00:05:45 fizzie: Well, SIFR seems to do absolutely nothing here. 00:06:10 Ah well. Means that things won't find a way to fuck with my font setting. 00:06:41 "Actionscript inside of each Flash file then draws that text in your chosen typeface at a 6 point size and scales it up until it fits snugly inside the Flash movie. " <-- Why not do it in INTERCAL instead. Really it would make perfect sense for an esolang. But for this?... 00:07:00 (a nice serif font with thin lines; rather nice to read at a decent DPI) 00:07:00 sIFR does nothing with noscript, either, but still. 00:07:15 I don't have noscript. Just AdBlock... 00:07:26 I use both noscript and adblock plus 00:07:35 and have no flash or java plugins 00:07:39 Noscript doesn't work with Conkeror. 00:08:00 "sIFR runs fine under other extensions like AdBlock"; that shouldn't be related. 00:08:11 pikhq, mhm 00:08:16 and why do you use that? 00:08:29 Mouseless browsing. 00:08:33 ok 00:09:04 And it's XULrunner based, so just about everything works anyways. 00:09:07 Conkeror is emacs for Firefox. 00:09:17 Emacs for Gecko, rather. 00:09:20 I am not sure why that would elicit a "why" from AnMaster. 00:09:29 It's not been a Firefox extension for a couple of years. 00:09:30 pikhq: It's rather more Firefox than just Gecko 00:09:36 * SimonRC <3 the future. 00:09:40 Anyway; sIFR is not the stupidest thing I've seen (I mean, it's not like placing body text in an image, for example; and they strongly advise against using it for body text, anyway), just silley. 00:09:44 It's XULrunner. 00:09:54 fizzie, agreed 00:09:58 Which is Gecko with the ability to load arbitrary XUL... 00:10:06 We do things by pulling little computer programs across the world 00:10:32 Soon we'll have intelligent agents running around! 00:11:09 SimonRC: wut 00:11:10 firefox use xulrunner... 00:11:28 ehirdghost, duh. read what he said above 00:11:40 23:09 SimonRC <3 the future. 00:11:41 23:10 SimonRC: We do things by pulling little computer programs across the world 00:11:42 I repeat: wt. 00:11:44 wut 00:11:46 it is crystal clear... 00:11:47 duh 00:11:49 night btw 00:11:53 Er, it is? 00:12:13 yeah, we live in the future 00:12:23 I am fairly sure we live in the present, SimonRC. 00:12:30 Fairly sure indeed. 00:12:38 ehirdghost, no. You are a ghost. You don't live 00:12:43 you lived in the past. 00:12:47 s/the future/The Future/ 00:12:50 AnMaster: By the way, we've been pulling little computer programs across the world for stuff since at *least* the invention of Javascript. ;) 00:12:59 pikhq, indeed. 00:13:07 Probably longer, if you count, say, UUCP. 00:13:10 AnMaster: I metalive. 00:13:18 I am not certain, but I think I was thinking this before Munroe made a comic about it 00:13:25 I never metalive I didn't like. 00:13:34 ehirdghost, I chose to ignore that pun and instead point out that is not valid grammar. 00:13:56 I will now metakill you. 00:14:28 also why is "I never meta I " supposed to be funny? It is a rather lame pun IMO. 00:15:39 Incidentally, how do the licensing terms go; if I have a copy of OS X, can I use some of the bundled fonts on a different computar? (I'm not sure I want to, just hypothetically speaking.) 00:15:46 computer* 00:16:04 Oh, sorry, I meant CANTOR-UPPER. 00:16:05 fizzie: I'm not sure. There's no DRM or anything; I don't think anyone cares. 00:16:28 fizzie, that made no sense. 00:16:44 it is noway near "computer" when you pronounce it 00:16:44 fizzie: I don't think there is anything in the EULA or whatnot. 00:17:00 fizzie: So it'd just be standard copyright law; if you're using it on another computer you own, fair use, probably. 00:17:06 fizzie: Assuming the computer in question supports TrueType or OpenType, yeah. 00:17:20 pikhq: he means can as in legally 00:17:22 "is it allowed" 00:17:25 ehirdghost, what if you use it in a document and that cause it to be bundled... 