00:05:42 -!- Bicyclidine has quit (Ping timeout: 264 seconds).
00:08:48 -!- heroux has joined.
00:18:00 -!- Bicyclidine has joined.
00:34:17 -!- adu has joined.
00:35:14 -!- NATT_SiM has quit (Remote host closed the connection).
00:35:20 <FreeFull> Melvar: Ogonek is used in other languages
00:35:38 <FreeFull> The name came from Polish though
00:37:46 -!- NATT_SiM has joined.
00:44:47 <Melvar> FreeFull: Yes, I know, I realize I kinda swallowed that.
00:51:18 -!- Phantom_Hoover has quit (Ping timeout: 264 seconds).
00:53:54 <FreeFull> Romanian has a weird comma thing
00:55:04 -!- shikhout has changed nick to shikhself.
00:55:33 -!- shikhself has changed nick to shikhin.
00:57:18 <HackEgo> [U+0218 LATIN CAPITAL LETTER S WITH COMMA BELOW] [U+021A LATIN CAPITAL LETTER T WITH COMMA BELOW]
00:57:19 -!- NATT_SiM has quit (Remote host closed the connection).
00:58:12 <Melvar> ĢĶĻŅ also have commas, but
00:58:21 <HackEgo> [U+0122 LATIN CAPITAL LETTER G WITH CEDILLA] [U+0136 LATIN CAPITAL LETTER K WITH CEDILLA] [U+013B LATIN CAPITAL LETTER L WITH CEDILLA] [U+0145 LATIN CAPITAL LETTER N WITH CEDILLA]
00:58:49 <Melvar> The unicode names are “wrong” on this.
00:59:00 <HackEgo> [U+0156 LATIN CAPITAL LETTER R WITH CEDILLA]
00:59:37 -!- NATT_SiM has joined.
01:00:23 <oerjan> `unicode LATIN CAPITAL LETTER C WITH CEDILLA
01:01:02 <HackEgo> U+000F <control> \ UTF-8: 0f UTF-16BE: 000f Decimal:  \ \ Category: Cc (Other, Control) \ Bidi: BN (Boundary Neutral)
01:01:09 <oerjan> `unicode LATIN CAPITAL LETTER C WITH COMMA BELOW
01:01:42 <oerjan> lifthrasiir: it seems to me that you have color codes in that?
01:01:54 -!- Bicyclidine has quit (Ping timeout: 245 seconds).
01:03:50 <HackEgo> [U+010F LATIN SMALL LETTER D WITH CARON] [U+0064 LATIN SMALL LETTER D] [U+030C COMBINING CARON]
01:03:54 <lifthrasiir> there are no official names for controls ("<control>", if you prefer that)
01:05:24 <HackEgo> U+0002 <control> \ UTF-8: 02 UTF-16BE: 0002 Decimal:  \ \ Category: Cc (Other, Control) \ Bidi: BN (Boundary Neutral)
01:05:53 <FireFly> It probably just uses UnicodeData.txt?
01:06:35 <HackEgo> #!/usr/bin/env python \ import os, sys \ import unicodedata \ s = u" ".join("[U+{0:04X} {1}]".format(ord(c), unicodedata.name(c, "DUNNO")) for c in " ".join(sys.argv[1:]).decode("utf-8")).encode("utf-8") \ if u"DUNNO" in s: \ os.execvp("multicode", ["multicode"] + sys.argv[1:]) \ else: \ print s
01:07:07 <oerjan> it has a fallback mechanism
01:08:19 <oerjan> the ďď above gets the default treatment, while FireFly's 0002 goes via the fallback
01:09:14 <HackEgo> #!/usr/bin/python \ \ \ import os, glob, sys, unicodedata, locale, gzip, re, traceback, encodings \ import urllib, webbrowser, textwrap \ \ # bz2 was introduced in 2.3, we want this to work also with earlier versions \ try: \ import bz2 \ except ImportError: \ bz2 = None \ \ # for python3 \ try: \ unicode \ except NameError: \
01:09:40 <FireFly> I thought it was just a matter of whether you supply only one or multiple characters
01:09:55 <HackEgo> [U+0061 LATIN SMALL LETTER A]
01:10:15 <HackEgo> U+0061 LATIN SMALL LETTER A \ UTF-8: 61 UTF-16BE: 0061 Decimal: a \ a (A) \ Uppercase: U+0041 \ Category: Ll (Letter, Lowercase) \ Bidi: L (Left-to-Right)
01:11:18 -!- olsner has quit (Ping timeout: 240 seconds).
01:13:35 <HackEgo> U+0061 LATIN SMALL LETTER A \ UTF-8: 61 UTF-16BE: 0061 Decimal: a \ a (A) \ Uppercase: U+0041 \ Category: Ll (Letter, Lowercase) \ Bidi: L (Left-to-Right) \ \ U+0002 <control> \ UTF-8: 02 UTF-16BE: 0002 Decimal:  \ \ Category: Cc (Other, Control) \ Bidi: BN (Boundary Neutral) \ \ U+0063 LATIN SMALL LETTER C \ UTF-8: 63 UTF-16BE: 00
01:22:13 <Dulnes> How does one go about hardening light?
