00:25:08 -!- zgrep has quit (Ping timeout: 246 seconds).
00:26:27 -!- zgrep has joined.
00:40:14 -!- hppavilion[1] has quit (Ping timeout: 244 seconds).
00:44:05 -!- tromp has joined.
00:48:13 -!- tromp has quit (Ping timeout: 246 seconds).
00:58:56 -!- Phantom__Hoover has joined.
01:03:22 -!- Phantom_Hoover has quit (Ping timeout: 272 seconds).
01:36:51 -!- Patashu has quit (Ping timeout: 255 seconds).
01:38:25 -!- boily has quit (Quit: MALARIA CHICKEN).
01:38:26 -!- Patashu has joined.
02:01:15 <zzo38> Do you like tsume shogi?
02:07:50 -!- Patashu has quit (Ping timeout: 240 seconds).
02:08:57 -!- Patashu has joined.
02:14:06 -!- Phantom__Hoover has quit (Read error: Connection reset by peer).
02:18:18 -!- tromp has joined.
02:20:06 <HackEgo> [wiki] [[Help, WarDoq!]] M http://esolangs.org/w/index.php?diff=43889&oldid=43885 * Dennis * (+1)
02:20:23 -!- Wright has quit (Read error: Connection reset by peer).
02:20:40 -!- Wright has joined.
02:33:00 -!- tromp has quit (Remote host closed the connection).
02:34:41 -!- tromp has joined.
02:41:56 <oren> I recently got the angre video game nerd video game for WiiU
02:51:02 -!- Patashu has quit (Ping timeout: 272 seconds).
02:52:39 -!- Patashu has joined.
02:59:28 -!- Patashu has quit (Ping timeout: 250 seconds).
03:00:12 -!- Patashu has joined.
03:04:49 -!- Patashu has quit (Ping timeout: 256 seconds).
03:05:05 -!- Patashu has joined.
03:19:07 -!- Patashu has quit (Ping timeout: 260 seconds).
03:21:21 -!- Patashu has joined.
03:28:05 -!- tromp has quit (Remote host closed the connection).
03:39:42 -!- Patashu has quit (Ping timeout: 255 seconds).
03:41:18 -!- Patashu has joined.
03:44:58 -!- tromp has joined.
03:54:30 -!- Patashu has quit (Ping timeout: 250 seconds).
03:55:16 -!- Patashu has joined.
04:24:37 -!- MDude has changed nick to MDream.
04:36:18 <pikhq> Oh for the love of fuck WHAT IS GOING ON. LANG is somehow getting explicitly set to "C" somewhere between my profile getting loaded and my X session getting loaded.
04:57:07 -!- Wright has quit (Ping timeout: 260 seconds).
05:03:51 <zzo38> I am making a robotfindskitten game with the new VM that I had recently invented; please tell me if you have some more idea what not-kitten message to add
05:03:56 -!- oerjan has quit (Quit: nuit).
05:06:52 <Hoolootwo> you should add all of the not-kitten messages from the standard game, plus suggestions
05:07:37 <zgrep> zzo38: http://lpaste.net/raw/139475
05:08:43 <zgrep> Seems to also be here? https://github.com/oriontransfer/RobotFindsKitten-python/blob/master/messages.rfk
05:09:40 <zzo38> I have some of my own to add too though. But, I intend to put mine in alphabetical order, and also to put them in uppercase due to the way that the text packing of this VM is working, and limit the text length since this VM uses a screen size of only 80x25 tiles (the rest of the screen is not used by this VM)
05:10:07 <zzo38> That's OK though I have started to add them
05:10:59 <zgrep> "It's cute and furry, but not kitten."
05:11:50 <zgrep> "'Meow', says the office chair and it rolls around."
05:12:35 <zzo38> The pieces don't move in this game
06:04:31 <zzo38> This is the program so far: http://sprunge.us/TIDA
06:25:06 <zzo38> [:counter [!_:a :target :spell]; :replacing [:match [:zone-change [:to :graveyard]]; :if-reason [:counter _:a]; :replace [:zone-change [:to :library; :who [:owner-of _:a]; :position :top]]] # Or maybe not quite like that
06:49:26 -!- Patashu has quit (Ping timeout: 240 seconds).
07:30:44 <myname> why did you invent a vm?
08:10:15 -!- AnotherTest has joined.
08:42:38 -!- salvob has joined.
08:43:32 -!- salvob has quit (Client Quit).
09:00:45 -!- Trioxin has quit (Read error: Connection reset by peer).
09:02:01 -!- Trioxin has joined.
09:25:13 <zzo38> What should be the distribution of the line, char, and byte offset numbers in the error message of "isutf8 /dev/urandom"?
09:42:22 <zzo38> zgrep: I did add "It's cute and furry, but not kitten." though, but the second one I don't want to because the pieces won't move other than the main robot
09:43:40 <zzo38> O, for some reason the first time I run this VM after xterm opens the margins don't get set properly but afterward it works.
09:43:56 -!- |f`-`|f has quit (Ping timeout: 246 seconds).
09:44:27 -!- |f`-`|f has joined.
09:58:18 -!- Patashu has joined.
10:03:55 -!- AnotherTest has quit (Ping timeout: 244 seconds).
10:05:02 -!- AnotherTest has joined.
10:15:23 -!- lleu has quit (Quit: That's what she said).
10:15:42 -!- lleu has joined.
10:19:51 -!- Phantom_Hoover has joined.
11:31:09 -!- nortti has changed nick to lawspeaker.
11:35:40 -!- lawspeaker has changed nick to nortti.
11:43:46 -!- AnotherTest has quit (Ping timeout: 246 seconds).
11:45:37 -!- AnotherTest has joined.
11:53:18 -!- J_Arcane has joined.
11:54:15 <fizzie> "Alas, the customer fell into one of the common gotchas when writing p/invoke: They confused BOOLEAN and BOOL. BOOL is a 32-bit integer, whereas BOOLEAN is an 8-bit integer." But of course.
11:58:46 <fizzie> http://blogs.msdn.com/b/oldnewthing/archive/2015/08/17/10635549.aspx
12:08:27 -!- b_jonas_ has changed nick to b_jonas.
12:09:39 <HackEgo> [wiki] [[Talk:???]] http://esolangs.org/w/index.php?diff=43890&oldid=43769 * Rdebath * (+333) Reduce apostrophes?
12:16:42 -!- AnotherTest has quit (Ping timeout: 244 seconds).
12:17:45 -!- AnotherTest has joined.
12:19:33 -!- boily has joined.
12:55:18 -!- Patashu has quit (Ping timeout: 250 seconds).
12:56:28 <HackEgo> [wiki] [[Talk:Help, WarDoq!]] N http://esolangs.org/w/index.php?oldid=43891 * Rdebath * (+181) /* Fail. */ new section
12:58:23 <HackEgo> [wiki] [[Talk:Pb]] http://esolangs.org/w/index.php?diff=43892&oldid=43853 * Rdebath * (+90) /* Visibility as it's running. */
13:05:26 * ashl wonders if boily has highlights for those greetings
13:07:26 <quintopia> if he doesnt by now, its a major oversight
13:15:14 -!- copumpkin has joined.
13:41:03 -!- copumpkin has quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…).
13:46:46 <boily> I'd like to, but I don't know how.
13:48:16 <boily> I should breakfast too. had coffee, did laundry, time to eat.
13:50:34 <quintopia> weechat seems hard to find documentation for
13:57:16 -!- AnotherTest has quit (Ping timeout: 250 seconds).
13:58:35 -!- AnotherTest has joined.
14:00:21 -!- AnotherTest has quit (Read error: Connection reset by peer).
14:14:04 -!- AnotherTest has joined.
14:22:32 -!- AnotherTest has quit (Ping timeout: 246 seconds).
14:23:56 -!- AnotherTest has joined.
14:31:03 <HackEgo> [wiki] [[.box]] http://esolangs.org/w/index.php?diff=43893&oldid=14509 * 174.93.62.143 * (-1)
14:43:20 <izabera> https://github.com/rdebath/Brainfuck/blob/master/bitwidth.b how long is this supposed to take on an interpreter with 8bit cells?
