00:02:23 <cheater-> so there's been a birthday today in the kitchen area upstairs
00:03:04 <cheater-> and i'm hanging out with the girl doing the drinks and i tell her there's tango in the ballroom downstairs and we go downstairs to check it out
00:03:10 <cheater-> and everyone's dressed like a pirate. wtf?
00:13:27 -!- Phantom_Hoover has quit (Quit: Leaving).
00:21:19 <elliott> 14:56:09 <lament> fortunately integers are not 'real' in any way :)
00:21:19 <elliott> 14:56:24 <lament> PUN NOT INTENDED AND I SHALL SMITE YOU IF YOU MAKE IT
00:21:19 <elliott> 14:58:40 <oklopol> lament: ACTUALLY INTEGERS ARE REALS, THEY'RE A SUBSET OF REALS
00:21:19 <elliott> 15:01:33 --- mode: ChanServ set +o lament
00:21:20 <elliott> 15:01:37 --- kick: oklopol was kicked by lament (lament)
00:21:22 <elliott> 15:01:37 --- join: oklopol (n=villsalo@194.251.102.88) joined #esoteric
00:21:26 <elliott> 15:01:43 --- mode: lament set -o lament
00:28:02 -!- zzo38 has joined.
00:28:35 -!- elliott has quit (Remote host closed the connection).
00:28:37 -!- elliott_ has joined.
00:29:07 <zzo38> Dangerous aardvarkoids?
00:39:26 <zzo38> If it wasn't for C, we'd be using BASI, PASAL and OBOL.
00:54:08 -!- FireFly has quit (Quit: swatted to death).
01:01:14 -!- TLUL has joined.
01:15:13 -!- SgeoN2 has joined.
01:15:14 -!- SgeoN1 has quit (Read error: Connection reset by peer).
01:34:04 -!- pikhq_ has joined.
01:34:15 -!- pikhq has quit (Ping timeout: 250 seconds).
01:36:47 -!- elliott_ has quit (Remote host closed the connection).
01:37:21 -!- elliott has joined.
01:41:21 <Sgeo> Is zzo38 hallucinating?
01:42:15 <zzo38> I do not have any hallucination mushrooms.
01:42:15 -!- SgeoN2 has quit (Read error: Connection reset by peer).
01:42:41 -!- SgeoN1 has joined.
01:44:16 -!- SgeoN1 has quit (Read error: Connection reset by peer).
01:46:45 -!- poiuy_qwert has quit (Quit: Leaving).
01:47:08 <zzo38> Is there some kind of HTTP response code to tell it not to try to download favicon.ico file multiple times?
01:56:55 -!- wth has joined.
01:57:54 -!- wth has left (?).
02:00:04 * oerjan detects a chinese ipv6'er approaching a crash course in english acronyms...
02:01:56 <elliott> i guess it might mean what the hell
02:02:12 <zzo38> No, it means "with" but they forgot the "i"
02:09:00 -!- F019 has joined.
02:14:52 <elliott> oerjan: monad comprehensions in ghc 7.2!
02:16:07 <elliott> oerjan would be better if his name was jeroan
02:16:27 <zzo38> Is "najreo" better?
02:17:20 <elliott> oerjan: your name is now oerjan
02:17:26 <elliott> i hope you like this new name!
02:17:38 <oerjan> it's so new and shiny!
02:18:40 <F019> lee monad in ghc 7.19
02:19:22 <elliott> i think we have another markov bot
02:19:34 <elliott> * [F019] (~molly@212.203.98.114): polly
02:19:36 <elliott> the last one was molly too
02:19:38 <F019> markov chained
02:19:41 <pikhq_> F019: How Markovian of þee
02:19:49 <elliott> F019: YOUR FUHRER IS DEAD!!
02:19:54 <elliott> F019: i'm all for nazism. except when... chips
02:19:58 <elliott> F019: peckity peckity peckity roo
02:20:01 <elliott> F019: camber amper amper sand
02:20:03 <elliott> F019: ffffffffffffffffffffffffffff
02:20:08 <elliott> F019: DEPLOY THE HATCHLINGS
02:20:10 <pikhq_> elliott: BUT MEIN FÜHRER LIVEÞ
02:20:15 <elliott> oerjan: if it's a human we'll get "...wtf /parts" in a few seconds
02:20:27 <F019> heim!! Fourreur
02:20:27 <oerjan> F019: congratulations on finally breaking elliott's brain
02:20:29 <elliott> F019: TARSKI WAS AN ONLY CHILD
02:20:36 <elliott> F019: BUT YOU ARE HIS BROTHER
02:20:44 <F019> Tasrski and Hultchm
02:20:58 <elliott> i made that up, but F019 stole it
02:21:15 <elliott> F019: toblerone is the solution to all the world's problems
02:21:19 <elliott> F019: luke i am your toblerone
02:21:27 <elliott> oerjan: maybe it's not a markov bot
02:21:36 <elliott> at least, milka is also chocolate, and has none of the letters in toblerone
02:21:56 <F019> Triangle Toblerone is
02:22:19 <oerjan> elliott: lee monad was a pun. whether this is a very advanced ai i cannot say
02:22:32 <oerjan> (clearly punny ais are very advanced)
02:22:44 <F019> a lee monad pour une panachee beer
02:23:13 <F019> today..... not so bad gag coffee
02:23:29 <elliott> F019: Bonjour! Bienvenue à cette chaîne!
02:23:46 <F019> merci Monsieur Elliott
02:23:47 <elliott> F019: Hello! Tere tulemast kanal!
02:24:14 <elliott> F019: Malheureusement personne ne sait ici française, je suis juste en utilisant Google Translate.
02:24:25 <elliott> Toutefois, aussi, la bienvenue.
02:24:27 <zzo38> F019: What kind of language can you understand?
02:24:50 <pikhq_> elliott: Eh, it was *technically* correct. Just somewhat awkward-sounding.
02:24:52 <F019> hihi---- google-TR or RT .... that's a FAQ ?
02:24:55 <oerjan> Kefir mjølk, kefir ikkje kaffi
02:25:15 <elliott> F019: Qu'est-ce que la caféine?
02:25:27 <F019> kefir, is a yaourth ?
02:25:35 <oerjan> Je m'appelle tres bien aussi
02:26:00 <F019> a viens du caf... it saturate some coffee
02:26:26 <elliott> Un banquier fait chaussettes de roses alors que les chèvres spectatrice s'assit à laquelle le bonheur serait en décroissance conséquents bucolique.
02:26:57 <elliott> F019: J'ai une chèvre, une autre chèvre, et un plus de chèvre. Combien d'autres chèvres dois-je?
02:27:25 <elliott> De chèvre. De chèvre. De chèvre. Quel est le nombre de chèvres?
02:27:27 <F019> heu.... pour faire combien de chvre ?
02:27:45 <elliott> F019: Avec un engin de chèvre de décision.
02:28:11 <zzo38> F019: That is a bad UTF-8 code.
02:28:22 <F019> yes or not 2 be free
02:28:35 <F019> 3 ? .... three?
02:28:51 <zzo38> (Maybe it is the xchat plugin that automatically types messages for you while you are away?)
02:28:56 <elliott> F019: Ne laissez pas, même si vous êtes un robot de chaîne de Markov, vous êtes un remarquablement ingénieux. Nous soutenons l'égalité des droits de l'homme-robot.
02:29:17 -!- pikhq has joined.
02:29:22 -!- pikhq_ has quit (Ping timeout: 260 seconds).
02:29:28 <F019> yes, viva lay roboten, raaaa dioooooo activitaaaaate
02:30:28 <F019> or not bitall ?
02:30:44 <elliott> Pour bitall, ou de ne pas bitall? C'est la quesiton.
02:31:25 <F019> Pro didjy or not to be djiiii ?
02:31:41 <F019> l'hypothese du continue, l'axiome du choix
02:31:56 <elliott> F019: En vérité, et de ce que c'est faire comme lui non quand de l'intervalle sur le son et?
02:32:21 <oerjan> elliott: c'est le math
02:32:22 <elliott> F019: Quelqu'un at-il vraiment été bien même décidé d'utiliser même aller voulez faire ressembler davantage?
02:32:22 <F019> En vrai, j'en ai pas la mooindre idea
02:32:29 <elliott> oerjan: oh. your specialty!
02:32:42 <F019> Viva el mathos
02:32:43 <elliott> points to the first person to translate my last question without using google
02:33:11 <elliott> Les idées de la lune? La lune lunaire dans la poubelle lunaire huard?
02:33:47 <F019> repeat the question, please after me, learning britishka englishka
02:33:54 <oerjan> elliott: i guess it starts with "has anyone been" but i cannot recall the rest
02:34:15 <zzo38> F019: Is that why you cannot write clearly?
02:34:46 <elliott> F019: Désolé - Je dis simplement que des bêtises. Je ne peux pas parler français. Je suis acceptable en anglais, mais :-)
02:35:27 <F019> or not mee too
02:36:01 <zzo38> How are we going to type, English or French?
02:36:30 <elliott> Nous type de hybrid of Anglais and French.
02:36:38 <F019> we are going to huuuu..... i let Tell Guillaume :)
02:36:47 <elliott> Il sera the greatest language jamais inventé!
02:37:01 <F019> hype Bride of Fronkonstinne
02:37:06 <elliott> oerjan: Best idée, or best idée?!
02:38:06 <zzo38> Is Frenglish like this? http://en.wikipedia.org/wiki/Franglais
02:38:15 <F019> brk.... are we some kind of Hezo Terrific langage growing up to this sheltering sky or why not?
02:38:24 <oerjan> elliott: Har nokon verkeleg vore langt like så bestemt å bruka sjølv dra ville gjera sjå meir som?
02:39:39 <elliott> oerjan: Il me semble at sekoittaminen språk 運命づけられています to fail!
02:39:43 <F019> Gluk wunch Herltitzch fr our Mazeltoff FAQ
02:39:58 <F019> yes, a little morceau
02:40:47 <oerjan> Ich bin ein Berlinerkranz
02:40:58 <zzo38> Perhaps there is nobody at the computer, because I told them to fix the UTF-8 and no proper reply, just the same mess as before
02:41:09 <elliott> oerjan: Offensichtlich ist die deutsche the one true language!
02:41:10 <F019> the one of Berline
02:41:20 <elliott> zzo38: he's making sense in French
02:41:29 <elliott> I think his grasp of English is limited :-)
02:41:31 -!- BeholdMyGlory has quit (Remote host closed the connection).
02:41:44 <F019> hu..... maybe :)
02:42:17 <zzo38> elliott: That may also be the possibility, yes.
02:42:18 <F019> wir sind die robote
02:42:27 <zzo38> (I don't know because I am not French)
02:42:38 <elliott> oerjan: Kaffe laget av perle øyne.
02:42:45 <F019> svenskaia ? or ja-ja mo.... yoyoma?
02:44:21 <F019> i'm making some loosy blog in french
02:45:03 <elliott> oerjan: Totem av spir, et gripende syn, å bli sett kun med fly lys eller salt.
02:45:04 <oerjan> elliott: Høres merkelig ut
02:45:07 <F019> so i'm little busy working girl
02:45:34 <F019> ut the bruk of gluk duke marbadul ? or molly?
02:45:55 <elliott> oerjan: Ved krefter uten sidestykke, en demonstrasjon av vidd, i bytte, elementer av pengeverdien, som brukes som dekorasjon i en hall av forestillinger. Og så videre, men likevel så videre.
02:46:02 <zzo38> F019: Try the French channel?
02:46:19 <elliott> does norwegian not use ; or something?
02:47:23 <elliott> oerjan: "Soot av trær!" den blinde mannen utbrøt; sot av trær JA! Og det, kjære leser, var en spøk.
02:48:00 <oerjan> elliott: are you translating something? i cannot google it...
02:48:38 <oerjan> except soot would be either a misspelling or horribly archaic...
02:49:05 <elliott> Nei, yay, ja, men likevel ikke. etc.
02:49:20 <F019> but canadian's have kick me ban off, haha...
02:50:11 <elliott> oerjan: Canucks & kanoer, en roman av Yynn Brok.
02:50:13 <oerjan> elliott: i mean it looks like free form poetry of some sort
02:51:01 <oerjan> oh wait that soot is obviously a failed translation
02:51:57 <oerjan> elliott: dammit i cannot google it even if i try translating it back
02:53:21 <elliott> oerjan: I antikken var en kule laget, laget for å bli sparket bare av høyere medlemmer av karnevalet, jo høyere medlemmer av pavens karneval. Så tåpelig som folk som folk liker å være, så da det var uunngåelig, og på den fjerde juni, 1761, ble en kule skutt inn i skogen av ro.
