00:00:24 <oerjan> hm maybe it wasn't on the web
00:00:27 <oren> http://www.orenwatson.be/fontdemo.htm is updated
00:01:04 <oren> now with 100% as many snowmen
00:01:23 <ais523> oren: the same number of snowmen?
00:01:25 <oren> er. wtf did i just mean
00:01:27 <izabera> what's the last mathematical operator?
00:01:41 <oren> ais523: no i mean I added the snowman
00:02:38 <ais523> oren: 100% as many of something is the same number of that thing
00:03:20 <oren> i know, I know
00:03:36 <zzo38> I remember some time ago I needed to know some algorithm relating to Levenshtein distance and Damerau-Levenshtein distance but I do not remember what, and it wasn't simply calculating the value.
00:04:21 <HackEgo> U+22FF Z NOTATION BAG MEMBERSHIP \ UTF-8: e2 8b bf UTF-16BE: 22ff Decimal: ⋿ \ ⋿ \ Category: Sm (Symbol, Math) \ Bidi: ON (Other Neutrals) \ Character is mirrored
00:04:24 <oerjan> zzo38: was it that text generation for a console?
00:04:34 <zzo38> oerjan: Maybe; I do not remember
00:05:07 <zzo38> But I think it may have had something to do with optimizing Damerau-Levenshtein distance
00:05:19 <shachaf> unicode bag brackets are so good
00:06:03 <oren> I don't have those, what characters are they
00:06:28 <izabera> least useful characters ever
00:06:53 <oren> I'll upgrade that block's priority
00:12:44 <zzo38> Actually, I think what I needed was a way to sort a list of items such that the sum of the Levenshtein distances of pairs of adjacent elements is minimized. (Either that or maybe it was actually the Damerau-Levenshtein distance; I do not quite remember. Or perhaps with generalized costs.)
00:13:28 <oren> isn't that the travleing salesman problem?
00:13:55 <zzo38> I suppose it is similar at least
00:16:41 <oren> you can reduce it to the traveling salesman problem and then use an existing algorithm for that
00:17:55 <zzo38> O, OK, but still I do not quite remember what it is I have needed before, or what for, so I won't do it now.
00:19:05 <zzo38> oerjan: What text generation for a console do you refer to?
00:28:41 <oerjan> it was either famicom or z-machine, i think
00:28:43 -!- qwertyo has joined.
00:29:07 <oerjan> you were trying to do it with few instructions
00:30:40 <zzo38> I remember about Z-machine trying to encode text using as few Z-characters as possible, yes
00:31:30 <zzo38> The algorithm that Infocom has described in their documentation does not work, and anyways they did not use it (they simply avoided using any permanent shifts at all; Inform does the same).
00:32:01 <izabera> still related to my typo thingy: does this list cover pretty much all the commands you use frequently?
00:32:04 <izabera> bash sh git sudo python iptables uname node perl wget curl tmux screen tar grep find ssh sed awk vim emacs diff sort pwd gzip free htop top kill cp cat mv ls du dd df rm rmdir mount chmod chown passwd mkdir locate man tail head less ping date file apt apt-get aptitude pacman yum rpm which ifconfig
00:33:19 <zzo38> izabera: Not really, although I expect different people it will be pretty differently (it also depends what distro is in use, and other stuff)
00:33:26 <oren> almost: scp nano and svn are missing
00:34:11 <zzo38> It may be useful to allow customization though
00:34:13 <izabera> i'm caching the corrections for those commands
00:34:47 <zzo38> Since they are different for different pairs of people and computers they use
00:35:15 <izabera> oh yes ofc i'm just writing this to speed up the lookup for the common cases
00:36:32 <izabera> if it's not in that list i just perform a regular lookup from all the commands in your system and ask the user if the closest guess is what they meant
00:36:38 <zzo38> Ah, OK, I don't really know what is most common but that seem reasonable I suppose. Depending how vim is set up it may also be started just by typing "vi" instead, is one; also some (including myself) will use dc as well, commonly
00:36:59 <izabera> ok adding scp nano vi svn hg
00:36:59 -!- Slereah has joined.
00:37:00 <zzo38> izabera: Yes but you could have, if the user wants to change it then they can recompile it
00:37:22 -!- Slereah has changed nick to Guest71111.
00:37:32 <izabera> there's no real need to recompile, you can just add your own to the configuration file
00:38:06 <zzo38> izabera: OK, then, but it may depend how the lookup works exactly, if it needs to be reindexed or not.
00:38:13 -!- Slereah__ has quit (Ping timeout: 250 seconds).
00:39:00 <zzo38> oerjan: Note that even what I have for Z-machine (which is based on much of your help, but also many of my own things), is only optimal if no fwords are used; with them it is much more complicated since they can interfere with what permanent shifts should be used and not, therefore changing the length.
00:39:25 <izabera> the lookup stops when it finds a word that has a levenshtein distance < 2 from the command you typed
00:39:46 * oerjan recalls someone noting about spell checking that if you're a severe enough dyslexic, asking "did you mean this" is insufficient: they won't necessarily know even if they see the word
00:40:21 <zzo38> izabera: O, I suppose that can work, then.
00:40:34 <oerjan> well that was not the exact phrasing, mind you
00:42:13 <zzo38> There are strange things (although I have considered them and made the algorithm take them into account), for example in the string "01234X56789" it will be most efficient to encode the "X" using an ASCII escape even though "X" is in the Z-character set.
00:51:40 <zzo38> Temporary shifts from state 0 to state 1 or 2 are one Z-character long. Permanent shifts from state 0 to state 1 or 2 are two Z-characters long. Permanent shifts from state 1 or 2 to state 0 are one Z-character long. All other shifts are not allowed. Uppercase letters happen to belong to state 1, and digits happen to belong to state 2; also, state 2 is the only state where ASCII escapes can be used.
00:52:46 -!- Phantom__Hoover has joined.
00:54:11 -!- Phantom_Hoover has quit (Ping timeout: 256 seconds).
00:55:28 <izabera> git: 'statu' is not a git command. See 'git --help'.
00:55:30 <izabera> Did you mean one of these?
00:55:40 <izabera> ^ stupid git doesn't use the levenshtein distance
00:57:34 <zzo38> What does git use?
00:58:10 -!- Melvar` has joined.
00:59:43 <izabera> dunno, looks like they only check the beginning of the word
00:59:45 -!- Melvar has quit (Ping timeout: 244 seconds).
00:59:53 -!- Melvar` has changed nick to Melvar.
01:00:48 -!- idris-bot has quit (Ping timeout: 272 seconds).
01:01:24 -!- idris-bot has joined.
01:01:37 <izabera> but it could be smarter -_-
01:04:48 -!- mihow has quit (Quit: mihow).
01:12:18 <lambdabot> echo; msg:IrcMessage {ircMsgServer = "freenode", ircMsgLBName = "lambdabot", ircMsgPrefix = "Jafet!~jafet@unaffiliated/jafet", ircMsgCommand = "PRIVMSG", ircMsgParams = ["#esoteric",":@ecco the dolphin"]} target:#esoteric rest:"the dolphin"
01:13:42 <zzo38> I would want the operating system for portable computers which would consists mainly of only a Forth environment and a SSH client, and then some things can be compiled in or not based on the hardware you intend to use it on
01:13:47 -!- qwertyo has quit (Read error: Connection reset by peer).
01:15:32 <izabera> so you're rewriting unix in forth
01:18:05 <zzo38> It would be simpler than UNIX if you don't need a full system; you can SSH to your computer at home if you need UNIX
01:35:43 -!- Melvar` has joined.
01:36:23 -!- idris-bot has quit (Ping timeout: 260 seconds).
01:37:27 -!- Melvar has quit (Ping timeout: 252 seconds).
01:39:12 <pikhq> It is not as though SSH is all that UNIX-specific, except by happenstance.
01:39:53 <ais523> pikhq: apparently Microsoft have claimed that ssh support in Powershell is a priority for the Powershell team
01:39:59 <ais523> I'm not sure it actually does support it yet though
01:40:21 <pikhq> IIRC it shipped in Windows 10
01:40:34 <pikhq> Ah, no, not yet but they damned well intend to.
01:40:40 <pikhq> Both in client and server.
01:48:38 <zzo38> Yes it doesn't have to be a UNIX server, that is just one possibility
01:51:25 <pikhq> A very likely one, but not at all inherent in the scheme.
01:51:34 <pikhq> It could just as easily be SSHing into another Forth machine.
02:02:47 <tswett> /topic whereas with a 9 you're curling the defender to the left
02:04:57 <tswett> So what's the correct way to write a 9, exactly?