14:44:05 <izabera> mine displays the text pretty quickly, then sleeps for a while, then prints the newline and exits
14:48:13 <quintopia> wow. no one has ever successfully run that program to completion before!
14:49:03 <quintopia> (i'm kidding. are you using the fastest interpreter according to rdebath's table?)
14:50:52 <izabera> i'm using one i wrote myself
14:51:49 <quintopia> then it takes how long it took! but probably longer than the optimized interpreters
14:52:41 <fizzie> https://esolangs.org/wiki/User:Rdebath#Performance_Matrix
14:53:28 -!- AnotherTest has quit (Ping timeout: 246 seconds).
14:54:27 -!- AnotherTest has joined.
15:01:05 -!- AnotherTest has quit (Read error: Connection reset by peer).
15:06:00 -!- AnotherTest has joined.
15:07:50 <ashl> izabera: is yours written in bash
15:09:30 -!- copumpkin has joined.
15:32:00 <J_Arcane> https://github.com/philipl/pifs
15:33:55 -!- impomatic_ has joined.
15:36:05 -!- copumpkin has quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…).
15:42:49 -!- AnotherTest has quit (Read error: Connection reset by peer).
15:45:43 -!- AnotherTest has joined.
15:52:24 -!- boily has quit (Quit: AFFILIATED CHICKEN).
16:13:54 <tswett> Hey all y'everyone. I came up with a cryptic crossword clue.
16:14:08 <tswett> Job description: Endless mixed investigations (6).
16:14:55 <tswett> Oh crap, that doesn't work. Lemme release a bug fix.
16:15:29 <tswett> Job description: A variety of endless investigations (6).
16:19:27 <oren> holy shit I actually beat a level!
16:22:13 <oren> izabera: how optimized is it, and what language?
16:30:04 <ashl> a level of what?
16:34:19 <oren> the AVGN video game
16:39:15 <oren> Maybe I should fix up my bf interpreter so it works and test it
16:47:26 -!- Froox has quit (Ping timeout: 246 seconds).
16:59:47 -!- atrapado has joined.
17:00:07 <izabera> ashl: yes oren: it reduces: ++[-]++ to 1 instruction, >>--++<< to nothing, [->+>++>+++<<<] to 1 instruction, [>+<[-]] to an if
17:01:36 -!- ais523 has joined.
17:07:26 <izabera> (1 instruction in bash can assign more than a variable)
17:09:53 -!- sebbu2 has joined.
17:10:18 -!- Wright has joined.
17:10:33 -!- sebbu has quit (Ping timeout: 252 seconds).
17:10:40 -!- sebbu2 has quit (Changing host).
17:10:40 -!- sebbu2 has joined.
17:17:19 -!- sebbu2 has changed nick to sebbu.
17:19:51 <ashl> izabera: you seem to be fond of bash
17:20:46 <int-e> oren: what kind of level did you beat?
17:25:47 <izabera> did you know that 10 to the 12th power microphones = 1 megaphone
17:31:30 -!- hppavilion[1] has joined.
17:35:33 -!- boily has joined.
17:36:56 <boily> quintopia: QUNTHELLOPIA!
17:37:09 <oren> "blizzard of balls" level
17:39:30 <boily> quintopia: security upgrading my laptop. will be available tantôt.
17:39:41 <boily> helloren. that sounds squishy and painful.
17:39:47 <HackEgo> [wiki] [[Talk:Folder]] http://esolangs.org/w/index.php?diff=43894&oldid=43887 * Hppavilion1 * (+130) /* A Folder RPG */ new section
17:40:40 <mauris> tswett, i'm thinking "synonym for investigation -> chop off final letter -> pluralize -> anagram" and it's a 6-letter job description?
17:41:10 <tswett> mauris: you're very much on the right track.
17:41:18 <mauris> so without the plural -s or -es, it's 4 or 5 letters, and then with the final letter, the original word is 5 or 6 -- probe, search?
17:42:17 <tswett> I could tell you the actual interpretation.
17:43:01 <mauris> "searcs" doesn't seem to anagram to anything. eh maybe give me a push in the right direction
17:44:14 <mauris> (cryptic crosswords are kind of bullshit imo. it's like word games crossed with mind reading)
17:44:30 <boily> obiously it's an "acress": a woman who acres.
17:49:45 <mauris> tswett, i feel like "chop off final letter" needs to be before "pluralize" or they'll just cancel each other out, and "-> ... -> ... -> plural" doesn't work or else the answer is a plural... hm
17:50:49 <tswett> They don't always cancel out. If you take "acress", then pluralize and chop off the last letter, you'll end up with "acresse".
17:51:59 <mauris> yeah, earlier i was looking at "study" -> "studies" -> "studie"... -> "duties"?
17:52:36 <mauris> that's almost a job description, in a very abstract way
17:53:22 <mauris> maybe, of course, the word i need isn't on here: http://www.thesaurus.com/browse/investigation?s=t
17:53:55 <ais523> tswett: how do you know what the plural of "acress" is, given that it isn't a real word?
17:54:34 <tswett> mauris: that's the answer. DUTIES.
17:54:56 <tswett> ais523: I assume "acress" is pronounced as I would expect and pluralizes regularly.
17:55:13 <mauris> ais523, clearly tswett passes the Wug Test
17:56:10 <boily> quintopia: AAAAAAAAAAAAAAAAURGH. "Could not connect to the Steam Network."
17:56:49 <boily> tswellott. "acress" is a made up word that I made up on the spot of making it up.
17:56:55 <b_jonas> an acress is an actress without a tee
17:57:20 <boily> b_jhellonas. I wouldn't be surprised if you were dictionarily right. English is weird.
17:58:13 <boily> quintopia: huh. just had to insist and be intimidating. it logge in.
17:59:35 -!- Phantom_Hoover has quit (Ping timeout: 265 seconds).
18:02:53 <int-e> boily: I get that a lot, too; it takes me two or three time to finally connect, typically. Which means I have to look up the password because the stupid client "forgets" it the first time connection fails.
18:03:21 <int-e> s/"forgets"/discards/
18:08:43 <HackEgo> [wiki] [[Special:Log/newusers]] create * Beta-Decay * New user account
18:11:39 -!- hppavilion[1] has quit (Ping timeout: 244 seconds).
18:15:47 -!- augur has quit (Quit: Leaving...).
18:18:34 <HackEgo> [wiki] [[Fourier]] N http://esolangs.org/w/index.php?oldid=43895 * Beta-Decay * (+2000) First post
18:23:30 -!- hppavilion[1] has joined.
18:27:47 <HackEgo> [wiki] [[Fourier]] http://esolangs.org/w/index.php?diff=43896&oldid=43895 * Beta-Decay * (+280)
18:28:30 <oren> Oh, now it work!
18:29:40 <oren> http://www.orenwatson.be/bfim.htm
18:31:45 <izabera> also what's the advantage of using loop{ }
18:31:54 <oren> looks better to me
18:32:38 <oren> this interpreter for brainfuck uses space below the call stack as code, loop stack and data space
18:32:51 <izabera> which compiler uses #include "" for its core libraries?
18:32:52 <zgrep> My eyes seem to avert themselves from that font...
18:33:22 <oren> uh... all of them
18:33:36 <oren> well it comiles
18:33:41 <ais523> izabera: although #include includes header files, not libraries
18:33:47 <ais523> the libraries are linked by seeing which functions you actually use
18:33:53 <fizzie> "(char*)&c-1000" I'm weirded out.
18:33:55 <ais523> this means that if you declare the functions yourself, and use no header files
18:34:02 -!- Phantom_Hoover has joined.
18:34:10 <ais523> it'll still work as long as the declarations are correct
18:34:14 <b_jonas> oren: though there's a problem with the font as used there, at least one my side:
18:34:27 <oren> i mean there are way worse things going on here than bad includes
18:34:29 <izabera> at least highlight loop and ei
18:34:46 <b_jonas> oren: it's subpixel anti-aliased, probably because the whole thing is shifted a non-integer number of pixels horizontally
18:34:52 <ais523> fizzie: return (void *)(uintptr_t)(offset + __builtin_offsetof(struct user, regs));
18:35:07 <b_jonas> oren: it's visible only on the white letters
18:35:15 <b_jonas> oren: is there something you can do to fix that?