02:54:11 <elliott> oerjan: En ensom Filosofen gikk inn i en bygning. "Au!"
02:55:36 <elliott> oerjan: it's just randomness :D
02:55:47 <elliott> tell me you laughed at that last line
02:55:49 <elliott> <elliott> oerjan: En ensom Filosofen gikk inn i en bygning. "Au!"
02:56:19 <elliott> oerjan: think i could get a book of my freeform norwegian poetry published?!?!??!?!?!?!?!
02:56:26 <oerjan> absurd humor. a slight misspelling.
02:56:52 <elliott> more like a misgrammaring then?
02:57:24 <oerjan> a lonely the philosopher
02:59:08 <zzo38> Does J timezone exist? I found a list of timezone with all letters A-Z except J.
03:00:38 <oerjan> was it http://www.timeanddate.com/library/abbreviations/timezones/military/ ? that also lacks J
03:00:44 <pikhq> zzo38: Huh, it doesn't exist.
03:01:00 <pikhq> I is UTC+9, K is UTC+10, and there is no J.
03:01:13 <elliott> hey oerjan told tswett oklopol's name but not me
03:02:02 <oerjan> elliott: if you look carefully at something you pasted above you will find all but one letter of oklopol's name in it
03:02:15 <elliott> 19:01:01 <oerjan> tswett: see msg (he doesn't really want it in the open iirc)
03:02:22 <elliott> oerjan: you are now legally obligated to tell me
03:02:27 <pikhq> F019: Irrelevant to nautical timezones, which are based entirely on latitude.
03:02:35 <zzo38> pikhq: Why is there no J? +14 is M+ and has no letter by itself?
03:02:57 <oerjan> elliott: i gave you a damn good hint.
03:03:44 <oerjan> elliott: i meant consecutively, in such a way that you might guess it's his name
03:04:30 <elliott> oerjan: you realise i cannot actually read the norwegian output :)
03:04:34 <oerjan> also i don't think i'm going to confirm it
03:04:47 <oerjan> elliott: er i meant something you pasted much further up
03:04:58 <elliott> <elliott> oerjan: Il me semble at sekoittaminen språk 運命づけられています to fail!
03:05:09 <oerjan> elliott: in a previous discussion today
03:05:25 <elliott> so much easier to tell me, man!
03:05:29 <oerjan> and if you find it, you will know.
03:05:36 <elliott> i talk so much in a day, oerjan.
03:05:51 <elliott> also, why do you think i have magical oklopol-name-recognition powers X_X
03:06:29 <elliott> 15:24:49 <elliott> oerjan: sheesh, stop ruling the channel by way of homoeroticism
03:06:33 <elliott> oerjan: is oklopol's name Homoqeroticism
03:07:01 <F019> wath for a Mook ?
03:07:03 <oerjan> elliott: the thing you pasted was _by_ oklopol
03:07:16 <elliott> i quoted oklopol today? :/
03:07:22 <elliott> 16:21:19 <elliott> 14:58:40 <oklopol> lament: ACTUALLY INTEGERS ARE REALS, THEY'RE A SUBSET OF REALS
03:07:50 <elliott> unless i pasted something BY oklopol without the <> line
03:07:52 <elliott> which i don't think i did.
03:08:17 <elliott> oerjan: did I _know_ oklopol wrote it?
03:08:32 <oerjan> define "wrote" but otherwise yes
03:08:54 <elliott> I fucking hate you, oerjan
03:09:34 <oerjan> i thought i was making it simple for you XD
03:09:38 <elliott> oerjan: wait is this an actual-day or an oerjan-day
03:09:44 <elliott> do i have to grep two days back in the logs
03:09:57 <oerjan> oh possibly oerjan-day
03:10:15 <elliott> oerjan: wait ville sallo is actually his real name?
03:10:33 <elliott> oerjan: i distinctly recalled we googled for that ages ago and concluded it couldn't possibly be that
03:10:37 <zzo38> I found another list, M for +12, M' for +13, and M'' for +14, M^ for +12:45
03:11:04 <oerjan> ville is a common finnish name
03:11:11 <F019> this time.... it's rainy again ?
03:11:22 <elliott> http://www.facebook.com/Villyyns <-- undoubtedly oklopol
03:11:37 <elliott> damn, there's like fifteen ville salos :(
03:12:13 <oerjan> maybe it's the finnish version of john smith
03:12:18 <zzo38> I found an answer why there is no J. J is used to indicate local time.
03:12:27 <elliott> 08:38:47 <nooga> jix: can't u understand that we don't have any Apples or any PearPCs or elsethingys
03:12:48 <elliott> Wow, Gregor used J ... like an astronaught.
03:13:29 <elliott> oerjan: would it be excessively snarky if I made a website that kept track of Ørjan Standard Time?
03:13:42 <elliott> either just going forwards by one hour each day, or using your irc talking as a guide
03:13:48 <elliott> oerjan: would it be ban-worthy?
03:14:08 -!- pikhq has quit (Ping timeout: 246 seconds).
03:14:16 -!- pikhq has joined.
03:14:45 <Gregor> elliott: Ohyeah, the editor.
03:14:48 <Gregor> elliott: Yuh, I used it.
03:15:18 <F019> SPACE is HIGH LOCK
03:15:51 <elliott> 15:00:06 <GregorR-W> Actually, you're generalizing a lot, the US is a big place :P
03:15:51 <elliott> 15:00:17 <GregorR-W> And England is ... well, not ;)
03:16:22 <elliott> Gregor: what you need to realise in the past is that the US is basically 20 Waleses or something
03:16:26 <elliott> we prefer to keep our Wales
03:16:26 <oerjan> the sun never sets on the us empire
03:16:28 <elliott> and therefore our sheep-fucking
03:16:40 <elliott> you're all about the sheep-fucking.
03:16:45 <elliott> Gregor Richards, sheep-fucker, everybody.
03:18:39 <oerjan> "But you fuck _one_ sheep..."
03:19:09 <F019> swiss-sheep are beating my a$$ till i'm dude
03:20:57 <pikhq> oerjan: If you count military bases, it really doesn't.
03:21:31 <pikhq> There are really US military bases in every time zone.
03:21:52 <pikhq> And this is fucking nuts.
03:22:11 <F019> adiom is the one and base idiom of all idioms of and only if it is
03:22:31 <pikhq> oerjan: Yes, the Army is headed by a rabid squirrel ATM.
03:22:37 <F019> Viva el Squirrelsinkalinka
03:23:02 <F019> Army Swiss Kniff :)
03:23:02 <pikhq> It's convinced that there's nuts in Iraq.
03:23:48 <F019> so bombarding with squirrels on Nutzy Pootzy
03:24:22 <elliott> eir sentience has been demonstrated
03:24:39 <elliott> you're talking about the squirrel :D
03:26:58 <F019> petit rongeur arboricole
03:30:15 <elliott> 19:12:47 <ihope_> Okay, the Recent Changes consists ENTIRELY of spam and reverting of spam.
03:30:15 <elliott> 19:13:04 <ihope_> *All* of this spam was done by unregistered users.
03:30:15 <elliott> 19:14:16 <GregorR> You're right - this situation must be resolved.
03:30:15 <elliott> 19:14:19 <GregorR> I shall log in and spam.
03:31:31 <oerjan> Gregor, the always helpful one
03:34:00 <elliott> 18:13:33 <graue> kipple: just now I forbade the string '<div' and required edit summaries from anonymouses...
03:35:00 <F019> <div> Squirrels are like a fool </div>
03:35:38 <oerjan> elliott: i'm starting to lean towards sentience myself
03:36:04 <elliott> oerjan: he was coherent in French, and made reasonable replies to statements messages after they were made.
03:37:38 <pikhq> Just non-fluent in English, one presumes.
03:37:54 <F019> good presumption
03:38:23 <F019> the river is fluent
03:38:23 <oerjan> to flu or not to flu, that's the question
03:38:46 <F019> pop.... oho i'd flu
03:44:39 -!- oerjan has quit (Quit: Good night).
03:48:24 <zzo38> pikhq: Yes, and was banned on the French channels (which is why they are on this channel)
03:52:10 <elliott> oh Rune on the wiki is kipple?
03:53:23 <elliott> 09:20:31 --- quit: ChanServ (ACK! SIGSEGV!)
03:57:34 <zzo38> pikhq: I can type only ASCII in my client. (However, other messages with Unicode work OK.) And at least now I know how to pronounce because I looked it up in WWWJDIC
03:58:37 <F019> it is true.... pikhq: Yes, and was banned on the French channels (which is why
03:59:01 -!- elliott has quit (Ping timeout: 255 seconds).
04:04:18 -!- pikhq_ has joined.
04:04:24 -!- pikhq has quit (Ping timeout: 252 seconds).
04:05:10 <F019> or i'm a Fighter FA/19 ??
04:17:05 -!- azaq23 has quit (Ping timeout: 240 seconds).
04:30:54 -!- F019 has quit (Quit: Leaving).
05:34:00 -!- pikhq has joined.
05:34:12 -!- pikhq_ has quit (Ping timeout: 250 seconds).
05:35:41 -!- augur_ has quit (Ping timeout: 264 seconds).
05:42:13 <pikhq> So. The Libyan rebels have declared a republic.
05:53:13 -!- asiekierka has joined.
05:58:12 -!- wth has joined.
05:59:55 -!- wth has left (?).
06:00:53 <zzo38> I invented some 'patamagic feats.
06:12:32 <zzo38> Do you think you can use Maximize Spell on Teleport to make it always do ten damage whenever there is a mishap?
06:15:41 -!- augur has joined.
06:23:04 <Ilari> Heh. Write a brainfuck interpretter in pointer-B. Swapping between the code and data makes that real fun.
06:25:17 <zzo38> What is pointer-B?
06:25:30 <Ilari> One esolang I designed.
06:25:46 <zzo38> I cannot find article.
06:26:39 <Ilari> http://esolangs.org/wiki/PointerB
06:27:03 <zzo38> O, there is no hyphen. Why did you type hyphen, then?
06:27:44 -!- augur has quit (Remote host closed the connection).
06:28:33 <Ilari> Maybe I misremembered the name... Or something.
06:28:50 <zzo38> If you can write the brainfuck interpreter, and you can see it is proper, then you can know it turing complete.
06:30:06 <Ilari> I think it is turing complete via self-modifying code (instructions 52, 53 and 54 of main instruction set access code space).
06:30:52 <Ilari> Without self-modification, no way it is turing complete (one stack and finite storage space).
06:31:13 <Ilari> That is '4', '5' and '6'.
06:32:13 <zzo38> Yes I can see that.
06:33:10 -!- TLUL has quit (Quit: *disappears in a puff of orange smoke*).
06:34:33 <Ilari> No conditional branch. But one can compute addresses to jump to.
06:36:12 <zzo38> How many other programming languages have that feature of no condition branch (but possibly can compute address to jump)?
06:41:16 -!- TLUL has joined.
06:42:59 -!- Patashu has quit (Ping timeout: 246 seconds).
06:44:44 <pikhq> Yeah, quite sure that computed goto is sufficient for TC-ness.
06:44:56 <zzo38> pikhq: Yes, I think so, too.
06:45:05 <pikhq> Erm, assuming infinite storage space, of course.
06:45:15 <zzo38> pikhq: Yes I was just about to mention that.
06:47:15 -!- augur has joined.
07:09:31 <Sgeo> LF codepoints?
07:32:25 <zzo38> I have managed to make up some of the \if... conditions in TeX without using \if...
07:33:23 <Sgeo> zzo38, can you manage to cause me to desire food?
07:33:41 <zzo38> Sgeo: Are you going to die from not eating?
07:34:12 <Sgeo> No, but I tend not to eat more than I should. I tend to eat a bit less than I should.
07:34:36 <zzo38> Then, no, I don't.
07:51:06 -!- cheater- has quit (Ping timeout: 255 seconds).
07:52:54 -!- cheater- has joined.
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:18:47 -!- pikhq has quit (Ping timeout: 250 seconds).
08:18:52 -!- pikhq has joined.
08:19:12 <Ilari> Sgeo: Just look some Epic Meal Time videos?
08:20:56 -!- asiekierka has quit (Ping timeout: 246 seconds).
08:32:56 -!- cheater- has quit (Ping timeout: 255 seconds).