01:27:52 <Taneb> Limestone impurities
01:28:45 <Dulnes> Hm? What about limestone
01:34:08 <Bike> i hear light is a wave
01:36:46 <Dulnes> www.princeton.edu/main/news/archive/S41/10/73O28/index.xml?section=topstories
01:37:08 <Dulnes> Idk if it's true or not but you decide
01:37:55 -!- madbr has joined.
01:39:15 -!- Sprocklem has quit (Ping timeout: 264 seconds).
01:40:52 <Bike> http://journals.aps.org/prx/abstract/10.1103/PhysRevX.4.031043 gee
02:03:31 -!- NATT_SiM has quit (Read error: Connection reset by peer).
02:03:53 -!- NATT_SiM has joined.
02:06:17 <Taneb> Dulnes, limestone is how you make water hard
02:06:40 -!- NATT_SiM has quit (Remote host closed the connection).
02:10:38 -!- NATT_SiM has joined.
02:11:52 -!- Sprocklem has joined.
02:12:20 -!- Sprocklem has changed nick to Guest97875.
02:14:07 -!- Guest97875 has quit (Changing host).
02:14:07 -!- Guest97875 has joined.
02:14:26 -!- Guest97875 has changed nick to Sprocklem.
02:16:35 -!- shikhout has joined.
02:19:34 -!- shikhin has quit (Ping timeout: 244 seconds).
02:25:17 -!- NATT_SiM has quit (Remote host closed the connection).
02:25:18 <adu> `unidecode 💩
02:25:19 <HackEgo> U+1F4A9 PILE OF POO \ UTF-8: f0 9f 92 a9 UTF-16BE: d83ddca9 Decimal: 💩 \ 💩 \ Category: So (Symbol, Other) \ Bidi: ON (Other Neutrals)
02:39:20 -!- olsner has joined.
02:44:20 -!- AndoDaan has quit (Ping timeout: 258 seconds).
03:14:53 -!- NATT_SiM has joined.
03:42:16 -!- hjulle has quit (Ping timeout: 256 seconds).
03:59:38 -!- Sprocklem has quit (Ping timeout: 240 seconds).
03:59:54 -!- MoALTz_ has joined.
04:02:55 -!- MoALTz has quit (Ping timeout: 258 seconds).
04:42:47 -!- aloril has quit (Remote host closed the connection).
04:55:30 -!- vanila has quit (Quit: Leaving).
04:59:25 -!- oerjan has quit (Quit: leaving).
05:33:00 -!- aloril has joined.
06:02:18 -!- adu has quit (Quit: adu).
06:08:21 -!- aloril has quit (Remote host closed the connection).
06:11:51 -!- ZombieAlive has quit (Remote host closed the connection).
06:13:25 -!- aloril has joined.
06:14:03 <zzo38> Again, he lost before he could pick his last remaining side card. How common is such a thing?
06:21:13 <zzo38> (I mean he ran out of cards.)
06:22:49 <int-e> Argh. "only". "Mark Zuckerberg estimated earlier this year that the company’s U.S. users spend a total of nine hours a day on digital media, but only 40 minutes of that on Facebook."
06:26:05 <zzo38> int-e: What company is that? How was the data estimated?
06:26:57 <int-e> I'm reading about "Facebook at Work", and I'm appalled at the idea.
06:27:43 <int-e> (The initial article I read about it said it would use "the familiar Facebook interface" and to me there is no such thing.)
06:27:52 -!- NATT_SiM has quit (Remote host closed the connection).
06:28:32 -!- MDude has quit (Ping timeout: 250 seconds).
06:32:21 <int-e> zzo38: Actually Facebook will have hard numbers on that (the statement is about their users), with some errors because just because a browser tab is visible doesn't mean anybody is using it, and because some of the people with US IPs are not actually in the US.
06:35:46 -!- MoALTz has joined.
06:36:25 <int-e> One of the biggest issue I have with Facebook is that they make Google look like the good guys.
06:36:55 -!- MoALTz_ has quit (Ping timeout: 264 seconds).
06:54:55 <int-e> I guess fungot would know...
07:15:10 -!- madbr has quit (Quit: Pics or it didn't happen).
07:24:39 -!- Patashu has joined.
07:29:18 -!- Patashu has quit (Remote host closed the connection).
07:29:29 -!- Patashu has joined.
07:31:16 -!- drdanmaku has quit (Quit: Connection closed for inactivity).
08:20:30 -!- shikhout has quit (Ping timeout: 255 seconds).
08:29:56 <fizzie> The MD for 4000 was quite suspicious too.
08:39:01 <zzo38> How to make error diffusion without "ghost images" appearing?