18:35:37 <oren> I don't know. the font is set to 12px
18:35:42 <b_jonas> oren: as for style, the equal sign glyph looks strange, the two bars are too far from each other
18:35:50 <fizzie> ais523: Well, okay, but presumably you know what you're pointing at. This is just a regular int c;.
18:35:53 <ais523> tbf that's a little less insane than it looks, because it's dealing with values in registers
18:35:57 <izabera> did you forget to implement , ?
18:36:07 <ais523> yeah that's ridiculous (and also violates the standard)
18:36:30 <b_jonas> izabera: loop{} looks like a rubyism to me
18:36:30 <ais523> fizzie: the pointer doesn't actually point at anything really, it typically comes out to a small integral void * value
18:36:56 <ais523> but this is because there are a ton of different address spaces involved (at least three of them), plus the code doesn't have enough information to know what's integers and what's pointers
18:37:00 <oren> b_jonas: really? I don't know ruby
18:37:33 <HackEgo> [wiki] [[Fourier]] http://esolangs.org/w/index.php?diff=43897&oldid=43896 * Beta-Decay * (+49)
18:37:36 <b_jonas> oren: in ruby, the standard library has a function called loop, so you can write infinite loops as loop { puts "hello, world"; };
18:38:29 <b_jonas> (you can use a while loop instead, of course)
18:40:06 <izabera> how does that allocate memory?
18:40:08 <oren> anyway the main conceit of this is to take the address of sthg on the stack, subtract 1000, then use the memory below that as we please
18:40:15 <quintopia> boily: ok. had to reboot and install updates ,but i'm connected again
18:40:33 <b_jonas> oren: as for the font, it also looks a bit vertically crowded for me, that is, there's too few space between lines
18:40:56 <oren> izabera: what what
18:41:23 <b_jonas> oren: also, isn't the dot and comma to similar, and the comma and semicolon too simialr?
18:41:39 -!- boily has quit (Ping timeout: 250 seconds).
18:41:59 <oren> on my screen , looks twice the size of .
18:42:22 <b_jonas> oren: well, the comma is seven pixels and the dot is four
18:42:25 <b_jonas> so there is some difference
18:42:28 <b_jonas> but I'm used to a bit more
18:42:53 -!- augur has joined.
18:43:03 <oren> my font, and also the bf interpreter I wrote
18:43:06 <b_jonas> in my font, the dot is four pixels, the comma is ten
18:43:16 <b_jonas> though that's a larger font, admittedly
18:43:17 <oren> http://www.orenwatson.be/bfim.htm in case you missed it
18:44:00 <quintopia> wow, i can see why we're discussing your font. i can hardly read it.
18:44:10 <b_jonas> oren: dunno, it's possible that the font is just too small for me, which is why I find some characters too similar
18:44:37 <oren> izabera: Anyway, the stack grows downward, and none of the functions use more than 1000 bytes of it, so the memory i'm using is safe
18:45:10 <oren> b_jonas: do you by any chance have a high dpi monitor?
18:45:48 <quintopia> is it usual for a bf interpreter to use a stack to track bracket locations?
18:45:51 <b_jonas> afterall, this is only 9x16, whereas my font is 10x20
18:46:06 <b_jonas> oren: should I send a screenshot?
18:46:37 <oren> that would help.
18:47:09 <oren> quintopia: I don't know. I came up with it a while ago but assumed it was obvious idea
18:48:29 <quintopia> oren: i agree with izabera that that data=code=(char*)&c-1000 looks dangerous.
18:48:30 <oren> about the space between lines, I deliberately made that small for this font because I like having a lot of SLOC onscreen at once
18:49:06 <oren> it is entirely illegal and only works under certain cpus and OS's
18:49:36 <oren> specifically linux x86 and linux x64 it works
18:50:10 <quintopia> so no intention of creating cross-platform interpreter then
18:50:29 <oren> it is completely undefined behaviour to acess data in this manner
18:50:32 <quintopia> i feel like one could write malicious code to break it
18:51:10 <oren> although i'm not sure how
18:51:27 <fizzie> I'm not weirded out by nasty tricks, I'm weirded out because there seems to be no reason for it in this case.
18:51:39 <b_jonas> oren: http://www.math.bme.hu/~ambrus/pu/oren-neoletters-ss0.png
18:52:07 <oren> the specific hack here is that calloced memory doesn't auto-extend on write, whereas the stack does
18:52:45 <b_jonas> fizzie: what? yuo're on #esoteric
18:52:51 <oren> so you can keep moving right and the tape will autoextend without copying anything with realloc
18:53:18 <b_jonas> oren: you could malloc a new auto-extending section, though that's also hard to do portable among systems,
18:53:33 <oren> yeah there is definitely antialiasing going on there
18:53:40 <oren> which I don't hav
18:53:51 <b_jonas> oren: and in both that case and this case you have to make sure to occasionally access the data if the code contains like a hundred thousand > commands followed by a +
18:54:00 <b_jonas> oren: but only subpixel antialiasing
18:54:08 <b_jonas> oren: the non-white characters are fine
18:54:26 <b_jonas> no wait, yellow aren't good either
18:55:03 -!- hppavilion[1] has quit (Ping timeout: 244 seconds).
18:55:04 <b_jonas> well, it looks like it's shifted to the right by part of a pixel
18:55:10 <b_jonas> probably the html is layed out that way
18:55:12 <quintopia> b_jonas: what if the code contains simply [>+]
18:55:27 <b_jonas> quintopia: that would eventually run out of memory, sure
18:55:32 <fizzie> The read(0,&c,1); is another thing that seems like gratuitously designed to break on non-little-endian systems, compared to just calling getchar.
18:55:37 -!- MDream has changed nick to MDude.
18:56:13 <fizzie> The stack generally doesn't auto-extend very much, compared to how much memory there would be available via regular memory allocation.
18:56:30 <b_jonas> quintopia: what mallocing a new area could help in is catching accesses to the left of the starting point of the tape
18:56:33 <oren> http://www.orenwatson.be/wskfkflasndvfals.png
18:57:31 <quintopia> b_jonas: aye. it's good to know where the bounds are.
18:57:59 <oren> it doesn't let you go beyond the bounds
18:58:12 <oren> there is ifs to catch that
18:58:32 <oren> }ei(*ip=='<'){if(dp<data)dp++;
18:59:32 <b_jonas> oren: did you manage to add more characters to the font though?
18:59:54 <oren> There are 2080 or so
19:00:31 <int-e> } else if (*ip == '<') { if (dp < data) dp++;
19:01:03 <oren> int-e: you don't like my spacing?
19:01:18 <b_jonas> that's way more than mine. I guess it adds a lot that you have cyrillic or greek characters (iirc)
19:01:40 <oren> http://www.orenwatson.be/fontdemo.htm
19:02:31 <b_jonas> that's a lot of latin characters
19:02:49 <oren> yeah I did LAtin1,b and IPA
19:03:34 <b_jonas> I don't have all the crazy characters from the 0x0200 block, the ones used only for transcribing African languages or whatever stuff
19:03:50 <b_jonas> do you have all letters for Vietnamese?
19:04:20 <oren> Does that require latin C or D?
19:04:23 <b_jonas> I don't have those either, though it wouldn't be vary practical either, if I wanted Vietnamese I'd need a font with slightly more space above characters
19:04:25 <oren> I didn't do those
19:05:05 <b_jonas> also, you have basic Greek
19:06:11 <b_jonas> not all the accented greek letters though. I think those add a lot, if you're counting characters.
19:06:12 <oren> and hebrew but no vowel points
19:07:25 <oren> but for example you can write א₀
19:07:29 <b_jonas> hmm, do you have \x{2009} ?
19:07:45 <b_jonas> I don't see it listed, but maybe it's hiding invisible
19:08:09 -!- boily has joined.
19:08:12 <oren> my punctuation is still spotty
19:08:48 <b_jonas> I'd recommend adding \x{2009} and other spaces. they're very easy to implement in a monospaced font.
19:09:08 <b_jonas> I see you have lots of punctuation and graphics characters
19:09:19 <b_jonas> do you have all of cp437, including the control characters?
19:10:24 <zzo38> Do you have the DEC technical set?
