'd expect to be solvable with only a subset of the legal moves
01:17:27 <b_jonas> currentliy featured question on StackOverflow asks how to do loops in metafont/ancient-APL style https://stackoverflow.com/q/53689001/
01:17:45 <b_jonas> wait, a rubik's cube question?
01:17:56 <b_jonas> oh right, it's by izabera
01:17:56 <izabera> ais523: i don't necessarily think that that is a good way to estimate what i want
01:18:03 <ais523> b_jonas: that doesn't look featured to me
01:18:28 <b_jonas> "66 bits are enough to encode all of them" => hmm, didn't it fit in 65 5bits?
01:19:09 <izabera> a little less than ~65.23 bits
01:19:15 <b_jonas> ais523: => wow, that's an interesting frequency list
01:19:47 <ais523> yes, I thought it was interesting too
01:20:19 <ais523> apparently this sequence is not on OEIS yet
01:20:44 <ais523> (and we'd probably want an up-to-date version of the corpus before adding it, mine's a bit old)
01:20:52 <b_jonas> ais523: does that count occurrences with or without multiplicity throughout books?
01:20:59 <ais523> it's also one of the longest inherently finite sequences I've seen
01:21:02 <ais523> b_jonas: with multiplicity
01:21:24 <ais523> as in, a number that's used once in 40 books and a number that's used 40 times in a single book will both have a frequency of 40
01:22:00 <b_jonas> ais523: how many times does 244823040 appear?
01:22:17 <shachaf> ais523: Was your general monad ? operator delimited by function?
01:23:08 <ais523> shachaf: it needs some definition of what scope it runs over; Rust's definition is "a function, or a block marked 'catch'" which seems reasonable; having it apply to a single block is also reasonable I think
01:23:29 <shachaf> But that makes {}/; not associative, right?
01:23:31 <ais523> b_jonas: less than 40
01:23:34 <izabera> this question is because a friend of mine wanted to come up with some position that was considered "hard" in any possible method
01:24:10 <ais523> shachaf: it's OK for {} to not be associative, I think; they aren't in Rust anyway because of lifetimes
01:24:16 <ais523> so Rust doesn't care about that
01:24:28 <ais523> and I'm not sure they inherently need to be associative in other languages either
01:24:31 <shachaf> Right, like C++.
01:24:56 <shachaf> Well, Haskell requires do { do { a; b }; c } = do { a; do { b; c } }
01:25:53 <b_jonas> "one in which one of the sets allows the three slice moves" => that would be very small and so almost useless
01:25:56 <b_jonas> IMO
01:26:08 <ais523> b_jonas: it's not that small if you allow quarter turns
01:26:26 <b_jonas> ais523: sorry, it's not featured. it's hot.
01:27:31 <ais523> b_jonas: here's a featured challenge on PPCG at the moment: https://codegolf.stackexchange.com/questions/105127/fastest-mini-flak-quine
01:27:43 <ais523> which will end up having to be featured three times, there was a 1100 rep bounty on it
01:27:48 <ais523> (which takes a while to hand out…)
01:28:30 <b_jonas> "one of the longest inherently finite sequences I've seen" => no way. there are interesting mathematically defined sequences that go to at least tetration sized numbers, perhaps more. most of them are more interesting as sets than as sequences, but those are still often put into OEIS.
