00:01:00 -!- roasted42 has joined.
00:01:29 -!- roasted42 has quit (Changing host).
00:01:29 -!- roasted42 has joined.
00:01:29 -!- roasted42 has changed nick to TheM4ch1n3.
00:02:56 <fizzie> int-e: I had an INVISIBLE TIMES there.
00:03:16 <int-e> fizzie: oh. I didn't see that :P
00:03:27 <HackEgo> [U+0040 COMMERCIAL AT] [U+2062 INVISIBLE TIMES] [U+0066 LATIN SMALL LETTER F] [U+006F LATIN SMALL LETTER O] [U+006F LATIN SMALL LETTER O]
00:04:23 <HackEgo> [U+0069 LATIN SMALL LETTER I] [U+006E LATIN SMALL LETTER N] [U+0020 SPACE]
00:05:20 <oerjan> stupid irssi sticks the zero-width characters too closely to what's on their left
00:05:39 <oerjan> i vaguely thought DEL bypassed that, but apparently not
00:06:47 <zzo38> I only use the C locale because all of the others are no good
00:09:13 <int-e> oerjan: tricky. I had to select several lines in emacs to get this to paste. I saw something funny - emacs' cursor becomes narrow (1 pixel wide) when it's on a zero-width thing.
00:10:24 <oerjan> ok it was easy enough with gvim
00:10:25 <int-e> And I guess the point is to distinguish axe from axe...
00:10:42 <oerjan> if only because it makes no attempt to make that char actually zero-width...
00:10:44 <int-e> (The former meaning a*x*e)
00:11:39 <int-e> actually emacs is slightly bad - it makes the character one pixel wide.
00:13:53 <zzo38> I saw the newest Magic: the Gathering new keywords, megamorph and exploit, and then, I try to think how to make up stuff that is using that.
00:17:20 <zzo38> My opinion is whether or not it makes it one pixel wide ought to depend on the font in use; that data should be included in the font metrics.
00:17:57 <int-e> it's a fixed width font, I expect all characters to have a multiple of the base width as their width
00:18:57 <zzo38> In such case they ought to do that then and not misalign due to being one pixel wide
00:19:06 <int-e> (it's bad enough that double-width and zero-width characters exist, and don't bring up the half-wit space)
00:19:55 <FireFly> I would imagine half-width spaces are fullwidth in most monospace fonts?
00:20:07 <izabera> boredom leads to this https://gist.github.com/izabera/818223f3ff7081052ed6
00:20:09 <FireFly> I think double-width isn't that bad, it's easy enough to deal with
00:20:38 <FireFly> izabera: you are a bad person :(
00:21:17 <izabera> that's not even the worst code i ever wrote
00:22:51 <izabera> how about this https://gist.github.com/izabera/0205fe4507db17bd93f9
00:23:50 -!- TheM4ch1n3 has quit (Ping timeout: 248 seconds).
00:27:20 <FireFly> too bad it isn't a polyglot :p
00:28:16 <izabera> well i guess one could just add #declare then { #declare fi }
00:30:34 -!- roasted42 has joined.
00:30:40 <izabera> ah dammit, the first line will never be valid c
00:30:59 -!- roasted42 has quit (Changing host).
00:30:59 -!- roasted42 has joined.
00:30:59 -!- roasted42 has changed nick to TheM4ch1n3.
00:31:17 -!- J_Arcane has quit (Read error: No route to host).
00:33:13 <FireFly> You could remove the shebang and invoke it with "bash" or "clang" directly
00:36:32 <pikhq> Also, without the shebang it's still executable for most purposes.
00:37:17 <pikhq> (with a call to execlp or execvp, which is what's *mostly* used, then if the binary can't be understood then the libc passes it to the shell)
00:39:03 <izabera> it's the kernel that returns ENOEXEC
00:39:11 <izabera> and the shell tries to eat it
00:39:26 <izabera> it would also try to execute perl or python :\
00:46:45 <pikhq> izabera: No, what happens is: the kernel returns ENOEXEC if it's not a valid executable, and the libc implementation of execvp or execlp specifically calls /bin/sh $file.
00:47:08 <pikhq> Perl or Python can't be run via this execution path, except if the shell script happens to say to.
00:47:33 <izabera> are you sure? i thought it was in a subshell
00:48:39 <pikhq> I am confident, by which I mean "that's what the spec says, and also what's in the sixth edition Unix source".
00:48:44 -!- heroux has quit (Ping timeout: 264 seconds).
00:49:31 <pikhq> Once a while back (it's somewhere in esoteric logs) I actually looked up how far back this behavior goes.
00:49:41 <pikhq> I stopped caring when I hit 6th edition.
00:50:02 <pikhq> "Meh, it's basically forever"
00:50:09 -!- heroux has joined.
00:50:43 <FireFly> "Do NOT hardcode terminal escape sequences. Use tput with the cap-names from the table below to get the right code for your terminal." b-but.. all terminals today ought to support ECMA-35 anyway, so it shouldn't be a problem
00:51:04 <int-e> Oh, it's the exec*p functions that do this.
00:51:24 <pikhq> FireFly: Yes, IMO you should just use the terminal codes these days.
00:51:38 <pikhq> I've been half-thinking of doing a simple-curses that just does ECMA-35.
00:52:29 <pikhq> ECMA-35 is the charset switching junk.
00:52:41 <pikhq> ECMA-48 is the terminal codes.
00:53:04 <pikhq> The answer to *ECMA-35* is "Just use UTF-8." :)
00:54:19 <int-e> I didn't know that... "The name "ANSI escape sequence" dates from 1981 when ANSI adopted ECMA-48 as their standard, ANSI X3.64".
00:56:59 <zzo38> Use VT100 character graphics for the G1 set and ASCII for the G0 set.
00:57:33 <FireFly> I wish the terminal would somehow get metadata about from which process/task some particular output originated, instead of getting the output from the shell/all programs running in it interleaved in one stream
00:58:05 <pikhq> Kinda tricky though.
00:58:22 <pikhq> The design of the Unix terminal system is entirely around old serial terminals.