19:10:46 <b_jonas> zzo38: I think I have most of it, probably all
19:10:54 <b_jonas> I'm not sure if I've added the snowman
19:11:16 <b_jonas> zzo38: what character is the DEC snowman?
19:11:24 <fizzie> I've mentioned it quite a few times, but since it's kind of topical to the discussion, there's a bitmap font at http://zem.fi/rfk86/ too. No attempt to get good character coverage, though, since it's only for that page.
19:11:26 <b_jonas> and is that officially in the set?
19:11:54 <b_jonas> zzo38: I think I automatically have most of it though from just cp437
19:11:54 <MDude> I shoudl make a font on English characters where each one is altered to look more like the most ancient known ancector glyph of it.
19:11:55 <oren> I have most of them
19:12:12 <MDude> So an A would look like an upside down ox head.
19:12:58 <oren> I think I may have forgotten to add some to the demo page
19:13:23 <b_jonas> oren: what? I thought the demo page was automatically generated from all characters you have
19:13:31 <MDude> But then, I also want to make the most ridiculous text encoding system.
19:13:38 <b_jonas> I should make a demo screenshot automatically generated by the wya
19:13:50 <oren> I just keep adding them
19:14:32 <oren> yeah the editor says I have 2086 in total
19:14:44 <MDude> Where isntead of supporting fonts or character sets at all, it has each document start with a pallete of glyphs.
19:14:54 <b_jonas> hehe, you have all the obsolate greek letters, but not the accented greek letters
19:15:01 <MDude> So even switching between fonts qould need a conversion table.
19:15:24 <oren> Mdude: that would be nice for cross platform compatibility
19:15:26 <zzo38> I don't see any snowman in the DEC technical set?
19:15:46 <b_jonas> zzo38: I think it's on the letter G
19:16:04 <oren> yeah I did forget to add some
19:16:04 <MDude> It'd pretty much by an image file format that's biased toward typed documents.
19:16:11 <b_jonas> zzo38: do you mean the DEC special character set which terminal emulators have?
19:16:40 <oren> 239b thru 23ae are supported
19:16:57 <zzo38> There is the VT100 character graphics set and the technical set, they are different
19:16:59 <MDude> oren: Yeah, but it'd also have a really difficult relationship with anything that has to support identical looking but technically different characters.
19:17:26 <zzo38> (Although much of them can be found in the PC set as well, some aren't. Some of the technical characters are not even in Unicode either actually)
19:17:59 <oren> there I added it to the demo
19:18:51 <MDude> I guess the main difference between it and systems made for 8-bit tables is that it wouldn't specify an upper limit to the number of characters.
19:19:26 <zzo38> The pieces of the large sigma aren't in Unicode, according to a webpage I found
19:20:35 <zzo38> MDude: You still might want to specify a limit of 32-bits or whatever; the DVI format has a limit of 32-bits per character code
19:20:44 <zzo38> (Although TeX can use only 8-bits)
19:22:38 <MDude> I'll consider it, though I think what I was thinking at the time was making something for weird made-up and procedurally generated runes.
19:22:46 <b_jonas> zzo38: I'm actually missing a lot of the DEC special set: ▚☃␉␌␍␊␋
19:22:53 <MDude> For use with comptuers that I pretend are made by aliens.
19:23:20 <HackEgo> [wiki] [[Fourier]] http://esolangs.org/w/index.php?diff=43898&oldid=43897 * 50.170.122.255 * (+4) Fixed markdown syntax for italics
19:23:26 <HackEgo> [wiki] [[Truth-machine]] http://esolangs.org/w/index.php?diff=43899&oldid=43814 * Undergroundmonorail * (+41) added pb
19:24:26 <b_jonas> I don't use many programs using these, but the few I've seen use only the line drawing characters like ↑↓→←█┘┐┌└┼⎺⎻─⎼⎽├┤┴┬│·
19:25:49 <oren> `unidecode⎺⎻─⎼⎽
19:25:56 <oren> `unidecode ⎺⎻─⎼⎽
19:26:12 <HackEgo> [U+23BA HORIZONTAL SCAN LINE-1] [U+23BB HORIZONTAL SCAN LINE-3] [U+2500 BOX DRAWINGS LIGHT HORIZONTAL] [U+23BC HORIZONTAL SCAN LINE-7] [U+23BD HORIZONTAL SCAN LINE-9]
19:26:13 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: unidecode⎺⎻─⎼⎽: not found
19:26:14 <b_jonas> ok, probably not all of those
19:26:25 <zzo38> I cannot get DEC technical characters to work on xterm
19:26:29 <oren> I have the light horizontal
19:26:38 <oren> but not the scan line ones
19:27:45 <b_jonas> zzo38: what font encoding in xterm?
19:27:50 <b_jonas> zzo38: and what terminal encoding?
19:28:21 <b_jonas> oren: the full DEC set is this: ↑↓→←█▚☃ ◆▒␉␌␍␊°±␋┘┐┌└┼⎺⎻─⎼⎽├┤┴┬│≤≥π≠£·
19:28:41 <zzo38> I am using the "fixed" set
19:28:42 <b_jonas> some of those might be non-portable extensions
19:28:48 <zzo38> (That is what the font is called.)
19:29:13 <zzo38> The VT100 graphics work, but when I try to use DEC technical characters I get ASCII instead
19:30:19 <ais523> zzo38: the way it works is, xterm (and most other terminals) have two character sets "G0" and "G1", and each of those sets are customizable
19:30:30 <b_jonas> zzo38: um, but what's the full name of the font? eg. -misc-fecupboard-medium-r-normal--20-200-72-72-c-0-iso10646-1 vs -misc-fecupboard-medium-r-normal--20-200-72-72-c-0-iso8859-2 are my fonts but depend on font encoding (first is utf-8, second is legacy)
19:30:35 <ais523> you use the SI and SO characters to switch between them , and escape codes to set them up
19:30:48 <ais523> I suspect that what's happening is that they're both set to ASCII, so switching between them isn't doing anything
19:30:55 <zzo38> ais523: Yes that is how I did it!
19:31:05 <b_jonas> zzo38: um, what do you mean by "VT100 graphics" if not the DEC charset?
19:31:20 <b_jonas> ais523: oh right, I should have asked you rightaway
19:31:29 <ais523> zzo38: try sending "ESC ) 0", i.e. \x1b)0
19:31:36 <b_jonas> ais523: are those characters I mentioned portable, or are they extensions in urxvt?
19:31:40 <ais523> that'll change G1 to be DEC characters
19:31:49 <zzo38> ais523: That works, but when I do "ESC ) >" then it does not work.
19:31:55 <ais523> b_jonas: they're sort-of quasi-portable
19:32:01 <ais523> zzo38: that's because xterm doesn't implement the > character set
19:32:16 <ais523> b_jonas: basically most terminals implement them but they don't all implement them quite the same way
19:32:31 <zzo38> The xterm documentation mentions it, and anyways it ought to implement it!
19:32:45 <ais523> after running my terminal testsuite on a bunch of terminals, I decided that making it work consistently on all terminals would fail
19:32:51 <ais523> maybe it does, I thought it didn't though
19:33:16 <b_jonas> I believe xterm with ixo-8859-1 or similar font encodings uses glyphs in control characters positions to support some of the DEC stuff,
19:33:22 <zzo38> (Clearly a font that supports it is needed, such as by extending the "fixed" font with these characters
19:33:27 <b_jonas> which is something of an X11 "standard" about font encoding
19:33:40 <zzo38> b_jonas: Yes, for the VT100 character graphics it does, I have looked at the fonts
19:33:47 <b_jonas> and xterm doesn't really like when the font encoding differs from the terminal encoding, though it does allow so
19:33:48 <ais523> (I was checking linux console docs as they're the most easily accessible for me, I have the xterm docs but only the source code for them and it's written in a *roff dialect, which means that the interpretation of the input depends heavily on command-line options and all the ones I've tried look ugly)
19:34:03 <b_jonas> I don't know what xterm does for utf-8 font encoding with utf-8 terminal encoding though
19:34:28 <ais523> the Unicode version of xterm doesn't support 8-bit character sets at all
19:34:38 <ais523> any bytes with the high bit set are interpreted as UTF-8
19:34:56 <b_jonas> Urxvt is different, it supports proper unicode fonts (utf-8 encoded x11 bitmap fonts, as long as they're monospaced) with fallbacks to multiple fonts, and uses proper unicode characters for all the DEC line-drawing stuff, plus some fallbacks.