01:31:07 <ais523> yes, but I haven't seen them
01:31:08 <b_jonas> ais523: re 1100 rep bounty, I used to think such couldn't be awarded at all, at least by a single user, because SE docs "https://codegolf.stackexchange.com/help/bounty" say "Additionally, if you offer multiple bounties on the same question, the minimum spend doubles with each subsequent bounty"
01:31:16 <ais523> the doubling caps at 500
01:31:26 <ais523> but someone else is awarding the spare 100 to reach a number that isn't divisible by 500
01:31:33 <b_jonas> ais523: "seen" in what sense? they wouldn't be listed explicitly on your screen obviously
01:31:41 <ais523> the definition
01:31:41 <b_jonas> ais523: yeah, I know
01:32:00 <ais523> I guess it's finite and growing
01:32:14 <ais523> any day, someone could publish a book that contains the fourtieth copy of a brand new number :-D
01:33:07 <ais523> (and of course, the original source material has to go beyond the things with 40 copies into things with 39 copies, and so on down to 1; it's just that the list I have handy stops there)
01:33:34 <b_jonas> ais523: Baez probably knows more about this, but I think there's some small congruance class, one of 6k+1, 6k-1, 4k+1, 4k-1, such that most small primes fall into one, but eventually the complementary class takes over at some very large threshold, and of course around that it switches a lot
01:34:23 <ais523> b_jonas: I know about that too (4n-1 vs. 4n+1 is the best known)
01:34:28 <ais523> but it doesn't really define a finite sequence
01:34:44 <ais523> because it swaps back at some point
01:35:05 <ais523> it defines a large /number/, but that's different
01:35:12 <ais523> I guess you could call the digits of that number a sequence
01:36:17 <b_jonas> ais523: it probably swaps out at many points, so the set of numbers n such there are more primes less than n in that congruence class would be a set, not just a number
01:36:39 <b_jonas> the set of natural numbers, say
01:36:48 <b_jonas> uh no
01:36:50 <ais523> yes, the interesting number is the lowest element of that set, which is still quite large
01:36:51 <b_jonas> that's stupid
01:36:58 <b_jonas> sorry
01:37:02 <b_jonas> I meant my own definition
01:37:16 <b_jonas> anyway, yes, that's probably not one of the more interesting large finite sequences
01:38:37 <ais523> oh, it's not that big, 26833
01:39:41 <ais523> big enough that it'd take a while to discover without the help of computers, though!
01:44:33 <ais523> I think there's a related problem where it's much larger
01:45:20 <b_jonas> ais523: I think it was a different sequence then
01:45:39 <b_jonas> ais523: hmm, but isn't the largest such number much bigger?
01:45:47 <b_jonas> it's the largest such number that's really interesting, not the smallest one,
01:46:00 <b_jonas> because there can be "random" anomalies around small primes
01:46:40 -!- MDude has joined.
01:47:21 <ais523> there are infinitely many such numbers, so there isn't a largest
01:49:33 <b_jonas> um, the largest number outside that set then
01:50:14 <ais523> there are infinitely many of those, too
01:53:25 <b_jonas> hmm, https://en.wikipedia.org/wiki/Optimal_solutions_for_Rubik%27s_Cube says the number of states with the maximal number of turns for the quater-turn metric is not known, but only very few such positions are known.
01:53:46 <b_jonas> not surprising, but I don't really pay attention to the quater-turn metric
01:54:03 -!- hexfive has joined.
01:55:46 <b_jonas> I was surprised back in 2010 when the 20 move upper bound proof for the face turn metric was announced, on how early it's been found
02:01:23 -!- b_jonas has quit (Quit: leaving).
02:24:21 -!- tng has joined.
02:30:16 -!- FreeFull has quit.
02:38:06 <esowiki> [[Talk:Real Fast Nora's Hair Salon 3: Shear Disaster Download]] https://esolangs.org/w/index.php?diff=58648&oldid=33787 * BradensEsolangs * (+149) /* Name */ new section
02:41:07 <esowiki> [[Talk:Real Fast Nora's Hair Salon 3: Shear Disaster Download]] https://esolangs.org/w/index.php?diff=58649&oldid=58648 * Ais523 * (+432) /* Name */ some history
02:43:05 <esowiki> [[Talk:Real Fast Nora's Hair Salon 3: Shear Disaster Download]] https://esolangs.org/w/index.php?diff=58650&oldid=58649 * Ais523 * (+118) OK, so now the spambot's made me visit IMDB to verify the details
03:08:46 -!- tng has quit (Quit: Mutter: www.mutterirc.com).
03:18:22 -!- Lord_of_Life has quit (Ping timeout: 246 seconds).
03:21:13 -!- Lord_of_Life has joined.
03:21:14 -!- Lord_of_Life has quit (Changing host).
03:21:14 -!- Lord_of_Life has joined.
03:57:57 -!- xkapastel has quit (Quit: Connection closed for inactivity).
04:00:00 -!- Taneb has quit (Quit: I seem to have stopped.).
04:02:26 -!- Taneb has joined.