00:59:11 <FireFly> You could probably still multiplex it in one stream in the end, but annotate with escape sequences somehow saying "the following is from the process with PID foo"
00:59:28 <zzo38> Use pipes if you want to add such annotations in, is one way
01:02:18 -!- vodkode has joined.
01:08:48 <oerjan> well it's simultaneous in most of europe
01:09:12 <lambdabot> Local time for izabera is Sun, 29 Mar 2015 03:11:16 +0200
01:10:02 <izabera> it just sent a ctcp request >.>
01:10:09 -!- vodkode has quit (Quit: Leaving).
01:10:23 <oerjan> THAT'S JUST WHAT IT WANTS YOU TO THINK
01:10:41 <izabera> well then it worked cause i'm totally thinking it :P
01:12:24 <zzo38> Is there such a thing as a television show called "This is not a pipe"?
01:12:44 <izabera> http://uploads7.wikiart.org/images/rene-magritte/the-treachery-of-images-this-is-not-a-pipe-1948(2).jpg
01:13:03 <oerjan> i assumed zzo38 already knew about the painting
01:13:17 <zzo38> Yes I know about Magritte's paintings
01:13:32 <oerjan> we have an esolang based on it, after all.
01:14:54 <zzo38> Yes, I saw that too
01:14:58 <izabera> is there a bot command to search the wiki?
01:15:12 <zzo38> Not as far as I know.
01:15:20 <oerjan> zzo38: apparently there's at least a tv show _episode_ of "switched at birth"
01:15:36 <oerjan> @google magritte site:esolangs.org
01:15:59 <zzo38> To fix it you would make a SQLite interface for searching the wiki and then make that extension loadable by SQLite on HackEgo, and now you can search the wiki using SQL commands.
01:16:11 <izabera> well that's an impressive language :)
01:16:33 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
01:21:44 -!- lleu has quit (Quit: That's what she said).
01:26:27 <FireFly> Here I was hoping to go to bed at a reasonable time, and then I suddenly lose an hour
01:30:08 <oerjan> alas, we were merely fooling ourselves
01:31:43 -!- Phantom_Hoover has joined.
01:37:26 <fizzie> We were just surprised by the DST switch happening at 01am instead of the 03am it does in Finland.
01:41:22 <int-e> so same time, really
01:43:42 <fizzie> Yes, from the "same instant" point of view.
01:44:28 <fizzie> Hmm, this client isn't doing DST.
01:44:49 <fizzie> Oh, it's in "Etc/UTC" timezone.
01:44:57 <zzo38> Astrolog requires setting DST manually
01:45:21 <oerjan> fizzie: a danger of living in britain, surely
01:46:12 -!- int-e has left ("TIME TRAVELING CHICKEN").
01:46:12 -!- int-e has joined.
01:46:59 <fizzie> Yes. I set this VPS in wintertime, and just assumed it was following local time. Though I guess for a VPS "Etc/UTC" is a sensible default.
01:56:48 -!- boily has joined.
02:01:27 -!- boily has quit (Client Quit).
02:04:14 -!- TheM4ch1n3 has quit (Quit: bye).
02:19:01 -!- oerjan has quit (Quit: COPYCAT CHICKEN).
02:34:42 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
02:45:30 <zzo38> How many things are wrong with this picture? http://zzo38computer.org/img_12/vid2.jpg
02:52:19 <zzo38> There are two contradictory specifications of the duration, for one thing. The title, pictures, and description are from three different movies.
03:03:48 -!- nisstyre has quit (Quit: WeeChat 1.1.1).
03:06:37 -!- Zefphex has joined.
03:06:50 <Zefphex> I aam att emerald city comic con
03:19:19 -!- nys has quit (Quit: sleep).
03:23:51 -!- heroux has quit (Ping timeout: 265 seconds).
03:24:51 <tswett> Hm. If I'm not mistaken, there's only one totally ordered set S, up to order isomorphism, with the following two properties.
03:25:54 <tswett> Every subset of S has a least upper bound; and there exists a countable subset T of S such that given two distinct points in S, there is an element of T between them.
03:26:03 -!- heroux has joined.
03:29:31 <tswett> I say this because Wikipedia says this:
03:29:33 <tswett> "It is a theorem that any linear continuum with a countable dense subset and no maximum or minimum element is order-isomorphic to the real line."
03:30:24 -!- adu has joined.
03:41:58 -!- Sgeo_ has quit (Read error: Connection reset by peer).
03:42:08 <elliott> izabera: these bash things are disgusting. I love it
03:44:40 -!- Sgeo has joined.
03:48:29 <zzo38> tswett: Is there an element of T between them?
03:55:08 -!- dianne has joined.
03:58:25 -!- Zefphex has quit (Read error: Connection reset by peer).
03:58:46 -!- Zefphex has joined.
04:03:45 -!- Zefphex has quit (Ping timeout: 264 seconds).
04:16:31 -!- nisstyre has joined.
04:16:49 -!- GeekDude has quit (Quit: {{{}}{{{}}{{}}}{{}}} (www.adiirc.com)).
04:47:22 -!- nisstyre has quit (Changing host).
04:47:22 -!- nisstyre has joined.
05:36:41 -!- adu has quit (Quit: adu).
05:44:36 <zzo38> I have the picture of a grid of Java/C/PHP/Ruby/Haskell as seen by the fans of those programming languages; now make up the new one, to also include more programming language, such as: C++, SQL, INTERCAL, Perl, Python, BASIC, Forth, Lisp, Scheme, assembly language, TeX, LaTeX, and possibly a few others.
05:50:56 -!- zadock has quit (Quit: Leaving).
06:00:32 -!- zadock has joined.
06:03:09 -!- ^v has quit (Read error: Connection reset by peer).
06:03:35 -!- ^v has joined.
06:16:21 <zzo38> http://zzo38computer.org/img_10/quickbasicerror.png
06:24:30 -!- vodkode has joined.
06:49:48 -!- heroux has quit (Ping timeout: 256 seconds).