19:35:26 <b_jonas> But I for one just don't use xterm these days, and if I have to use it as a fallback, then I don't much care how the DEC specials come out, as long as they don't completely mess up the screen layout
19:35:40 <ais523> I have two uses for xterm
19:35:46 <ais523> one is if I need to test how something looks on xterm
19:36:00 <b_jonas> the other is to bootstrap another terminal?
19:36:01 <ais523> the other is if I need to run a terminal /inside/ an insane environment (sandbox, container, valgrind etc.) for some reason
19:37:06 <zzo38> I use only xterm; the others have various problems
19:37:18 <zzo38> (I don't have any others installed)
19:37:32 <b_jonas> by the way, I have a modified terminfo I used to use for xterm, because the default xterm terminfo claims the backspace key generates \x08, whereas I configure xterm to make it generate \x7f (xterm allows either), and some programs actually believe that terminfo entry and won't accept \x7f.
19:37:45 <b_jonas> so my modified terminfo differs only in that it tells (correctly) that \x7f is generated
19:38:12 <b_jonas> (a possible workaround is to just use TERM=vt100 instead of TERM=xterm of course)
19:38:46 <b_jonas> Currently my only use for xterm is bootstrapping till I install urxvt
19:39:18 <b_jonas> (or possibly fallback if somehow I really break all installed copies of urxvt, which hasn't happend yet, but could happen)
19:39:23 <ais523> b_jonas: Debian did a lot of good for the Linux ecosystem by asking their maintainers to forcibly configure every terminal mode program to \x7f = delete-backwards
19:39:24 <zzo38> I would want a bit different kind of xterm, where some features are removed and others are added
19:39:32 <b_jonas> oren: anyway, that's a nice font
19:39:57 <b_jonas> ais523: most programs could just accept both
19:40:20 <b_jonas> (emacs is the exception, which uses control-h for a different purpose)
19:40:29 <b_jonas> and indeed, some programs do accept both, regardless terminfo
19:40:33 <ais523> b_jonas: NetHack also uses control-h for a different purpose
19:40:38 <ais523> although then, it doesn't use backspace for anything
19:41:09 <b_jonas> yeah, there's also the control-j vs control-m thing, which has sort of the same historical confusion
19:41:44 <ais523> b_jonas: its really interesting seeing the output from web of lies
19:41:49 <ais523> because it has \r\n newlines
19:42:13 <pikhq> Because terminal. :)
19:42:49 <b_jonas> as in iirc, on PC bios, enter generates \x0a and control-enter generates \x0d, but on linux traditionally enter generates \x0d (because of ancient history reasons) which the terminal layer translates to \x0a unless you turn that off in the stty settings
19:43:00 <zzo38> Features to add can include the DEC technical set (the ESC ) > set), ANSI music, support for scroll lock, status bar support, resource containing escape sequences to initialize the terminal with (whenever it is reset), some of the more obscure DEC features, complete Tektronix emulation, X mouse cursor selection, etc
19:43:03 <ais523> it's simulating a terminal atm and when a program outputs \n, the terminal driver sends an \r\n code to the actual "terminal" (that Web of Lies is emulating), at least with the configuration I'm using
19:43:30 <b_jonas> and what control-enter, control-backspace, control-space, control-2, control-3, control-4, control-5, control-6, control-7 generate is sort of random among terminals
19:43:56 <ais523> hmm, I'm not sure I tried control-digit on test_uncursed
19:44:19 <pikhq> Which means it's merely *almost* possible to implement a terminal-generic termcap/terminfo.
19:44:21 <b_jonas> zzo38: do you have a simple list of character numbers (say as hexadecimal) in your font somewhere, so I can quickly compare with mine?
19:44:39 <b_jonas> oren: or I can send the list for what my font has instead
19:44:59 <b_jonas> pikhq: well, ais523 sort of does that, but without the termcap/terminfo part
19:45:10 <b_jonas> pikhq: and implements the generic _library_, without the description stuff
19:45:21 <ais523> pikhq: I do have the "NumLock/F1" key
19:45:22 <pikhq> b_jonas: If you're happy not implementing termcap/terminfo *in particular* it's totally doable.
19:45:46 <pikhq> The main issue with termcap/terminfo is the API assumes a one-to-one mapping between code sequence and keys.
19:45:46 <ais523> and yes, why would I go via termcap/terminfo when they solve the wrong problem?
19:46:26 <b_jonas> pikhq: that's only one of the problems, but sure
19:46:29 <pikhq> There's a few different places where terminals disagree, but there's at least enough agreement (and fixed number of ways they disagree) such that it's totally possible to be terminal agnostic.
19:46:51 <pikhq> b_jonas: A lot of the other issues with being terminal agnostic are purely theoretical.
19:46:56 <pikhq> (though, sigh, not all of them.)
19:47:11 <b_jonas> pikhq: sure, though there are really broken terminals that you just can't support, like that one that generates the exact same escape sequence for the home and end keys
19:47:25 <pikhq> Yes, but that's just broken.
19:47:44 <zzo38> Features to remove can include TrueType, Unicode, toolbar, resource to change the text of the menus, widget sets other than Athena, etc
19:47:55 <pikhq> Let's be real, though, ain't nobody going to care about a terminal that breaks with ANSI color escape sequences. :)
19:48:36 <ais523> pikhq: those even work in DOS
19:48:42 <ais523> although basically nothing else does
19:48:48 <ais523> I managed to run my terminal testsuite in DOSbox
19:49:01 <pikhq> Yes, the ansi.sys logic is a very unusual *subset*.
19:49:13 <ais523> amazingly, it (i.e. DOSbox) supports enough of the codes I need that libuncursed output should theoretically look fine there
19:52:18 <pikhq> I've still got half a mind to try and write a mostly-terminal-agnostic libcurses, though.
19:53:14 <ais523> pikhq: why not just use libuncursed
19:53:26 <pikhq> Because a lot of stuff uses libcurses.
19:53:29 <ais523> or would it do something that libuncursed doesn't? (e.g. support for physical VT100s)
19:53:37 <ais523> libuncursed implements curses' API, mostly
19:53:48 <b_jonas> pikhq: do you need curses/libterminfo compatibility? do you need the input side (keyboard), the output side (drawing), or both?
19:53:51 <ais523> I guess there are a few exceptions, e.g. it has a separate init function
19:54:00 <ais523> the idea was to make it easy to port curses programs, rather than to be a drop-in replacement
19:54:11 <b_jonas> ais523: what? why would physical vt100s not work?
19:54:26 <pikhq> ais523: It'd be rather nice to be a drop-in replacement for some stuff, though.
19:54:26 <ais523> b_jonas: they need the other end of the connection to intentionally slow it down
19:54:39 <b_jonas> but can't the terminal layer do that?
19:54:39 <pikhq> Admittedly, libcurses sucks and no new application should use it.
19:54:46 <pikhq> Yes, but it doesn't.
19:54:56 <pikhq> Yes, but it doesn't.
19:55:02 <ais523> it's in the curses layer for reasons I don't fully understand
19:55:04 <zzo38> I have just written the escape codes directly
19:55:26 <ais523> probably whoever first wrote it had perms to change curses (which is user-mode) but not the (kernel-mode) settings stty changes
19:55:35 <b_jonas> ais523: maybe because the terminal layers handles it only when you use it as a dumb terminal, but doesn't know about how much to slow down for escape sequences
19:56:08 <b_jonas> I mean, it still has to work in non-curses applications that just use the canonical translation mode
19:56:11 <ais523> ISTR some of the delay calculatoins are really complex
19:57:02 <b_jonas> though at least you don't have to model how the carriage accelerates
19:58:02 <oren> a bunch of invisible characters
19:58:24 <HackEgo> [U+2000 EN QUAD] [U+2001 EM QUAD] [U+2002 EN SPACE] [U+2003 EM SPACE] [U+2004 THREE-PER-EM SPACE] [U+2005 FOUR-PER-EM SPACE] [U+2006 SIX-PER-EM SPACE] [U+2007 FIGURE SPACE] [U+2008 PUNCTUATION SPACE] [U+2009 THIN SPACE] [U+200A HAIR SPACE]
19:59:12 <ais523> hmm, my client doesn't display them (as in, not that it's drawing blank space to the screen: the characters aren't there at all)
19:59:24 <ais523> (I'm verifying via trying to select the text in question)
19:59:38 <ais523> presumably something in the display layer is trimming off trailing whitespace
20:03:41 <oren> It seems that some programs will display the glyphs that I drew for each one, others will just draw space.