02:04:57 -!- Melvar` has changed nick to Melvar.
02:05:09 -!- idris-bot has joined.
02:05:16 <tswett> The other day, I decided to try to come up with exact descriptions of the uppercase Latin letters.
02:05:20 <tswett> I wasn't quite sure how to describe C.
02:06:22 -!- mauris has quit (Ping timeout: 244 seconds).
02:07:44 <tswett> "There are three strokes. The first stroke is a line from top left to bottom left. The second stroke is a clockwise arc, beginning at top left, proceeding through mid-top right, and ending at mid left. The third stroke is also a clockwise arc, beginning at mid left, proceeding through mid-bottom right, and ending at bottom left. The third stroke may extend slightly farther to the right than the second stroke."
02:09:14 <tswett> "There is only one stroke, which is a counterclockwise arc, beginning at mid-top right, proceeding through top center, then through bottom center, and ending at mid-bottom right."
02:09:52 <zzo38> Then you can put them into METAFONT
02:10:56 <tswett> Describing G is gonna be tough.
02:11:01 <tswett> Nobody knows what a G is really shaped like.
02:12:17 -!- idris-bot has quit (Ping timeout: 265 seconds).
02:12:32 -!- Melvar` has joined.
02:13:38 -!- Melvar has quit (Ping timeout: 272 seconds).
02:15:08 <zzo38> pikhq: Yes you could even be SSHing into another Forth machine, if it has a SSH server, but as far as I know Forth machines don't usually run SSH servers, isn't it?
02:16:10 -!- Wright has quit (Ping timeout: 240 seconds).
02:16:42 <zzo38> Some device drivers and filesystem drivers (and software to use them) may be provided optionally for such thing as a camera, harddrive, floppy/CF/SD/MMC/USB, CD/DVD, fax modem, sound card.
02:17:44 <zzo38> Even if you don't have a fax modem you could try to send a fax using the audio port (you may use an external device to connect it to a telephone line, such as the data port of a telephone found at a hotel), so that if you are working on something and you want to print it out, then you can do so.
02:18:52 <zzo38> If the file is on a remote server then they may also have a program on the remote server to select printer controller mode and then convert DVI to Sixel graphics and then the local program can then convert the Sixel graphics into fax machine format.
02:21:32 <zzo38> Another optional device might be a mouse which can be used to send xterm mouse commands and to copy/paste text on the screen. And then, you would have either wired ethernet or wi-fi or both, if you have wi-fi then you would also have a program to find the access point and to enter the password and so on.
02:22:58 -!- Wright has joined.
02:23:36 -!- mauris has joined.
02:27:11 <zzo38> Based on my ideas of how such a system would be used you probably don't need support for the filesystems of CD and DVD, although you might want to play an audio CD.
02:31:08 -!- mauris has quit (Ping timeout: 265 seconds).
02:31:27 -!- Phantom__Hoover has quit (Read error: Connection reset by peer).
02:31:51 <oren> There are three strokes. the first is a straigt line from top right to middle right. the second is a counterclockwise arc from top middle through middle left, bottom middle, ending at middle right. the third is a straight line from middle right to the centre
02:32:29 <oren> There are three strokes. the first is a straigt line from top right to top middle
02:32:50 -!- Melvar` has quit (Ping timeout: 240 seconds).
02:33:12 <oren> the second is a counterclockwise arc from top middle through middle left, bottom middle, ending at middle right. the third is a straight line from middle right to the centre
02:33:15 <zzo38> You should make the description by mathematical notations, or by using METAFONT
02:33:40 <oren> zzo38: that isn't as cool as having the system parse English text
02:33:49 -!- Melvar has joined.
02:35:56 <zzo38> You could have it to parse English texts too perhaps and convert them into METAFONT codes that can call the use of some macros; that way the thickness and other variation can be change by changing the macros and/or by changing the value of global variables
02:39:19 -!- MDude has changed nick to MDream.
02:41:52 <pikhq> zzo38: Correct, Forth machines generally don't run SSH servers. This is a hypothetical possibility, not something I anticipate anyone actually doing.
02:42:44 -!- Melvar` has joined.
02:45:00 -!- Melvar has quit (Ping timeout: 245 seconds).
02:45:42 <oren> wouldn't forth be more useful if the underlying cpu architecture was stack-based somehow
02:46:25 <zzo38> oren: It may if the CPU instruction set is designed to run Forth; a few do have some instructions like that, but even if it doesn't you can still implement Forth
02:46:29 <zzo38> X forwarding doesn't seem needed here (it would just complicate it), but what may be useful might be "Plan9 forwarding"; is such a thing possible to do with SSH?
02:46:51 <pikhq> zzo38: You can forward arbitrary ports over SSH, so "yes".
02:47:12 <pikhq> Additionally IIRC the protocol can be expanded to cover unforeseen needs.
02:49:09 <zzo38> If you do that then you could use it to access a CF card on the client rather than being physically connected the device to the server
02:51:18 -!- mauris has joined.
02:52:45 -!- copumpkin has quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…).
02:54:49 -!- Melvar`` has joined.
02:55:32 -!- mauris has quit (Ping timeout: 246 seconds).
02:55:56 -!- mauris has joined.
02:55:56 -!- mauris has quit (Changing host).
02:55:56 -!- mauris has joined.
02:57:01 -!- Melvar` has quit (Ping timeout: 252 seconds).
03:00:11 -!- Melvar has joined.
03:02:24 -!- Melvar`` has quit (Ping timeout: 272 seconds).
03:02:41 -!- Melvar` has joined.
03:05:08 -!- Melvar has quit (Ping timeout: 260 seconds).
03:11:24 <HackEgo> [wiki] [[GolfScript]] http://esolangs.org/w/index.php?diff=43913&oldid=43403 * Oerjan * (-25) Seems alive again
03:23:46 <HackEgo> [wiki] [[K-on Fuck]] M http://esolangs.org/w/index.php?diff=43914&oldid=43411 * Oerjan * (-6) /* Perl Implementation */ Mediawiki markup needs only one space, and doesn't ignore extras
03:24:57 -!- mauris_ has joined.
03:24:57 -!- mauris_ has quit (Changing host).
03:24:57 -!- mauris_ has joined.
03:28:26 -!- mauris has quit (Ping timeout: 246 seconds).
03:56:46 -!- mauris_ has changed nick to mauri.
03:57:09 -!- mauri has changed nick to mauris.
04:04:57 -!- hppavilion[1] has joined.
04:05:45 <pikhq> Fond of the HP Pavilion or something?
04:07:14 -!- oren has changed nick to LenovoThinkpadT5.
04:07:19 -!- LenovoThinkpadT5 has changed nick to ThinkpadT540p.
04:08:13 <HackEgo> [wiki] [[Refract]] http://esolangs.org/w/index.php?diff=43915&oldid=43454 * Oerjan * (-13) Some proofreading
04:11:15 <zzo38> Do you know of a program to read and show where all of the backreferences are in a DEFLATE stream? If it doesn't exist I could write one though
04:19:14 <HackEgo> [wiki] [[N]] http://esolangs.org/w/index.php?diff=43916&oldid=43434 * Oerjan * (+11) wikify
04:26:18 <HackEgo> [wiki] [[Jumper]] http://esolangs.org/w/index.php?diff=43917&oldid=43446 * Oerjan * (+1) tag error, section case
04:26:47 <Sgeo> Is my solution for a Hack ASM project readable or terrible? http://pastie.org/private/zvdtyldnacegggvd95v4og
04:27:12 <HackEgo> [wiki] [[Jumper]] M http://esolangs.org/w/index.php?diff=43918&oldid=43917 * Oerjan * (-12) /* Examples */ oh and more of these markdown spaces
04:28:29 <ais523> zzo38: I wrote one of those for my custom compression format, but it doesn't do DEFLATE
04:28:50 <oerjan> ais523: can you add a spam detection for when people start lines with 4 spaces twh
04:29:17 <zzo38> ais523: What is your custom format?
04:29:33 <ais523> oerjan: are actual spambots doing that? or is it humans who don't know how to wikify?
04:30:08 <ais523> zzo38: it's basically a recursively-defined dictionary, where strings are defined in terms of shorter strings
04:30:24 <ais523> until eventually you can write the entire string you care about concisely
04:30:50 <ais523> in my tests it had much better compression than gzip even with no attempt at optimization, and when heavily optimized, was slightly better than bzip2
04:31:02 <ais523> unfortunately, "slightly better than bzip2" isn't really a selling point when there are other compression formats around which are much better
04:32:20 <zzo38> Well, I need a nonrecursive dictionary for use with Z-machine text compression, because Z-machine does not support recursive dictionaries.