08:34:01 -!- Phantom_Hoover has joined.
08:35:11 -!- cheater- has joined.
08:44:28 -!- impomatic has quit (Quit: ChatZilla 0.9.86 [Firefox 3.5.17/20110121150729]).
08:47:05 -!- GreaseMonkey has quit (Quit: The Other Game).
08:55:30 -!- Patashu has joined.
09:01:11 -!- zzo38 has quit (Remote host closed the connection).
09:42:13 -!- MigoMipo has joined.
10:13:35 -!- pikhq has quit (Read error: Operation timed out).
10:13:56 -!- pikhq has joined.
10:28:12 <Phantom_Hoover> I just loaded YouTube's homepage to discover that they had a pretty explicit thumbnail on it, *when they don't give me access to anything deemed inappropriate*.
10:29:28 -!- pikhq_ has joined.
10:29:56 -!- pikhq has quit (Ping timeout: 276 seconds).
10:30:04 <olsner> they're just out to get you, and have already e-mailed your parents about your transgressions
10:40:03 -!- asiekierka has joined.
11:21:34 -!- FireFly has joined.
11:26:32 <Vorpal> <Phantom_Hoover> I just loaded YouTube's homepage to discover that they had a pretty explicit thumbnail on it, *when they don't give me access to anything deemed inappropriate*. <-- just for you. I don't see anything such
11:27:05 <Phantom_Hoover> Vorpal, did you know: YouTube's front page is generated dynamically!
11:27:37 <Vorpal> Phantom_Hoover, yes I know. But if I refresh it like a minute later it still looks the same
11:27:44 <Vorpal> Phantom_Hoover, it takes a while to change :P
11:28:21 <Phantom_Hoover> a) I was logged in, dramatically influencing the layout. b) I am in a different country than you, altering the featured videos enormously.
11:28:45 <Vorpal> a) ah, I was not. b) oh that explains why they featured a Swedish video
11:29:12 <Vorpal> though all but one video on the front page for me is actually English
11:29:30 <Vorpal> Phantom_Hoover, anyway I can only say it's a pity I missed out on the explicity :P
11:33:37 <Phantom_Hoover> It's still there if you want the link to satisfy your sick cravings.
11:34:48 -!- leonid has left (?).
11:37:33 <Phantom_Hoover> Ah, there's a "5" in the top-left corner of the thumbnail.
11:43:27 <Vorpal> Phantom_Hoover, does it?
11:43:35 <Vorpal> Phantom_Hoover, is that some TV channel over there or?
11:44:01 <Phantom_Hoover> Vorpal, yes, and it's owned by Richard Desmond, porn baron extraordinare.
11:44:45 <Vorpal> by the way, I checked on xkcd today, to see if it was still as bad. The last comic was so meta it wasn't funny.
11:45:48 <Vorpal> Phantom_Hoover, do you agree?
11:58:27 -!- Patashu has quit (Quit: MSN: Patashu@hotmail.com , Gmail: Patashu0@gmail.com , AIM: Patashu0 , YIM: patashu2 .).
12:18:04 -!- TLUL has quit (Quit: *disappears in a puff of orange smoke*).
12:24:32 -!- pikhq has joined.
12:24:55 -!- pikhq_ has quit (Ping timeout: 276 seconds).
12:55:33 -!- azaq23 has joined.
12:56:37 <fizzie> Nothing Finnish on youtube front page for me, but I guess that's not terribly surprising.
12:59:52 -!- asiekierka has quit (Ping timeout: 250 seconds).
13:21:37 -!- asiekierka has joined.
13:25:43 -!- BeholdMyGlory has joined.
13:59:19 -!- pikhq_ has joined.
13:59:53 -!- pikhq has quit (Ping timeout: 276 seconds).
14:01:40 <asiekierka> so i want to make Brainf**k in my sandbox game, 64pixels
14:01:53 <asiekierka> the point is the only current possible tape design only lets me do a uni-directional tape
14:02:10 <asiekierka> it's of a limited length so i could time it, but still
14:18:55 <asiekierka> so i'm looking for a very minimal esolang
14:20:30 <cheater-> is unidirectional tape even TC
14:21:10 <asiekierka> look, i can't make it infinite because of the limitations of the game
14:21:21 <asiekierka> and i'm not adding yet ANOTHER block type just to make it TC
14:42:43 -!- pikhq_ has quit (Read error: Operation timed out).
14:44:21 -!- pikhq has joined.
14:44:27 <fizzie> It's not much of a "tape" if you can only move in one direction; it's more of a single cell-sized variable that you can reset to zero (assuming a zero-initialized "tape").
15:06:12 -!- poiuy_qwert has joined.
15:10:18 -!- sebbu has quit (Read error: Connection reset by peer).
15:11:09 -!- sebbu has joined.
15:11:09 -!- sebbu has quit (Changing host).
15:11:09 -!- sebbu has joined.
16:02:06 -!- poiuy_qwert has quit (Quit: Leaving).
16:02:35 -!- calamari has joined.
16:19:13 -!- pikhq has quit (Ping timeout: 252 seconds).
16:19:19 -!- pikhq_ has joined.
16:24:12 -!- ais523 has joined.
16:24:57 <ais523> hmm, this week's mystery: why does Konversation always load with Akregator's icon, even if Akregator isn't running?
16:25:06 <ais523> I think I'll go reinstall it, be back in a moment
16:25:07 -!- ais523 has quit (Remote host closed the connection).
16:26:10 <Sgeo> Today I Learned: ais523 is a KDE person.
16:27:03 -!- ais523 has joined.
16:27:17 <ais523> no idea how that happened...
16:30:54 -!- asiekierka has quit (Read error: Operation timed out).
16:31:04 <ais523> calamari: apparently there's a maintained fork of KDE 3
16:32:56 -!- zzo38 has joined.
16:37:09 <ais523> Gregor (and everyone else who cares): I was fed up with the egobot BF Joust hill still being buggy and all the apparently working impls either not working or being vaporware or being egojsout and so not running as part of the hill
16:37:16 <ais523> so I wrote this: http://sprunge.us/JRMI
16:37:58 <ais523> I'm not sure if it works for nested (({{}})%)%-style things (it should work but is untested on that), but I've tested everything else, and it's producing the same results as bugfixed-egojoust but a lot faster
16:41:44 -!- asiekierka has joined.
16:42:12 <calamari> I looked at the joust description but couldn't figure out how you actually got points lol
16:42:37 <ais523> for beating other programs
16:43:01 <ais523> there's a hill which contains all the best submissions of the past
16:43:14 <ais523> when you submit a program, it's run against all the submissions on the hill
16:43:32 <ais523> and gets points according to what it beats, and on what proportion of tape lengths and polarities
16:43:48 <calamari> apparently my communications skills are lacking.
16:43:54 <ais523> (better opponents give more points)
16:44:03 <ais523> oh, you mean how are individual runs judged?
16:44:07 <ais523> the two programs run on the same tape
16:44:30 <ais523> and a program loses if it tries to move off the end of the tape, or its flag (the tape element it starts on) becomes 0 for two turns in a row
16:45:16 <ais523> so generally speaking, good programs try to do that to their opponent
16:45:35 <ais523> trying to zero the enemy flag is the usual strategy
16:45:40 <calamari> cool, thanks for the explanation
16:46:44 <ais523> if you look at the animations linked from the wiki strategies page, most of them will try to find the enemy flag and set it to 0, and most of the strategies are based around a) doing that more efficiently, or b) making it harder for the opponent to do that
16:47:09 <ais523> and the major issue is that the flag looks much the same as any other tape cell
16:48:36 <Gregor> ais523: The hill is currently running fizzie's cranklance, which is not known to be buggy. If I snag a few minutes I'll see if this gets different results.
16:49:03 <ais523> Gregor: waterfall3 vs. allegro was giving me the wrong results when I tried it a couple of days ago
16:49:11 <ais523> http://esolangs.org/wiki/BF_Joust
16:49:37 <Gregor> ais523: Talk to fizzie :P
16:50:03 <ais523> as in, different from egojsout, bugfixed-egojoust, and juiced (the name of my impl)
16:50:22 <Gregor> ais523: Talk to fizzie :P
16:50:49 <ais523> indeed, I pinged him then and didn't get a response (I don't actually /know/ that fizzie is male, but I'm guessing)
16:51:46 <Gregor> To quote a high-school English teacher "English is a male language, if you don't know somebody's sex then you should use 'he', deal with it."
16:52:43 <ais523> from 10 to 30 cells long, games are done on all lengths
16:52:54 <ais523> it's important that the programs don't know how long it is because then finding the flag would be trivial
16:53:13 <ais523> so you say, for instance, that one program beats another on 15 out of 21 tape lengths
16:53:54 <ais523> when comparing two programs, you also do a second run where + and - are swapped in one of the programs, to prevent degenerate behaviour based on polarity dependence
16:53:58 <ais523> so there are 42 runs in all
16:55:13 <pikhq_> Gregor: English is not a male language! It is the language of males and the sexless!
16:55:36 <pikhq_> Gregor: You see, the English were discriminatory towards *women*. Those devoid of gender are perfectly fine by them.
16:55:52 <ais523> pikhq_: *against women, surely?
16:56:10 <Gregor> pikhq_: And yet, we have no gender-neutral living pronouns...
16:56:12 <pikhq_> I still haven't had my morning cup of coffee.
16:56:25 <pikhq_> Gregor: We have a male-or-genderless pronoun.
16:56:30 <pikhq_> Gregor: Which is my point.
16:56:40 <Phantom_Hoover> pikhq_, the English were discriminatory towards women because they were SUCH BETTER PANSIES
16:56:52 <Gregor> pikhq_: OK, fair point :P
16:57:41 <calamari> babies are sometimes referred to as "it"
16:58:04 <Gregor> Pfff, babies aren't living until they're one year old.
16:58:26 <Gregor> This is why abortion is allowed until 21 months after pregnancy.
16:58:39 <Gregor> s/pregnancy/conception/
16:59:11 <pikhq_> Gregor: I'm pretty sure that abortion after birth will get you tarred and feathered in most jurisdictions.
16:59:43 <Gregor> But merely tarred and feathered :P
16:59:53 <pikhq_> Oh, no, that's just the start.
17:00:00 <pikhq_> You'll then be aborted.
17:00:20 <ais523> I thought being tarred and feathered was generally fatal
17:00:56 <pikhq_> ais523: No, just permanently scarring.
17:02:58 <Gregor> MERELY permanent scarring.
17:14:44 <ais523> hmm, I conclude that the vast majority of YouTube comments are actually parodies of each other
17:15:08 <ais523> the proportion of people making fun of the typically stupid YouTube comments has actually risen above the proportion of people making stupid comments
17:17:29 -!- asiekierka has quit (Read error: Operation timed out).
17:22:45 <fizzie> Oh, my name's been mentioned.
17:23:25 <ais523> fizzie: I'm accusing your BF Joust interps of being buggy
17:23:37 <fizzie> Yes, I see. Well, it wouldn't be the first time.
17:23:39 <ais523> try running waterfall3 vs. allegro, and comparing to egojsout
17:24:18 <ais523> or to juiced, which I pastebinned in the channel a little earlier today
17:26:38 <fizzie> Yes, time to compare some traces, I guess.
17:31:35 <fizzie> (Your problematic programs are always awfully long.)
17:31:53 <ais523> my guess is a parsing problem related to ({}({})%)%
17:32:03 <ais523> as egojoust had one of those before I fixed it
17:32:48 <fizzie> Also I still can't read egojsout trace format.
17:34:20 <fizzie> What does it mean when it shows "(128 + )128 (9 > )9" where all three of ")128", "(9" and ">" are highlighted in red? That it "executed" all three during that cycle?
17:34:53 <ais523> I think it means that the ()*128 loop ended, the ()*9 loop started, and it executed the >
17:38:46 <fizzie> Oh yes, it is a parsing "bug".
17:39:10 <fizzie> For the value of "bug" that equals "does not ignore whitespace when looking for digits after *".
17:39:23 <fizzie> (I may have mentioned I'd really like to have a definitive spec for the language.)
17:40:18 <ais523> the exact parsing details after * and % are a little tricky, in the end I just aimed for compatibility with other interps
17:41:25 <fizzie> Well, I'll make {crank,gear}lance ignore spaces there. I do think I recall hearing that some other interps do.
17:41:52 <ais523> all others do because nearly all my defend programs have spaces there to make it clearer what they're doing
17:42:11 <ais523> I suppose I could remove them if really necessary, but it'd make them uglier for not much gain
17:42:12 <fizzie> You could just left-align your numbers instead of right-aligning them.