20:04:31 <oren> in some places looks like a compressed 4/EM and others just a space
20:04:37 <ais523> oren: you mean your glyphs aren't entirely blank?
20:04:37 <b_jonas> ais523: what if you put them between two characters, like A B?
20:05:08 -!- hppavilion[1] has joined.
20:05:08 <oren> ais523: yes. that way you can tell what it is
20:05:08 <ais523> b_jonas: they're hard to type, and I can't copy-and-paste oren's :-)
20:05:25 <b_jonas> ais523: that's why I said A B
20:05:43 <b_jonas> oh, you mean in some other program than the client
20:05:45 -!- atrapado has quit (Quit: Leaving).
20:05:54 <ais523> in oren's comment they're rendering as a single space
20:06:01 <ais523> and yet all those spaces should be much wider than one space
20:06:13 <ais523> some are wider, some are narrower
20:06:13 <oren> http://www.orenwatson.be/fontdemo.htm here thry are on a web page
20:06:37 <oren> the first line under general punctiation has a bunch of them
20:06:42 <b_jonas> oren: oh, they're not blank there
20:07:45 <zzo38> The program that renders them shouldn't decide to put spaces or not; that should be the job of the font metric data to decide
20:08:21 <oren> ENQUAD EMQUAD ENSP EMSP 3/EM 4/EM 6/EM FIGSPA PUNSPA THINSPA HAIRSPA
20:09:46 <oren> all of which are rendered as single spaces because unicode doesn't specify how to deal with monospace fonts
20:10:01 <oren> or ratehr specifies it stupidly
20:10:32 <oren> emquad should be specified to be 4 wide
20:10:38 <b_jonas> question. HTML treats four particular space characters in the text part of html source code specially, as defined in http://www.w3.org/TR/html401/struct/text.html#h-9.1 . where is the section defining that in the html5 standard? or is it not there because it's off-loaded to the CSS standard only?
20:13:13 <oren> Oh, so that's why my zwsp character isn't being displayed
20:14:14 <oren> wait no, it's in a pre tag....
20:14:30 <b_jonas> maybe zwsp is treated specially in a different way?
20:14:49 <oren> yeah probably. maybe it is zero width
20:15:03 <oren> even though my glyph for it isn't.
20:16:19 <pikhq> ais523: BTW, the trac for nethack4 seems to be giving a permission error.
20:16:45 <ais523> pikhq: it's on a separate server, which is having problems
20:17:01 <ais523> I think it was set to 403 out "temporarily" to avoid exacerbating them
20:17:07 <ais523> but we're not quite sure what to do next
20:17:09 <b_jonas> what's the CSS property that sets html-style whitespace squashing?
20:19:41 <b_jonas> answering own question: property is called white-space
20:20:06 <oren> http://www.w3schools.com/cssref/pr_text_white-space.asp
20:21:04 <b_jonas> http://www.w3.org/TR/css3-text/#white-space-property
20:21:29 <zzo38> There may be several webpages, some servers might not use SQLite internally but some might, but I want to access them using SQLite-based client too, so using virtual table module extension you can access them with SQLite instead of web-browser. How to do that?
20:27:47 <oren> I think next I'll try to put in as many of the math operators as I can draw
20:28:13 <oren> that should be able to bump me past 2200
20:31:50 -!- atehwa_ has quit (Ping timeout: 265 seconds).
20:32:46 <b_jonas> hmm, I don't have replacement character (\xfffd) for some reasons
20:32:46 -!- villasukka has quit (Ping timeout: 248 seconds).
20:32:51 <b_jonas> I thought I had that in the font
20:33:24 -!- villasukka has joined.
20:33:24 -!- atehwa has joined.
20:38:49 <b_jonas> oren: the hex code ranges of characters I have in my font are: {00:109, 10c:11f, 122:125, 128:131, 134:13e, 141:148, 14a:17f, 192, 218:21b, 232:233, 250:252, 254, 258:25c, 25e:25f, 261, 263, 265, 268:26a, 26f:270, 272:273, 275:276, 278:279, 27e, 282:283, 289:28a, 28c:28d, 28f:290, 292, 294:295, 29d, 2c6:2c8, 2cc, 2d0, 2d8:2d9, 2db:2dd, 391:3a1, 3a3:3a9, 3b1:3c9, 1e80:1e85, 1ef2:1ef3,
20:38:56 <b_jonas> 2000:200d, 2010:2016, 2018:2022, 2024:2026, 202f:2030, 2032:2037, 2039:203a, 203c:203d, 204e, 207f, 20a7, 20ac, 211f, 2122, 2190:2195, 21a8, 21d0, 21d2, 21d4, 2200, 2202:2204, 2206:2209, 220b:220c, 220f, 2211:221a, 221e, 2223:222b, 2234:2236, 223c, 2248, 2260:2261, 2264:2267, 2282:2283, 2286:2287, 2295:2299, 22a0, 22a2:22a5, 22bb:22bd, 22c2:22c5, 22ee:22f1, 2300, 2302, 2308:230b, 2310, 2320:2321, 23b8:23bd, 2500, 2502, 250c, 2510, 2514, 2518, 251c, 2524,
20:39:18 <b_jonas> if you want, compare that to yours, or maybe give a full list of character codes you have so I can compare
20:39:39 <b_jonas> (wait, was that too long for a line?)
20:39:52 <b_jonas> {00:109, 10c:11f, 122:125, 128:131, 134:13e, 141:148, 14a:17f, 192, 218:21b, 232:233, 250:252, 254, 258:25c, 25e:25f, 261, 263, 265, 268:26a, 26f:270, 272:273, 275:276, 278:279, 27e, 282:283, 289:28a, 28c:28d, 28f:290, 292, 294:295, 29d, 2c6:2c8, 2cc, 2d0, 2d8:2d9, 2db:2dd, 391:3a1, 3a3:3a9, 3b1:3c9, 1e80:1e85, 1ef2:1ef3,
20:40:07 <b_jonas> 2000:200d, 2010:2016, 2018:2022, 2024:2026, 202f:2030, 2032:2037, 2039:203a, 203c:203d, 204e, 207f, 20a7, 20ac, 211f, 2122, 2190:2195, 21a8, 21d0, 21d2, 21d4, 2200, 2202:2204, 2206:2209, 220b:220c, 220f, 2211:221a, 221e, 2223:222b, 2234:2236, 223c, 2248, 2260:2261, 2264:2267, 2282:2283, 2286:2287, 2295:2299, 22a0, 22a2:22a5, 22bb:22bd, 22c2:22c5, 22ee:22f1,
20:40:13 <b_jonas> 2300, 2302, 2308:230b, 2310, 2320:2321, 23b8:23bd, 2500, 2502, 250c, 2510, 2514, 2518, 251c, 2524, 252c, 2534, 253c, 2550:256c, 2580, 2584, 2588, 258c, 2590:2593, 25a0, 25ac, 25b2, 25ba, 25bc, 25c4, 25c6, 25cb, 25d8:25d9, 263a:263c, 2640, 2642, 2660, 2663, 2665:2666, 266a:266b, 27e6:27e9}
20:40:47 <oren> do you know wat program I can use to find out?
20:40:57 -!- carado has quit (Remote host closed the connection).
20:41:07 <b_jonas> oren: dunno, in what formats do you have your font or how do you create it?
20:41:37 <oren> http://fontstruct.com/
20:42:50 <boily> quintopia: can't play yet, because it's missing Flash 32bit for Chrome. but it was mainly to answer the question "can I launch steam play stuff on Linux".
20:43:07 <boily> quintopia: the answer is yes I can, so I can buy that hero siege thing, and play too!
20:43:38 <quintopia> boily: well, don't bother. it's super beta and the server is down. i've been fighting it all morning
20:43:52 <boily> but but... there's a promotion and... :/
20:44:01 <boily> okay, won't buy it.