04:32:34 <zzo38> ais523: What kind of data have you tried to compress though?
04:32:44 <ais523> zzo38: I tested it on the GPL version 3 and on a C-INTERCAL paxball
04:34:20 <zzo38> Depending on the kind of data some preprocessing steps may be useful; much depends on the kinds of data I think, which is why I wanted to make the compression format where it joins together various blocks which may include compressed user-defined codes too
04:34:38 <pikhq> ais523: Is there a good reason to call it a paxball?
04:35:05 <pikhq> I mean, the pax format is basically ustar.
04:35:09 <ais523> pikhq: it's INTERCAL, thus it's arbitrarily different for no good reason
04:35:13 <ais523> other than being different
04:35:29 <ais523> also, it's INTERCAL, which means that it especially tries to follow standards if nobody else does
04:35:47 <ais523> (also, pax is backwards-compatible and so people can actually read it)
04:35:59 <pikhq> (because pax is literally ustar with extensions)
04:36:05 <ais523> being different for the sake of it is INTERCAL's actual reason for existence, though
04:36:13 <ais523> just like being hard to compile is Befunge's
04:36:22 <ais523> the other stuff, such as the actually-kind-of-neat control flow model, came later
04:36:31 <pikhq> That seems to understate the difficulty of compiling Befunge.
04:36:53 <ais523> I think it was the invention of NEXT FROM that really kicked it off
04:37:45 <ais523> having functions that can call themselves opens up a bunch of possibilities that would be awkward to do in other ways (I mean, without other code having to call into them first)
04:38:29 <HackEgo> [wiki] [[???]] http://esolangs.org/w/index.php?diff=43919&oldid=43768 * 108.35.44.196 * (+479) Explain the name “???”
04:39:34 <pikhq> Is C-INTERCAL compilable in a K&R C compiler?
04:39:38 <ais523> oerjan: anyway, wrt that detection thing, it would be possible to display a message that just explained how wikimarkup worked, which would just be the message (plus requiring a second confirm)
04:39:51 <ais523> pikhq: I don't think so, but I'm not sure; maybe if you used unprotoize first
04:40:19 <pikhq> Probably not worth the effort. Being pedantically correct C is itself probably worth it.
04:44:16 <zzo38> Do you know how to prove that the algorithm I am using for figuring out optimal shift sequences for Z-machine text is perfect?
04:47:57 -!- ThinkpadT540p has changed nick to oren.
04:53:49 <oerjan> zzo38: wasn't it based on finding the optimal sequences for initial substrings, given each final state? so assuming all states are accounted for, it should be just induction on adding another character...
04:55:00 <oerjan> basically, the sequences for each substring must be optimal for given initial and final state
04:55:33 <ais523> mauris: huh, I didn't realise you'd written a golfscript derivative
04:56:55 <zzo38> oerjan: That isn't the entire thing; there is other stuff too that I added, because it isn't quite as simple as I had explained to you at first; nevertheless I believe that your way would work but see the entire thing to learn how it works
04:57:14 <zzo38> It is http://zzo38computer.org/zmachine/doc/tricky.txt
04:58:16 <ais523> mauris: that customizable h command is genius :-)
04:59:20 <mauris> i like my f command too!! so good for code golf
04:59:43 <mauris> push a string containing all of fizzbuzz
04:59:46 <ais523> hmm, I still need to write an interpreter for that language where you give it the number of an anagolf problem
05:00:14 <mauris> http://golf.shinh.org/p.rb?FizzBuzz#ranking
05:00:15 <ais523> then it tries out the solutions to that problem that are visible in reveal source, and takes a majority vote
05:00:23 <ais523> is f customizable too?
05:00:33 <mauris> sadly no! wow, maybe it should be
05:00:46 <ais523> hmm, the logical consequence of this
05:01:00 <ais523> is to write a language that's TC using nothing /but/ standard programming examples as commands
05:01:08 <ais523> like, you can do control flow by feeding a hello world to a fizzbuzz, or something
05:02:03 <pikhq> print fizzbuzz[fibonacci(1)]
05:02:08 <mauris> that sounds interesting
05:02:45 <mauris> imo it should be TC but the only things it can actually output are standard programming example things
05:02:47 <ais523> it'd be great because it'd be really hard to show off its TCness without there being an easy way to write whichever example program you picked
05:02:57 <ais523> mauris: TC internals but only standard example as output?
05:03:06 -!- olsner has quit (Ping timeout: 244 seconds).
05:03:37 <mauris> so you can write a program that counts from 1 to 10, but only by counting in unary using "hello world"s as a unit
05:03:59 <mauris> separated by... "99 bottles of beer on the wall"s
05:04:00 <pikhq> That'd be a fun Brainfuck interpreter.
05:04:38 <ais523> mauris: I've recieved horrified complements for the way I did I/O in Radixal!!!!
05:05:28 <ais523> (ugh, that fourth exclamation mark /still/ gets me even nowadays, although it's not quite as mind-destroying as it once was)
05:05:30 -!- olsner has joined.
05:06:15 <mauris> mind-destroying how so
05:06:24 <ais523> "To avoid confusion, all numbers in this specification will be written in Roman numerals." is still one of my favourite sentences
05:06:37 <ais523> mauris: I decided that three was the highest justifiable number, which is why it uses four
05:07:17 <HackEgo> [wiki] [[Radixal!!!!]] M http://esolangs.org/w/index.php?diff=43920&oldid=36037 * Ais523 * (+0) /* Radixal!!!! integers */ fix accidental use of decimal, which would crash the reader if encountered
05:08:09 <mauris> oh nooooo i found https://esolangs.org/wiki/EsoInterpreters again
05:08:10 <oerjan> zzo38: as long as the cost table still is correct with your adjustments, and all characters either follow it or have cost entirely independent of state (i.e. they can be ignored), then it should work...
05:08:11 <ais523> <Esolang> (Yes, this means that although arbitrary codepoints can be input, it's not possible to output arbitrary codepoints; programmers are encouraged to find similar-looking codepoints that can be output and use them instead.)
05:08:21 <ais523> mauris: hmm, that makes me want to work on the Underlambda project again
05:08:28 <ais523> but I got bogged down in details
05:08:41 <myname> ais523: i like the idea of the new language
05:09:03 <pikhq> ais523: Gotta love INTERCAL.
05:09:05 <ais523> but it'd be hard to make it work, and I don't really want a hard-assed version
05:09:27 <mauris> i want to extend the chain. but how do you actually implement things like Lazy K
05:09:33 <ais523> like, how do we get control flow out of standard examples? clearly fizzbuzz can do some sort of loop, but…
05:10:44 <myname> string building can be made by accessing an index in 99 bottles of beer
05:11:05 <ais523> but what operation lets us index strings?
05:11:14 <zzo38> It certainly seem different than how Infocom intended it, but it is the way optimal for their design. Their design is strange in this way, it seem like; not only with text but also with instructions and other stuff, there are strange things you can do which are almost certainly not what Infocom had in mind.
05:11:47 <mauris> what about looping by passing callbacks for fizz() and buzz()
05:12:13 <myname> ais523: what about c like pointer arithmetics for that?
05:12:40 <myname> fizz and buzz should actuaöly be related to fizz and buzz
05:12:47 <myname> don't just rename commands
05:12:49 <zzo38> They almost certainly would not have purposely designed it such that these kind of algorithm would be in use.
05:12:51 <ais523> mauris: the problem there is that you'd probably just put all your logic in fizz and have a no-op for the buzz
05:13:12 <ais523> myname: what sort of keyboard lets you typo an ö?
05:13:38 <zzo38> (What their compiler actually does is not use any permanent shifts at all; only temporary shifts are used. They have put in other features too which they have never used.)
05:13:47 <ais523> hmm, I guess next to l is a reasonable position for that
05:13:53 <oerjan> almøst a norwegian too
05:14:17 <myname> ü is next to p and öä are next to l
05:14:47 <pikhq> https://cadie.googlecode.com/svn/trunk/INTERCAL-style-guide.html Clearly I should use INTERCAL at work.
05:14:56 <oerjan> huh the german keyboard is probably close to swedish then
05:15:17 <ais523> pikhq: fun fact, I have commit access to that repo
05:15:22 <oerjan> my keyboard has öä marks on those keys but their not active
05:15:46 <myname> i do think the swiss keyboard is weird
05:16:13 <zzo38> (Joystick support, XZIP menus, the ORIGINAL? instruction, etc. I have also seen various other things in their story files (from disassembling them in the debugger I wrote) that I have been able to optimize even more better, too.)