00:17:35 Oh, all fonts let you bundle them in PDFs and whatnot, I think/ 00:17:41 ehirdghost, say a document with the entire UTF-8 chart :D 00:17:45 Most any sane one, at least. 00:17:53 so end user can extract it all 00:17:57 Yay; if that's the truedness, there's some form of common sense left. 00:18:01 would that be legal? 00:18:03 fizzie: I've done it 00:18:08 I just used a converter of .dfont -> .ttf, iirc 00:18:11 and it worked fine 00:18:13 Called "fondu" 00:18:22 No DRM or anything; one command line invocation and an upload 00:18:27 BTW, fun fact. Typefaces are not subject to copyright. 00:18:27 Yes, well, I have it from reputable sources that you also download QuickBASIC copies of dubious legality. 00:18:34 pikhq: Really? 00:18:41 ehirdghost: sounds a bit cheesy 00:18:43 pikhq, what are they subject to then? And is that US only? 00:18:46 pikhq: Then why was Arial ever created? 00:18:50 The *computer code* describing them can be. 00:18:51 oerjan, augh 00:19:07 ehirdghost: Novel and non-obvious designs can be patented. 00:19:16 Helvetica is pretty "obvious"... 00:19:22 Yeah. 00:20:34 As a side note, does anyone actually use Zapfino? 00:20:36 opentype contains parts under patent 00:20:37 It's utterly unreadabl 00:20:38 e 00:21:03 Other countries have typeface copyright. 00:21:10 Uh, I'm not sure "file" is correct here: "Monaco.dfont: MS Windows icon resource" 00:21:14 AnMaster: Yeah; mostly the hinting algorithms, IIRC. 00:21:16 pikhq: Does Europe? 00:21:19 fizzie: :-D 00:21:24 pikhq, others == non-US or? 00:21:24 Monaco is love. 00:21:32 Others == non-US. 00:21:44 "Unfortunately, just before the project was completed, Siegel wrote a letter to Zapf, saying that his girlfriend had left him, and that he had lost all interest in anything. Thus Siegel abandoned the project and started a new life, working on bringing color to Macintosh computers, and later becoming an Internet design expert. " 00:21:45 pikhq, who cares about US? 00:21:46 XDD 00:21:48 I don't 00:22:00 Sorry; hard to get out of US-centric phrasing sometimes. 00:22:00 AnMaster: Yes, I know you take pride in your rabid hate of the US (because IT'S POPULAR or something). 00:22:06 Most of us don't. 00:22:31 Also, Zapfino? Unreadable? 00:22:42 Ehm, yes. 00:23:04 No, it looks like calligraphic text. Rather readable, though probably not the best for long works. 00:23:17 well, it's hard for me to read… 00:23:35 OS X specific? 00:23:41 I can't find it here 00:23:47 AnMaster: OS X builtin font. 00:23:53 screenshot? 00:24:00 entirely separate from os x 00:24:04 but bundled with OS X 00:24:14 saying "OS X specific" is Wrong; I don't know of any OS X only fonts. 00:24:21 http://upload.wikimedia.org/wikipedia/commons/c/c7/Zapfino.svg 00:24:23 maybe Geneva 00:24:39 that's nice 00:24:48 readable for being a calligraphic one 00:24:49 AnMaster: Yes, but try reading text in it 00:25:03 ehirdghost, It would be good for logos or such 00:25:08 not for long works no 00:25:25 ehirdghost: Calligraphic fonts aren't meant for long works... 00:25:28 but why so many variants there 00:25:31 Of course. 00:25:35 But even for titles. 00:25:43 AnMaster: it has over 18,000 ligs 00:25:45 and how do you select which one? 00:25:48 ehirdghost, wow.... 00:25:50 and manually 00:25:52 afaik 00:25:56 XeTeX. ;) 00:26:04 er wait 00:26:06 1,400 ligs 00:26:09 still insane 00:26:12 ah ok 00:26:26 I'ma download MacTeX 00:26:34 Does it still use that ugly iInstaller crap, I wonder. 00:27:34 Oh lord, FTP. 00:27:39 Why, why FTP. 00:27:55 Ah, an HTTP mirror. 00:28:01 700KB/sec. Most excellent. 00:28:03 Thank you Germany. 00:28:16 ehirdghost, what is wrong with ftp.... 00:28:21 Everything 00:28:38 well, passive ftp works fine in my experience 00:28:57 Aside from it's connecting back to the initiator of the connection for the transfer link, it's a decent protocol. 00:29:03 Perhaps a bit overengineered, though. 