07:07:10 -!- heroux has joined.
07:08:18 -!- izabera has changed nick to darthvader.
07:08:45 -!- darthvader has changed nick to izabera.
07:51:44 -!- vodkode has quit (Ping timeout: 252 seconds).
08:02:07 -!- Lymia has quit (Quit: Hug~♪).
08:02:21 -!- Lymia has joined.
08:22:45 -!- J_Arcane has joined.
08:45:28 -!- heroux has quit (Ping timeout: 244 seconds).
08:47:32 -!- heroux has joined.
09:15:51 <zzo38> Have you ever heard of a TV show called "Slug Terra"?
09:25:03 <zzo38> I have never heard of such thing until a week ago, and that was at the night time, which enabled me to notice something that I probably wouldn't have otherwise noticed.
09:26:38 -!- shikhin has joined.
09:31:03 <zzo38> Do you know what kind of stuff I would notice due to such thing?
09:41:24 <izabera> i promise i will stop spamming this but i made it so much better https://gist.github.com/izabera/818223f3ff7081052ed6
09:44:22 -!- shikhin has quit (Ping timeout: 245 seconds).
10:06:43 -!- Patashu has quit (Remote host closed the connection).
10:06:55 -!- Patashu has joined.
10:36:24 -!- Phantom_Hoover has joined.
10:40:32 -!- shikhin has joined.
10:40:50 -!- shikhin has quit (Client Quit).
11:09:36 -!- oerjan has joined.
11:11:18 <oerjan> <Zefphex> and iiiam lost <-- just look for the foglios hth
11:15:16 <int-e> I'm wondering how helpful this advice can be, 8 hours after the plea for help :)
11:15:34 <oerjan> and while he's not on the channel
11:15:50 <b_jonas> fungot, how can you fit 500 sheep in a truck?
11:15:51 <fungot> b_jonas: have you ever been to europe?
11:16:34 <fungot> b_jonas: perhaps in one of my favourites!!! eheheheheeh and which were exact. expect you have to
11:16:36 <oerjan> but hey, maybe _someone_ here is there. although why they'd then be reading the channel messages i don't know.
11:17:11 <oerjan> fungot: hey not _all_ europeans know how to handle sheep badly.
11:17:11 <fungot> oerjan: but what about ( for a hcoop account, in case you have destroyed the fnord of a
11:17:15 -!- FreeFull has quit (Quit: Gotta go for now).
11:18:00 <oerjan> destroying fnords, a life mission
11:18:20 <int-e> b_jonas: maybe... http://www.theage.com.au/victoria/sheep-rain-down-on-cars-as-truck-tips-over-highway-20120531-1zl0w.html
11:18:22 <b_jonas> seriously, the news says that today the trailer of a truck on the highway has tipped to the side, and that half of the 500 sheep the truck had carried has died
11:18:45 <b_jonas> int-e: ah, a different similar accident
11:19:17 <b_jonas> int-e: http://www.hirado.hu/2015/03/29/kamionbaleset-miatt-lezarjak-az-m7-est-zalakomarnal/ is today's news
11:19:29 <int-e> It'd be easier with dead sheep, I presume. Though perhaps you want to cool the bodies then, and then it becomes harder again.
11:19:54 <b_jonas> int-e: the news specifically says the sheep died in the accident, so they were probably alive before
11:20:05 <b_jonas> but maybe the trailer tipped because there were too many sheep
11:20:33 <b_jonas> it's already scary how a double-decker bus can carry 80 humans, and that's without a trailer
11:20:39 <oerjan> google translate, do not let me down now
11:21:46 <oerjan> i saw in the newspaper about modern cruiseships, putting 5000 people on a boat no matter how safe sounds like tempting fate to me
11:22:38 -!- H2O1 has joined.
11:22:45 <b_jonas> oerjan: yeah, it's scarier with humans these days because with humans you assume the intention is to deliver all of them alive
11:22:54 <b_jonas> whereas with sheep or slaves you can't be sure about that
11:23:57 * oerjan whacks HackEgo upside the head
11:24:29 <HackEgo> H2O1: Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: <http://esolangs.org/>. (For the other kind of esoterica, try #esoteric on irc.dal.net.)
11:24:51 <b_jonas> oerjan: assume because if a human dies in such an accident, their insurance company will sue the transport company
11:25:07 -!- H2O1 has left.
11:26:02 -!- SopaXorzTaker has joined.
11:26:43 <int-e> oerjan: so what do you think about putting more than 500 people into a flying aluminium can with wings?
11:27:29 <int-e> b_jonas: I've found a number of similar incidents in to 100-200 sheep range. I guess 500 is rather high.
11:31:25 <int-e> haha. "Researchers have managed to get two computers to communicate using heat and thermal sensors."
11:32:41 <int-e> (from https://www.schneier.com/blog/archives/2015/03/yet_another_com.html )
11:37:57 <oerjan> next up, alien computers communicating through biological evolution
11:38:59 <oerjan> or maybe sirens of titan already did that
11:39:55 * oerjan sometimes wonders if it's a bad idea to read all these plot synopsises on wikipedia
11:40:55 -!- Patashu has quit (Ping timeout: 256 seconds).
11:41:53 <oerjan> i suppose the next step is to use big bangs
11:43:25 <oerjan> wait, this is clearly the _real_ purpose of the bible code hth
11:43:36 -!- MoALTz has quit (Quit: Leaving).
11:43:56 <int-e> oerjan: so that's why they built the LHC?
11:44:28 <int-e> to communicate with extra-universal(uh, is that right?) aliens?
11:45:34 <oerjan> i think extra-universal is about as technically incorrect as extra-dimensional
11:46:32 <oerjan> i'm not sure there _is_ an agreed scientific term for such a thing.
11:47:10 <FireFly> What is the relationship to inter-universal Teichmüller theory?
11:47:19 -!- erdic has quit (Remote host closed the connection).
11:47:38 <oerjan> i do not know, i do not understand inter-universal Teichmüller theory.
11:47:47 <Koen_> wait, so the universal turing machine was not about communicating with aliens?