05:16:17 <myname> especially since caps lock + shift + x = x does not hold
05:16:37 <mauris> It's not difficult to write interactive programs in Lazy K. However, you should be aware that doing so is, technically speaking, a sin.
05:16:42 <ais523> myname: I was thinking "that doesn't work on my keyboard" then realised I rebound caps lock ages ago and sort-of forget its original purpose
05:16:57 <ais523> I had an alternative binding (shift-shift) to turn on caps lock for a while
05:17:09 <ais523> (if you can modify other keys with shift, why not shift?)
05:18:19 <ais523> pikhq: huh, I've just noticed that that style guide doesn't outright say "alternate sparks and ears at alternate levels of grouping", which is normally considered the #1 INTERCAL readability tip
05:18:34 <ais523> it sort-of gets close but doesn't outright state it
05:18:47 <ais523> and the examples seem to be written in that form (because what sane programmer wouldn't?)
05:18:56 <myname> there is an à key that does ä on shift, À on caps lock and Ä on shift+caps lock
05:19:17 <ais523> myname: ugh, most keyboards that need that sort of thing use altgr for that purpose
05:20:05 <ais523> (although on a UK keyboard, altgr is used for two keys: the second |, which is different from | on many OSes but not on Linux/X by default, and € which was a relatively recent addition)
05:20:22 <ais523> I'm not quite sure why we had an altgr key, actually, other than futureproofing
05:20:46 <myname> i don't even know how to make the second |
05:20:48 <ais523> the secondary | can't be that important, as shown by the fact that the two are frequently merged
05:20:57 <ais523> myname: it's altgr-` on a UK keyboard
05:21:05 <ais523> sometimes one is drawn solid and the other as a broken line
05:21:09 <ais523> or one is thicker, or the like
05:21:17 <ais523> but both map to the ASCII version with the config I have at the moment
05:22:06 <shachaf> myname: I think some Hebrew layouts use caps lock + shift + numbers for the what's-it-called.
05:22:10 <ais523> anyway, reading that style guide again
05:22:13 <shachaf> Diacritics? Whatever the name is in English.
05:22:23 <ais523> I'm amused at how ".5 is always a boolean" has become such a standard in the INTERCAL community
05:22:36 <ais523> it just sort-of happened with the system library's convention, then sort-of expanded from there
05:22:58 <pikhq> ais523: As I understand it, AltGr mostly came into being in the UK as a consequence of using the ISO mechanical layout rather than the ANSI mechanical layout.
05:23:12 <ais523> I suspect this is at least half to do with a really common technique of using an undocumented part of the system library internals to save one line of code
05:23:37 <ais523> (to the extent that when I wrote syslibc, I had to preserve that undocumented behaviour because too much existing code broke otherwise)
05:24:22 <ais523> also, PLEASE NOTE is my standard comment syntax, but only because it looks better than the alternatives
05:25:04 <pikhq> *Aaaah*. The UK keyboard is designed to also be usable for Irish.
05:25:24 <ais523> that makes sense, although in that case I'd expect Irish letters to appear on the keycaps
05:25:53 <pikhq> I don't know why they don't, but apparently the vowels have accented versions selected with Alt-Gr.
05:25:56 <ais523> "Con: Allowing extra groups sometimes allow you to avoid having to exchange all the rabbit-ears in a subexpression with sparks (and vice versa) when editing code."
05:26:05 <ais523> OK, they've come /really really close/ to insisting that they always alternate
05:26:07 <ais523> and still haven't stated it
05:26:14 <pikhq> That might be intentional. :)
05:26:15 <ais523> I guess the author must just have assumed they said that already
05:26:28 <ais523> æe→ø↓ (altgr-vowels on my keyboard)
05:26:55 <ais523> is that e the same as an ASCII e? they look identical to me but that'd seem like a huge waste of an altgr binding
05:26:57 <oren> left arrow and doiwn arrow?
05:27:25 <ais523> I know altgr-e is é on Windows, and is sometimes even written as such on the keycap
05:27:33 <oren> try the e and then `
05:27:41 <pikhq> In a way I'm spoiled by the ANSI keyboard. It's friggin' great for English and programming.
05:27:49 <ais523> oren: I can type é as altgr-; e
05:27:52 <pikhq> Probably something to do with those being all they do. :P
05:28:06 <ais523> and that's my muscle-memorized keybinding even though é can be typed as compose, e, '
05:28:23 <ais523> pikhq: I think languages tend to be designed around keyboards, especially more recently
05:28:52 <ais523> ¬ would presumably be used more if the UK keyboard was also used in the US
05:29:11 <pikhq> Also historically; the US keyboard layout is very close to the typical typewriter layout.
05:29:42 <ais523> I've seen typewriter layouts with no 0 or 1
05:29:53 <ais523> I think you were supposed to use O and I instead
05:31:02 <pikhq> Actually, quite specifically the US keyboard layout is *basically* the Selectric keyboard layout (which was slightly different from traditional in where some of the special keys got located).
05:33:12 <oren> shachaf: I think they are called vowel pointing
05:33:30 <oren> (wikipedia says so)
05:33:48 <mauris> AZERTY (which i usually use) has a separate key for a letter that exists in exactly one french word
05:33:57 <mauris> some other junk is on it if you hold shift. but wow
05:34:43 <shachaf> oren: That's a verb, it sounds like.
05:34:50 <pikhq> ais523: Huh. The actual *spec* for the UK keyboard doesn't actually specify any use for AltGr, it just specifies it'll be there.
05:34:56 <shachaf> According to https://en.wikipedia.org/wiki/Niqqud the name is just transliterated.
05:35:06 <shachaf> But I don't know whether to use q or k or what, and a double one or a single one.
05:35:11 <ais523> pikhq: I'm glad I have an altgr, anyway
05:35:20 <ais523> but this does not make it any easier to explain the key to Americans
05:35:42 <pikhq> I've taken to binding right-win to Compose, which at least gets faintly similar flexibility.
05:35:54 <ais523> I guess it was just future expansion? € is pretty important nowadays and it was nice that we had somewhere to put it once we discovered we needed it?
05:36:05 <ais523> (there's also an iso-8859 variant which has a euro in and is otherwise -1)
05:36:26 <shachaf> I used to bind the menu key to Compose.
05:36:34 <oren> that's windows 1252
05:36:37 <shachaf> This keyboard has neither a menu key nor a right Windows key.
05:36:41 <pikhq> Also a consequence of being near to the EU -- note that *physically* all the European keyboards are actually identical.
05:37:02 <pikhq> The differences are just in the keycaps and what the OSes do with 'em.
05:37:10 <b_jonas> ais523: yes, there's two or three such variants:
05:37:40 <ais523> pikhq: huh, US keyboards are physically different from UK ones?
05:38:02 <pikhq> https://en.wikipedia.org/wiki/File:KB_United_States-NoAltGr.svg This is a normal US keyboard.
05:38:13 <ais523> some of the keys are different shapes but I thought that was just done via the use of keycaps that didn't plug into all of the (whatever you call the thing that keycaps plug into)
05:38:24 <b_jonas> because iso-8859-15 is the modification of iso-8859-1 with euro, and iso-8859-16 is the modification of iso-8859-2 with euro and the disunified Romanian letters, these two don't get used much because they came late and unicode was already well-spread by the time;
05:38:55 <pikhq> You've *buttons* we don't have.
05:39:17 <pikhq> But yeah, no. Each keycap has its own hole.
05:39:18 <oren> I also have a numpad
05:39:30 <oren> and arrow keys
05:39:40 <b_jonas> and cp1252 is the modification of the original cp1252 without renaming it, with euro added, and cp1250 is the modification of the original cp1250 with euro and the disunified Romanian letters added, without renaming, and these do get used a lot because some Windows programs use locale byte encodings by default and these often are the locale's byte encodings
05:40:10 -!- hppavilion[1] has quit (Ping timeout: 240 seconds).
05:40:24 <shachaf> But that's a bizarre place to put \
05:40:30 <ais523> pikhq: I was doing that on the basis that you can easily switch a UK keyboard over to being interpreted as US
05:40:35 <ais523> and then you just have some duplicate keys
05:40:56 <ais523> (the main one being that our # key becomes a duplicate \)
05:41:07 <pikhq> Though, L-Alt and R-Alt generate different codes, so you *can* switch to a more general sort of layout. Namely, US-International, which has R-Alt as Alt-Gr.