00:29:05 pikhq, indeed 00:29:18 pikhq, parallel transfers are nice though 00:29:20 It's completely insecure in every way, and for passive basic file downloads it has 0 advantages compared to HTTP 00:29:22 True, true. 00:29:26 And several disadvantages 00:29:30 Such as? 00:29:40 ehirdghost, apart from connect back? 00:29:55 pikhq: have you ever _used_ FTP/ 00:29:57 ? 00:30:03 ehirdghost, I have 00:30:08 for both up and download 00:30:24 Yes, I have. 00:30:27 works well, apart from the insecure bit and separate data channel. 00:30:29 Also, the server software required for it is bit too large... 00:30:33 -!- Corun has joined. 00:30:45 Ilari, err. ISS is a bit insecure. Lets drop http 00:30:56 Ilari, there are small FTP servers. 00:30:57 ... 00:31:00 ftp(1) is a rather nice program. 00:31:24 pikhq, indeed. I prefer sftp though mostly. For security 00:31:28 Ilari: Uh, ftpd probably comes in under a megabyte. 00:31:32 AnMaster: Well, yeah. 00:32:08 AnMaster: Implementing HTTP server is probably smaller task than implementing FTP server... 00:32:13 At least with FTP you can, on host C, transfer data between servers A and B without things going through C. 00:32:29 Ilari, "probably"? 00:32:29 It is horribly complicated due to historical raisins, though. 00:32:29 -!- Corun has quit (Client Quit). 00:32:41 fizzie, I managed that with scp iirc 00:33:04 I find it very unlikely that scp can do it. 00:33:12 Sftp can, scp can't. 00:33:15 fizzie, maybe sftp then 00:33:16 ok 00:33:43 How do you do it with sftp, then? 00:34:15 sftp foo:bar baz: 00:34:32 * ehirdghost boo 00:34:36 Sorry, had to fill my ghost quota. 00:34:50 ehirdghost, so what was the issue you had with ftp? 00:34:56 AnMaster: Whooooo 00:35:00 WHOAOOAOOOOOO 00:35:02 you never answered what apart from separate control channel 00:35:12 s/sftp/scp/; sorry. sftp is a ftp-style program, and current scp programs are just sftp frontends. 00:35:13 pikhq: That writes to local file "baz:" here. 00:35:20 Really?!? 00:35:26 "and current scp programs are just sftp frontends." 00:35:28 Wait what? 00:35:31 no 00:35:38 Sorry. That was an epic thinko. 00:35:45 they are not 00:35:53 THAT WAS DUMB. 00:35:59 :| 00:36:19 ehirdghost, still. I'm waiting for an answer 00:36:19 I enjoy infuriating you by withholding it. :P 00:36:19 you seem to avoid answering the question 00:36:20 WHOAOAOAOA 00:36:39 ehirdghost, well I just think it means you can't think of any rational reasons 00:36:40 :P 00:36:44 so your loss 00:36:49 That's your prerogative 00:37:59 that comment made me even more sure about what I just said 00:38:12 I get "Permission denied, please try again. Permission denied, please try again. Permission denied (publickey,password,hostbased)." for a two-host scp thing. That is a bit strange. 00:38:13 If you haven't realised yet, I really don't care what you think about me. 00:38:28 Gawd, MacTex is 1GB… I don't even know how that's possible… still in awe 00:38:35 ehirdghost, I think you do actually. Just are afraid to admit it 00:38:46 AnMaster: hahahaaha; you wish, maybe 00:39:04 ehirdghost, you feel insecure in yourself 00:39:20 "you show all the signs of not caring what I think about you; therefore you are insecure and secretly desire my confirmation but are too scared to seek it" 00:39:29 Yes, I'm sure. Solid reasoning there. 00:39:53 ehirdghost, except you don't "show all signs of not caring". You rather try to show that but fail. 00:40:01 Which is very different 00:40:23 You're opening up my heart and showing me my deepest desires. It would be heartbreaking if it wasn't bullshit. 00:41:04 ehirdghost, Yes you are scared to admit it. The more you deny it, the more you prove it. ;P 00:41:23 I am not a ghost. I am not a ghost. I am not a ghost. I am not a ghost. I am not a ghost. I am not a ghost. I am not a ghost. I am not a ghost. 00:41:26 Now you have to believe I am. 00:41:50 ehirdghost, Um, how does that follow logically 00:42:06 -!- FireyFly has quit ("Later"). 00:42:09 "X is true. Evidence: You are denying X many times." 