04:05:25 -!- ais523 has quit (Quit: quit).
04:24:05 <zzo38> Does any web browser have a "save form" command?
04:28:56 <zzo38> Is there any proposal for a PNG chunk for XPM palettes?
05:10:00 -!- doesthiswork has quit (Quit: Leaving.).
06:06:01 -!- doesthiswork has joined.
06:37:40 <zzo38> Now I wrote a specification for using XPM colours in PNG. This allows you to define the preferred mono or grey representation of a colour picture, as well as to do such things as match text colours and background colours of a document.
07:06:02 -!- doesthiswork has quit (Quit: Leaving.).
07:42:33 <esowiki> [[Hello world program in esoteric languages]] https://esolangs.org/w/index.php?diff=58651&oldid=58001 * Ais523 * (+1186) /* The Waterfall Model */ Hello, world!
07:55:06 -!- imode has quit (Ping timeout: 250 seconds).
07:59:41 <esowiki> [[The Waterfall Model]] https://esolangs.org/w/index.php?diff=58652&oldid=58558 * Ais523 * (+105) /* External resources */ link Ratiofall
08:11:46 -!- tromp has quit (Remote host closed the connection).
08:12:02 -!- tromp has joined.
08:59:20 -!- b_jonas has joined.
08:59:46 <b_jonas> zzo38: I could be wrong, but I sort of have the impression that graphic designers are very different from us, and they don't want that kind of tool
09:00:00 <b_jonas> including the ones that design game graphics
09:00:59 <b_jonas> also, not many people use grayscale graphics on displays anymore, it's mostly used in print
09:20:42 -!- tromp has quit (Remote host closed the connection).
09:21:17 -!- tromp has joined.
09:28:36 -!- b_jonas has quit (Quit: leaving).
09:34:17 -!- arseniiv has joined.
10:06:13 -!- xkapastel has joined.
10:54:05 -!- tromp has quit (Remote host closed the connection).
10:54:18 -!- tromp has joined.
11:25:17 -!- tromp has quit (Remote host closed the connection).
11:25:55 -!- tromp has joined.
11:30:19 -!- tromp has quit (Remote host closed the connection).
11:30:34 -!- tromp has joined.
11:30:52 -!- hexfive has quit (Quit: WeeChat 2.2).
11:31:58 -!- tromp has quit (Remote host closed the connection).
11:32:30 -!- tromp has joined.
11:46:08 -!- tromp has quit (Remote host closed the connection).
11:46:21 -!- tromp has joined.
11:46:43 -!- wob_jonas has joined.
11:48:10 <wob_jonas> ais523: re primes mod 4, let's see in https://arxiv.org/abs/math/0408319
11:49:16 <wob_jonas> ais523: however, it also says that I was wrong, the sequence is infinite
12:01:05 -!- wob_jonas has quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client).
12:02:54 -!- wob_jonas has joined.
12:03:20 <wob_jonas> that article probably has the plots were made by Mathematica, because it uses the Mathematica font
12:17:09 <wob_jonas> `? nobel
12:17:11 <HackEso> nobel? ¯\(°​_o)/¯
12:17:12 <wob_jonas> `? noble
12:17:13 <HackEso> noble? ¯\(°​_o)/¯
12:18:10 <wob_jonas> slashlearn Nobel//"Nobel" is Norwegian for "noble", a title of honor awarded by kings.
12:18:35 <myname> i wonder if there are people who call the nobel price "fields medal of chemistry/physics/..."
12:19:45 <wob_jonas> I dunno. but there are people who point out that there are at least three awards that are sometimes called the "Nobel prize of mathematics"
12:21:32 <myname> great for the mathematitians
12:30:38 <Taneb> I think we should call more things Nobel prizes of their respective categories
12:31:01 <Taneb> Like "The Nobel prize of sheep shearing in Northumberland"
12:33:11 <myname> the nobel prize of sports
12:47:36 -!- Melvar has quit (Ping timeout: 250 seconds).
13:01:20 -!- Melvar has joined.