05:41:18 <pikhq> https://en.wikipedia.org/wiki/File:KB_US-International.svg
05:41:21 <oren> OH so that's why my friend's laptop has two \ keys
05:41:24 <shachaf> US International (AltGr Deadkeys) is the right one.
05:41:43 <pikhq> Very few actually use the thing.
05:41:58 <shachaf> I've been using it for a long time.
05:42:18 -!- hppavilion[1] has joined.
05:43:58 <pikhq> And then there's the rarely used Canadian multilingual keyboard. :)
05:44:44 <pikhq> https://commons.wikimedia.org/wiki/File:KB_Canadian_Multilingual_Standard_comment-en.svg Yeah, that's right. Two Alt-Gr buttons.
05:46:22 <Taneb> That was a nap and a half
05:46:28 <Taneb> Well, more like about 12 naps
05:46:34 <Taneb> Well, more like a solid night's sleep
05:46:44 <Taneb> I just accidentally slept for about 12 hours
05:47:33 <ais523> Taneb: my record is something like 18, that's only happened something like twice though
05:47:41 <ais523> and I was very tired for other reasons beforehand
05:47:55 <ais523> more worrying is when I'm really tired, go to bed, wake up feeling refreshed
05:48:00 <ais523> then realise I've only been asleep three hours
05:48:01 <Taneb> ais523, I was not expecting to go to sleep at all
05:48:13 <Taneb> I just intended to have a short nap
05:48:21 <ais523> then go around trying to establish the date because three and twenty-seven seem about equally believable
05:48:32 <ais523> (and the most accessible clocks show time but not date)
05:49:46 <oren> pikhq: yeah. my firend has a Acer Aspire 6920G (giant media centre lap-crusher) with the canadian multilanguage layboard
05:49:51 <b_jonas> I sometimes sleep short and feel refreshed, or sleep very long, but I don't get confused about the time, I almost always sleep in a way that I know approximately how much time I slept, without a clock
05:50:38 <oren> wait that's not the right model...
05:50:50 <Taneb> But I had a really bad night on Sunday night and was tired for most of yesterday
05:52:06 <b_jonas> apparently the keyboards sold in Sweden have two Swedish keyboard layouts printed on them, which differ only in two keys: the ; and ' have ö and ä swapped
05:52:55 <zzo38> I have made up a VM as well as Minesweeper and Robotfindskitten and Munching Squares and a few other examples
05:52:58 -!- mauris has quit (Read error: Connection reset by peer).
05:53:22 -!- mauris has joined.
05:54:08 <zzo38> http://zzo38computer.org/prog/quackvm.zip - Source-codes is provided, and needs xterm and Linux to run; CTRL+Q quits and CTRL+R restarts.
05:55:57 <oren> right. it's acer aspire 6920ZG, note the Z, for, uh... really really bog
05:57:46 <oren> http://regmedia.co.uk/2008/03/14/acer_6920_1.jpg
06:04:21 * Sgeo wonders if zzo38 would like Hack ASM
06:05:57 <izabera> zzo38: some browser don't support gopher -__-
06:06:22 <Taneb> Trivia: I have a gopher-accessible selfie
06:06:30 <Taneb> It is the only thing I have put in the gopherspace
06:07:55 <zzo38> izabera: Yes, some don't, but some do and there are also standalone gopher clients, and extensions to support gopher in some browsers
06:09:10 <HackEgo> [wiki] [[Brainfuck Sharp]] M http://esolangs.org/w/index.php?diff=43921&oldid=43670 * Hppavilion1 * (+46) Fixed formatting (whoops)
06:10:14 <Taneb> Oh right, my esolangs page is also accessible via gopher
06:10:21 -!- evalj has joined.
06:10:26 <Taneb> And an interpreter fro brainfuck in Snap
06:10:27 <izabera> http://www.anybrowser.org/campaign/ you two read this
06:11:44 <Taneb> I believe I've seen that before, and agree with the sentiment
06:12:34 <b_jonas> Yeah... back ages ago I once had a cgi page that printed a Best Viewed With marker for the other browser than the one you were using
06:13:46 <b_jonas> back in those days sites had said “Best Viewed By Internet Explorer at resolution 800x600” or something. These days they don't say that, they're just silently broken on most browsers and window sizes and font sizes other than what they were testing for.
06:13:52 <lifthrasiir> (you don't need parallax to view this page)
06:14:21 <Taneb> "This is best viewed in a browser"
06:14:37 <Taneb> "Stop reading it using curl"
06:14:40 <b_jonas> lifthrasiir: no, I don't want that either. a webpage should be readable without eyes too, on a Braille terminal or screen reader or with the bytes relayed directly to my mind or whatever
06:14:42 <zzo38> I just don't do anything that cares about such things; leave the client to decide the best way to render it. Gopher requires this.
06:14:58 <b_jonas> zzo38: yes, that's the correct approach
06:15:23 <Taneb> I have no idea if my esolangs page (the only page I really maintain) is readable with a screen reader or Braille terminal
06:15:29 <b_jonas> and it's generally easy too, with only a few cases where it's difficult because of some feature I want not accessible in an easy portable way
06:15:45 -!- Wright has quit (Ping timeout: 252 seconds).
06:16:22 <b_jonas> Taneb: what matters most for that is that you don't put important information (like text of the links on a menu listing the pages of your site) in images showing text only, but also as text (possibly alt text)
06:16:45 <lifthrasiir> b_jonas: I tried to add "best viewed with one or more computing apparatuses" and gave up
06:17:01 <Taneb> b_jonas, the only image is a little public domain button at the end
06:17:21 <izabera> many people use an image representation of their email address
06:17:51 <Taneb> No, mine is in plain text and uses no techniques to forestall spiders
06:17:51 <izabera> http://aiju.de/ but there's this other approach
06:17:52 <zzo38> But you can also use a text description of your email address, that is also used
06:17:53 <b_jonas> izabera: I use both an image representation and an obfuscated text on my homepage
06:18:13 <oren> most of my pages can be read using curl
06:18:26 <izabera> b_jonas: obfuscated like "feed this to a brainfuck interpreter"?
06:18:36 <b_jonas> izabera: as in Contact e-mail: <img src="http://www.math.bme.hu/~ambrus/sc/email-c.png" alt="ambrus@@mmaatthh..bbmmee..hhuu" style="position: relative; top: 4px;">
06:18:51 <zzo38> I mostly just use plain text files anyways, so you can easily just use command-line program anyways
06:18:58 <oren> specifically, if you leave off the .htm, you will typically get a plain text version
06:19:09 <oren> because of some cgi i wrote
06:19:20 <b_jonas> but that's from way back. these days I don't hav to matter because I get lots of academic spam from my email appearing in metadata of scientific publications in plain text unobfuscated
06:19:53 <izabera> just thought i'd mention that my pastebin, http://arin.ga serves a raw version when viewed with curl and wget, or a html version with a browser
06:20:46 <lifthrasiir> I have a TeX representation for my email address: \sideset{someone}{\!mearie}{\mathop{@}}\!_\cdot org
06:21:09 <b_jonas> zzo38: I use HTML even for most of the simple pages (eg. http://www.math.bme.hu/~ambrus/pu/verst and the other two poems quoted on my homepage): that makes it slightly easier to provide metadata, and people can still read or copy the text easily.
06:21:22 <lifthrasiir> (of course, I have also noted that any email pointing to the present domain will reach me)
06:21:27 <Taneb> Hmm, my esolangs page doesn't get through w3c's validator unscathed
06:21:59 <zzo38> I put metadata only using <LINK> tags in the root page (that's all it contains), since that is the only page that needs metadata.
06:22:26 <b_jonas> zzo38: sure, that works too, at least if you make sure the http headers are correct
06:22:42 <b_jonas> there are some metadata you can give as either a http header or html elements
06:22:53 <b_jonas> the html elements were first, but then they invented a solution for non-html data
06:23:08 -!- Sgeo has quit (Ping timeout: 265 seconds).
06:23:15 <zzo38> Well, the stuff I have tried it with (only OpenID actually, so far) do work
06:23:47 -!- mauris has quit (Read error: Connection reset by peer).
06:25:27 <b_jonas> But basically I thought, if the html page contains mostly text without tags, then people can read the source even without a html viewer, so I don't bother keeping it text
06:25:41 -!- mauris has joined.
06:25:41 -!- mauris has quit (Changing host).
06:25:41 -!- mauris has joined.
06:26:28 <zzo38> Yes, that does help too.
06:27:46 <b_jonas> There's a HTTP headers for giving the language of the text, and I think there's some for giving the canonical url (same as <link rel="canonical"> in html) and instructions for crawlers to index or not index it
06:28:03 -!- Guest71111 has changed nick to Slereah.