00:42:17 I didn't say everything was opposite of what you said. 00:42:51 ehirdghost, also just ask random $person with a tinfoil hat! The gov denies it so it must be tru! 00:42:53 true!* 00:42:56 ~ 00:43:04 9/11 was an inside loeb 00:43:10 loeb? 00:43:27 []([]P -> P) -> []P 00:43:39 ehirdghost, ...? 00:43:45 haskell I see 00:43:48 http://en.wikipedia.org/wiki/Löb's_theorem 00:43:55 hm no 00:43:59 hm yes. 00:44:01 Wikipedia does not have an article with this exact name. Please search for Löbs theorem in Wikipedia to check for alternative titles or spellings. 00:44:02 err 00:44:03 hm no 00:44:09 I didn't see what you asid 00:44:14 AnMaster: you forgot the apostrophe 00:44:35 ehirdghost, my irc client did 00:44:42 click fail 00:45:52 "There is a paraconsistent version in Carl Hewitt [2008]." <-- ? 00:45:59 wth does that mean 00:46:07 don't you have a dictionary? 00:46:24 not here, and parents are sleeping in the room with it 00:46:26 so no 00:46:43 Your computer blocks all dictionary sites? 00:46:44 How queer. 00:46:58 except the gay ones, obviously 00:47:01 ehirdghost, that was a result of your curse before 00:47:08 since you typoed it 00:47:32 T'was no typo; was the speak of thine ghosts. 00:47:52 ehirdghost, well that made it misfire 00:48:03 We have an extravolutionary version of the language communicasystem; for extra extrapossibilities with which to extrapolate. 00:48:06 so now it blocks me googling for anything you mentions 00:48:09 for life 00:48:10 fizzie appears to be fluent in it while alive; though. 00:48:16 Or maybe he is a ghost. 00:48:24 AnMaster: Wikipedia mentioned paraconsistent, not I. 00:48:35 ehirdghost, directly or indirectly 00:48:40 so it affects this 00:48:42 AnMaster: Everything. 00:48:47 Now you can never use Google. 00:48:59 ehirdghost, not for the word "everything" no 00:49:10 this is only literal phrases like that 00:49:14 AMD. Intel. x86. x86_64 00:49:17 and I already know that word 00:49:18 Befunge. 00:49:20 cfunge. 00:49:21 and those 00:49:24 so it won't affect 00:49:27 AnMaster: You can never look up info about them 00:49:28 only unknown ones 00:49:36 ehirdghost, wrong. Only unknown words 00:50:00 That is particularly arbitrary. 00:50:05 ehirdghost, plus due to the misfire it is time limited. Lasts about 1-1.5 weeks in average :/ 00:50:20 ehirdghost, well if you haven't typoed it, it wouldn't have misfired 00:50:34 If this sentence is true, then AnMaster is cursed. 00:50:38 Logical bomb in your face. 00:51:02 ehirdghost, fail to see the logical bomb there... 00:51:16 http://en.wikipedia.org/wiki/Curry's_paradox 00:52:42 ok. That is just a false statement due to the A then B not being a casual connection 00:53:00 I see you didn't read the article. 00:53:46 not the whole yet. But false was the wrong word. The right word would be: logical nonsense not connected with the real world. 00:54:09 Yeah, uh, keep reading. 00:55:18 "In formal languages, we sometimes interpret "If X then Y" as a material conditional. On this reading, it simply means "Y, or else not X". Here we would read the sentence as "Santa Claus exists, or this sentence is false". On this reading, Curry's paradox is simply a variant on the liar paradox. However, in natural language this is not usually what we mean by "If X then Y". For instance, "if 6*7=42, t 00:55:18 hen the moon exists" is true as a material implication, but is generally not considered true in natural language, because the moon's existence does not seem to be related to this fact of arithmetic." 00:55:20 ehirdghost, ^ 00:55:35 yes 00:56:10 in other words. the claim describes a non-existent causal connection. 00:56:18 or: logical nonsense 00:58:41 modusPonens :: (p -> q, p) -> q 00:59:19 ehirdghost, natural language isn't an exact science 00:59:24 languages* 00:59:35 Shush, I'm doing logics in mah type system. 