13:11:01 <wob_jonas> and the IOCCC judges award the Nobel prize of obfuscated programming
13:13:51 <myname> obfuscated C programming
13:14:28 <FireFly> or, really, just the Nobel prize of C programming
13:14:38 * FireFly ducsk
13:14:42 <FireFly> ducks, even
13:18:39 <myname> i like the contest to create the largest c++ error message better
13:19:34 <wob_jonas> how about the http://www.underhanded-c.org/ _
13:28:21 -!- AnotherTest has quit (Ping timeout: 252 seconds).
13:29:33 <wob_jonas> myname: https://tgceec.tumblr.com/ ?
13:45:29 -!- AnotherTest has joined.
14:00:25 -!- doesthiswork has joined.
14:13:41 <wob_jonas> AHAHAHA apparently mailing list web archives aren't the only ones affected by overzelous "email address protection" filters:
14:14:13 <wob_jonas> "https://buy.garmin.com/en-US/US/p/119594" is a vendor's page with specs of a camera they make, and it says "Photo burst: 3/[email protected], 5/[email protected] or 10/[email protected]"
14:14:32 <FireFly> Heh
14:15:50 <wob_jonas> probably automatically decodes it in client side if you allow javascript or something
14:15:55 <wob_jonas> I didn't try
15:20:34 -!- Lord_of_Life has quit (Ping timeout: 250 seconds).
15:22:03 -!- Lord_of_Life has joined.
15:22:03 -!- Lord_of_Life has quit (Changing host).
15:22:03 -!- Lord_of_Life has joined.
15:27:40 -!- Cale has joined.
15:31:58 -!- sleepnap has joined.
16:07:58 -!- tromp has quit (Remote host closed the connection).
16:08:34 -!- tromp has joined.
16:22:49 -!- Essadon has joined.
16:37:19 -!- tromp has quit (Remote host closed the connection).
16:37:32 -!- tromp has joined.
16:47:27 -!- AnotherTest has quit (Ping timeout: 240 seconds).
16:54:11 -!- tromp has quit (Remote host closed the connection).
16:54:43 -!- tromp has joined.
16:57:03 -!- imode has joined.
17:01:36 -!- tromp has quit (Remote host closed the connection).
17:01:51 -!- tromp has joined.
17:05:34 -!- AnotherTest has joined.
17:16:08 -!- nfd9001 has quit (Ping timeout: 246 seconds).
17:16:58 -!- tromp has quit (Remote host closed the connection).
17:17:32 -!- tromp has joined.
17:34:19 -!- tromp has quit (Remote host closed the connection).
17:34:33 -!- tromp has joined.
17:47:24 -!- wob_jonas has quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client).
17:50:38 -!- tromp has quit (Remote host closed the connection).
18:12:27 -!- AnotherTest has quit (Ping timeout: 240 seconds).
18:19:15 -!- MDude has quit (Remote host closed the connection).
18:32:14 -!- zzo38 has quit (Disconnected by services).
18:32:20 -!- zzo38 has joined.
18:36:45 <esowiki> [[Hurgusburgus]] N https://esolangs.org/w/index.php?oldid=58653 * BradensEsolangs * (+3687) Created page with "'''Hurgusburgus''' is a programming language by [[User:BradensEsolangs|Braden]], to be very confusing. ==The basics of Hurgusburgus== Hurgusburgus is based on a [[queue]]. A p..."
18:37:13 <esowiki> [[User:BradensEsolangs]] https://esolangs.org/w/index.php?diff=58654&oldid=58644 * BradensEsolangs * (+43)
18:39:02 -!- zzo38 has quit (Ping timeout: 250 seconds).
18:45:02 -!- MDude has joined.
18:46:11 <esowiki> [[Brainfuck]] https://esolangs.org/w/index.php?diff=58655&oldid=57519 * BradensEsolangs * (-11) Hmmm, brain is not a swear word
18:47:27 -!- zzo38 has joined.
18:51:24 <esowiki> [[Language list]] https://esolangs.org/w/index.php?diff=58656&oldid=58566 * BradensEsolangs * (+19) /* H */
18:57:57 <esowiki> [[Special:Log/newusers]] create * Baidicoot * New user account
18:58:58 -!- FreeFull has joined.
19:06:06 -!- b_jonas has joined.