06:28:08 <b_jonas> oh, I think one for the title too
06:28:28 <b_jonas> because plain text is plain text and doesn't have a general way to represent any of these
06:28:31 <zzo38> I just can use robots.txt to tell crawlers not to index it (although I allowed wget)
06:29:34 <b_jonas> zzo38: yes, robots.txt tells about the noindex, but you may also want to tell about the nofollow, for user-submitted material that contains spam
06:30:11 <b_jonas> to indicate that even though it contains links to buy-rolex-sunglasses-cheap-online.com, you're not trying to say that's a good site
06:30:27 <b_jonas> but you still want to allow the robots to crawl and make that page searchable
06:31:27 <b_jonas> practically speaking, the encoding and language are the most important for plain text
06:31:35 <zzo38> I don't like to use nofollow
06:31:56 <zzo38> I just use ASCII encoding so it doesn't matter.
06:32:45 <lifthrasiir> (that's also why I think rel=nofollow etc. is a bit unfortunate, since it really has to be a container attribute affecting all descendants)
06:33:28 <zzo38> Even for untrusted HTML, I don't like to use nofollow, or to obey it except for internal links.
06:49:54 <zzo38> I just use plain text mostly, it works fine
06:51:56 <b_jonas> zzo38: sure, but obviously ASCII encoding won't work for me for those three poems, they contain non-ascii, so I have to specify an explicit encoding
06:52:09 <zzo38> Yes, in that case you do.
06:52:10 <b_jonas> (namely utf-8 or iso-8859-2)
06:52:38 <zzo38> But ASCII should be use when possible. Of course in some cases it isn't, so you use other encodings
06:54:28 -!- x10A94 has joined.
06:55:04 -!- evalj has quit (Remote host closed the connection).
06:56:42 <oerjan> hm, hungarian poetry with only ascii characters
06:57:13 <oerjan> that's probably a lot harder than with norwegian
06:58:01 -!- AnotherTest has joined.
06:58:27 <zzo38> Yes, but it is one kind of restrictive writing to try I suppose, like they make poetry with other kinds of restrictive writing too (no "e", word lengths must follow a predefined pattern, initial letters of each word must spell out a prefix of the full poem, etc)
07:03:10 -!- hppavilion[1] has quit (Ping timeout: 240 seconds).
07:12:08 -!- J_Arcane_ has joined.
07:13:50 -!- J_Arcane has quit (Ping timeout: 246 seconds).
07:14:03 -!- J_Arcane_ has changed nick to J_Arcane.
07:24:19 <mroman> oerjan: being fit probably includes surfing .
07:25:09 -!- J_Arcane has quit (Ping timeout: 244 seconds).
07:27:10 <oerjan> if you're australian, yes. if you're norwegian, somewhat more doubtful.
07:43:06 <zzo38> Do you any like a tsume shogi game?
07:46:08 -!- nisstyre has quit (Ping timeout: 272 seconds).
07:50:00 -!- oerjan has quit (Quit: Z?).
08:01:56 <zzo38> Now I am finding level 63 difficult
08:02:30 <Taneb> I am going to head off now, I think
08:05:08 -!- ais523 has quit (Ping timeout: 272 seconds).
08:07:28 <zzo38> The king is going to escape
08:10:31 -!- AnotherTest has quit (Ping timeout: 256 seconds).
08:11:16 -!- AnotherTest has joined.
08:12:02 <b_jonas> "<oerjan> hm, hungarian poetry with only ascii characters" -- that's hard, because á and é are very common letters
08:14:25 -!- nisstyre has joined.
08:19:00 -!- ais523 has joined.
08:20:32 <izabera> someone hacked an italian gov site http://www.giustiziaminorile.it/
08:21:21 <izabera> they wrote a couple of meaningless sentences with several spelling mistakes in a green monospaced font + low res jpeg picture as a background
08:32:59 -!- Patashu has joined.
08:34:46 -!- J_Arcane has joined.
09:16:14 -!- FireFly has quit (Ping timeout: 244 seconds).
09:28:15 -!- sebbu has quit (Ping timeout: 240 seconds).
09:40:25 -!- llue has joined.
09:40:52 -!- lleu has quit (Ping timeout: 260 seconds).
09:59:08 -!- aretecode has quit (Remote host closed the connection).
10:02:23 -!- aretecode has joined.
10:04:36 -!- rdococ has quit (Read error: Connection reset by peer).
10:04:56 -!- rdococ has joined.
10:09:42 -!- ais523 has quit (Ping timeout: 260 seconds).
10:10:29 -!- rdococ has quit (Ping timeout: 244 seconds).
10:12:16 -!- rdococ has joined.
10:18:01 -!- rdococ has quit (Ping timeout: 256 seconds).
10:18:24 -!- boily has joined.
10:19:31 -!- rdococ has joined.
10:25:14 -!- ais523 has joined.
10:57:38 -!- FireFly has joined.
11:00:10 -!- Frooxius has joined.
11:17:35 -!- boily has quit (Quit: Poulet!).
11:22:00 -!- FireFly has quit (Ping timeout: 264 seconds).
11:22:12 -!- sebbu has joined.
11:22:56 -!- sebbu has quit (Changing host).
11:22:57 -!- sebbu has joined.
11:24:20 -!- Patashu has quit (Remote host closed the connection).
11:24:23 -!- ais523 has quit.
11:25:12 -!- Patashu has joined.
11:29:27 -!- copumpkin has joined.
11:44:28 -!- MDream has quit (Read error: Connection reset by peer).
12:02:22 -!- TieSoul has joined.
12:05:12 -!- Melvar` has changed nick to Melvar.
12:07:19 -!- TieSoul has quit (Ping timeout: 260 seconds).
12:09:04 -!- TieSoul has joined.
12:09:18 -!- idris-bot has joined.
12:20:25 <Taneb> izabera, I can't work out what that translates to... Children's Justice?
12:20:35 <Taneb> ("giustiziaminorile" that is)
12:24:11 -!- FireFly has joined.
12:28:24 <oren> They also put that consipracy theory music on the sitr
12:29:18 <Taneb> Like, the child courts?
12:30:14 -!- Phantom_Hoover has joined.
12:33:20 <oren> apparently yes
12:33:36 <oren> https://it.wikipedia.org/wiki/Dipartimento_per_la_giustizia_minorile
12:34:01 -!- Patashu has quit (Ping timeout: 260 seconds).
13:21:27 -!- heroux has quit (Ping timeout: 252 seconds).
13:22:03 -!- heroux has joined.
13:28:56 -!- `^_^v has joined.
13:44:21 -!- ProofTechnique has joined.
13:47:46 -!- aloril has quit (Ping timeout: 272 seconds).
13:48:53 -!- Sgeo has joined.
13:49:32 -!- aloril has joined.
13:57:27 -!- gonad has joined.
14:00:51 -!- llue has quit (Ping timeout: 265 seconds).
14:02:09 -!- augur has quit (Ping timeout: 252 seconds).
14:04:44 -!- AnotherTest has quit (Ping timeout: 246 seconds).
14:13:43 -!- augur has joined.
14:42:19 <izabera> Taneb: yes, sorry for not explaining it
15:00:07 <izabera> http://www.yoda.arachsys.com/pics/calling/ these are actual pictures of jon skeet
15:00:16 <izabera> highest rep on stackoverflow
15:12:00 <Taneb> He looks like someone who has fun
15:13:25 <izabera> there's a link to his christian preaching site
15:13:50 <myname> the girl with the leash looks like she has fun
16:09:09 <int-e> tsk, why does Isabelle make it so hard to define functions that aren't actually functions ;-)
16:10:21 <izabera> this is not the first time i ask it
16:13:20 -!- ProofTechnique has quit (Quit: My Mac has gone to sleep. ZZZzzz…).
16:21:15 -!- Wright has joined.
16:28:11 -!- tromp__ has quit.
16:35:36 -!- Wright has quit (Ping timeout: 272 seconds).
16:47:15 -!- hppavilion[1] has joined.
16:49:20 <int-e> sorry, I should probably try to start a conversation on #isabelle instead
16:49:59 <fizzie> Grumble grumble websites that disable pasting into password boxes.
16:53:10 <ashl> i wonder what they think they're achieving with that
17:02:45 -!- tromp_ has joined.
17:15:25 -!- ProofTechnique has joined.
17:16:30 -!- hppavilion[1] has quit (Ping timeout: 250 seconds).
17:31:27 <izabera> oren: are you erik demaine?