00:59:47 ehirdghost, well not abov 00:59:50 above' 00:59:51 * 00:59:55 when you said it 01:00:00 Anyway, a proof of the above proposition: 01:00:03 blergh this kbedor 01:00:05 modusPonens (f, x) = f x 01:00:06 keyboard* 01:00:44 ehirdghost, as I said. Why would it be related 01:00:52 you have to prove to me there is such a connection first 01:01:02 I was making a cheap joke. 01:01:03 Chill. 01:01:37 ehirdghost, I mean I think this paradox is rather lame. It forgets about this think called "false claim" even "lie" 01:01:50 not much of a paradox 01:02:00 if you enter garbage you will get garbage back 01:02:05 Err, I don't think you understood it. 01:02:11 oerjan: care to explain it to him? 01:03:14 ehirdghost, 01:03:38 try telling someone on the street 01:04:01 "if this sentence is true, then you must give me all your money" 01:04:07 see what reaction you get 01:04:21 Bweheheh. 01:04:37 Yes, because going up to a random person on a street is an environment of complete logic and formal reasoning. 01:04:48 That happens to be the most retarded reasoning I've heard today, though. I'll give you that. 01:05:25 It is, of course, sanest to observe that there is nothing compelling anyone to give you money, therefore the sentence is quite false. 01:05:35 ehirdghost, Yes it is the same because natural languages allow this thing called lie. You have to prove your "if A then B" really is a connection that exists 01:05:51 did you actually look at the formal language section 01:06:03 oerjan: please relieve the strain from my being and explain it to him 01:06:04 ehirdghost, natural languages isn't a formal language 01:06:19 ehirdghost, and you said it in natural language first 01:07:18 Say it in Lojban. 01:07:23 ehirdghost, so tell me, why do you think the initial assertment: if true then P is valid? 01:07:23 ;) 01:07:45 oerjan; I'm tired of this idiot, plz take him 01:08:01 I'm just saying if you put in garbage you get garbage back 01:08:07 try being practical 01:08:29 pikhq, I don't know lojban 01:09:19 Practical or logically correct. 01:09:23 I take the latter. 01:09:36 I'd prefer to combine them 01:09:46 You let the former take precedence, evidently. 01:09:51 a down to earth approach 01:10:00 that actually gives useful results 01:10:17 I might think you less of an idiot if your only argument wasn't "that's wrong, ask a random person on the street" 01:10:33 ehirdghost, I have given a lot of other arguments 01:10:45 To the ether, maybe—certainly not here. 01:10:48 and that one was mostly a joke 01:11:06 ehirdghost, ........ read scrollback 01:11:14 I did; maybe you're hallucinating. 01:11:38 ehirdghost, I told you that you need to verify your claims are relevant before you use them 01:11:54 if pink then blue 01:12:16 Try making logical sense; or don't because I can't be arsed, you're clearly not interested in actual logic more than fuzzy human intuitive belief bullshit 01:12:17 now what does that mean? does it make a lot of sense? No 1) it is out of context. 01:12:45 2) even if it was in context, how could you know that this implication is really true 01:13:14 logic is a useful tool only when you put useful input into it 01:13:29 if you just feed it random data you will get garbage back 01:13:53 You know, that makes absolutely no sense whatsoever. 01:13:56 ehirdghost, clearly you fail to see you need to verify the initial assertions to be able to extrapolate from them 01:13:58 Do you actually understand what you'r esaying? 01:14:02 ehirdghost, yes 01:14:06 you don't I see 01:14:14 I think you should read the article very carefully again. 01:14:26 ehirdghost, I did. And you saw I quoted a bit ? 01:14:38 "For instance, "if 6*7=42, then the moon exists" is true as a material implication, but is generally not considered true in natural language, because the moon's existence does not seem to be related to this fact of arithmetic." 01:14:42 read that again please 01:14:57 Yes, you missed the bit that came next. Anyway, fuck off, this is boring and you clearly have no grasp of logic whatsoever. 