20:44:11 <quintopia> well, fine, buy it, but don't expect to be able to multiplayer today
20:44:25 <quintopia> maybe the promotion is taxing their servers
20:44:41 <b_jonas> oren: are the characters listed in the examples page all in the font?
20:45:11 <quintopia> i'm still up for some BL2, but I need to go get a walk in before it starts raining again
20:45:28 <oren> fc-query gives me
20:45:44 <b_jonas> oren: wait, in the demo page, why do you have a semicolon char in the place of the greek question mark char (;) ?
20:45:59 <oren> because that is what it looks like
20:46:09 <boily> quintopia: no trouble. I was asleep up until 15 minutes ago, and still in an not quite optimal brain phase.
20:46:10 <oren> according to wikipedia anyway
20:46:11 <b_jonas> oren: yes, but the html actually has a semicolon
20:46:36 <b_jonas> sorry, I have the sometimes annoying tendency to see the bytes, not the content
20:46:53 <b_jonas> I know they should look similar
20:47:02 <quintopia> boily: yeah i'm having the same problem. i arrived here 3 minutes after you pinged out earlier. (if i'd not accidentally hit the "restart now" button, I would have been here 3 minutes sooner)
20:47:36 <oren> should be fixed now
20:47:44 <b_jonas> oren: also, you have \x{3a2} in the demo page, which is a non-existant character
20:48:20 <b_jonas> http://www.orenwatson.be/fontdemo.htm doesn't seem to have changed
20:48:35 <b_jonas> (I mean, that part hasn't changed, I didn't cmp)
20:50:08 <oren> Hmm I'll try using a different editor to add it. It's possible that nano fucks it up
20:50:46 <oren> nah it's 0x37e
20:51:22 <oren> maybe browsers reduce everything to a base form?
20:51:44 <b_jonas> oren: possible, let me curl it
20:55:01 <b_jonas> oren: ok wait, let me say something positive too. I see you have the Romanian characters ș and similar, which is great (though obviously these used to be missing in old fonts and encodings for historical reasons, there's no reason why you wouldn't have them these days)
20:55:47 <oren> Well mostly I've been trying to cover entire unicode blocks at a time
20:56:13 <b_jonas> oren: your Ŵ looks strange, it has no circumflex visible
20:56:53 <b_jonas> oren: yes, but exactly for historical reasons, ș isn't in a much used block
20:58:14 <oren> which w is that?
20:58:30 <HackEgo> [U+0174 LATIN CAPITAL LETTER W WITH CIRCUMFLEX]
20:58:42 <oren> probably a mistake
20:59:10 <oren> thanks for pointing that out I would never have noticed
20:59:48 <b_jonas> oh, you probably have it but it doesn't display in the browser
21:01:33 <oren> NEITHER APPROXIMATELY NOR ACTUALLY EQUAL TO
21:01:54 <oren> the perfect operator for asshole mathematicians proving someone wrong
21:01:59 <b_jonas> this font seems nicely drawn, even if some of how it looks isn't to my tastes
21:03:59 <b_jonas> I know I said this already, but the ð looks strange
21:04:32 <oren> can you give an example picture of how it should look
21:04:54 <oren> I think I drew that one from memory or intuition
21:06:22 <b_jonas> oren: https://upload.wikimedia.org/wikipedia/commons/thumb/a/aa/Latin_letter_%C3%90.svg/1280px-Latin_letter_%C3%90.svg.png from https://en.wikipedia.org/wiki/File:Latin_letter_%C3%90.svg
21:06:33 <b_jonas> the lowercase one (the uppercase is fine)
21:07:39 <b_jonas> oren: now as for more letters, is there a reason why you drew ď Ľ ľ with the normal caron, rather than the apostrophe-shaped caron that's often used for those letters?
21:08:51 <b_jonas> And same for ť (I forgot there's a fourth)
21:08:59 <oren> I wasn't aware a caron can look that an apostrohe
21:09:26 <b_jonas> oren: I think it can look like both in those characters, but the apostrophe is more common
21:09:45 <b_jonas> so in a monospaced font like this, you _could_ use the vedge shaped caron, it's just uglier
21:10:08 -!- Patashu has joined.
21:10:54 <b_jonas> oren: your ĝ is totally wrong
21:11:00 <b_jonas> it appears to have a caron, not a circumflex
21:11:35 <b_jonas> also, why the heck am I proofreading a terminal font in the middle of the night?
21:11:50 <oren> Uh... I don't know but thanks
21:14:50 <boily> it's neither unapproximatively the middle of the night.
21:15:00 <b_jonas> oren: I'm not quite sure in this, because they're used only in Latvian or something, but I think ģ should have an accent that looks like an over-comma, not a cedilla, just like ș (Romanian s with comma) as opposed to ş (Turkish s with cedilla)
21:15:23 -!- TieSoul has quit (Remote host closed the connection).
21:16:16 <b_jonas> oren: the cedilla and the comma below was originally considered the same, then got disunified, which is why we have the encoding problems with the Romanian ș and ț, but if I understand right, ģ ķ ļ ņ ŗ only ever existed with a comma below/above, never a real cedilla
21:17:09 <b_jonas> I'm not sure I understand this correctly though
21:18:04 <oren> adding it to the list anyway, i'll research it
21:18:32 -!- oerjan has joined.
21:18:48 <b_jonas> oh, you have the maltese H with bar
21:19:04 <b_jonas> (as opposed to the physicist's h with bar)
21:21:39 <b_jonas> (they're actually the same character, though there's a compatibility alias for the latter, just like for the dreaded kelvin sign, if I understand correctly)
21:22:25 <b_jonas> oren: İ (Turkish capital I with dot) looks completely wrong
21:23:02 <b_jonas> oren: it's missing the dot in your font
21:23:36 <b_jonas> oren: ı looks wrong too I think
21:23:42 <oren> I see a dot there
21:23:49 <b_jonas> wait, maybe I'm confusing the characters here
21:24:53 <b_jonas> yes, the Turkish İ and ı look right, sorry
21:28:38 <b_jonas> oren: ŷ \x{177} looks wrong
21:29:16 <b_jonas> it should have a circumflet
21:30:33 <b_jonas> oren: is it deliberate that the letter z part of ž differs from that in ż ?
21:30:38 <b_jonas> it's only like a pixel or something
21:33:02 <oren> yeah the small z is supposed to be rounded more than the large z ehen both are the same size but I guess I wasn't consistent
21:33:18 -!- AnotherTest has quit (Quit: ZNC - http://znc.in).
21:35:36 -!- ais523 has quit.
21:36:27 <b_jonas> oren: I think your ˛ \x{2db} is wrong: it should be a standalone ogonek, but it looks like a cedilla
21:36:33 -!- augur has quit (Ping timeout: 255 seconds).
21:37:18 <oerjan> > var.fix$("The channel where "++).show
21:37:20 <lambdabot> The channel where "The channel where \"The channel where \\\"The channel whe...
21:39:09 <b_jonas> hmm, I don't know how these crazy standalone IPA accent characters are supposed to work
21:39:13 <HackEgo> [wiki] [[Esolang:Sandbox]] http://esolangs.org/w/index.php?diff=43900&oldid=43805 * Nooodl * (-612321) clean this mess of a sandbox up for once
21:44:26 -!- Patashu has quit (Ping timeout: 240 seconds).
21:44:28 <b_jonas> oren: do you have the crazy Welsh rare letters: ẁ ẃ ẅ and their uppercase?
21:44:31 <b_jonas> they're not in the demo page
21:46:42 <boily> I thought Welsh only ŵed the w. ẁ, ẃ and ẅ exist?
21:47:24 <b_jonas> boily: I don't really know
21:47:29 <b_jonas> boily: that's why they're crazy letters
21:47:44 <b_jonas> boily: I have some evidence that at least one of the four exist, at least in some conventions
21:47:55 <b_jonas> doesn't mean it's always written
21:48:06 <b_jonas> boily: https://en.wikipedia.org/wiki/Grave_accent#Length
21:48:45 <oren> which block are they in?
21:49:36 <b_jonas> well, the block is probably \x{1e00}
21:50:02 <boily> b_jonas: interesting. they do eẍist and are used!