17:42:31 <fizzie> Anyway, I'll conform to the quasi-spec soon enough.
17:47:06 <ais523> anyway, now I know what I need to do to fix my program for the egobot hill
17:48:40 <ais523> !bfjoust waterfall3 http://sprunge.us/IaPS
17:49:00 <EgoBot> Score for ais523_waterfall3: 61.4
17:49:43 <ais523> 28 | + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + | 61.4 | 27.9 | 28 | ais523_waterfall3.bfjoust
17:49:51 <fizzie> Gregor: http://git.zem.fi/chainlance/blob_plain/HEAD:/gearlance.c is the fixed version.
17:50:03 <ais523> amazingly, it isn't first, because allegro's wins are more convincing
17:50:37 <fizzie> A straight-"+" row is impressive none-the-less.
17:56:03 <ais523> there are a few programs where the result depends a lot on constant tweaking, I used a program to automatically tweak the constants for those
17:57:18 -!- zeotrope has joined.
18:01:31 <fizzie> http://zem.fi/egostats/ updated with that latest hill; based on http://zem.fi/egostats/plot_dpoints.png you indeed have some quite close cases (almost grey squares) there.
18:02:04 <fizzie> Also seems to do most of the winning on longer tape lengths.
18:02:21 <ais523> most of the close places are polarity dependences
18:02:31 <ais523> the strategy's rather different on the two polarities
18:04:53 <ais523> if you want a hilarious run to watch, watch one of its wins on longer tapes against lead_acetate_philip
18:05:08 <fizzie> Waterfalls 2 and 3 have an interesting valley near the home flag in http://zem.fi/egostats/plot_tapeabs.png
18:05:43 <ais523> is that to do with time spent on the cell?
18:06:11 <fizzie> It's an average of the absolute value at that point at the end of the match.
18:06:38 <ais523> ah, that makes sense, because the cell 2 away from the flag is left deliberately blank
18:06:59 <ais523> waterfall actually uses the tape for calculations, which is almost unheard of in BF Joust
18:07:04 <fizzie> There's also quite a regular structure in the http://zem.fi/egostats/plot_p28_ptapemax.png plot, which is "maximum value in the cell when the program >d or <d away from it" averages.
18:08:09 <ais523> yep, that's showing the decoy/tripwire structure
18:08:35 <zeotrope> zzo38: what is the default starting position and direction of a Memfractal program?
18:18:02 -!- Mathnerd314 has joined.
18:49:12 <ais523> the other interesting thing about waterfall3 is that, despite its length, it was basically written entirely by hand
18:49:27 <ais523> although a few of the constants were overfitted to the hill by computer in order to get that perfect record
18:50:50 -!- sebbu2 has joined.
18:51:19 -!- sebbu has quit (Ping timeout: 252 seconds).
18:53:32 -!- oerjan has joined.
18:55:20 <fizzie> Well, it's still no FFSPG when it comes to size.
18:55:29 <fizzie> [('Gregor_furry_furry_strapon_pegging_girls', 87949),
18:55:29 <fizzie> ('Gregor_furry_furry_leather_discipline_girls', 50264),
18:55:29 <fizzie> ('quintopia_space_elevator', 15601),
18:55:29 <fizzie> ('ais523_definder2', 5536),
18:55:29 <fizzie> ('ais523_waterfall3', 5488),
18:55:47 <fizzie> (Program length using a rather arbitrary measure.)
18:56:14 <ais523> !bfjoust waterfall3 http://sprunge.us/hHGW
18:56:20 <ais523> OK, that's only a very marginal improvement
18:56:25 <ais523> but I want to get above allegro somehow
18:56:35 <ais523> perhaps I'll just submit a program that beats allegro and a bunch of other things, but loses to waterfall
18:56:41 <EgoBot> Score for ais523_waterfall3: 62.1
18:57:32 <fizzie> Deewiant: Ping, pong, your leadership is being threatened.
18:58:12 <fizzie> I'm sure he's got some sort of a "poll bfjoust report, send SMS when other programs get threateningly close" system set up, though.
18:59:04 <fizzie> (Note to self: a possible new service for fungot?)
18:59:04 <fungot> fizzie: i'm not up to coding a 4k. :p ( copyright should expire soon.)" yzi/ fit, on unrealistic demands? that sounds as a likely translation, considering the beep, not the
18:59:17 -!- elliott has joined.
18:59:35 <ais523> !bfjoust fast_rush_slow_clear <
18:59:39 <EgoBot> Score for ais523_fast_rush_slow_clear: 0.0
18:59:52 <fizzie> Close, but not *quite*.
19:00:03 <ais523> I'll put it back when I'm done, presumably
19:00:35 <ais523> !bfjoust decoybooster2 <
19:00:42 <EgoBot> Score for ais523_decoybooster2: 0.0
19:00:58 <ais523> now I'll put those two back again
19:01:16 <ais523> (yes, looking for programs of mine that waterfall3 does worse against than allegro does and deleting them is a cheap tactic)
19:01:21 <fizzie> This was a... curious exercise.
19:01:33 <ais523> !bfjoust fast_rush_slow_clear >>>>>>>>(>[+++++[-.]])*21
19:01:36 <EgoBot> Score for ais523_fast_rush_slow_clear: 19.2
19:02:01 <ais523> !bfjoust decoybooster2 (>)*7++<(-)*85(<(-)*85<(+)*85)*3(-)*43(>)*8(>[(+)*5[-.]])*21(+(.)*5)*10000
19:02:03 <EgoBot> Score for ais523_decoybooster2: 21.5
19:02:11 <ais523> fizzie: I wanted to get onto the notable programs list on the wiki
19:02:18 <ais523> which is for former hill leaders only
19:02:25 <fizzie> Ah, there's a rule like that.
19:03:01 <ais523> I've been busy winning the egojoust (well, cranklance) hill
19:03:18 <ais523> also, I wrote my own BF Joust interp because lance is still vaporware
19:03:21 <ais523> (it's linked in the logs)
19:05:25 <elliott> Gregor just didn't respond to the ping and integrated cranklance before he noticed
19:05:33 <elliott> and it's less buggy than cranklance too
19:05:53 <fizzie> You just missed the latest "bug".
19:06:15 <fizzie> (I'm pretty sure that would be UNDEF in mycology terms, though.)
19:06:26 <ais523> elliott: have you posted it anywhere?
19:06:32 <ais523> elliott: cranklance didn't like * whitespace number
19:06:45 <ais523> and I left-justified the numbers in waterfall3 as a workaround
19:06:46 <elliott> ais523: no, because what's the point if cranklance is already integrated?
19:06:55 <ais523> because more interps publically available is a good thing?
19:07:03 <elliott> ais523: There's one parser bug I have to fix to make it "work", I was going to integrate it after the fixed-point scoring system was done.
19:07:13 <elliott> ais523: Since Gregor jumped the gun on cranklance without seeing my ping, I haven't bothered to do so.
19:07:30 <ais523> anyway, waterfall3 is even creatively named!
19:07:39 <ais523> I could have just called it defend18 or something...
19:07:46 <ais523> you should be happy for me
19:08:18 <ais523> it was top of the hill for a moment (because I totally just suicided a couple of my programs that allegro beat by more than it did), and still beats all opponents in their individual matches
19:08:49 <ais523> it's going to take a while to explain how it works, though, as it uses pretty much every defensive technique in existence, as well as several new ones
19:09:05 <ais523> if you want to see hilarious, watch one of the matches it wins against lead_acetate_philip on a longer tape
19:09:16 <ais523> which it does by detecting it and deleting its own decoys in order to prevent it changing strategy
19:11:12 -!- elliott_ has joined.
19:11:37 -!- elliott has quit (Read error: Connection reset by peer).
19:21:12 <Gregor> I don't give one flying fuck which BFJ engine is in EgoBot, but please god PLEASE somebody else integrate it :P
19:21:19 <Gregor> Gimme a hg bundle and I'll use whatever.
19:21:29 <elliott_> Gregor: If you do the fixed-point scoring, maybe. :p
19:21:46 <ais523> it's less urgent now that fizzie told me how to work around the bug
19:22:28 <elliott_> Buying a month of reddit gold for someone with 51 hours to live... makes sense
19:22:47 <ais523> that's an oddly precise numbre
19:23:34 <ais523> two significant figures seems surprisingly accurate for a survival time estimate
19:23:41 <elliott_> ais523: "On Tuesday I'll finally end my battle with cancer thanks to Oregon's Death with dignity act."
19:24:36 <fizzie> Gregor: There's an updated gearlance.c if you like to fix that latest incompatibility. (I still can't bring myself to call it a bug without putting quotes around it.) Incidentally, would you want me to change the sign of the printed score value in my official copy too?
19:25:10 <elliott_> ais523: dude, yell at fizzie how it's not a bug. i got enough abuse for opposing it :-P
19:25:13 <ais523> I was annoyed enough at having to make juice handle ({})* (interpreting it as ({})%) to handle one of Gregor's programs
19:25:34 <elliott_> ais523: template for you: "BRAINFUCK WHITESPACE INSENSITIVE"
19:25:35 <ais523> even though that was actually done by removing rather than adding code
19:25:47 <elliott_> ugh @ ({})*, we agreed that wasn't equal
19:25:47 <ais523> elliott_: do you mean, how it is a bug?
19:25:55 <elliott_> ais523: no, I mean, fizzie is MOCKING its bugliness
19:26:06 <elliott_> when I said it was WONTFIX, you yelled at me for hours :P
19:26:09 <ais523> BECAUSE SINCE WHEN WAS BF SENSITIVE TO WHITESPACE
19:26:10 <elliott_> he needs the same treatment! equality!
19:26:15 <ais523> yep, but I've done it once alreayd
19:26:18 <elliott_> fizzie: See, feel bad. You've made ais yell. He never does that.
19:26:23 <elliott_> You should probably just cry a bit now.
19:26:41 <fizzie> ais523: Ignoring whitespace (and whitespace only) seems very un-bf thing to do.
19:26:58 <ais523> fizzie: I was arguing that it should delete everything that wasn't numbers
19:27:03 <ais523> but that would be incompatible with everything
19:27:17 <elliott_> we agreed on an interpretation
19:27:21 <ais523> for bonus points, I had to implement *-1 even though that's ridiculous parsewise
19:27:23 <elliott_> fizzie: all the digits must be together
19:27:26 <elliott_> but anything can come before the digits
19:27:27 <ais523> that's why I used the past tense
19:27:44 <ais523> oh, juiced allows only whitespace before the digits atm, although that's easy enough to change
19:27:47 <elliott_> this is because a number is an atomic code element, and also because it's convenient
19:27:51 <ais523> I was trying to remember what the agreed-on interpretation was
19:28:04 <elliott_> ais523: well, it's the one i unilaterally decided and everyone else was ok with :)
19:28:08 <ais523> I'll fix it at whatever point it matters
19:28:26 <fizzie> Crank/gear also allows only whitespace too, since that's what I was complained about.
19:28:51 <ais523> it was actually made out of my bug fixes and enhancements to egojoust
19:28:55 <ais523> I just changed the engine underneath them
19:29:10 <fizzie> Arguably though you could say (...)*[]42 should still be not-fine.
19:29:33 <elliott_> The grammar is '*' comment digit+.
19:29:46 <elliott_> comment is any character apart from +-<>.*%()[]
19:29:48 <ais523> that needs a special case in juiced's parser, and I'm not at all convinced it's useful
19:29:53 <ais523> given that - is a command in its own right
19:29:57 <elliott_> The grammar is '*' comment (optional '-') digit+.
19:30:15 <ais523> at least, - being context-sensitive like that irritates me a bit
19:30:44 <ais523> it's inelegant in a different sense
19:30:56 <ais523> if there was some other representation of negative numbers, it would be elegant
19:31:04 <ais523> perhaps I should insist on a Unicode minus rather than a hyphen there
19:31:10 <ais523> but that would mean parsing Unicode
19:31:40 <fizzie> ais523: Well, (42) is -42 in some (accounting-related?) contexts.
19:31:46 <fizzie> But that's even worse, I guess.
19:32:28 <zzo38> fizzie: If it is red...
19:32:28 <oerjan> <elliott_> ais523: template for you: "BRAINFUCK WHITESPACE INSENSITIVE" <-- it would be a bit annoying if it interpreted *5 blah blah 10 as *510 though...