08:43:06 <int-e> I suppose one should not distribute errors across edges in the image, but how to implement that, I don't know.
08:43:33 <fizzie> Gimp has a "reduced color bleeding" Floyd-Steinberg mode, but I'm not entirely sure what it does.
08:43:58 <fizzie> And I vaguely remember something about scanning every other line left-to-right and every other right-to-left, but I'm not sure what sort of artefacts that is supposed to reduce.
08:47:26 <fizzie> You could apply some general image segmentation algorithm (to the original or the quantized form), and then do color conversion and dithering within each segment, but I wouldn't be surprised if all that accomplishes would be to just make segmentation errors be visible as artefacts.
08:49:42 <zzo38> I thought of something else to try let's see how well it works.
08:51:12 <zzo38> No, what I tried doesn't work.
09:01:51 <zzo38> I am getting "ghost images" which are pretty far away from the original and have a similar shape.
09:04:18 -!- Phantom_Hoover has joined.
09:06:35 -!- AndoDaan has joined.
09:06:52 <zzo38> I think I found the mistake now.
09:07:10 <int-e> how do you distribute the error among neighbouring pixels? I don't see this happening...
09:08:06 <zzo38> I fixed it; the mistake was in a different calculation.
09:54:20 -!- AndoDaan_ has joined.
09:56:58 -!- AndoDaan has quit (Ping timeout: 250 seconds).
10:12:39 -!- mihow has quit (Read error: Connection reset by peer).
10:14:07 -!- mihow has joined.
11:00:32 -!- shikhin has joined.
11:13:33 -!- boily has joined.
11:32:11 -!- Phantom_Hoover has quit (Ping timeout: 258 seconds).
11:35:43 -!- Patashu has quit (Ping timeout: 264 seconds).
11:40:33 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: tomroman: not found
12:18:52 <boily> I can't segment “tomroman”. Is it “Thomas Roman”, or “to mroman”?
12:25:08 -!- drdanmaku has joined.
12:25:19 <AndoDaan_> mroman, nearly done with the MNNBFSL interpreter in Burlesque
12:25:34 -!- boily has quit (Quit: EMULSIFIED CHICKEN).
12:27:17 <AndoDaan_> Finally when i'm figuring out how to simulate variables and functions in burlesque, you add them.
12:30:23 -!- AndoDaan_ has changed nick to AndoDaan.
12:30:52 <mroman> AndoDaan: usually you just keep a list somewhere laying around and use DimAccess or DimSet
12:31:05 <mroman> or BlockAccess and setat
12:31:31 <mroman> the other approach is to translate the input program to Burlesque and then eval it
12:31:35 <mroman> possibly a combination of both.
12:33:35 <AndoDaan> i'm figuring if you have an array with something like {{"a"vv 3} {"add2"vv 2+.}} you can search for "add2" and evauate the block it's in.
12:37:30 <mroman> there's cn which does exactly that btw.
12:37:39 <mroman> well... not exactly that but similar
12:37:58 <AndoDaan> You think Anarchy Golf will update when you're done upgrading?
12:38:13 <mroman> !blsq 9"a"{{"a"==} {vv?*}}cn
12:38:16 <mroman> !blsq 9"a"{{"a"==} {vv?*}}cne!
12:38:16 <blsqbot> | ERROR: Burlesque: (.*) Invalid arguments!
12:38:25 <mroman> !blsq 9"a"{{"a"==} {vvS[}}cne!
12:38:25 <blsqbot> | ERROR: Burlesque: (S[) Invalid arguments!
12:38:31 <mroman> !blsq 9"a"{{"a"==} {vv^^?*}}cne!
12:38:31 <blsqbot> | ERROR: Burlesque: (.*) Invalid arguments!
12:38:31 <blsqbot> | ERROR: Burlesque: (^^) Stack size error!
12:38:34 <mroman> !blsq 9"a"{{"a"==} {vv^^?*}}cne!
12:38:35 <blsqbot> | ERROR: Burlesque: (.*) Invalid arguments!
12:38:35 <blsqbot> | ERROR: Burlesque: (^^) Stack size error!
12:38:38 <mroman> !blsq 9"a"{{"a"==} {vv^^?*}}cn
12:38:49 <mroman> !blsq 9"a"{{"a"==} {^^?*}}cne!
12:38:53 <AndoDaan> I'm using cn for the MNNBFSL instructions
12:39:06 <mroman> !blsq 9"b"{{"a"==} {^^?*} {"b"==} {^^?*^^?*}cne!
12:39:07 <blsqbot> | ERROR: (line 1, column 43):
12:39:07 <blsqbot> | expecting "%", "g", "s", "S", "m{", "q", "{", "\"", "-", digit, "'", "(", "y" or "}"
12:39:13 <mroman> !blsq 9"b"{{"a"==} {^^?*} {"b"==} {^^?*^^?*}}cne!
12:41:46 <AndoDaan> still missing the while condition, and will be reduced some what, but that's what it's shaping up to be.