21:50:09 <b_jonas> I think that block contains partly letters for some transcription of Chinese with the tones shown as accent marks
21:50:25 <b_jonas> boily: I don't know if all four of those letters exist in Welsh
21:50:43 <oren> latin extended additional is on my list
21:50:45 <b_jonas> boily: they exist in _some_ language because they're in unicode, but I suspect Chinese transliteration for that
21:52:26 <b_jonas> iso-8859-14 encodes all four
21:52:49 <b_jonas> that means they're used for _some_ celtic language, whether Welsh or some other I can't tell
21:53:36 <oren> I have all those
21:53:49 <b_jonas> boily: what the heck are those?
21:54:02 <oren> I think those are used for mandarin pinyin
21:54:22 <oren> yeah. u with umlaut, with tones
21:54:56 <oren> (on the internet most people use v instead of umlaut u IIRC)
21:54:59 <b_jonas> I _think_ ǘ (u with diaresis and acute) appeared in some very few very rare Hungarian text, instead of ű, back when the writing conventions weren't figured out
21:55:06 <b_jonas> but I'm not quite sure in this
21:55:48 <b_jonas> but it's not clear if that use deserves an encoding separate from ű
21:55:57 <boily> oren: on my keyboard layout, when I type [Dead ¯] then [v] I get ǖ. same for the others.
21:56:24 <b_jonas> it might because ű has some uses other than Hungarian, historically speaking
21:58:44 <b_jonas> oren: you don't have ≈ (approximately equal)
21:59:07 <oren> yeah I'm working on that block right now, it will be fuller when I update it
22:00:24 <b_jonas> oren: also \x{223c} ∼ (single middle squiggle relation operator)
22:01:18 <b_jonas> also \x{2282} ⊂ (subset of)
22:02:01 <b_jonas> though of course it's questionable how useful these maths operators are in a monospaced font
22:02:29 <oren> bah, that's a question I never consider!
22:03:49 <b_jonas> oren: ⊥ would probably be genuinely useful even in monospaced, because some crazy haskellers who like unicode symbols can type it
22:04:24 <b_jonas> you know, the kind of haskellers who replace all instances of \ -> <- => :: ++ with unicode symbols
22:05:58 <boily> hellochaf. I take it you're not a crazy Haskeller?
22:06:10 <b_jonas> I see you have all the line drawing characters, even the ones not in cp437
22:06:15 <HackEgo> [wiki] [[JSFuck]] M http://esolangs.org/w/index.php?diff=43901&oldid=43255 * Nooodl * (-11) genderrrrr
22:06:24 <oren> there are non-crazy haskellers?
22:06:42 <b_jonas> (eg. you have the round ones that I think are in the Commodore 64 character set)
22:08:30 <oren> I'm adding all the math operators that I can actually draw at the size of this font
22:08:41 <b_jonas> oren: shouldn't ■ \x{25a0} be a smaller box than what you have? That's the character that encodes the cp437 byte \xfe which is used in some dos text mode GUI programs, and I think in vga fonts it's a smaller box.
22:09:18 <b_jonas> ok, maybe not in all vga fonts, just the ones I'm used to
22:11:22 <b_jonas> oren: you don't have \x{2302}. maybe you should check if you have all the cp437 chars in the ascii control places (I think I have added all of those to my font specifically)
22:11:36 <b_jonas> you seem to have most of them
22:12:36 <oren> Oh, that one is in the font but not on the page. i'll fix it
22:12:47 <b_jonas> good night now, and keep up the good work
22:15:51 <b_jonas> for the record, 0b_jonas is not me, it's my evil twin
22:16:19 <b_jonas> he's sometimes called ₀b_jonas or ₀₈_jonas
22:19:10 <boily> 0b_jonas is to b_jonas what oerjan is to int-e. understood.
22:19:29 <HackEgo> ¯\(°_o)/¯ is a misspelling of ¯\(°_o)/¯
22:20:50 <tswett> I'm trying to think of a mathematically nice formal system which corresponds to Rust.
22:21:21 <tswett> Now, it kind of seems like if Rust were purely functional, then T, &T, and &mut T would all be essentially the same thing.
22:23:32 <tswett> T and &T would be the same thing. But passing a &mut T to a function would correspond to passing a T and then expecting another T back.
22:27:30 <shachaf> That gets even more interesting when you think about subclasses in C++.
22:27:54 <shachaf> Say you have struct A { int x; }; struct B : A { int y; };
22:28:43 <shachaf> void f(A *a) { a->x *= 10; }, and then you pass a B * to f.
22:29:11 <shachaf> What is the actual relationship of A and B, from this "purely functional" perspective?
22:29:55 <shachaf> You can turn B to A, when you pass it into f. And, given the new A that f gives you, you can "put it back" into the original B.
22:30:17 <shachaf> So you have (B -> A, B -> A -> A)
22:30:38 <shachaf> this is a familiar structure hth
22:33:44 <HackEgo> os is the accusative plural of us.
22:36:16 <tswett> Sure enough. First-person plural accusative: "os".
22:36:25 <tswett> The nominative is "vi".
22:36:50 <tswett> Now, I suppose there's no reason not to simply take STLC and add & and &mut to it.
22:37:10 <tswett> First problem. &T and &mut T aren't actually types.
22:37:26 <tswett> &'a T and &'a mut T are types, given a lifetime 'a.
22:37:38 <oren> math operators updated
22:39:23 <tswett> It doesn't seem to make much sense to have references when there are no memory locations for them to reference.
22:41:12 <tswett> I guess that both "immutable borrow" and "mutable borrow" constructs seem to make sense. As I envision it, I can't see how the "immutable borrow" construct would be useful at all, so let's look at the mutable borrow construct instead.
22:42:34 <tswett> If you have a T, then you can perform a mutable borrow of T. Within the mutable borrow, you get a &mut T that you can pass around willy-nilly. At the end of the borrow, the same &mut T must be returned back.
22:43:26 <tswett> Now, &mut Ts are uncopyable, so we need to use linear logic.
22:44:08 <tswett> The same &mut T must be returned back? What have we accomplished if the &mut T has to be given back?
22:44:13 <tswett> We could have just used a T this whole time.
22:44:58 <tswett> &mut is starting to feel like it's just syntactic sugar.
22:45:00 <quintopia> boily: i'm gonna take a shower. message me on steam when you're ready.
22:52:37 -!- augur has joined.
22:59:37 <tswett> All right, &mut feels a lot like syntactic sugar. But let's consider linear typing now. If T is your average linear type, then T can't be copied, but &T can.
23:01:28 <tswett> I'm pretty sure that in general, the behavior of &T can't really be predicted by the behavior of T.
23:04:01 <tswett> The thing about a &T is that when you do things with a &T, you're doing stuff with the T and then returning the *same* T back.
23:10:53 <boily> quintopia: IEUAAAAAAAAAAAAHRGHGHGHGHGHFLBLBLBLBLBLBLBLBLBLBLBL!
23:11:03 <tswett> So I think the moral of the story is that a &A -> B is essentially the same as an A -> (A, B) with the property that the function's first output always equals its input.
23:11:15 -!- Phantom_Hoover has quit (Ping timeout: 245 seconds).
23:11:55 <tswett> It's not really &A -> B; it's really &'x A -> B. 'x is a lifetime symbol, and the caveat is that 'x must not appear anywhere in B.
23:15:04 -!- Phantom_Hoover has joined.
23:30:26 -!- Phantom_Hoover has quit (Ping timeout: 244 seconds).
23:31:24 -!- Phantom_Hoover has joined.
23:37:19 <tswett> So far, &mut isn't very fancy.
23:38:06 <tswett> It seems that a "&'x mut A -> B" is essentially the same as an "A -> (A, B)" (without the equality requirement), with the caveat that 'x must not appear anywhere in B.
23:39:08 <tswett> It seems like the big difference between T and &mut T is that if you have a &mut T, you can drop it, but the underlying value will come back later.
23:39:44 <tswett> I don't know if that's actually theoretically useful.
23:52:27 -!- boily has quit (Quit: asdfasdfasdf).
23:58:24 <oerjan> `unicode NEITHER APPROXIMATELY NOR ACTUALLY EQUAL TO