11:48:09 <oerjan> Koen_: only a small step towards it hth
11:49:00 <oerjan> although there has been some scifi where you can to weird stuff with math
11:49:34 <oerjan> i never finished reading that one that was linked here
11:50:13 <oerjan> (mind you, that goes for most books these days)
11:50:52 <Koen_> vernor vinge has some interesting stuff about communication and intelligence
11:53:45 <b_jonas> nah, there are much better ways to communicate with parallel universes
11:53:52 <oerjan> god i hate muffled loud bass seeping through the building
11:55:53 <oerjan> b_jonas: you need a communication channel, which would tend to be highly dependent on exactly how you're connecting to the parallel universe...
11:56:02 <b_jonas> for example, you can swap people between the universes through transporter accidents,
11:56:39 <b_jonas> or you can leave a bowl of tungsten on your desk and hope that psychic aliens from an alternate universe will repalce it with a message written in some radioactive isotope.
11:56:57 <b_jonas> either of those seem more efficient than big bangs.
11:57:31 <b_jonas> well, unless you're the kind of guy referenced in the title text of the butterflies xkcd strip
12:02:38 <FireFly> oerjan: if you have fish seeping through the building I think you have bigger problems
12:03:41 <oerjan> FireFly: hey i already made that joke yesterday
12:04:41 <int-e> FireFly: http://www.walltanks.com/ hth
12:04:45 <oerjan> 's ok, i was hoping someone would do that
12:05:22 <int-e> now to figure out the context...
12:07:06 <int-e> <oerjan> <orin> god damn it these earbuds have too much bass <-- bass are inferior to babel fishes, anyhow <-- that's a multitude of several fish?
12:07:26 -!- erdic has joined.
12:17:03 <izabera> spot the idiot https://bugzilla.redhat.com/show_bug.cgi?id=1202858
12:17:21 <int-e> @google plural of fish
12:17:22 <lambdabot> http://grammarist.com/usage/fish-fishes/
12:17:22 <lambdabot> Title: Fish vs. fishes - Grammarist
12:17:59 <int-e> izabera: . o O ( Why does my monitor turn into a mirror when I visit that page? )
12:20:41 <int-e> funny nick, superusersdo
12:30:38 <b_jonas> would you like mustard with your sausage, fungot?
12:30:38 <fungot> b_jonas: ok, depending on what you mean by persistance...) ( span...)) is
12:30:47 <b_jonas> oh no, he's having a lisp again
12:34:09 <fungot> oerjan: functions, macros, that can be done with the engine... the scsh wiki doesn't even have a compare-and-swap? :)
12:44:55 -!- MoALTz has joined.
12:54:33 -!- SopaXorzTaker has quit (Remote host closed the connection).
13:02:47 <HackEgo> [wiki] [[Special:Log/newusers]] create * Aldaron * New user account
13:06:18 -!- boily has joined.
13:13:53 <oerjan> tswett: sounds like that has to be isomorphic to a closed interval
13:14:29 <oerjan> @tell tswett sounds like that has to be isomorphic to a closed interval
13:15:44 <lambdabot> Foldable t => (a -> b -> b) -> b -> t a -> b
13:15:56 <lambdabot> Foldable t => (a -> b -> b) -> b -> t a -> b
13:19:35 <oerjan> and a simple way to prove it is to find an isomorphism between your T and the rationals in some interval
13:20:28 <oerjan> a stern-brocot tree, say
13:21:20 <oerjan> (the "some" is because you didn't specify whether T might contain the maximum or minimum)
13:21:22 -!- vodkode_ has quit (Ping timeout: 256 seconds).
13:22:30 -!- vodkode_ has joined.
13:22:54 <tswett> Actually, a one-element set has both those properties as well.
13:23:27 <tswett> So, say also that the minimum and the maximum must be distinct.
13:23:53 <oerjan> "at least two elements"
13:24:11 -!- SopaXorzTaker has joined.
13:25:27 <tswett> Now, the "given two points in S, there's an element of T between them" thing lets you come up with a monomorphism from the rationals in [0, 1] to T, assuming that T has at least two points.
13:28:47 <oerjan> because T is countable, it's not hard to make it onto as well
13:30:35 <oerjan> in fact, if T is a sequence you get a constructive map
13:35:53 -!- roasted42 has joined.
13:37:20 -!- roasted42 has quit (Changing host).
13:37:20 -!- roasted42 has joined.
13:37:20 -!- roasted42 has changed nick to TheM4ch1n3.
14:06:43 -!- heroux has quit (Ping timeout: 264 seconds).
14:08:35 -!- heroux has joined.
14:11:40 <boily> `relcome TheM4ch1n3
14:11:51 <lambdabot> oerjan said 14h 43m 58s ago: <boily> fungot: I wouldn't mind cats or dogs or diamondies pouring down. as long as it isn't any more snow. <-- i suspect diamondies would be even harder on the car drivers. btw oslo was also unexpectedly hit by snow yesterday, complete chaos at the traditional start of easter vacation week.
14:11:52 <HackEgo> TheM4ch1n3: Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: <http://esolangs.org/>. (For the other kind of esoterica, try #esoteric on irc.dal.net.)
14:12:07 <oerjan> i thnk TheM4ch1n3 has been here before
14:12:25 <boily> it may be the case, but `relcomes are idempotent.
14:12:48 <lambdabot> CYUL 291400Z 22012KT 30SM SKC M05/M13 A3021 RMK SLP233
14:13:10 * boily is still naïvely waiting for normal temperatures...
14:13:19 <lambdabot> ESSB 291350Z 14015KT 7000 -RA BKN007 03/02 Q0989 R12/29//54
14:13:21 <boily> btw, hellørjan hth.