12:43:52 <mroman> line 13 looks like a syntax error
12:45:33 <AndoDaan> fat pinky. fixed http://codepad.org/Y1WCJjFj
12:50:04 <AndoDaan> p is close to { on my keyboard.
12:51:27 <AndoDaan> and the constant {PP ... Pp}e! doesn't help with the muscle memory.
12:53:48 <AndoDaan> Gonna finish it up later todayor tomorrow. I just wanted to say it was fun. Thanks for suggesting it.
12:54:04 -!- AndoDaan has quit (Quit: Sleep? What is sleep?).
12:56:03 <mroman> What a weird keyboard.
12:56:23 <mroman> Everybody knows P is not even close to {
13:01:11 <fizzie> p is right below } (altgr-0) here.
13:03:18 -!- Frooxius has quit (Quit: *bubbles away*).
13:04:08 <elliott> 12.5 gigabytes of swap used is not a good statistic...
13:07:37 -!- Frooxius has joined.
13:12:56 <FireFly> P would be close to { if I didn't use dvorak
13:33:29 <mroman> FireFly: Exactly. But who the hell doesn't use Dvorak.
14:02:52 <fizzie> I can't find any reasonable statistics quickly, but I'd wager a guess that several people.
14:03:40 -!- S1 has joined.
14:05:21 <fizzie> While looking for said statistics, though, I did find incontrovertible proof that learning Dvorak makes you smart: "-- I could almost feel my synapses firing faster and new neuron connections being made. I felt like I got a turbo boost to my IQ during that period where I thought fast, was more creative and just generally sharper."
14:07:11 <elliott> sounds like the results people report for trepanning
14:08:09 <elliott> I read a great webpage once that was like [pages of description of how he got a buddy to drill into his skull] [hyperbolic praise over how amazing he feels in the immediate aftermath and how great an idea it was] [pause] [entry admitting that the gains didn't last and probably never existed that much and that he regrets doing it]
14:08:27 <elliott> sorry about the hole in your head, dude
14:09:46 <fizzie> That sounds vaguely familiar. (And also a bit more extreme than typing with a different layout.)
14:11:08 <elliott> look, the Keyboard Fascists will never get me to trepannify myself
14:18:34 -!- NATT_SiM has joined.
14:25:18 -!- hjulle has joined.
14:36:43 -!- NATT_SiM has quit (Remote host closed the connection).
14:51:03 -!- NATT_SiM has joined.
14:51:49 -!- NATT_SiM has quit (Remote host closed the connection).
15:08:57 -!- `^_^v has joined.
15:41:51 <Bike> that sounds like a good webpage
15:46:21 <J_Arcane> Scheme in Arabic. https://github.com/nasser/---
15:58:40 -!- hubs has joined.
15:58:53 -!- S1 has changed nick to S0.
16:01:16 -!- drdanmaku has quit (Quit: Connection closed for inactivity).
16:01:43 -!- hubs has quit (Quit: hubs).
16:04:13 -!- S0 has changed nick to S1.
16:13:17 -!- drdanmaku has joined.
16:30:18 <HackEgo> [U+002D HYPHEN-MINUS] [U+002D HYPHEN-MINUS] [U+002D HYPHEN-MINUS]
16:30:47 <Melvar> Does github not allow arabic repositories or something?
16:34:24 <J_Arcane> maybe because the reponames are used as default directory names?
16:35:03 <Bike> i'm pretty sure i have a couple folders named in cyrillic
16:35:24 <Melvar> J_Arcane: I’m not sure how you mean that would affect it.
16:36:13 <J_Arcane> I don't know the state of support for unicode filenames in various OSes.
16:36:28 <J_Arcane> Perhaps they default to 'ascii safe' names for them.
16:37:30 <HackEgo> [U+061F ARABIC QUESTION MARK]
16:39:58 -!- MDude has joined.
16:41:50 <Melvar> “first:hrf rest:(hrf / [٠١٢٣٤٥٦٧٨٩] / '-' / '؟' / 'ـ' )* space?” ← This line looks very funny with proper bidi layout.
16:42:07 <Bike> is that phoenician
16:42:47 <Bike> oh. numerals. not used to those.
16:43:00 -!- S1 has quit (Quit: S1).
16:44:47 <Melvar> So, of that line, “٠١٢٣٤٥٦٧٨٩] / '-' / '؟' / 'ـ” is right-to-left (including mirroring the ‘]’), and then “٠١٢٣٤٥٦٧٨٩” is a left-to-right subchunk because that’s what arabic does with numerals.
16:46:36 <elliott> Melvar: do you have an, uh, reference rendering of that line
16:47:34 <Melvar> I could screenshot it from my browser … it’s at https://github.com/nasser/---/blob/master/peg/qlb.peg under “symbol =”.
16:47:57 <Bike> hm. the [ ends up next to the underscore for me.
16:48:12 <Bike> thaaaaat's probably not right
16:48:33 <Melvar> It’s not an underscore.