19:08:47 <esowiki> [[Esolang:Introduce yourself]] M https://esolangs.org/w/index.php?diff=58657&oldid=58635 * Baidicoot * (+253) /* Introductions */
19:46:51 <esowiki> [[Butng]] N https://esolangs.org/w/index.php?oldid=58658 * Baidicoot * (+1317) Created page with "== What is Butng? == Butng was originally created by [User:Baidicoot|Baidicoot] as an extension of lambda calculus. However, it has since grown, and, with the help of [User:Go..."
19:59:49 <esowiki> [[Butng]] https://esolangs.org/w/index.php?diff=58659&oldid=58658 * Baidicoot * (+820)
20:08:25 <esowiki> [[Brain-Flak]] https://esolangs.org/w/index.php?diff=58660&oldid=58536 * BradensEsolangs * (+4)
20:11:29 <esowiki> [[Truth-machine]] https://esolangs.org/w/index.php?diff=58661&oldid=58606 * BradensEsolangs * (+47) /* Implementations */
20:21:37 <esowiki> [[Butng]] M https://esolangs.org/w/index.php?diff=58662&oldid=58659 * Baidicoot * (+10) /* Placeholder Law */
20:22:17 <esowiki> [[Butng]] https://esolangs.org/w/index.php?diff=58663&oldid=58662 * Baidicoot * (+14)
20:29:28 -!- AnotherTest has joined.
20:42:16 <mniip> HELO
20:42:25 <mniip> help me design a language
20:44:38 -!- arseniiv has quit (Ping timeout: 245 seconds).
20:48:01 -!- imode has quit (Quit: WeeChat 2.3).
21:07:56 <zzo38> What language?
21:09:56 <esowiki> [[Butng]] https://esolangs.org/w/index.php?diff=58664&oldid=58663 * Zzo38 * (-14) User links weren't working; the user pages still do not exist yet
21:10:16 <mniip> something that is the next step above assembler
21:10:33 <mniip> with a real simple compiler
21:10:45 <mniip> I'm thinking forth-like
21:11:38 <mniip> but with functions, stackframes, and "backwards compatibility" with machine code in terms of data/text mixins
21:22:01 <b_jonas> mniip: by `"backwards compatibility" with machine code", you mean a well-defined ABI?
21:22:21 <b_jonas> or a foreign call interface or both?
21:22:38 <b_jonas> possibly a trivial one
21:22:44 <mniip> inline machine code
21:23:33 <b_jonas> ah
21:23:35 <b_jonas> ok
21:23:37 <b_jonas> so that too
21:24:15 <b_jonas> mniip: and what platform/architecture does the compiler compile to?
21:25:35 <mniip> 8086
21:26:21 <mniip> I am willing to ignore segmentation and consider that everything is bound to the CS segment
21:27:42 <zzo38> Would it be capable of making DOS programs?
21:27:44 <b_jonas> mniip: the original 8086? 16 bit mode?
21:27:51 <mniip> zzo38, COM programs yeah
21:28:18 <zzo38> OK. COM already sets the code and data segments same automatically, so you do not have to do that.
21:28:47 <mniip> zzo38, what I meant is I don't have to design the language around far pointers and the like
21:28:51 <b_jonas> all four of the original 8086 segment registers in fact
21:30:08 <b_jonas> which is useful because 8086 instructions implicitly refer to all four in some cases
21:30:25 <b_jonas> so if you stick to a single segment, then you never have to add segment overrides
21:36:08 <b_jonas> mniip: and what do you want the compiler to run on?
21:36:18 <b_jonas> and implemented in?
21:36:29 <mniip> 8086
21:36:30 <mniip> 8086
21:39:05 <b_jonas> mniip: there are some small forth native compilers for 8086, colorforth being one of the more infamous ones, you could look at those.
21:39:19 <b_jonas> though those are partly implemented in themselves, rather than raw 8086
21:42:44 <mniip> I wonder if I need keywords
21:44:19 <zzo38> There is possibility to have keywords but that aren't reserved words.