19:32:36 <ais523> I've seen superscript minus for unary minus before now
19:32:46 <ais523> oerjan: indeed, I think that was one of the counterarguments
19:32:51 <ais523> especially as I use numbers in comments a lot
19:33:04 <ais523> elliott_: I suppose I need to update esolangs.el's syntax highlighting too
19:33:15 <ais523> atm it only does )* whitespace number (and the equivalent with %)
19:34:23 <elliott_> where number = digit+ | '-' digit+
19:35:21 <oerjan> <elliott_> oerjan: RTFrest of the talk <-- NO! BACKSCROLL ALL THE WAY!
19:35:38 <ais523> elliott_: for bonus points, is , a comment in that context?
19:35:45 <ais523> I think it's technically still reserved
19:35:51 <ais523> just in case someone other than zzo38 finds a use for it
19:35:58 <ais523> (zzo38's definition never really caught on)
19:36:35 <ais523> it reads output from the opponent's . or a random number if they haven't output anything
19:36:51 <ais523> it could actually be brokenly good depending on how the randomization works
19:36:58 <ais523> due to being able to change a value at faster than lightspeed
19:37:53 <zzo38> Because of the way of working now, just ignore the randomization and have it just do the same as . if there is no opponent output
19:38:15 <ais523> hmm, the issue is still that using . would be dangerous on a low-valued cell
19:38:25 <ais523> in case its value was used to instantly zero your flag
19:38:39 <zzo38> (The randomization can be used for a kind of 2-player game instead)
19:39:35 <ais523> a randomization command would actually be genuinely useful in writing programs (one potential semantics is to reverse all + and - in the program with a 50% chance), but a complete pain to run on hills because results would no longer be deterministic
19:39:47 <fizzie> elliott_: Coincidentally, your quasi-formal grammar doesn't work: if "comment is any character apart from [that list]", then numbers are comments too, and "* comment number" will have comment matching those numbers. Not to nitpick on the fact that you could read that as single-character-only comments.
19:39:57 <elliott_> fizzie: That list included numbers.
19:40:01 <elliott_> Or if it didn't it was a mistake.
19:40:04 <zzo38> ais523: Exactly what is my thoughts on it. Randomization would only be used in 2-player game, not in hill game.
19:40:07 <fizzie> <elliott_> comment is any character apart from +-<>.*%()[]
19:41:35 <fizzie> If that's the case, then it's a bit strange that you can put in random numbers when there's no preceding * or % for them.
19:42:09 <ais523> I should go and write a wiki description of waterfall3
19:42:16 <ais523> also of space_elevator, given that I understand how it works quite well by now
19:42:19 <ais523> even though I didn't write it
19:43:12 <elliott_> ais523: here's a fun segfault for you: http://sprunge.us/CgJK
19:43:14 <ais523> I doubt I'll be able to add an egojsout animation, though, it takes a huge amount of time to beat simple
19:43:25 <elliott_> (the alternative is *subscribing to a gnu.org mailing list*, and that's terrifying enough that I'm just going to prevent you with a trace first)
19:43:31 <elliott_> also, my theory down at the bottom must be wrong
19:43:36 <elliott_> because even if expr were changed
19:43:44 <ais523> what's that a segfault in?
19:43:51 <elliott_> ais523: mcmap, when calling Guile
19:43:53 <ais523> it looks like a scheme interp
19:43:57 <elliott_> on scm_c_eval_string(foo) where foo is "2"
19:44:23 <ais523> try using valgrind, I find it gives better explanations of segfaults than gdb does
19:44:58 <elliott_> ais523: the segfault is in Guile code, which makes me suspect that Shit is Verily Up
19:44:58 <ais523> you've never used valgrind? seriously?
19:45:26 <ais523> oh, and had a Mac for ages, on which it doesn't run
19:45:29 <ais523> that probably has something to do with it
19:45:30 <zzo38> I find gdb works for fine for me, I never use valgrind.
19:45:34 <elliott_> 08.02.04:13:22:18 <ais523> I wonder if it's possible to use valgrind as a garbage collector?
19:45:39 <elliott_> first time you mentioned valgrind in here
19:45:45 <elliott_> I was hoping it would be "valgrind? what's that?!"
19:46:09 <elliott_> ais523: I was probably put off using Valgrind by Vorpal's religious devotion to it
19:46:29 * ais523 tells egojsout to do waterfall3 vs. simple on tape length 25, at the cost of most of eir computer's CPU cycles
19:47:21 <ais523> gah, it's only up to 1500 or so, and I suspect it goes near the cycle limit to beat it
19:47:33 <elliott_> Want me to run it on this SUPERCOMPUTER :P
19:47:39 <ais523> it takes something like 99700 cycles to beat allegro on tape length 30
19:48:12 <ais523> in fact, it even changes to a rush strategy at one point when it has a perfect lock, just to stay within the cycle limit
19:48:23 <ais523> (it uses the lock to set up decoys, which is much faster than a full-tape clear)
19:48:32 <elliott_> meanwhile, I've been itching to make a language that's like the union of Haskell, Lisp and C#
19:49:35 <elliott_> actually one thing I've realised is that you can easily do Lisp-style macros in a language with complex syntax, as long as it has pattern matching
19:49:45 <elliott_> because any sane such language will have quotation and unquotation
19:50:09 <elliott_> invert {if ,x then ,y else ,z} := {if !,x then ,y else ,z}
19:51:11 <elliott_> do I need any special compiler flags to use valgrind?
19:51:42 <ais523> no, although -g produces better output
19:51:46 <ais523> because then it can give line numbers
19:52:37 <elliott_> ==2849== Use of uninitialised value of size 8
19:52:37 <elliott_> ==2849== at 0x5B4CB40: GC_mark_and_push_stack (mark.c:1396)
19:52:41 <elliott_> oddly mcmap has exited immediately
19:52:48 <elliott_> but I can see the boehm GC is going to cause some noise...
19:53:02 <ais523> how does a GCed program segfault anyway?
19:53:23 <elliott_> ais523: Guile isn't really a "GC'd program"
19:53:32 <elliott_> it's a Scheme implementation that uses the boehm GC to
19:53:38 <elliott_> *Boehm GC to implement Scheme garbage-collection
19:53:48 <elliott_> ==2849== Invalid free() / delete / delete[]
19:53:48 <elliott_> ==2849== at 0x4C27D71: free (vg_replace_malloc.c:366)
19:53:48 <elliott_> ==2849== by 0x6536A0A: free_mem (in /lib/libc-2.12.1.so)
19:53:48 <elliott_> ==2849== by 0x65365A1: __libc_freeres (in /lib/libc-2.12.1.so)
19:53:48 <elliott_> ==2849== by 0x4A2366B: _vgnU_freeres (vg_preloaded.c:62)
19:53:53 <elliott_> ==2849== Address 0x4045980 is not stack'd, malloc'd or (recently) free'd
19:53:57 <elliott_> mcmap doesn't run in Valgrind!
19:54:11 <ais523> valgrind segfaults-fast
19:54:25 <ais523> but I've never tried it on a garbage-collected program
19:54:30 <ais523> I imagine the result would be kind-of messy
19:54:52 -!- impomatic has joined.
19:55:05 <elliott_> Which probably means that it's incredibly tedious :-P
19:55:21 <elliott_> oh, it does indeed segfault immediately
19:55:27 <impomatic> elliott_: are you still working on your Forth?
19:55:42 <ais523> well, it's nice to get the segfault where the bug is rather than later
19:55:43 <elliott_> impomatic: Yes, though right now I'm busy integrating mcmap with Guile.
19:55:56 <elliott_> ais523: I don't think there's a bug, I think it's just Boehm GC doing undefined things that Valgrind decides it doesn't like
19:56:11 <elliott_> "The Boehm GC performs all manner of dark magic, most of which valgrind doesn't like. It's normal." --Google
19:57:43 <ais523> hmm, /me exits egojsout
19:57:49 <ais523> it'd only gone up to 6000 cycles or so
19:58:04 <ais523> it can more or less handle the awesomeness of triplock3 in terms of reporting results, but not in terms of reporting debug info
19:58:15 <ais523> it's a netbook, what do you expect?
19:58:20 <impomatic> elliott_: I'm still in the race then :-P I just started my top-down implementation :-) http://twitcode.org/show/251/forth-outer-interpreter
19:58:32 <ais523> impomatic: is it being written in Redcode?
19:58:47 <elliott_> impomatic: Top-down Forth... what madness!
19:58:58 <elliott_> impomatic: also, error reporting?? that's ridiculous!
19:59:03 <elliott_> just crash on invalid words, it saves bytes
19:59:16 <elliott_> impomatic: well, assuming you're trying to do this in 510 bytes without an oS
19:59:59 <ais523> oh, that reminds me, I hit upon an improvement to triplock3 while writing waterfall3
20:00:09 <impomatic> 8086 implementation to get it running followed by MSP430 as soon as the hardware arrives :-)
20:00:23 <elliott_> impomatic: But in one sector? :p
20:00:35 <elliott_> Maybe if you stored that interpreter word pre-compiled.
20:01:14 <ais523> heh, triplock3 times out on one polarity against simple modified to be vulnerable against its strategy
20:01:32 <impomatic> elliott_: we'll see. I'll be happy if it's under 1K. I think the full ANS core would take approx 4K
20:01:38 <ais523> impomatic: I've been busy dominating the egobot BF Joust hill
20:01:45 <elliott_> impomatic: don't bother with ANS compliancy
20:01:50 <elliott_> impomatic: not even Chuck Moore does
20:01:58 <ais523> 28 | + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + | 62.1 | 28.3 | 28 | ais523_waterfall3.bfjoust
20:04:02 <ais523> thanks for writing sexyghoul/spookygoth, beating them was actually really intellectually interesting
20:04:13 <ais523> especially as I mostly try to beat the strategy rather than the individual program
20:04:48 <impomatic> ais523: http://bit.ly/cwyE4s = MSP430 :-)
20:05:08 <ais523> hmmm, what's with the URL shortener?
20:05:14 <elliott_> x86 needs ten times more registers
20:05:32 <ais523> floating-point systems have thousands of registers
20:05:39 <impomatic> Sorry :-) http://processors.wiki.ti.com/index.php/MSP430_LaunchPad_(MSP-EXP430G2)
20:05:55 <elliott_> ais523: why did RISC lose again?
20:05:56 <ais523> so that they can context-switch quickly
20:05:59 <ais523> there's a different set of registers for each thread
20:06:02 <olsner> infinite registers? but that'd probably make it turing complete!
20:06:08 <ais523> so there's no need to actually swap them out
20:06:13 <ais523> olsner: not if you need to name them by hand
20:06:35 <ais523> on the other extreme, I've seen systems that have only one register, or from another point of view the whole memory is registers
20:06:39 <ais523> and they work quite well too
20:07:47 * elliott_ tries to figure out how to organise memory
20:08:10 <ais523> what about one of those ringbinders full of plastic wallets you can put paper in
20:08:16 <olsner> elliott_: you can think of memory as an array
20:08:26 <elliott_> i don't want things to overlap
20:08:33 <elliott_> and i want certain things to be able to grow to fill all of memory
20:08:37 <elliott_> but i can't just pick random constants
20:08:41 <elliott_> because (1) low enough, there's the bios and stuff
20:08:45 <elliott_> (2) high enough, you might not have that kind of memory
20:10:06 <elliott_> hmm, ok, how to do the packing
20:10:25 <elliott_> subtract 64, and with 0x111111, and then pack
20:11:22 <ais523> I thought you were taking every fourth bit for some hashing reason, or whatevert
20:12:47 <calamari> design a file system with no dedundancy so that fsck is never needed
20:12:58 <Vorpal> <elliott_> Well, Vorpal did it. <-- did what?
20:13:06 <Vorpal> elliott_, run a gced program in valgrind?
20:13:10 <Vorpal> elliott_, depends on which GC
20:13:19 <Vorpal> elliott_, you can do it for something using libpython.
20:13:20 <olsner> calamari: you mean one that is guaranteed not to leave anything fsck can salvage?
20:13:22 <Vorpal> elliott_, can't be done
20:13:27 <Vorpal> elliott_, completely incompatible
20:13:33 -!- sebbu2 has quit (Read error: Connection reset by peer).
20:14:04 -!- sebbu has joined.
20:14:05 -!- sebbu has quit (Changing host).
20:14:05 -!- sebbu has joined.