16:49:10 <Melvar> It’s next to the opening [ because it’s the last thing in the right-to-left chunk.
16:50:41 <elliott> Melvar: I was just curious as to whether my terminal is doing it correctly.
16:52:18 -!- hjulle has quit (Ping timeout: 265 seconds).
16:56:08 -!- shikhout has joined.
16:56:50 <Melvar> elliott: http://i.imgur.com/Hz2WcBD.png
16:57:17 <elliott> Melvar: cool, it survives irssi + mosh + terminal for me and displays the bidi correctly
16:57:29 <elliott> (in layout order, not logical order)
16:57:53 <Melvar> Well, that should only be a matter of your terminal, since it’s about character properties.
16:58:12 <elliott> you trust programs that layout unicode characters on screen to not mess it up?
16:58:21 <elliott> er, programs that run in a terminal I mean
16:58:42 <Bike> my ignorant question is now preserved forever
16:59:01 <Melvar> I mean, mosh possibly could confuse the terminal about – ohgodIhavetobeelsewhere
16:59:16 <elliott> https://mosh.mit.edu/#techinfo I think it's more likely to be right with mosh than without
16:59:22 -!- shikhin has quit (Ping timeout: 255 seconds).
16:59:22 <elliott> since it does things with unicode
17:05:50 <Melvar> elliott: If you select the line, it should select according to logical order, so if you get discontinuous-looking selections, your terminal is doing it right and I would like to know which it is.
17:06:22 <elliott> my selection are continuous but weird
17:06:39 <elliott> if I select from the third character after the second displayed [ to the ' then I get '؟' / 'ـ'
17:07:06 <elliott> it generally handles unicode way better than the usual X11 terminals
17:07:21 -!- NATT_SiM has joined.
17:07:31 <elliott> by "then I get", I mean it looks like I've selected what I said, but it copies as what I pasted.
17:08:48 <Melvar> Huh, maybe it applies the selection to the logical order but shows it superimposed on the layout order?
17:09:17 <elliott> https://letsencrypt.org/ huh.
17:09:23 <elliott> Melvar: right, that's what I think
17:09:35 <Melvar> Anyway, really need to leave now.
17:09:35 <elliott> it visually selects like any other line but copies weirdly
17:11:58 -!- NATT_SiM has quit (Ping timeout: 255 seconds).
17:19:12 -!- NATT_SiM has joined.
17:19:38 -!- NATT_SiM has quit (Remote host closed the connection).
17:20:18 -!- NATT_SiM has joined.
17:41:18 -!- S1 has joined.
17:51:46 -!- S1 has changed nick to |S}.
17:57:07 -!- NATT_SiM has quit (Remote host closed the connection).
17:58:45 <fizzie> elliott: But will it work in IE?
17:59:00 <elliott> fizzie: they have an existing root CA co-signing for them, it seems
17:59:13 -!- augur has quit (Remote host closed the connection).
17:59:57 <fizzie> Oh, IdenTrust is a CA? Okay, with that sort of name, shouldn't have been surprised.
18:00:07 <elliott> well, they sign the letsencrypt.org key.
18:01:37 -!- NATT_SiM has joined.
18:02:58 -!- MoALTz has quit (Quit: Leaving).
18:17:34 -!- NATT_SiM has quit (Remote host closed the connection).
18:18:57 -!- NATT_SiM has joined.
18:19:53 -!- |S} has changed nick to S1.
18:29:26 <HackEgo> [wiki] [[Talk:My Unreliable Past]] http://esolangs.org/w/index.php?diff=41028&oldid=41025 * BCompton * (+712) IO Questions
18:36:04 -!- NATT_SiM has quit (Remote host closed the connection).
18:37:15 -!- NATT_SiM has joined.
18:54:30 -!- NATT_SiM has quit (Remote host closed the connection).
19:00:03 -!- Sprocklem has joined.
19:01:20 -!- NATT_SiM has joined.
19:05:34 -!- NATT_SiM has quit (Remote host closed the connection).
19:05:47 -!- NATT_SiM has joined.
19:14:14 -!- shikhout has changed nick to shikhin.
19:19:39 -!- shikhin has quit (Ping timeout: 240 seconds).
19:20:48 -!- shikhin has joined.
19:29:31 -!- NATT_SiM has quit (Remote host closed the connection).
19:29:50 -!- Sprocklem has quit (Ping timeout: 245 seconds).
19:32:16 -!- NATT_SiM has joined.
19:39:41 -!- Patashu has joined.
19:49:46 -!- augur has joined.
19:51:19 -!- augur has quit (Remote host closed the connection).
19:57:48 -!- NATT_SiM has quit (Remote host closed the connection).
20:01:24 -!- S1 has left.
20:05:19 -!- NATT_SiM has joined.
20:07:43 -!- zzo38 has quit (Remote host closed the connection).
20:10:28 -!- augur has joined.