17:44:35 <int-e> fizzie: that sounds like a ridiculously bad idea
17:45:18 <int-e> "my password is so much safer now that I had to paste it to an editor window in order to copy it"
17:49:40 <Slereah> Eh, you are more at risk from password bruteforcing than you are of hackers getting in your computer
17:52:22 -!- idris-bot has quit (Ping timeout: 265 seconds).
17:52:23 -!- Melvar has quit (Ping timeout: 244 seconds).
17:52:26 -!- Melvar` has joined.
17:52:31 <myname> bruteforcers won't use copy&paste
17:52:40 <ashl> citation needed.
17:53:03 <ashl> presumably the reason for pasting is that they use randomly generated password
17:53:06 -!- mihow has joined.
17:53:11 <myname> doing stuff by hand takes way too much time and work
17:53:20 <ashl> seems much more likely that such passwords would be stolen than brute-forced
18:10:35 -!- Melvar` has changed nick to Melvar.
18:11:14 -!- hppavilion[1] has joined.
18:18:46 -!- hppavilion[1] has quit (Ping timeout: 260 seconds).
18:19:03 -!- Wright has joined.
18:20:33 -!- hppavilion[1] has joined.
18:38:42 <b_jonas> Has anyone written self-fulfilling statements about controversial incompatible grammar changes, such that eg. in the standard grammar X the statement parses as (I hope that the gods forbid me that (I have used grammar Y, which will not happen).),
18:38:59 <b_jonas> but in the heretic grammar Y it parses as (I hope that ((the gods forbid me that I have used grammar Y), which will not happen))?
18:39:14 <b_jonas> I'm trying to compose a nice statment like that about the heretic cucmen variant of the lojban grammar.
18:40:09 <b_jonas> Making such a statent is possible, but I'd also like to make it nice, so that it doesn't sound too contreived, and is somewhat robust.
18:40:47 <Hoolootwo> that almost sounds like INTERCAL's abstain
18:48:38 -!- hppavilion[1] has quit (Ping timeout: 260 seconds).
18:51:22 -!- hppavilion[1] has joined.
18:54:50 <mauris> actually, what on earth is a "heretic grammar" even
18:56:41 <mauris> <fizzie> Grumble grumble websites that disable pasting into password boxes. <-- at least with browsers you can probably just fiddle with the input box using dev tools
18:56:51 <shachaf> mauris: i know someone who's even maurer than you
18:58:41 <mauris> yes i believe i've seen this someone in #haskell
19:12:04 -!- TieSoul has quit (Remote host closed the connection).
19:18:17 <b_jonas> mauris: by heretic variant, I mean it is a language that I disdain and claim shouldn't ever be used, because it is so incompatible with lojban that both speaking it will confuse listeners who expect lojban and speaking lojban will confuse listeners who expect cucmen, and even if you are aware of both of them, it is often hard to write sentences in a way that they have identical meaning in both. unlike, say, C and C++, where it is easy to write C code tha
19:18:29 <b_jonas> unlike, say, C and C++, where it is easy to write C code that also works in C++ if you know the basic differences.
19:19:17 <mauris> if i google "cucmen +lojban" i get zero results; where is this a thing?
19:20:19 <mauris> but i guess you mean: a sentence that is true in both languages but parses differently
19:20:28 <b_jonas> mauris: by cucmen in particular, I mean what's sometimes called "cmevla-brivla merger" or something like that (but that's a bad name). it's the grammar variant that removes the (sumti-6 -> LA [relative-clauses] CMENE) production and adds the (tanru-unit-2 -> CMEVLA) production.
19:21:11 <b_jonas> some people speak this variant and even try to advocate it.
19:22:16 <mauris> haha. i love how lojban is supposedly perfectly unambiguous, but it really isn't, because people keep pulling shit like this
19:22:26 -!- impomatic_ has joined.
19:23:38 -!- hppavilion[1] has quit (Ping timeout: 260 seconds).
19:23:46 <b_jonas> mauris: yes, and that modified language doesn't even deserve to be called lojban, unlike some extensions that are actually good.
19:24:26 <b_jonas> I want to make a public statement that if someone knowingly and deliberately keeps using cucmen even after he is taught of the errors in his ways, he is a heretic and shall be excummunicated, and shall ask the gods for foregiveness by repeating that prayer, in lojban, and understand it.
19:24:42 <mauris> like apparently they were changing what "lo" means, but all of the tutorials still explained old "lo", and it made me realize the language was silly and i sort of lost interest.
19:25:22 <b_jonas> And the prayer shall be self-fulfilling in the sense that they can't speak it in cucmen and still hope the gods will foregive them.
19:25:51 <b_jonas> But if they see the light and speak it in lojban, the gods will already have forgiven them.
19:26:27 <b_jonas> Does this make sense, at least in #esoteric ?
19:27:30 <b_jonas> So I'm asking if anyone has already done some statement similar to this, about possibly other topics than cucmen.
19:28:36 <b_jonas> The goal is that the statement shouldn't seem to be contreived for this purpose, but should sound natural, so as to demonstrate in a self-contained way why cucmen is a bad idea and is very incompatible with lojban.
19:28:43 <mauris> (even when i google "cucmen" "lo" "mi" i find no sources calling this thing that. is it your own nickname for the concept, or?)
19:28:54 <b_jonas> mauris: it's my own nickname, yes
19:29:00 -!- hppavilion[1] has joined.
19:30:08 <b_jonas> the name others use is something like "cmevla-brivla merger" but that's a _bad_ name, because that makes it sound as if it's a conservative extension that permits you to use cmevla in more ways. in reality, it's changing cmevla so you usually can't use them in ways that you can in lojban.
19:31:42 <b_jonas> I needed a short nickname that's easy to remember and recognize, and I also wanted it to be a cmevla (or juxtaposition of multiple cmevla) because I think that helps compose this statement.
19:32:14 <b_jonas> As in, the statement will probably contain the word "cucmen" and use that to help the ambiguous parsing.
19:32:43 <b_jonas> But I haven't got the statement yet, so it's possible it won't actually have that word, or won't use it in a way significant for the parsing difference.
19:33:33 <b_jonas> It's also possible that what I wrote above won't be the meaning of the statement of course. I chose that particular meaning because I think it's possible to write the statement to mean roughly those two things, but I might have to change it to actually work.
19:42:00 <HackEgo> [wiki] [[User:Nooodl/vector.css]] N http://esolangs.org/w/index.php?oldid=43922 * Nooodl * (+48) Created page with "h1, h2, h3, h4, h5, h6 { font-family: Georgia; }"
19:43:28 <mauris> i... i like serif headers. what can i say
19:46:32 <b_jonas> I can make the declaration say that you have to pray to something like "lo la .lojban. cafyfraxu cevni" = the merciful gods of Lobjan, which is an example for a phrase that isn't trivial to modify in a way that it parses correctly in both lojban and cucmen.
19:47:07 -!- AnotherTest has joined.
19:51:42 -!- x10A94 has quit (Read error: Connection reset by peer).
19:53:30 -!- hppavilion[1] has quit (Ping timeout: 255 seconds).
19:59:19 -!- `^_^v has quit (Quit: This computer has gone to sleep).
20:02:15 <b_jonas> I think that in the prayer, I want to use a phrase with form similar to {lo nu co'e la .cucme. poi brodi poi brodu} where the {poi brodu} part is an inner qualifier for the whole phrase, but in cucme it is an outer qualifier for the {la .cucme.},
20:03:33 <b_jonas> only that way it doesn't have the right meaning, because it's not cucme that I want to qualify, but the forgiving.
20:04:15 <b_jonas> I'll have to think of how this works
20:41:54 <b_jonas> Lua has changed a lot since I last looked at it (at version 5.1). I should look at the details.
20:59:18 -!- ProofTechnique has quit (Quit: My Mac has gone to sleep. ZZZzzz…).
21:04:44 -!- AnotherTest has quit (Ping timeout: 246 seconds).
21:06:06 -!- AnotherTest has joined.
21:08:14 -!- Patashu has joined.
21:12:04 -!- hppavilion[1] has joined.
21:24:06 -!- hppavilion[1] has quit (Ping timeout: 240 seconds).
21:29:51 -!- llue has joined.
21:31:33 -!- AnotherTest has quit (Quit: ZNC - http://znc.in).
21:33:54 -!- gonad has quit (Ping timeout: 272 seconds).
21:38:06 -!- ProofTechnique has joined.
21:39:14 -!- trn has quit (Ping timeout: 244 seconds).