01:15:08 ... 01:15:25 ehirdghost, lets say you are coding in prolog 01:15:36 Not. Interested. 01:15:36 and listing initial "facts" or whatever 01:16:00 then don't those facts also have to be true for the problem you are trying to solve 01:16:03 to be useful? 01:16:14 Did you miss where I said not interested? 01:16:14 You can only build your your axiojms 01:16:17 axioms* 01:16:26 You're an idiot; you completely misunderstand Curry's paradox, and I am tired of talking. 01:16:27 if your axioms are false... tough luck 01:16:38 ehirdghost, good thing you are writing then 01:16:44 Curry's paradox has nothing to do with defining axioms whatsoever. Go. Away 01:21:30 I suggest a system with three truth values: true, false, EPARADOX (fatal error) 01:21:34 ~ 01:21:55 ehirdghost, ;Å 01:21:57 ;P* 01:22:06 * oerjan is glad he was afk 01:23:06 oerjan, do you agree that if you enter false initial "facts/axioms" in a theorem prover you will get a useless result back? 01:23:13 yes or no 01:23:22 sure 01:23:24 that's nothing to do with curry's paradox 01:23:34 from the wikipedia page it seems to be that 01:23:51 if foo then bar. Well sure. If that connection actually holds. 01:23:56 oerjan: can you explain curry's paradox to him… 01:24:02 but if it doesn't. Tough luck 01:25:04 no tonight my dear, i've got a headache 01:25:08 *not 01:25:09 now a really interesting paradox is Russel's paradox for example. 01:25:22 curry's paradox is a generalization of russell's paradox. 01:25:59 ehirdghost, well I have been discussing this in the context of the natural language case. Which is what we begin with. 01:26:15 natural language doesn't excuse you from using logic 01:26:43 no, but natural language is well known for not being a formally well defined language 01:27:52 By your logic, all representations of logical formula in natural language suddenly lose their attachment to logic because you change '->' to 'implies'. 01:30:14 For instance, consider the following sentence: 01:30:14 If a man with flying reindeer has delivered presents to all the good children in the world in one night, then Santa Claus exists. 01:30:14 Imagine that a man with flying reindeer has, in fact, done this. Does Santa Claus exist, in that case? It would seem so. <-- sounds probable yes. But it *could* be someone else doing it. It would need further investigation. Such an event would be circumstantial evidence. Not proof 01:30:21 ehirdghost, ^ 01:30:42 Say, remember when I said I don't give a shit that you're logically illiterate? 01:30:46 Guess what hasn't changed? 01:31:12 -!- calamari has quit (Read error: 104 (Connection reset by peer)). 01:31:22 ehirdghost, the fact that this is not really a paradox in natural languages. Just a nonsense statement 01:31:33 that hasn't changed 01:33:04 * SimonRC goes to bed. 01:33:09 same 01:33:10 night 02:27:56 -!- Slereah_ has joined. 02:35:33 -!- CakeProphet has joined. 02:38:59 -!- Slereah has quit (Read error: 110 (Connection timed out)). 02:42:23 -!- ehirdghost has changed nick to ehird. 02:47:47 -!- cherez has joined. 02:47:52 -!- cherez has left (?). 02:48:05 it's alive! 02:48:16 (BWAHAHA) 02:49:56 -!- Sgeo has joined. 03:07:01 -!- Slereah has joined. 03:24:02 -!- Slereah_ has quit (Read error: 110 (Connection timed out)). 04:12:34 -!- neldoreth has quit (Read error: 113 (No route to host)). 05:00:46 -!- GreaseMonkey has joined. 05:10:41 -!- CakeProphet has quit ("lol"). 05:45:09 -!- oerjan has quit ("leaving"). 05:47:35 -!- Sgeo has quit ("Leaving"). 07:59:59 -!- clog has quit (ended). 08:00:00 -!- clog has joined. 08:21:00 -!- Slereah_ has joined. 08:25:58 -!- Dewi has quit ("bbl - storm"). 08:31:33 -!- Slereah has quit (Read error: 110 (Connection timed out)). 08:43:03 -!- neldoreth has joined. 08:50:15 -!- psygnisfive has quit (Remote closed the connection). 09:26:17 -!- tombom has joined. 09:34:47 -!