20:14:11 <Vorpal> elliott_, as far as I understood it the incompatibility is on a fundamental level
20:14:19 <oerjan> <ais523> hmmm, what's with the URL shortener? <-- filthy [kg]ah?d*h?af*[iy] supporter!
20:14:27 <Vorpal> elliott_, as in, would require a complete redesign of either valgrind or boehm
20:15:33 <Vorpal> I can get "kadafi" out of that I think
20:15:39 <olsner> Vorpal: it's a regexp matching khaddaffi or whatshisname
20:15:46 <ais523> it's probably a reference to the .ly domain
20:15:57 <elliott_> Qadaffi is the preferred, I think.
20:16:01 <elliott_> But with some apostrophicals in there.
20:16:09 <Vorpal> elliott_, don't know how it is spelled in English
20:16:30 <Vorpal> elliott_, trans-whatever then
20:16:35 <Vorpal> elliott_, hm: http://en.wikipedia.org/wiki/Muammar_Gaddafi
20:16:49 <Vorpal> also http://en.wikipedia.org/wiki/Muammar_Gaddafi#Name
20:16:52 <olsner> http://en.wikipedia.org/wiki/Muammar_al-Gaddafi#Name the diagram there is awesome
20:16:54 <elliott_> Plenty transliterations are valid.
20:17:50 <oerjan> elliott_: i've seen k used, although you're right q should be there too
20:18:17 <elliott_> He's Khedhdhafy or Khuzzai from now on to me
20:18:31 <Vorpal> elliott_, I think the Swedish newspapers use some spelling with K
20:19:38 <elliott_> How to pack the bits with few code...
20:20:35 <elliott_> :(, 2*si isn't a valid address?
20:20:39 <oerjan> <Vorpal> elliott_, don't know how it is spelled in English <-- the whole point of my regexp there was that his name is spelled so many different ways it's a joke in itself
20:21:26 <olsner> it'd be something like (Q|[KG]h?)[aeu](...)aff?[iy], with (...) being a regexp for the variations between d and zz
20:21:26 <oerjan> i think maybe i should have used + on the d and f...
20:21:33 <Vorpal> elliott_, also doesn't guille suck?
20:21:42 <elliott_> And for what reason would it suck?
20:21:59 <Vorpal> elliott_, you said it did before. Compared to what is now racket
20:22:05 <elliott_> It used to be pretty bad at R5RS, but they've got a new release now.
20:22:08 <elliott_> Vorpal: Dude, it's for embedding into a program.
20:22:14 <elliott_> Racket is utterly inapplicable.
20:22:20 <olsner> it's almost (dh?|th|z){1,2}, but not exactly
20:22:39 <elliott_> NOW WHY ISN'T [fs:2*si] A VALID ADDRESS
20:22:55 <Vorpal> elliott_, in what system?
20:22:56 <oerjan> <elliott_> Gandalf. <elliott_> That's his name. <-- IT EXPLAINS SO MUCH
20:23:28 <oerjan> it had to happen eventually
20:23:54 <elliott_> <Vorpal> elliott_, in what system?
20:24:07 <olsner> elliott_: 64, 32 or 16-bit? I think you mean esi
20:24:24 <elliott_> If you complain about using readable Intel syntax rather than AT&T's ((fs(%si,*2,*2,), then I'll kill you.
20:24:24 <Vorpal> elliott_, oh registers?
20:24:27 <olsner> also, remember that the fs: segment override costs a byte
20:24:36 <elliott_> and it's OK, it's just VGA memory
20:24:46 <Vorpal> elliott_, what /are/ you doing?
20:26:06 <elliott_> olsner: are the lines that have invalid effective addresses.
20:26:32 <zzo38> Can you use the BIOS call for keyboard characters? And then use ASCII codes 0x20...0x5F? Does that works?
20:26:49 <elliott_> zzo38: That's my current plan, but actually the BIOS call gives me the scancode too!
20:26:56 <elliott_> Which means that I should be able to do the 5-bit packing I wanted to.
20:28:20 <ais523> !bfjoust waterfall2 http://sprunge.us/dSER
20:28:21 <impomatic> elliott: why not mov ah, 0x07 / mov word [fs:2*si], ax
20:28:24 <EgoBot> Score for ais523_waterfall2: 44.4
20:28:28 <ais523> may as well left-justify the numbers for posterity
20:28:49 <elliott_> impomatic: Because ah is the scancode that I would rather not clobber.
20:28:51 <ais523> even though it hurts waterfall3 somewhat
20:29:01 <elliott_> impomatic: I could copy it to another register, but I still don't think [fs:2*si] is an OK address
20:29:05 <ais523> waterfall2 vs. waterfall3 is incredibly close and constant-tweaking-dependent a lot
20:29:10 <ais523> so it's up to me to decide which way that match goes
20:29:37 <ais523> and waterfall2 seems to only have six losses and one tie, that's not bad
20:29:39 <elliott_> ais523: why keep multiple versions of a program on the board at once?
20:29:47 <elliott_> if they're similar and one is just an improvement I think it's bad to have both on the board
20:29:51 <ais523> the strategy is actually somewhat different
20:29:57 -!- augur has quit (Remote host closed the connection).
20:30:07 <ais523> waterfall3 attempts to adapt to the opponent's strategy
20:30:14 <EgoBot> Use: !bfjoust <program name> <program> . Scoreboard, programs, and a description of score calculation are at http://codu.org/eso/bfjoust/
20:30:14 <ais523> whereas waterfall2 doesn't, apart from rushing against nonattacking enemies
20:30:25 <ais523> waterfall1 isn't up there because it's an older version of 2
20:30:32 <fizzie> elliott_: AFAIK the SIB byte addressing modes (anything with *2/*4/*8) can be done in 32-bit code only.
20:30:40 <ais523> normally I have two versions of a program, the dumb version and the clever version
20:30:57 <elliott_> Hey, can I jump into a 32-bit code segment without going into protected mode? :-)
20:31:00 <ais523> like defend12/defend13
20:31:08 <ais523> or arguably defend7/defend9
20:31:09 <Vorpal> elliott_, isn't that unreal mode?
20:31:17 <elliott_> Vorpal: no, that's the opposite of unreal mode.
20:32:07 <fizzie> elliott_: The 16-bit mode operand encodings can do [{bx,bp,si,di}+N] (with no/8-bit/16-bit N) and also [{bx,bp}+{si,di}], but that's it.
20:32:33 <fizzie> You can shift di by one.
20:32:38 <fizzie> At least I think you can.
20:32:42 <olsner> elliott_: obviously you can't load a 32-bit code segment outside PM :)
20:32:49 <elliott_> olsner: I CAN DO WHATEVER I WANT
20:33:00 <ais523> <elliott_> Oh, right. <-- no, left
20:33:02 <Vorpal> elliott_, do it on an avr instead
20:33:09 <ais523> rightshift would divide by two
20:33:10 <olsner> elliott_: ... but you can't use a SIB byte in 16-bit code?
20:33:24 <fizzie> olsner: As far as I can tell, no.
20:33:49 <fizzie> olsner: My "ModRM Memory References, 16-Bit Addressing" table contains no encodings that would use a SIB byte.
20:34:01 <elliott_> which way around do I want to pack
20:34:07 <elliott_> I think starting at LSB for first char
20:34:09 <olsner> fizzie: just pointing out that elliott_'s powers were unable to fix x86's instruction encodings :)
20:34:18 -!- wareya_ has quit (Read error: Connection reset by peer).
20:34:19 <impomatic> Elliott_: forget the AVR, it's Harvard architecture.
20:34:28 <Vorpal> impomatic, oh good point
20:34:38 <fizzie> Vorpal: "Scale-index byte", the thing that's used to encode the "register*4 + register" sort of address modes.
20:34:53 <Vorpal> elliott_, nothing wrong with harvard. It is common on embedded systems
20:34:57 <olsner> not scale/index/base byte?
20:35:08 -!- wareya has joined.
20:35:31 <fizzie> olsner: Oh, right, yes, it has "base" in it too. Otherwise "SIB byte" would join the "PIN number" crowd. :p
20:35:32 <Vorpal> <elliott_> x86 is the worst. <elliott_> Harvard is the WORST
20:35:39 <elliott_> OK, so, anyone have a QWERTY scancode table?
20:35:41 <Vorpal> since "worst" is absolute
20:35:45 <elliott_> I wanna see how much I need to offset and the like to get a good range of chars :P
20:35:52 <olsner> fizzie: not that there's anything wrong with repeating the last component of an acronym, IMO
20:36:06 <calamari> elliott_: I think there's one in ralf browns interrupt list
20:36:20 <ais523> olsner: IMO /opinion/?
20:36:42 <Vorpal> fizzie, also "cd disc", not as common in English as "CD-skiva" is in Swedish though
20:37:07 <elliott_> just need to tweak ah to be right now then
20:37:11 <ais523> I think it depends on the acronym in question
20:37:28 <ais523> to be precise, repeating the last component makes sense if it's a general class of nouns and you use the acronym to say which in particular
20:37:34 <ais523> as in, "the sort of machine that is an ATM"
20:37:57 <Vorpal> ais523, Asynchronous Transfer Mode machine?
20:38:07 <olsner> elliott_: Esc was 1 in the table I found
20:38:17 <ais523> Vorpal: Automatic Teller Machine machine
20:38:38 <ais523> one of the most common repeated-last-component acronyms
20:38:39 <elliott_> Science has yet to device an Automatic Penn Machine.
20:38:48 <olsner> what's "teller"? a person who works in a bank?
20:39:07 <elliott_> impomatic: How are you going to pack word names?
20:39:17 <Vorpal> ais523, http://www.google.com/search?q=ATM <-- hit 1 is Asynchronous Transfer Mode, hit 2 is Automatic Teller Machine. And no I'm not logged in
20:39:34 <ais523> Vorpal: it customizes results even if you aren't logged in
20:39:38 <ais523> in fact, I fear it does even if you block cookies
20:39:46 <Vorpal> ais523, hm how. I have dynamic ip
20:39:58 <ais523> perhaps based on other people in the same IP range
20:40:01 <ais523> or maybe just geolocation
20:40:05 <Vorpal> ais523, also I block click tracking
20:40:08 <impomatic> Elliott_: just as an ASCII string. I'm aiming for a small forth, not minimal at any expense :-)
20:40:19 <Vorpal> ais523, geolocation fails badly for me. My ISP use a country-wide pool
20:40:24 <elliott_> impomatic: hey, but even colorForth packs names! and it takes up whole _kilobytes_!
20:40:24 <Vorpal> and it rotates quickly
20:40:29 <olsner> en... automatisk banktjänstemannamaskin
20:40:42 <ais523> the UK mostly geolocates quite well
20:40:42 <olsner> glad we picked bankomat instead
20:40:53 <ais523> Wikipedia knows I live in Birmingham, for instance, using its geolocation database
20:41:03 <ais523> (although that's hardly a secret given my email address)
20:41:05 <Vorpal> olsner, damn you beat me to it
20:41:16 <fizzie> elliott_: ESC is 1 in the "set 1" scancodes that the keyboard controller uses by default, I think.
20:41:22 <elliott_> olsner: "banktjänstemannamaskin", amazing
20:41:28 <fizzie> It's 110 in the "set 2" AT keyboard scancodes that go over the wire.
20:41:37 <calamari> elliott_: http://www.ctyme.com/intr/rb-0045.htm
20:41:43 <Vorpal> elliott_, it backtranslates to "bank official machine"
20:41:56 <olsner> and sounds like a machine that makes tellers
20:41:58 <zzo38> I expect a small system would fit in conventional memory even without packed names, although still make difficult to make the kernel in 510 bytes
20:42:01 <Vorpal> elliott_, and what is amazing about it
20:42:36 <elliott_> the maximum scancode I can store is 41
20:42:53 <elliott_> I need to filter out most of this punctuation
20:42:58 <olsner> would be fun if you could just strangle the staff when they're being difficult and the automatic teller machine would just make a new one
20:43:01 <elliott_> maybe packing ascii would be better :/
20:43:49 <calamari> just use morse code for data entry ;)
20:44:21 <elliott_> calamari: why, it's only logical.
20:44:22 <olsner> calamari: still has to be converted for display...
20:44:30 <fizzie> Even if you start from Q, un-remapped 5 bits only goes up to 2fh, so you'd miss the B, N and M keys from the alphabet set.
20:44:43 <elliott_> Great, "or ebx, al" isn't okay :P
20:44:52 <fizzie> What would that even mean?
20:44:55 <elliott_> Thankfully I can or bl instead.