14:13:23 <lambdabot> ENVA 291350Z VRB02KT 9999 VCSH FEW026 BKN045 08/M01 Q0982 BECMG 25012KT RMK WIND 670FT 07002KT
14:13:33 <FireFly> I realised I have an airport closer to me than ESSA
14:14:38 <lambdabot> CYOW 291400Z 23007KT 15SM FEW240 M03/M13 A3020 RMK CI1 SLP234
14:14:42 <lambdabot> CYYZ 291400Z CCA 23008KT 15SM SCT230 M01/M10 A3025 RMK CI3 CONTRAIL SLP253
14:14:50 <lambdabot> CYVR 291400Z 06011KT 12SM -RA FEW019 OVC042 09/07 A3019 RMK SF1SC7 SLP223
14:15:13 <boily> so you have to be on the other end of the country to enjoy something palatable.
14:20:23 <lambdabot> LOWI 291350Z VRB02KT 9999 FEW060 BKN100 15/05 Q1006 NOSIG
14:20:32 <int-e> nice spring weather
14:23:28 <int-e> . o O ( Oh yeah, the tell-tale sign of growing up: Doing household chores as a means of procrastination. )
14:26:34 -!- heroux has quit (Ping timeout: 264 seconds).
14:26:47 -!- boily has quit (Quit: HYPERREAL CHICKEN).
14:27:04 -!- heroux has joined.
14:27:09 <olsner> when you procrastinate the household chores, what's that a sign of?
14:30:19 -!- TheM4ch1n3 has left.
14:36:18 -!- TieSoul has quit (Quit: http://quassel-irc.org - Chat comfortably. Anywhere.).
14:39:28 <int-e> olsner: normal teenager behavior
14:43:47 <orin> Just only keep two of each type of utensil and two plates
14:44:24 <orin> then washing the dishes involves less work
14:44:39 <orin> especially if everything is steel or aluminum
14:47:54 -!- TieSoul has joined.
14:53:46 -!- FreeFull has joined.
15:12:20 -!- copumpkin has joined.
15:27:46 -!- SopaXorzTaker has quit (Read error: Connection reset by peer).
15:31:00 -!- shikhin has joined.
15:37:09 <tswett> I came up with an idea for how string literals could work in my amazing new programming language I'm creating.
15:37:33 <tswett> But this idea has dire consequences.
15:38:11 <tswett> Here's one way of writing the string literal for the word "melee": "melee"
15:38:17 <tswett> Here's another way: ""_melee_""
15:38:27 <tswett> And here's another way: ""emeeleeeee""
15:39:26 <tswett> Also, here's one way of writing the string literal for an apostrophe: "'"
15:39:30 <tswett> Here's another way: ''''''''
15:40:09 <oerjan> what string literal is AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA twh
15:40:31 <tswett> Well, the easiest way to write that would be "AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA".
15:40:39 <tswett> But another way would be ""AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA"".
15:40:43 <oerjan> no no, the other way around hth
15:40:54 <tswett> Oh. That's not a string literal; that's an identifier.
15:41:56 <tswett> Another way of writing "I came, I saw, I conquered" would be this: "",I came,, I saw,, I conquered,""
15:42:05 <tswett> That should give you the general idea of how these weird and silly string literals work.
15:42:31 <tswett> Let's see if lambdabot magically understands these.
15:42:33 <tswett> > ""AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA""
15:42:34 <lambdabot> data constructor ‘AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA...
15:44:49 <tswett> The nice thing about these string literals is that they can contain apostrophes and double quotes with no problems at all.
15:45:23 <tswett> ""_Martha said, "'Get over here,' he said."_""
15:45:31 <tswett> See, you're getting the hang of it.
15:45:38 <tswett> Come back tomorrow for lesson 2!
15:57:32 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: rerere: not found
16:03:07 <FireFly> tswett: have you considered lua-style [[strings]] or [==[strings]==] or ...
16:03:33 <tswett> FireFly: those were part of the inspiration for this idea.
16:08:39 <fungot> a1a12ab2b23bc3c34cd4d4d
16:10:13 <orin> I prefer hollerith
16:10:37 <b_jonas> how about C++ raw string literals?
16:11:35 <b_jonas> they're useful because they let you use multi-character delimiters, so you don't have to choose a single character that can't appear in the string
16:11:41 -!- oerjan has quit (Quit: ""alaatera"").
16:11:49 <tswett> b_jonas: I'm not familiar with those.
16:12:31 <b_jonas> tswett: the syntax is like this, R"(foo)" is equivalent to "foo", and so is R"bar(foo)bar"
16:12:55 <tswett> Tyhöistä ja alaatera olen.
16:13:35 <b_jonas> in raw string literals, nothing other than the delimiter has a special meaning. backslashes, parenthesis, double quotes, single quotes, question marks are all literal. newlines are allowed and aren't translated, the newlines gets in the value of the string as it's present in the source, which could be cr-lf or lf or whatever.
16:14:10 <b_jonas> backslash escapes or trigraphs can't be used in such string literals. you can concatenate multiple raw string literals by juxtaposing them, just like with ordinary string literals.
16:16:29 <tswett> I've also thought about doing that sort of thing.
16:17:40 <tswett> "" (Note, by the way, that the custom delimiter can't be whitespace. This message is the empty string, followed by a bunch of identifiers, followed by the empty string again.) ""
16:17:59 <tswett> Yes, "can't" is a valid identifier even though ' is a string delimiter. "(Note," is also a valid identifier.
16:18:25 <tswett> fungot: what's your opinion on all this?
16:18:25 <fungot> tswett: to return ( 1... 10) and then under my name point out the damage that regular bikes can do to prevent crimes from happening, and when
16:18:32 <orin> hollerith is easier to generate reliably
16:18:49 <orin> but harder for humans
16:19:32 <tswett> 'bash-style literals aren'\''t hard for programs to generate.'
16:20:00 <b_jonas> tswett: but then you can't denote the empty string by ""
16:20:18 <tswett> b_jonas: what do you mean?
16:21:08 <b_jonas> if you use two double quotes (rather than three like python does) as a delimiter, then you can't use two double quotes for the traditional meaning of a string literal with an empty string value
16:21:19 <orin> Hmm, how about a compromise in the style of Joel Spolsky's 14"fucked strings"
16:22:05 <orin> in other words, have a length, delimiter, the string and then a required end delimiter
16:22:58 <tswett> I'm never using two quote marks, alone, as a delimiter.