21:45:06 <mniip> nah I mean symbolic keywords vs alphanumeric keywords
21:45:22 <zzo38> O, OK
21:45:35 <mniip> "?" vs "if"
21:46:15 <zzo38> Yes, OK, I know now how you mean
21:46:20 <oren> mniip: if you have a stropping system, it makes it easier to have keywords
21:46:35 <mniip> a what system
21:47:03 <oren> mniip: like if all keywords start with a particualr symbol
21:47:11 <zzo38> Requiring apostrophes around all keywords
21:47:43 <oren> or in fortran, less than is .lt.
21:48:11 <zzo38> (You can also do it other way around, or require a prefix symbol for nonkeywords like LLVM does, or do like SQL does that you can optionally put quotation marks around a name if you want to make it to be treated as not a keyword)
21:49:35 <b_jonas> mniip: or, it's not a forth-alike, but you can look at http://www.vttoth.com/CMS/index.php/projects/49
21:49:47 <b_jonas> only you'd implement it in 8086 directly, not in itself
21:50:11 <b_jonas> it doesn't have inline machine code, only machine code as separate functions, but it wouldn't be hard to add inline machine code with some register use convention
21:51:22 <b_jonas> as in, add an emit statement that lets you directly emit bytes of machine code, 16-bit addresses of global variables or functions, and 8-bit address of local variables relative to BP
22:07:54 <b_jonas> apparently there's some disagreement in how the libc fscanf function should handle inputs that require more than one byte of lookahead, which comes up in the %g, %x and %i formats.
22:11:28 <mniip> you mean wrt the file pointer?
22:12:40 <b_jonas> mniip: that too, but more whether the partial input is accepted (as in, fscanf fills the variable with the parsed value and continues on with the pattern, as opposed to rejecting the partial input by not modifying the variable and stopping the pattern right there)
22:25:08 <esowiki> [[Hurgusburgus]] https://esolangs.org/w/index.php?diff=58665&oldid=58653 * BradensEsolangs * (+45)
22:25:44 <esowiki> [[Truth-machine]] https://esolangs.org/w/index.php?diff=58666&oldid=58661 * BradensEsolangs * (+0) /* Hurgusburgus */ Should be a zero, not a space.
22:27:26 <esowiki> [[Hurgusburgus]] https://esolangs.org/w/index.php?diff=58667&oldid=58665 * BradensEsolangs * (+5)
22:34:21 -!- sleepnap has quit (Quit: Leaving.).
22:41:27 -!- zzo38 has quit (Ping timeout: 240 seconds).
22:42:02 -!- zzo38 has joined.
22:46:15 -!- AnotherTest has quit (Ping timeout: 264 seconds).
22:59:08 <esowiki> [[Brainfuck]] https://esolangs.org/w/index.php?diff=58668&oldid=58655 * Rdebath * (+11) Undo revision 58655 by [[Special:Contributions/BradensEsolangs|BradensEsolangs]] ([[User talk:BradensEsolangs|talk]]) (Someone else didn't get the joke.)
23:08:48 -!- tromp has joined.
23:24:34 <mniip> ok, I think I came up with a specification
23:28:24 <int-e> scary
23:29:20 <mniip> :add :1{ @2 @1 + $: }
23:30:23 <mniip> no wait that doesn't work
23:30:27 <mniip> :add :0{ @2 @1 + $: }
23:30:29 <mniip> or
23:30:49 <mniip> :add :1{ @2 @1 + $=0 $: }
23:31:53 -!- tromp has quit (Remote host closed the connection).
23:32:29 -!- tromp has joined.
23:33:10 <int-e> mniip: are you inventing a stack-based virtual machine?
23:33:51 <mniip> int-e, a mix of forth and assembler
23:42:21 -!- Melvar has quit (Ping timeout: 244 seconds).
23:43:16 <b_jonas> mniip: that's not a specification
23:43:28 <mniip> b_jonas, correct, those were examples
23:43:48 <mniip> probably not useful to you but helps me clear up my mind!
23:44:27 -!- zzo38 has quit (Ping timeout: 240 seconds).
23:44:57 -!- zzo38 has joined.
23:47:30 -!- tromp has quit (Remote host closed the connection).
23:47:46 -!- tromp has joined.
23:48:16 -!- Essadon has quit (Quit: Qutting).
23:56:33 -!- Melvar has joined.
←2018-12-09 2018-12-10 2018-12-11→ ↑2018 ↑all