20:16:26 -!- augur has quit (Remote host closed the connection).
20:16:40 -!- augur has joined.
20:25:25 -!- Patashu has quit (Ping timeout: 258 seconds).
20:31:58 -!- NATT_SiM has quit (Remote host closed the connection).
20:37:34 -!- NATT_SiM has joined.
20:38:29 -!- augur has quit (Remote host closed the connection).
20:39:53 <newsham> static type checker for javascript http://flowtype.org/
20:55:54 -!- NATT_SiM has quit (Remote host closed the connection).
20:57:31 -!- NATT_SiM has joined.
21:03:33 -!- vanila has joined.
21:04:38 -!- NATT_SiM has quit (Remote host closed the connection).
21:08:44 -!- NATT_SiM has joined.
21:12:59 -!- b_jonas has quit (Ping timeout: 240 seconds).
21:15:49 <int-e> Is CSS property inheritance esoteric? http://int-e.eu/~bf3/tmp/test.html
21:16:17 -!- copumpkin has quit (Ping timeout: 256 seconds).
21:17:01 <Dulnes> i dont use firefox? and why do you need to know
21:17:16 <int-e> I'm honestly wondering whether it's supposed to look like that, according to the CSS specification.
21:17:42 <Dulnes> it does it on some chrome sites
21:18:05 <Dulnes> Firefox just overlaps based on Font style
21:18:22 <Dulnes> and kinda pushes it together
21:18:47 <Dulnes> So i guess? idk ask someone else i use noscript
21:19:04 <int-e> Ah but did you look at the source? The question is about the meaning of the relative line-height, 100%.
21:20:24 <Dulnes> Also newsham that will be very useful
21:20:28 -!- b_jonas has joined.
21:22:00 <Dulnes> Ask taneb or someone im going to bed
21:22:45 <int-e> Good night, Dulnes
21:22:51 <elliott> int-e: fwiw they overlap in my browser (chrome) too, but I suspect you know that
21:22:58 <vanila> it looks the same on chromium
21:23:46 <int-e> elliott: no, I only have this one browser on this computer. (and links, which won't exhibit this particular effect)
21:25:20 <int-e> I also haven't looked at the specification yet. It's quite possible that this is correct (though I have my doubts that it was intended.)
21:29:41 <int-e> uh, brx' solution will take a while to decipher...
21:29:48 <Dulnes> actually i might stay to help int-e
21:30:05 <Dulnes> waits for phone to die
21:30:24 <int-e> Dulnes: it's not important
21:30:32 <Dulnes> Use palemoon and see if it doesnt overlap
21:31:35 <elliott> compiler flags are not going to change firefox's layouting engine
21:33:17 <int-e> I can explain why the lines overlap easily -- the browser is inheriting the line-height as an absolute length, so we're rendering a 30pt font with a line height derived from a 10pt one.
21:33:38 <int-e> Dulnes: sleep seems more important to me
21:42:37 <int-e> okay... I got the gist of it. grep/1/&/2/&/3/&/4/,9..$$ computes permutations of 1..4. Then it's actually using 'eval' to evaluate expressions. And it has 14 expression shapes that it tries in a specified order ... fun.
21:43:06 <int-e> (This is brx' solution to Make 24 on anagol)
21:47:14 <int-e> and it doesn't work in my perl...
21:47:36 <int-e> (needs one more space)
21:50:20 <int-e> oh and I guess it needs $$ to be between 4321 and 9999.
21:55:50 -!- NATT_SiM has quit (Remote host closed the connection).
21:59:19 -!- NATT_SiM has joined.
22:10:51 <fizzie> Oh, Make 24 timed out?
22:16:38 -!- reynir has changed nick to reynur.
22:23:39 -!- augur has joined.
22:24:35 -!- augur has quit (Read error: Connection reset by peer).
22:24:44 -!- augur has joined.
22:28:54 <vanila> The shorted lambda term that no one knows if halts or not is probably collatz conjecture
22:29:09 <vanila> at least labmda function from church nat to church nat
22:29:27 <vanila> I do not know of closed lambda term whose halting is hard to decide
22:31:51 <vanila> I was just thinking 3x+1 and halving is quite easy and short with church numerals
22:31:54 <tromp_> i dont see how collatz translates to a (non)halting term
22:31:57 <vanila> and then you just have to loop it
22:32:11 <vanila> well w ecan't be sure yet that the lambda term halts on all church nat inputs
22:32:20 <vanila> when considered as a function N -> N
22:32:40 <tromp_> collatz conjecture is a forall exists statement
22:32:49 -!- NATT_SiM has quit (Remote host closed the connection).
22:33:08 <vanila> you just run the collatz 'step' function in a loop until you hit 1
22:33:27 <tromp_> goldbach is a single lambda term
22:33:31 <Jafet> The term that halts if and only if it can find an even integer > 2 not the sum of two primes
22:33:45 <vanila> you need primality checking for that so I think it would be longer
22:33:47 <elliott> you can write a program to search for collatz counterexamples
22:33:56 <elliott> it'll be shorter than goldbach
22:34:22 <Jafet> They would need to be periodic counterexamples
22:34:50 <vanila> goldbach will probbably be resolved before collatz
22:34:53 -!- NATT_SiM has joined.