16:24:23 <tswett> """This is a valid string literal consisting of the string S, a quotation mark, S again, another quotation mark, and a period, where the string S is "This is a valid string literal consisting of the string S, a quotation mark, S again, another quotation mark, and a period, where the string S is "."""
16:25:36 <orin> the empty string in hollerith is 0"
16:25:50 <orin> in quoted holerith it is 0""
16:41:28 -!- AnotherTest has joined.
16:42:17 <tswett> Which apparently I could type if I only had a <dead_greek> key.
16:52:00 <b_jonas> tswett: no, it's ε actually. "ɛ" is the pronunciation symbol that looks the same.
16:52:53 <b_jonas> (or at least may look the same, depending on the font you use for greek letters)
16:54:16 <b_jonas> come on people, don't keep mixing up similar looking characters
17:05:35 -!- ProofTechnique has quit (Ping timeout: 252 seconds).
17:15:13 -!- orin has quit (Quit: leaving).
17:16:12 -!- oren has joined.
17:18:24 <oren> hollerith strings have an additional problem. Consider 2"私 which is valid if the encoding is shift-jis, but the number needs to be 3 in utf-8
17:20:23 <zzo38> That isn't too much of a problem if one program uses one character encoding for one string literal.
17:20:43 <zzo38> And that changing the encoding isn't allowed.
17:21:11 <zzo38> (This is independent of whether or not the programming language cares about encoding or not.)
17:22:37 <oren> right. as long as code is never converted from one encoding to another there is no problem
17:23:02 <FireFly> tswett: you could get a <greek> modifier by abusing ISO_Group_Latch
17:23:58 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: α: not found
17:24:10 <tswett> Uh, what channel was that a command in.
17:24:16 <zzo38> Yes, and my opinion is that the code shouldn't be converted from one encoding to another.
17:25:05 <tswett> b_jonas: whoops. I meant to say ε, not ε.
17:25:21 <b_jonas> tswett: those are the same
17:25:36 <oren> they look the same to me
17:25:49 <tswett> I meant to say "I meant to say ε, not ɛ.", not "I meant to say ε, not ε."
17:26:30 <tswett> oren: shouldn't the number refer to the number of characters following rather than the number of bytes following?
17:26:40 <tswett> Everyone knows code is made of characters, not bytes.
17:26:57 <oren> that makes it rather hard to allow strings in several encodings?
17:27:30 <FireFly> I think lenght-prefixed isn't very user-friendly in a source language
17:27:48 <tswett> As in, in the middle of some source code that's in one encoding, embed a string literal that's in a different encoding in order to get those bytes?
17:28:11 <tswett> Everyone knows strings are made of characters, not bytes.
17:28:34 <b_jonas> tswett: now nest that two more levels exponentially, and put ∈ in somewhere!
17:28:42 <oren> yeah, consider a program that has to operate in China (with utf-16) and japan (with shift-jis)
17:29:20 <oren> there is no such thing as an encoding-independent "character"
17:30:19 <FireFly> Just mandate one of the Unicode encoding as the True Encoding for source files
17:30:35 <oren> Unicode has han unification
17:32:04 <oren> Firefly: and mandate conversion at run time from unicode to local encoding?
17:33:11 <oren> encodign conversions are not fast
17:34:29 -!- GeekDude has joined.
17:34:57 <oren> much better to keep string for each locale in its local encoding
17:36:50 <oren> note that many multibyte encodings such as UTF-16 can't be mixed with ascii properly without hollerith, because the byte " can occur as part of a character
17:40:51 <FireFly> I thought encoding conversion was fast enough usually
17:40:57 <FireFly> Depends on the encoding I suppose
17:41:40 <oren> Unicode normalization is I believe the main problem.
17:42:53 <oren> conversion of shift-jis TO unicode is easy
17:43:26 <oren> unicode in general can't be converted back
17:45:08 <oren> (which isn't usually a problem, but if you care about distinctions that unicode doesn't carea bout after normalization...)
17:46:27 <oren> I wonder what happens if I convert a string with french accents in it into shift-jis
17:48:08 <pikhq> Why not simply refuse to even deal in not-UTF-8?
17:48:23 <pikhq> And leave legacy charsets for legacy applications.
17:49:04 <oren> because there are uncoding used today which aren't utf-8
17:49:27 <oren> windows uses UTF-16 for non-english, for exaple
17:49:30 <zzo38> Normally you should not use this Hollerith notation anyways, so isn't too much problem
17:49:47 <pikhq> And that's a massive bug on Windows' part.
17:50:03 <oren> according to you
17:50:07 <pikhq> (keeping in mind, of course, that UTF-16 <-> UTF-8 is lossless for valid strings)
17:50:41 <oren> and much of the japanese internet uses Shift-JIS
17:50:44 <zzo38> I think you should mandate ASCII as the True Encoding for source files
17:50:58 <pikhq> Shift-JIS is also dying in Japan. Quite rapidly.
17:51:00 <zzo38> And allow UTF-8 and PC character set and stuff too, but disallow UTF-16
17:51:19 <FireFly> zzo38: that's easy if you're in an english-speaking country
17:51:29 <FireFly> still somewhat easy in a european country
17:51:41 <FireFly> (with a latin-based orthography)
17:51:52 <FireFly> Then it quickly gets more impractical..
17:53:19 <FireFly> oren: what kind of distinctions, by the way?
17:53:42 <zzo38> If the programming language isn't using non-English words in general, then you should use ASCII as the input character set. String literals and comments can use other encodings, or even just written in a ASCII representation and converted to other encodings using macros (you can even convert UTF-8 to UTF-16 using macros then). Also some things just aren't in Unicode anyways, and Unicode normalization and stuff confuses everything
17:54:34 -!- shikhin_ has joined.
17:54:38 <zzo38> Nevertheless if a comment is written in UTF-8 or Shift-JIS or ISO-8859-1 or whatever, the compiler shouldn't care. Same if it is a string literal it shouldn't care.