20:45:37 <elliott_> fizzie: Alas, 6 bits of ascii-64 does not include !.
20:45:56 <elliott_> It *does* include the uppercase alphabet, however.
20:46:00 <elliott_> Which I could possibly reuse...
20:46:05 <elliott_> Vorpal: I'm packing it into 32 bits.
20:46:15 <Vorpal> elliott_, the whole program
20:46:17 <zzo38> elliott_: Which is why I suggested starting at 0x20 instead of 0x40
20:46:40 <fizzie> elliott_: If you're going to use 6 bits, you could consider ascii minus 32.
20:46:42 <Vorpal> lambdabot, no I meant ‽
20:46:56 <zzo38> That way you have ! as well as other punctuation common in Forth. (The punctuation {|}~ is not as common in Forth)
20:47:03 <fizzie> elliott_: That one has everything that's printable in ascii except the backtick, a .. z, and {|}~.
20:47:13 <fizzie> Oh, zzo38 said it first.
20:47:23 <zzo38> And you can make Forth without doing ` or lowercase
20:47:25 <elliott_> It'll be uppercase-only, but who cares.
20:47:32 <elliott_> fizzie: What would five bits of ascii minus 32 give me?
20:47:48 <fizzie> elliott_: Lots of punctuation, the digits, no alphabetic characters or @. :p
20:47:57 <olsner> btw, when I suggested ascii-64, that was with 5 bits (giving you @, A-Z, [\]^ and _)
20:48:32 <zzo38> With 6 bits ASCII-32 you do get @, A-Z, [\]^, _
20:48:37 <Vorpal> elliott_, it is a 5 bit encoding
20:48:48 <zzo38> And I would expect using all of these in Forth.
20:49:09 <olsner> zzo38: yes, but using a whole extra bit too
20:49:10 <elliott_> fizzie: I don't need the digits :P
20:49:11 <fizzie> Vorpal: VGA doesn't have a Baudot font in ROM, though.
20:49:26 <zzo38> Vorpal: I know Baudot, too. Although you need a letter/figure shift..... otherwise you have only letters and no number/punctuation
20:49:42 <zzo38> elliott_: Yes, I would have put A immediately after 9 if I designed it
20:50:08 <elliott_> I'd have lowercase first and uppercase last, since uppercase is less common.
20:50:10 <Vorpal> elliott_, compact the range
20:50:13 <elliott_> To allow for storing fewer bits.
20:50:27 <Vorpal> elliott_, you can shift the high range down
20:50:32 <Vorpal> no need to use the same range
20:50:41 <elliott_> Vorpal: Show me the low-byte x86 for it :P
20:50:58 <Vorpal> elliott_, well no. I don't do much x86 asm
20:50:59 <zzo38> I wouldn't have lowercase first, still I would order things very differently and possibly have a few differences in control characters and stuff too
20:51:07 <Vorpal> elliott_, but I would start with using rax. Which you can't
20:51:15 <fizzie> elliott_: In any case, to summarize, with ascii you'd probably want either 6 bits starting from 32 (just about everything except lowercase stuff), or 5 bits starting from 64 (the alphabet and @[\]^_, but no other punctuation).
20:51:21 <elliott_> 64-bit code would be huge, even ignoring the long mode dance.
20:51:35 <elliott_> fizzie: I could always make ^ write to memory instead.
20:51:43 <fizzie> Yes, it's !y enough for that.
20:51:54 <zzo38> Also, like ASCII really is, still have ('0'&0x0F)==0
20:52:10 <Vorpal> zzo38, why is that good?
20:52:29 <olsner> 32-bit code in 64-bit mode is pretty much the same size as in 32-bit mode really (most of the stuff has the exact same encoding and meaning)
20:52:35 <elliott_> *Seems* like my word-reader is under 47 bytes.
20:52:43 <olsner> (but it definitely wouldn't be shorter)
20:52:46 <zzo38> Vorpal: Isn't it obvious why the low four bits of the code for '0' should be all zero?
20:52:58 <zzo38> Especially if 'A' would come immediately after '9'?
20:53:13 <Vorpal> zzo38, I don't do much crazy asm golfing
20:53:26 <calamari> lol complaining about ascii.. hen use ebcdic
20:53:27 <Vorpal> zzo38, I code in stuff that generally uses utf-8 anyway
20:53:28 <zzo38> Although if I did that, perhaps '0' would then come at 0x40 instead of 0x30
20:53:51 <Vorpal> zzo38, no it isn't obvious
20:53:51 <ais523> calamari: the "bcd" in "ebcdic" stands for "binary-coded-decimal"
20:54:11 <zzo38> Vorpal: It isn't because of golfing, it is just logical!
20:54:26 <calamari> just saying, there are much worse than ascii :)
20:54:41 <zzo38> calamari: Yes, I agree
20:55:23 <Vorpal> ah I'm not the only one who doesn't get it
20:55:34 <zzo38> Vorpal: Now you have bits 0b01000000=='0' and 0x4A=='A' it makes sense isn't it?
20:55:49 <elliott_> so that the alphabet looks right in hex
20:56:05 <elliott_> that's a pretty weak justification IMO
20:56:20 <zzo38> And since it is 0x40 instead of 0x30 that means the bits are aligned better for that purpose, too.
20:56:30 <zzo38> Including if you use the entire alphabet.
20:57:05 <zzo38> elliott_: Then 0x50=='G' and so on...
20:57:33 <Vorpal> zzo38, how would 0x50=='G' make sense?
20:57:45 <zzo38> Vorpal: Because it comes after 'F'.
20:57:54 <Vorpal> zzo38, yes I get how it happens
20:57:58 <Vorpal> zzo38, but shouldn't it be
20:58:20 <Vorpal> zzo38, that would be a LOT more logical :P
20:58:44 <zzo38> There is no 0x4G is not a hex number, you need to do it in binary, you use hex just to make shorter typing, instead of binary.
20:59:04 <zzo38> In case of computer with 9 bits in one byte, you should use octal instead.
20:59:11 <fizzie> Vorpal: You said "0x", though.
20:59:39 <Vorpal> elliott_, not in an alternate reality!
20:59:46 <Vorpal> where base 23 is called mex
21:00:09 <Vorpal> elliott_, I prefer 16#F00
21:00:55 <Vorpal> elliott_, awesome idea
21:01:07 <Vorpal> elliott_, actually wouldn't that be base 69?
21:01:17 <elliott_> Oh man, look at the comedy spewing out of Vorpal's mouth.
21:01:24 <zzo38> Sometimes they use things like $42_{\rm ten}$ for forty-two, and so on
21:01:24 -!- Lymia has quit (Quit: ==(>^w^)> ==(> >.<)>).
21:01:30 <Vorpal> elliott_, no I'm not claiming it is good
21:01:46 <Vorpal> zzo38, yes that happens in literature
21:02:21 <zzo38> fizzie: That is used too, sometimes.
21:02:56 <zzo38> elliott_: Which is why, to using the words...
21:03:22 <elliott_> But "ten" is just as ambiguous as 10.
21:03:34 <zzo38> If you mean the hexadecimal number 0x10 then you write "tex" not "ten"
21:03:36 <fizzie> Wikipedia's hex table uses "hex", "dec" and "oct" subscripts.
21:03:55 <Vorpal> zzo38, no then people would think you mean /usr/bin/tex
21:04:04 <zzo38> And "hundrek" for 0x100
21:04:23 <fizzie> So is 0o100 a "hundo"?
21:04:31 -!- elliott_ has quit (Remote host closed the connection).
21:04:34 -!- elliott has joined.
21:04:40 <Vorpal> fizzie, woudln't it be hundro?
21:04:42 <zzo38> I do not know if anyone made up words for saying octal numbers
21:04:48 <zzo38> But there is some for hexadecimal numbers.
21:05:01 <Vorpal> olsner, could be a dog too
21:05:03 <fizzie> Vorpal: I'unno, "hundo" sounds like a dialecty "hundred".
21:05:55 <zzo38> Perhaps today you can make up octal speeching. So, we can have decimal, hexadecimal, octal speeching, now.
21:05:56 <fizzie> Urban dictionary "hundo": "An increment of 100 dollars. Normally used when reffering to spending habits."
21:06:12 <fizzie> Also two other entries much like that.
21:06:18 <Vorpal> fizzie, so only incrementally used?
21:06:55 <fizzie> No, more generally too. Apparently "100%" can also be said "hundo".
21:07:00 <Vorpal> zzo38, what about septal and niertal (or whatever you call it?)
21:07:47 <zzo38> Vorpal: You can do that if you want, but probably is not quite commonly that it could be used enough.
21:08:12 <zzo38> While hexadecimal would be used secondly to decimal, and octal maybe thirdly?
21:08:37 <Vorpal> I have written 0b a lot more than 0 in C
21:08:43 <Vorpal> (that was C where I knew the compiler
21:08:48 <calamari> had to use octal the other day due to Java
21:09:17 <Vorpal> Specifically I used 0b prefix in C when dealing with IO register masking on some embedded systems
21:09:27 <zzo38> Yes octal is rare. However, a lot of TeX: The Program uses a lot of octal numbers, although I think it would be clearer if hex is used instead (WEB supports both... Pascal supports neither...)
21:09:35 <calamari> yeah I couldn't figure out how to exter a character in hex, so I used '\ooo'
21:10:13 <zzo38> I have, however, occasionally found octal useful. I find octal clearer than hexadecimal when entering the bit patterns for a seven-segment display, is one thing.
21:11:24 <zzo38> Do you understand?
21:12:21 <zzo38> calamari: Can you use \x like you can in C?
21:12:40 <calamari> I could do that in a string but it didn't like it in the char constant
21:13:01 -!- mycrofti1 has joined.
21:13:16 -!- nooga__ has joined.
21:14:02 -!- comex_ has quit (Ping timeout: 250 seconds).
21:14:02 -!- mycroftiv has quit (Ping timeout: 250 seconds).
21:14:03 -!- nooga has quit (Ping timeout: 250 seconds).
21:14:39 <olsner> couldn't you just have done char foo = 0xff; if it was a constant?
21:14:45 <olsner> or (char)0xff or whatever
21:14:52 <fizzie> Vorpal: Septenary and nonary.
21:15:01 <zzo38> olsner: I would think that too, at least in C, but maybe in Java it doesn't do?
21:15:03 -!- comex has joined.
21:15:48 <zzo38> However, it is possible to write a Java program in C, if you have a C compiler to target JVM
21:16:05 <olsner> wouldn't surprise me if conversions to/from char are needlessly limited
21:16:12 <fizzie> You can cast a number into char, but it does need the cast.
21:16:31 <fizzie> And character literals indeed only do \nnn with octal, or \uxxxx with hex.
21:16:43 <elliott> `addquote <zzo38> However, it is possible to write a Java program in C, if you have a C compiler to target JVM
21:16:46 <olsner> one funny thing is that Java allows \uxxxx escapes in code too - useful if someone uses µ in an identifier
21:17:18 <olsner> (blackberry has stuff like that in their api)
21:18:08 <zzo38> In my programs, if I want Greek letter in an identifier, I will do something like, @f mu TeX and now it will print in Greek, even though I did not type it in Greek.
21:19:23 <zzo38> And even Hebrew, a bit.....
21:20:07 <Vorpal> <fizzie> Vorpal: Septenary and nonary. <-- I prefer ninerary then
21:21:23 <oerjan> HOW DARE YOU MANGLE ALREADY MANGLED LATIN
21:21:36 <Vorpal> oerjan, because niner is an awesome spelling
21:21:45 <Vorpal> oerjan, http://en.wikipedia.org/wiki/ICAO_spelling_alphabet
21:21:58 -!- Patashu has joined.
21:31:23 -!- impomatic has quit (Quit: ChatZilla 0.9.86 [Firefox 3.5.17/20110121150729]).
21:32:03 -!- zeotrope has quit (Ping timeout: 276 seconds).
21:33:06 <zzo38> Just because I can, I have made up some conditional functions in TeX without using any of the primitive conditional commands at all.
21:34:42 <zzo38> Of course it is not how you would do in actual documents, it is just to show how it can be done.
21:35:10 <zzo38> Now see if you can figure out any of them by yourself, too.
21:44:13 -!- augur has joined.
21:50:07 * oerjan swats FireFly -----###
21:50:21 <oerjan> YOUR ESOLANGS WEBSITE IS BROKEN
21:50:32 -!- Lymia has joined.
21:50:32 <oerjan> elliott: NO THIS TIME I HAD EXCELLENT CAUSE
21:50:32 -!- Lymia has quit (Changing host).