22:35:01 <vanila> people are getting really good results on prime gaps
22:35:10 <Jafet> itt: an integer for which collatz termination is unprovable
22:35:21 <tromp_> no, elliot, collatz doesnt have finite refutation
22:35:41 <Jafet> Nonexistence of periodic collatz is refutable
22:35:45 <elliott> I was proposing a special case
22:35:50 <tromp_> and goldbach is quite small
22:36:23 -!- NATT_SiM_ has joined.
22:36:32 -!- NATT_SiM has quit (Read error: Connection reset by peer).
22:36:38 -!- NATT_SiM_ has quit (Remote host closed the connection).
22:37:24 <tromp_> existence of nontrivial collatz periods is certainly less interesting than goldbach
22:40:30 <tromp_> elliot, can you code a collatz period finder in 284 bits?
22:41:12 <elliott> it's a pretty simple recursive algorithm
22:42:03 -!- oerjan has joined.
22:42:30 <tromp_> 284 bits is not much. you have to divide by on church numerals, and search over list... it asdds up quickly
22:42:41 <elliott> https://en.wikipedia.org/wiki/Collatz_conjecture#As_an_abstract_machine_that_computes_in_base_two I like collatz-conjecture-as-halting-problem
22:42:45 <elliott> tromp_: you could do it on bit strings instead.
22:42:48 -!- reynur has changed nick to reynir.
22:43:16 <vanila> we don't need to find period??
22:43:21 <vanila> are there loops which never reach 0
22:43:28 <FreeFull> elliott: It's really easy to solve the Collatz conjecture if you've solved the halting problem
22:43:46 <FreeFull> elliott: It's easy to construct a program that halts if there is a counterexample
22:43:52 <FreeFull> Then you just feed it to your oracle :)
22:43:52 <vanila> i dont thinnk we care about cycles
22:44:30 <FreeFull> Actually, it's somewhat tricky
22:44:50 <tromp_> elliot: ok, that looks doable in 284 bits
22:44:51 <FreeFull> You actually want to use the oracle inside your program
22:45:12 <FreeFull> For each number, check if the program that computes the sequence halts
22:45:18 <vanila> what are you counting the bits of?
22:45:24 <vanila> is that that weird lambda language
22:45:28 <FreeFull> If so, continue with the next number
22:45:36 <FreeFull> Then, use the oracle on that program, to check if it halts
22:45:50 <FreeFull> If it doesn't, the collatz conjecture is true
22:46:13 <vanila> why not just count characters of a lambda calculus program or so?
22:46:37 <tromp_> because bits is the most neutral measure
22:50:10 -!- ZombieAlive has joined.
22:51:19 -!- Koen__ has joined.
22:54:03 -!- NATT_SiM has joined.
22:55:12 <oerjan> some of us seem to have become a bit addicted to code golfing
22:55:39 <vanila> im too bad at perl to golf
22:55:59 <oerjan> vanila: i golf in haskell
22:56:09 -!- shikhout has joined.
22:56:11 <oerjan> anarchy golf has a bunch of languages
22:56:29 <Bike> i golf on a fairway you sick fuckers
22:56:37 <Koen__> you should make a language where every program is 140 characters
22:56:38 <oerjan> you can even golf in several esolangs
22:56:50 <oerjan> (a number here do burlesque)
22:57:08 <Bike> 140 characters and infinite alphabet. funes the memorious
22:57:27 <Dulnes> my eyes hurt i must change the colour of my client
22:58:51 <oerjan> ah make 24's deadline has expired
22:59:17 <oerjan> i didn't try that one though, looked too arduous to do as anything other than compression, which doesn't interest me.
22:59:40 -!- shikhin has quit (Ping timeout: 264 seconds).
23:00:08 -!- NATT_SiM has quit (Remote host closed the connection).
23:01:45 -!- NATT_SiM has joined.
23:02:48 <int-e> oerjan: you can still look at the solutions. brx' is the closest to a proper one.
23:04:55 <oerjan> oh i did look at that one
23:07:35 <oerjan> i have no real idea how either works, though.
23:07:44 <int-e> Apparently I've been the only one who encoded the solutions as operator pattern and permutations.
23:09:41 <oerjan> Dulnes: http://golf.shinh.org/p.rb?Make+24
23:10:58 <int-e> oerjan: I have a function f that produces a stream of numbers 0..3; the key is this pattern: (-1FGF)>1G(>2F)<1GF)>2 <-- each F is expanded to $F[f] (and perl -a conveniently places the 4 fields of each input line in @F), and each G is replaced by $G[f], and G is the array of operators. Finally, (xx and )xx are replaced by ( if $vxx evaluates to true; $v is another number between 0 and 3. So that results in the various ways...