17:55:38 -!- AnotherTest has quit (Remote host closed the connection).
17:55:43 <oren> But some string literal syntaxes don't work with certain encodings
17:56:46 -!- AnotherTest has joined.
17:57:28 -!- shikhin has quit (Disconnected by services).
17:57:31 -!- shikhin_ has changed nick to shikhin.
18:01:06 <tswett> Does Shift-JIS make distinctions that Unicode doesn't?
18:01:24 <pikhq> No known charsets do.
18:02:07 <tswett> Then I don't see any obvious problem with saying that your source code consists of a sequence of Unicode characters.
18:02:09 <pikhq> And if they do, "a charset makes this distinction" is reason enough to add a char to Unicode.
18:04:02 <tswett> Yeah, my impression is that Unicode cares a lot about translation into Unicode being reversible.
18:04:29 <oren> conversion of unicode back to shift-jis, is the problem, because in orer for source code to convert properly, it involves lots of normalization
18:05:03 <oren> which isn't necessary if you KNOW your code consists only of sjis compatible characters
18:05:28 <oren> but the compiler can't know that
18:06:27 <oren> pikhq: yes, because unicode contains lots of characters that sjis doesn't distinguish
18:07:21 <oren> forget kanji, consider math-greek vs. greek?
18:07:26 <pikhq> It is incredibly unlikely in the process of compilation that the compiler magically introduced a bunch of characters that aren't in SJIS.
18:07:53 <pikhq> (or that it's freaking relevant -- why does the compiler have to be converting jack shit back to SJIS?)
18:08:31 <oren> otherwise you are converting at run time... which needs to be even more general
18:09:01 <pikhq> You... probably do otherwise.
18:09:27 <pikhq> Unless you are providing a damned weird execution environment where you're compile-time Unicode but run-time SJIS.
18:09:38 <tswett> oren: are you saying that conversion from SJIS to Unicode is not reversible?
18:10:04 <pikhq> tswett: They are saying that SJIS->Unicode->COMPILER MAGIC->SJIS may not work.
18:10:17 -!- ProofTechnique has joined.
18:10:25 <pikhq> Because a compiler is *really* going to magically introduce a bunch of random Unicode codepoints that aren't in SJIS.
18:10:28 <oren> correct, not if you want to use the same system for general unicode-to-sjis as for "unicode from sjis"->sjis
18:11:15 <oren> otherwise it is like you need a data type for "unicode, but guearanteed to have only the chars from sjis in it"
18:11:28 <pikhq> I'm quite amused, BTW, that you're arguing this with SJIS of all things. That is perhaps the single worst legacy charset in existence.
18:13:00 <tswett> oren: I don't suppose you could give an example of some text where the SJIS-to-Unicode-to-SJIS conversion would fail?
18:13:11 <oren> SJIS is USED. WIDELY. I don't know where you're getting your information, but the PC games I bought last year don't work under wine becasue I need SJIS.
18:13:36 <pikhq> *siiigh* Windows boogs.
18:13:58 <pikhq> SJIS is *used*, yes. SJIS is also *objectively terrible*.
18:14:05 <oren> tswett: that's not the problem. the problem is that it would be slow because unicode is complicated
18:14:43 -!- AnotherTest has quit (Remote host closed the connection).
18:15:01 <pikhq> http://git.musl-libc.org/cgit/musl/tree/src/locale/iconv.c I dunno, that doesn't look that bad.
18:16:14 -!- AnotherTest has joined.
18:17:48 <pikhq> (by the by, here is Windows' fundamental execution model: everything, *everything* is Unicode. For purposes of legacy compat it has a number of legacy-charset functions that convert strings to/from Unicode before calling the real ones.)
18:18:14 <oren> iconv: illegal input sequence at position 0
18:18:48 <pikhq> (that Japanese programs happen to only ever use the legacy functions is, well, a bug.)
18:19:03 <oren> pikhq: according to you.
18:19:55 <pikhq> Lemme find the MSDN page.
18:20:38 -!- AnotherTest has quit (Remote host closed the connection).
18:21:33 <pikhq> (though it ends up being really obvious this is the way it *has* to work if you're familiar at all with NT internals)
18:21:40 -!- AnotherTest has joined.
18:21:42 <oren> iconv -fUTF-8 -tSHIFT-JIS | iconv -fSHIFT-JIS -tUTF-8 fails on a lot of stuff
18:22:05 <pikhq> It's more than likely the first bit that fails.
18:22:25 <pikhq> iconv -fSHIFT-JIS -tUTF-8 | iconv -fUTF-8 -tSHIFT-JIS should round-trip.
18:22:28 <oren> it fails on stuff that is supported in both charsets
18:24:38 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: unidecoe: not found
18:25:04 <HackEgo> U+E60A - No such unicode character name in database \ UTF-8: ee 98 8a UTF-16BE: e60a Decimal:  \ () \ Uppercase: U+E60A \ Category: Co (Other, Private Use) \ Bidi: L (Left-to-Right)
18:25:16 <pikhq> WTF char is that supposed to be?
18:26:38 <pikhq> That... is not even remotely valid shift-jis.
18:27:18 <oren> how about hisone 誧
18:27:33 <oren> http://rikai.com/library/kanjitables/kanji_codes.sjis.shtml
18:28:31 <pikhq> (for clarification, the char you output the first time was 0xEE 0x98 0x8a. Which is *not a legal encoding in Shift-JIS*)
18:29:05 <HackEgo> [U+8AA7 CJK UNIFIED IDEOGRAPH-8AA7]
18:30:25 <pikhq> The fuck is wrong with glibc iconv?
18:30:52 <pikhq> The fucking Unicode tables have that mapping.
18:32:47 <pikhq> Oh for the love of
18:33:08 <pikhq> "Shift-JIS" to glibc iconf refers to an old version of Shift-JIS that did not have the character.