21:39:20 <HackEgo> [wiki] [[Special:Log/newusers]] create * IfThenElse * New user account
21:40:28 -!- gonad has joined.
21:41:41 -!- gonad has changed nick to dagon.
21:41:41 -!- trn has joined.
21:42:40 -!- oerjan has joined.
21:44:12 -!- llue has quit (Ping timeout: 264 seconds).
21:45:56 -!- Patashu has quit (Ping timeout: 272 seconds).
21:47:27 <oerjan> <b_jonas> "<oerjan> hm, hungarian poetry with only ascii characters" -- that's hard, because á and é are very common letters <-- my guess is it's still easier than doing without e in norwegian, and possibly even english?
21:48:09 <oerjan> as in, what i recall from hungarian grammar is that the most basic suffixes don't have long vowels unless the root ends with a vowel.
21:51:50 <fizzie> Somehow I thought an "IfThenElse" account already existed, it sounded familiar.
21:51:51 <oerjan> in norwegian nearly all verb suffixes (except "strong" verbs past tense) contain e, as do all the articles.
21:52:41 <oerjan> and most of the noun endings too
21:53:09 <oerjan> but nouns don't have suffixes in the indefinite singular.
21:53:45 -!- hppavilion[1] has joined.
21:54:50 <oerjan> nynorsk fares a little better on the verbs, they often have a instead
21:55:14 <oerjan> oh no:har = en:has works
21:55:31 <oerjan> (either bokmål or nynorsk)
21:56:12 <oerjan> i suppose poetry can get away with leaving out articles a lot
21:56:54 <fizzie> oerjan: Poetry can get away with not using any real words, I think, if you're flexible enough.
21:57:36 -!- dagon has quit (Read error: Connection reset by peer).
21:57:37 <oerjan> yep http://www.nale.no/nale/favourites/jevkulturuke.html
21:58:23 <oerjan> only the title and final word are real (although the final word needs an extra l)
21:59:02 <oerjan> and yet the meaning of the poem is perfectly clear
21:59:57 -!- lleu has joined.
21:59:57 -!- lleu has quit (Changing host).
21:59:57 -!- lleu has joined.
22:00:07 <shachaf> oerjan: the meaning is not perfectly clear hth
22:00:21 <fizzie> "Cannot read property 'row' of null" it's not good.
22:00:36 <b_jonas> oerjan: no, I'm still awake
22:00:58 <b_jonas> oerjan: yes, it's probably easier than writing without "e" in English
22:01:04 <mauris> oerjan: what do the title and final word +l mean?
22:01:04 <oerjan> shachaf: it is if you know the meaning of those two words hth
22:01:25 <oerjan> mauris: culture week (a kind of festival) and cowpie nonsense
22:01:44 <oerjan> (a _rural_ festival, generally)
22:02:16 <shachaf> oerjan: is one of them something a rooster says twh
22:02:18 <oerjan> (well, frequently, anyway)
22:02:25 <mauris> here's something similar http://www.libellulus.com/article-ulcerations-111381103.html
22:02:39 <oerjan> shachaf: that would be "kykelipi", except for the final two letters
22:02:57 <oerjan> (i.e. the title of the entire poetry collection)
22:03:22 -!- impomatic_ has quit (Quit: http://corewar.co.uk).
22:03:34 <shachaf> in hebrew a rooster says "kukuriku"
22:03:46 -!- hppavilion[1] has quit (Ping timeout: 240 seconds).
22:03:49 <shachaf> apparently that's true in croatian too??
22:05:42 <shachaf> mauris: have you considered changing your name to maurest
22:06:09 <oerjan> hm there's a single word in that word that isn't an anagram of the rest
22:06:26 <mauris> shachaf: someone's gotta be the positive degree
22:11:31 <oerjan> here's a live reading by the poet hth https://www.youtube.com/watch?v=Tz8hxErxhBI
22:11:50 <b_jonas> oerjan: I just did a quick search (may have bugs) among Arany János's poems in Hungarian for consecutive sequences of entire lines containing only ascii.
22:12:03 <b_jonas> The longest there is is only 10 words.
22:12:37 <oerjan> so, not something you do by chance
22:14:21 <b_jonas> In case you're curious, tied for the longest is the last two lines of the famous ballad Ágnes Asszony.
22:16:09 <b_jonas> And in this book of his poems, en-dashes, quotation marks, and italicized text (all three of which he uses regularly) are all represented as ASCII characters, so this is a very charitable interpretation of ASCII-only.
22:18:56 -!- hppavilion[1] has joined.
22:20:51 <b_jonas> Wait, my search has a bug.
22:27:47 <b_jonas> Longest found now is 18 words.
22:35:59 <oerjan> why does google search only give you the choice between "all languages" and "whichever language is your setup"? there's no fucking way to say i want anything _but_ hungarian.
22:36:41 <oerjan> (trying to find any english description about the poem)
22:37:26 <mauris> try https://www.google.com/preferences
22:38:04 <mauris> err, https://www.google.com/preferences#languages
22:38:07 <shachaf> Or https://www.google.com/advanced_search ?
22:38:49 <oerjan> hm advanced search looks more promising
22:39:24 -!- boily has joined.
22:39:49 <oerjan> the trouble is, google.no doesn't have advanced_search listed anywhere.
22:40:18 <shachaf> i found it via https://encrypted.google.com/search?q=advanced+search hth
22:41:33 <oerjan> bah i still get only album listings and videos
22:43:06 <oerjan> i guess this will have to do http://virginmediashorts.com/film/4478/the-ballad-of-mistress-agnes
22:46:08 * oerjan wonders if google is removing and hiding all the really good search options because they're too computation intensive
22:46:41 * oerjan watches fizzie and pikhq accusingly
22:46:58 <shachaf> hah, out of the accusation zone
22:47:20 <oerjan> oh did you use to work at google
22:47:48 <oerjan> i'm sure this is extensively described in the google non-disclosure agreement
22:56:20 <fizzie> oerjan: There was a third one too, I forget who.
22:57:37 <fizzie> FWIW, if I search something at google.com, then the SRP's "cogwheel" menu does have an "Advanced search" option.
22:57:53 <fizzie> Where you can set any particular language.
22:58:24 <fizzie> Uh, where SRP is the "search results page", sorry.
22:58:24 <shachaf> But you can't exclude languages.
22:58:57 <fizzie> Is it something else? I don't really remember, it's just "the serp".
23:00:01 <shachaf> I've forgotten so many Google code names and things.
23:00:22 <fizzie> They certainly like having code names and terms, I've noticed that much.
23:02:29 <fizzie> Aw, the let-sen-crypt launch was pushed back.
23:02:38 <fizzie> "General availability: Week of November 16, 2015"
23:02:46 <fizzie> I think it was supposed to be either August or September.
23:03:34 <shachaf> I do remember a build system with an unfortunate name.
23:03:40 <shachaf> So unfortunate that they renamed it.
23:04:35 <fizzie> Well, they've got http://bazel.io/ open-sourced now.
23:05:32 <shachaf> Not that one, the thing with the web UI.
23:05:33 -!- rdococ has quit (Ping timeout: 252 seconds).
23:05:55 <fizzie> If they've renamed it, I might not have heard of the previous name.
23:06:19 <shachaf> It has a very fast new name.
23:07:53 <shachaf> whoa, speaking of Google websites, did you come across that one great site?
23:07:58 <shachaf> I probably shouldn't talk about it in here.
23:08:42 <fizzie> Sorry, go/that-one-great-site didn't go anywhere.
23:10:01 <shachaf> Hmm, is go/<confidential-deployment-system>guy still around?
23:12:38 -!- rdococ has joined.
23:16:08 <olsner> oh, did shachaf stop working at google?
23:16:50 <HackEgo> 304) <Gregor> Google Maps has options for "avoid highways" and "avoid tolls", but no "avoid Chicago" \ 672) <Vorpal> <ais523> northern ireland is quite a way to drag someone from scotland <-- not really. I just checked in google earth <ais523> Vorpal: but dragging people across water's a bit tricky \ 769) <oerjan> Gregor: hey no fair doing ungoogl
23:16:57 <shachaf> I haven't worked at Google for over a year.
23:19:27 -!- rdococ has quit (Ping timeout: 256 seconds).
23:24:01 <fizzie> shachaf: There are 128 different "go/<X>guy"s, as far as I can tell. Plus the web-based list searches only by prefix, not by regex, which is silly.
23:29:25 -!- rdococ has joined.
23:30:03 -!- FireFly has quit (Changing host).
23:30:03 -!- FireFly has joined.
23:35:11 -!- rdococ has quit (Ping timeout: 260 seconds).