21:50:32 -!- Lymia has joined.
21:51:20 <elliott> Can you make it with macros? --Zzo38 03:29, 11 February 2011 (UTC)
21:51:20 <elliott> We couldn't implement the functionality of macros in this first version due to time constraints. But we thought about it and we suggested to allocate a special buffer in the infinite tape to host the macros. May be that will be for future versions of the ClearBF Compiler. Yasser 21:08, 03 March 2011 (UTC)
21:51:25 <elliott> MACROS HAVE NOTHING TO DO WITH THE TAPE
21:53:09 <ais523> WHAT IF YOU WANT TO CHANGE THEM AT RUNTIME
21:53:43 <elliott> ais523: WELL IT ALL MAKES SENSE NOW
21:54:13 <FireFly> * oerjan swats FireFly -----###
21:54:36 <ais523> I like the way oerjan manages to do that more or less completely out of the blue every now and then
21:55:11 <oerjan> http://firefly.nu/diverse/esolangs/migol09/bf.mgl in particular
21:56:20 <oerjan> also http://firefly.nu/diverse/esolangs/Migol09/miGoL.mgl and for that matter the entire diverse/ directory afaict
21:56:35 <elliott> Phantom_Hoover: "Two ladders rest in opposite sides of a room. One is 10 feet long, while the other is 12 feet long."
21:56:38 <FireFly> Yep, cause it's actually Diverse, but it used to be on a shitty windows server
21:57:00 -!- zzo38 has quit (Quit: HA HA HA).
21:57:00 -!- Wamanuz has joined.
21:57:05 <FireFly> now the former should work
21:57:18 <FireFly> and the latter, if you change the casing of the Migol09 directory
21:57:53 <elliott> Phantom_Hoover: "I tried it again. This time I measured the apparent parallax movement of my thumb to an object twelve feet away. Extrapolating from that ratio in arcseconds, I calculated that the opposite wall is 0 AU away. X = 0 AU."
21:57:55 -!- Wamanuz5 has quit (Read error: Operation timed out).
21:58:28 <elliott> Phantom_Hoover: http://i.imgur.com/QXqoj.png
22:01:19 <ais523> elliott: you have to admit, that value of 0 AU is probably accurate to quite a lot of decimal places
22:03:20 <ais523> hmm, BF Joust is stuck in my head
22:03:23 <ais523> much the same way songs can be
22:03:38 <ais523> even though I'm mostly done with waterfall3, as I can't think of much of a way to improve it
22:03:47 <Phantom_Hoover> http://www.reddit.com/r/math/comments/fy9gq/rmath_im_learning_my_first_language_this_semester/
22:03:48 <ais523> special-casing short tapes doesn't help as it weakens the program on everything else
22:04:14 <ais523> and generally speaking, given a sufficiently long tape it beats more or less every other strategy
22:04:19 <ais523> except for lead_acetate_philip on one polarity
22:04:29 <ais523> and some draws against shudders
22:04:34 <Ilari> Pfft... Solving 4th degree equation numerically. 4th degree equations can be solved symbolically as well. :-)
22:04:47 <elliott> "I thought FORTRAN was *the* science language, but I'm confused why it can't handle as big as 3^10."
22:05:01 <ais523> 59049 should be in range, surely?
22:05:16 <ais523> and yes, I know that number off by heart because of Malbolge
22:05:24 -!- TLUL has joined.
22:05:26 <ais523> although it comes up in TriINTERCAL too
22:05:28 <oerjan> FireFly: ok i fixed the links on EsoInterpreters, Migol and your user page, unfortunately i have no idea how to search for the rest efficiently if there are any
22:05:32 <elliott> I like how they "calculated" the maximum FORTRAN integer to be 2^31 - 1.
22:05:58 <elliott> "Haskell handles the so-called "bignums" transparently and its syntax is probably as close to mathematics as you can get."
22:06:04 <elliott> HASKELL'S SYNTAX IS NOT MATHEMATICS THAT DOESN'T EVEN MEAN ANYTHING
22:06:17 <elliott> ais523: Hey, uh, how do you enable core dumps.
22:06:18 <FireFly> oerjan, oh, thanks. No more that I know of, at least
22:06:29 <ais523> elliott: in bash, ulimit -Sc 1024000 or whatever number you like
22:06:39 <ais523> it's in kilobytes, IIRC
22:06:45 <elliott> <mark_weaver> enable core dumps and then look at the backtrace in gdb
22:06:48 <elliott> Backtrace of... the core dum,p?
22:06:51 <ais523> using -Scnot just -c lets you change your mind leater
22:06:54 <elliott> IUNNO WHAT I'M MEANT TAH DO
22:07:01 <ais523> Phantom_Hoover: it allows you to increase the limit
22:07:08 <ais523> normally, ulimit limits are set hard and unchangeable
22:07:13 <ais523> so you can use them for security
22:07:17 <ais523> and core dumps have backtraces
22:07:27 <ais523> that you can see by opening them in gdb and running the backtrace command
22:09:24 <Phantom_Hoover> The Casio FX-83ES is more sophisticated than the 85ES.
22:09:45 <elliott> I like how core dumps go to core
22:10:03 <elliott> in this modern gnu world I would expect mcmap.2897598734545.elf-dump or something else similarly ENTERPRISEY
22:10:36 <elliott> i'll let ais523 explain the whole
22:10:45 <fizzie> There's a kernel config thing where you can write a pattern for the
22:11:00 <elliott> I thought you were SLEEPING, anyway.
22:11:11 <fizzie> It can have all kinds of %x expandables too.
22:11:39 <fizzie> Pids, uids, gids, name of executable, hostname, so on.
22:11:45 <Phantom_Hoover> Ah, the classic "This calculator only does FRACTIONS AAAH I HATE IT"
22:11:48 <fizzie> So that you can be the enterprise.
22:12:53 <elliott> Be the enterprise you wish to see in the world.
22:14:37 <fizzie> echo '%h.%e.%t.%p.%u:%g.%s.core' > /proc/sys/kernel/core_pattern
22:14:59 <fizzie> Though the expanded name might be longer than the limit of 64 chars.
22:15:09 <Phantom_Hoover> I like the fact that Amazon includes a scientific calculator as an "office product".
22:16:09 -!- MigoMipo has quit (Remote host closed the connection).
22:29:17 -!- Phantom_Hoover has quit (Remote host closed the connection).
22:31:05 <elliott> pikhq_: Hey look, it's like that stock market thing! http://en.wikipedia.org/wiki/MERS
22:34:26 -!- EgoBot has quit (Ping timeout: 276 seconds).
22:35:28 -!- HackEgo has joined.
22:35:29 -!- EgoBot has joined.
22:39:41 -!- pikhq has joined.
22:39:48 -!- pikhq_ has quit (Ping timeout: 260 seconds).
22:53:41 -!- azaq23 has quit (Quit: Leaving.).
22:54:07 -!- azaq23 has joined.
22:56:36 -!- azaq231 has joined.
22:57:13 -!- azaq231 has quit (Client Quit).
22:57:40 -!- azaq231 has joined.
22:58:26 -!- azaq23 has quit (Ping timeout: 250 seconds).
23:02:18 <elliott> "HarperCollins says US libraries can lend its ebooks only 26 times as print books have to be replaced after that"
23:04:02 <copumpkin> I had no idea print books were replaced
23:04:18 <elliott> copumpkin: "Basically this asshat pulled that number out of his butt. I used to work in at a library and books easily last way more than 26 borrowings (e.g., bestellers get borrowed over 20 times in just their first year). A library couldn't economically survive if books didn't last more than 26 borrowings, which btw is why libraries only stock hardback versions."
23:04:48 <elliott> copumpkin: Not only is it yet another water-is-not-wet argument for copyright (are there any others?), but it's based on false premises too.
23:04:56 <elliott> Well, not that I don't expect HarperCollins really believes that shit.
23:05:48 -!- ais523 has quit (Remote host closed the connection).
23:21:17 -!- zzo38 has joined.
23:22:21 <cheater-> show me a water-is-not-wet argument based on correct premises.
23:22:30 <zzo38> I tried to make up a code table to see how I might have done it if I had made a code instead of using ASCII and it is somewhat like that - http://sprunge.us/cOaE
23:23:26 <zzo38> How can you decide how many times you lend the books based on whether or not water is wet?
23:24:05 <zzo38> cheater-: Do you think it is a bit better job than ASCII?
23:24:25 <zzo38> (It is not even complete yet, however)
23:24:43 <cheater-> now here's an additional problem
23:25:02 <cheater-> make sure that it withstands adding random noise to the signal
23:25:13 <cheater-> or rather, that it's more resilient than ascii
23:25:51 <zzo38> To do that you need to add an extra error correction, it is not part of the character coding. That would be a separate thing.
23:26:00 <cheater-> this means for one thing being able to easily notice erroneous characters
23:26:09 <cheater-> no, no need for error correction, just sanity correction
23:26:28 <cheater-> e.g. words usually don't have capital letters in the middle unless they're all caps
23:26:51 <cheater-> or the fact that words are made out of letters or that in a computer language certain characters come in pairs
23:27:07 <zzo38> Hopefully you should be able to tell, but some of redundancy in language is also depending whether or not you write in English.
23:27:54 <cheater-> i think the most basic thing is, looking at those characters that come in pairs, that flipping just several bits doesn't give you another character of that type
23:28:04 <cheater-> and certainly that the distance to the dual character is maximized
23:28:23 <cheater-> so you'd probably want [ to be ] with all bits flipped or something like that
23:28:25 <zzo38> Maybe you can try making a simulation to see what will happene
23:28:56 <zzo38> cheater-: Doing some of those things, however, might mess up other aspects of the code
23:33:42 <zzo38> Like, if [ is ] with all bits flipped, then it might disrupt other patterns in this code.
23:34:27 <zzo38> I did not design it for noise correction.
23:40:02 <Patashu> onew of my programs is on the hill still
23:43:01 <elliott> Patashu: it's been like that for years
23:43:31 <Patashu> unless you think bf joust has depth?
23:43:35 <elliott> BF Joust has depth, certainly.
23:43:39 <elliott> You are not aware of the new developments?
23:43:45 <elliott> Incredibly advanced hybrid-defence programs have been written.
23:43:48 <elliott> That is, part defence, part attack.
23:43:48 <Patashu> but, stuff like fyb and core wars
23:43:53 <EgoBot> Use: !bfjoust <program name> <program> . Scoreboard, programs, and a description of score calculation are at http://codu.org/eso/bfjoust/
23:43:57 <elliott> Patashu: FYB has no depth; it is fundamentally broken.
23:44:02 <elliott> Was demonstrated on this channel a while back.
23:44:10 <elliott> Gregor admitted it, at least.
23:44:20 <elliott> I firmly believe that BF Joust has the same depth as Core War.
23:44:31 <elliott> Patashu: Here's some good reading material.
23:44:38 <elliott> http://codu.org/eso/bfjoust/in_egobot/quintopia_space_elevator.bfjoust
23:44:38 <elliott> http://codu.org/eso/bfjoust/in_egobot/ais523_defend9_75.bfjoust
23:44:44 <elliott> http://codu.org/eso/bfjoust/in_egobot/ais523_waterfall3.bfjoust
23:45:11 <elliott> Also the completely insane http://codu.org/eso/bfjoust/in_egobot/Gregor_furry_furry_strapon_pegging_girls.bfjoust, which contains its own generator program.
23:47:42 <elliott> Patashu: Note that all the ones linked are partially computer-generated (well, dunno about waterfall3).
23:48:11 <elliott> space_elevator because it's basically a bunch of mechanical repetitions of the same basic strategy, defend9.75 because it's incredibly complex, and FFSPG because... I don't know, Gregor is a madman.
23:49:38 <oerjan> AT LEAST BFJOUST KEEPS HIM OFF THE STREETS
23:51:07 <zzo38> I noticed someone edited my Wikipedia userpage by adding the text "Why the heck did MFGGer link to page of this?" with the summary "This page has DRM in it" and then changed it back in five minutes. What kind of stuff is that?
23:52:01 <zzo38> If by "has DRM in it", you mean that the letters "DRM" appear like that, then it is correct it probably does.
23:53:39 <zzo38> Do you like to use redundant userboxes that are redundant?
23:59:16 -!- FireFly has quit (Quit: swatted to death).
23:59:18 <Sgeo> I once made a userbox that was supposed to be randomly colored
23:59:24 <Sgeo> Then #rand was eliminated