18:33:23 <oren> that explains that
18:33:29 <pikhq> iconv -fUTF-8 -tshift_jisx0213
18:36:19 <oren> that's a weird name, since the official sources call it shift-jis-2004
18:36:45 <oren> oh well, it makes sence
18:36:57 <pikhq> It's a standard from 2004, entitled JISx0213.
18:37:29 <pikhq> I don't think glibc much coordinates with your official source.
18:39:40 <oren> http://x0213.org/. JISx0213 is a character set, Shift-jis is a method of encoding it. JISx0213 is like unicode, Shift-jis is like uft-8
18:39:57 <oren> sorry, shift-jis-2004
18:40:47 <pikhq> JISx0213 does include a definition of Shift-JIS IIRC.
18:40:59 <pikhq> Well, suppose I could just check, you did link to it. :)
18:41:44 <oren> I looked at the encoding table. they apparently have added a lot of characters to unicode to ensure compatibility
18:42:32 <pikhq> Yep. Unicode policy is that charsets should roundtrip, and country charsets cooperate to ensure this.
18:42:44 <pikhq> Yep, quite clearly defines Shift-JIS but also as "Shift-JIS-2004".
18:43:42 -!- Melvar has quit (Quit: storm).
18:43:46 -!- idris-bot has quit (Quit: Terminated).
18:55:25 <oren> I dunno if they are cooperating. The jis-213 plane 2 characters 85-84 to 85-87 (that is, four variants of the grass radical) appear to spell out "fuck you han unifaction" to me
18:57:39 -!- AnotherTest has quit (Ping timeout: 265 seconds).
18:58:20 -!- Melvar has joined.
19:01:42 -!- heroux has quit (Ping timeout: 250 seconds).
19:01:42 -!- idris-bot has joined.
19:03:29 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
19:06:44 -!- Phantom_Hoover has joined.
19:09:06 -!- heroux has joined.
19:23:39 -!- mitchs has quit (Quit: mitchs).
19:25:57 -!- Koen_ has quit (Quit: The struct held his beloved integer in his strong, protecting arms, his eyes like sapphire orbs staring into her own. "W-will you... Will you union me?").
19:37:13 -!- shikhin has quit (Ping timeout: 250 seconds).
19:37:27 -!- mitchs has joined.
19:39:16 <zzo38> Do you write any UNIX-style filter programs much, or do you do other ways?
19:43:15 -!- shikhin has joined.
19:44:34 -!- Vorpal has quit (Quit: ZNC - http://znc.sourceforge.net).
19:44:58 -!- Vorpal has joined.
19:44:58 -!- Vorpal has quit (Changing host).
19:44:58 -!- Vorpal has joined.
19:51:22 -!- shikhin has quit (Quit: leaving).
19:55:28 -!- TieSoul has changed nick to TieSleep.
19:59:22 -!- Patashu has joined.
20:05:10 -!- dianne has quit (Remote host closed the connection).
20:07:38 -!- dianne has joined.
20:11:03 -!- heroux has quit (Ping timeout: 248 seconds).
20:17:56 -!- heroux has joined.
20:22:22 -!- Patashu has quit (Ping timeout: 264 seconds).
20:29:05 <oren> I usually write programs as filters if they only have one input and one output.
20:32:29 <zzo38> Yes, but it seem as suitable for many things, to have one input and one output.
20:35:56 <oren> What. What does virtual void shade(Ray3D&) = 0; mean in C++
20:36:30 <oren> i'm initializing the memeber funtion to zero?
20:36:55 <HackEgo> [wiki] [[~ATH]] http://esolangs.org/w/index.php?diff=42240&oldid=42226 * SuperJedi224 * (+0)
20:37:56 <pikhq> oren: It means it's a pure virtual function.
20:38:06 <oren> Oh, it means the virtual function MUST be overridden?
20:38:48 <pikhq> It's a bit weird, but more-or-less what that's doing is setting the vtable entry for the class to 0.
20:39:01 <oren> ok. that is a dumb syntax, but in a language where you bit-shift files by varibles, I can deal.
21:03:47 -!- AnotherTest has joined.
21:09:03 <zzo38> Do vtables in C++ have to store functions or can you put other stuff in?
21:13:37 <zzo38> Such as constants, I suppose
21:20:29 -!- copumpkin has quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…).
21:28:08 <zzo38> You could even in C, make something a bit like QueryInterface and using interface IDs made with macros and stuff. Still this is different than IUnknown; the pointer here is only to a constant vtable, or null if the interface is not supported; actually much simpler than Microsoft's version. But, I would be using some GNU extensions here, such as zero-length arrays.
21:48:58 -!- AnotherTest has quit (Ping timeout: 256 seconds).
21:52:00 -!- oerjan has joined.
22:05:35 <oerjan> <fungot> tswett: to return ( 1... 10) and then under my name point out the damage that regular bikes can do to prevent crimes from happening, and when <-- this sounds reasonable but what happened to _our_ regular bike?
22:05:35 <fungot> oerjan: are those all volunteers? " smerdyakov" " richter") t
22:06:17 <olsner> fungot: do you hae a regular bike?
22:06:18 <fungot> olsner: err never mind. :) :( :( :( :(
22:07:23 <oerjan> fungot: WHAT DID YOU DO TO BIKE
22:07:24 <fungot> oerjan: the only problem was not showing any of the other
22:12:33 <olsner> ooh, bike is a person!
22:12:40 <olsner> or was, depending on fungot
22:12:40 <fungot> olsner: unless after all these months i cant stand 4chan was the first step in converting infix to prefix notation
22:30:05 <oren> I thought the first step in converting infix to prefix is the google "[name of language] precedence table"
22:30:51 <oren> but I suppose if I posted it to 4chan Hitler will answer my quaestion
22:36:24 <FireFly> fungot: what did you do to Bike?
22:36:24 <fungot> FireFly: c) execution time exceeds some limit on identifier length, there's a pdf somewhere with it.
22:38:15 <fizzie> Sounds like a government conspiracy.
23:41:02 <zzo38> This is a C code to work objects supporting multiple interfaces: http://sprunge.us/aUEL
23:56:46 -!- oerjan has quit (Quit: nuit).