23:11:04 <int-e> ...of placing () and there are only 4 of those total.
23:11:37 <int-e> s/(\W)(..)/"$v$2?\$1:''"/gee does the (xx and )xx replacement; and s/(\w)/$$1[f]/g takes care of the F and G.
23:11:45 <oerjan> Dulnes: it's in theory about expressing 24 as an expression in given numbers, but it's ruined by the fact you have to get _exactly_ the same answers as listed, with no real rule for which one to choose.
23:12:28 <oerjan> so everyone who solved it has to encode the wanted answers in the program.
23:13:01 <int-e> no. brx hasn't, he's specified a search order.
23:13:23 <oerjan> and it gives the right result?
23:13:33 <int-e> (and the search isn't exhaustive)
23:13:50 <Dulnes> May i share this problem with some people
23:14:11 <oerjan> of course. the deadline is expired though, although you can still submit additional answers.
23:16:28 <oerjan> there are several currently open, see http://golf.shinh.org/
23:18:38 <int-e> oerjan: as far as I can see he's searching for the lexicographically latest permutation such that one of the following pattern (with numbers omitted) works: -*(-) -*(+) -** -*- +*(-) +*(+) +** +*- (++)* (*-)* +++ ++- /++ +-- with later ones taking priority.
23:19:14 <int-e> so it's a very restricted search but you can, for example, permute the inputs and it'll still work.
23:20:07 <oerjan> int-e: ok so he's encoded the wanted _priorities_ in an ad-hoc way rather than the answers
23:21:58 <HackEgo> http://codu.org/projects/hackbot/fshg/index.cgi/file/tip/bin/toroman
23:22:31 <int-e> oerjan: it's not pretty but far better than I dreamed possible.
23:22:56 <Dulnes> oerjan: btw codegolf ? thats competitive right
23:26:39 <int-e> oooh fizzie exploited the trailing whitespace, too
23:27:36 <int-e> fizzie: the thing is a bash, you could've used $[] for $(())
23:28:10 -!- `^_^v has quit (Ping timeout: 265 seconds).
23:28:35 <Dulnes> 37*6e124 * 556( 56 / 34 ) = ¿?
23:30:40 -!- Phantom_Hoover has joined.
23:31:39 <HackEgo> [wiki] [[My Unreliable Past]] M http://esolangs.org/w/index.php?diff=41029&oldid=41027 * BCompton * (+1) /* Semantics */ Data is plural
23:36:22 <Dulnes> Whats that language that makes everything look like binary
23:37:52 <tromp_> binary lambda calculus?
23:38:05 <oerjan> Dulnes: of course, the goal is to have the shortest program (in your language, or sometimes at all)
23:39:03 <Dulnes> anylanguage? even if its one of the joke ones
23:41:33 <oerjan> sure. some of the esolangy ones are pretty good for golfing, with very short commands. others are hard to golf in, though.
23:41:55 <oerjan> as i said, burlesque is popular here.
23:42:42 <oerjan> i've submitted a few in unlambda. not very good for numbers, that.
23:42:45 <Sgeo> Does an esolang that is easier to implement a REPL for than any other mode including file interpretation sound like a thing that could be made?
23:42:51 <Sgeo> I blame FreeFull for giving me the idea
23:43:53 <oerjan> if you can think of a language where that actually _is_ easier, sure.
23:43:54 <FreeFull> You have to keep in mind, stdin is practically a file
23:44:13 <FreeFull> Maybe if you tied the output to the input
23:44:26 <FreeFull> So the user has to use the output to change their input to be suitable
23:44:33 <Dulnes> Ive had alot of short commands in python and javascript++ and C# idk if i can make "short" commands in an esolang i like to make commands long and stringy for no reason
23:44:43 <FreeFull> And you can use randomisation to make sure the output doesn't stay the same
23:44:54 <FreeFull> So you can't just feed in a file and have it work
23:46:48 <int-e> fizzie: there. 180 :)
23:47:01 * Dulnes jokingly makes it in huh?++
23:47:07 <FreeFull> I was thinking earlier, it'd be neat to have a system where you can feed in stuff like i² = j² = k² = ijk = -1; and have it automatically derive how to work with quaternions
23:48:13 <int-e> fizzie: (2 bytes from the $[]; the remaining bytes came from recompressing the data (1 byte) and then truncating the file (ok, so zcat is now somewhat unhappy, but who cares...))
23:48:45 -!- NATT_SiM has quit (Remote host closed the connection).
23:54:09 -!- NATT_SiM has joined.
23:56:27 <Dulnes> 37*6e124+*+556(+56+%2F+34+)+&cad=h = 2.032998e+129
23:59:30 -!- NATT_SiM has quit (Remote host closed the connection).
23:59:40 <Jafet> You can't make a system that takes arbitrary rewrite rules or group presentations and works out the correct normal forms