- psygnisfive has joined. 09:34:56 hm 09:35:07 natural language quantifiers are AWESOME. 09:35:12 i just feel you should know this. 10:05:13 -!- GreaseMonkey has quit ("Quit: Quit: Quit: Quit: Quit: Quit: Quit: Quit: Quit: Quit"). 10:34:44 -!- oklofok has joined. 10:43:31 -!- Slereah has joined. 10:53:41 -!- Slereah_ has quit (Read error: 110 (Connection timed out)). 10:56:00 -!- Slereah_ has joined. 11:01:59 -!- Slereah has quit (Read error: 145 (Connection timed out)). 12:19:16 -!- Dewi has joined. 13:34:03 -!- neldoret1 has joined. 13:43:35 -!- neldoreth has quit (Read error: 113 (No route to host)). 14:01:28 -!- bsmntbombdood has quit (Read error: 113 (No route to host)). 14:02:31 -!- bsmntbombdood has joined. 14:38:59 -!- KingOfKarlsruhe has joined. 15:09:34 -!- KingOfKarlsruhe has quit (Remote closed the connection). 15:13:56 -!- ehird has left (?). 15:14:45 -!- ehird has joined. 15:16:03 hi 15:20:40 -!- zzo38 has joined. 15:21:05 Nobody is ever actively on #anagol even though some names are listed 15:21:12 Why is that? 15:21:59 it's not too popular a channel 15:22:02 Sometimes people talk 15:22:20 A lot of people leave their IRC clients on to read what people said when they're not away, for channels that aren't logged 15:22:21 I do that 15:22:29 Maybe some of them will have away set in /whois 15:23:43 I tried whois shinh and stuff like that but I'm not sure if that means they are away or not 15:26:48 -!- KingOfKarlsruhe has joined. 15:27:41 I tried various names with whois command but I can't see anything about away, is there some code for being away that I forgot about? 15:27:50 It shows up if they are away 15:28:05 My IRC client says shinh hasn't talked for 59 hours 15:29:35 in my client I have to do /wii for that 15:29:52 I can only get idle time for whois on myself. And if it shows up when they are away, which line does it show up on, the 311 line or the 320 line or some other line? 15:30:12 What does "/wii" means 15:30:16 You need to ask the remote server if you want idle-time information. /wii is a common alias, "/whois nick nick" usually works too. 15:30:46 "/wii nick" => "/whois nick nick", which means "ask nick's whois-info from the server nick is on". 15:31:05 O, thanks I did "whois shinh shinh" and I got the idle time for shinh (214878 seconds) 15:31:16 That's a lot of seconds. 15:31:53 fizzie: you know TeX, right? How do you put a \ in the document? 15:31:56 Well yes, the IRC server returns it in seconds I did the calculation it is approx 59.7 hours 15:32:30 ehird: You can do $\backslash$ although it might look non-text-like since it's math-mode-fluff. 15:32:36 Ah. 15:32:43 Is there a \rawcodeystylething{} block thang? 15:32:48 That would work 15:33:37 There seems to be a \textbackslash command, according to some reference. 15:33:37 Thanks for telling me I need to indicate the name twice if I want the 317 line (although I'm not sure why the server shouldn't figure that out automatically?) 15:34:09 zzo38: IRC is weird 15:34:15 Probably it was done this way for backwards compatibility 15:34:48 And another place says "\char`\\", which is a piece of raw TeX, should also work. 15:34:59 ehird: OK. However I can get the 317 line for myself without needing to type my name twice. 15:35:03 If you want a large block of verbatim text, there's of course \begin{verbatim} ... \end{verbatim}. 15:35:12 You get the 317 line for everyone who happens to be on the same server as you. 15:35:16 At least you should. 15:35:57 * AnMaster wonders how to get GCC to generate an integer constant without $ in inline asm 15:35:57 O. So does it do that to save bandwidth from accessing other servers when it doesn't have to? 15:36:05 Could be that. 15:36:22 There's also a \verb=xyz= command which does xyz verbatim, but maybe \ is too extra-magical even for that. 15:36:37 Is there a superscript/subscript combiner in unicode? 15:36:52 (You can freely use any delimiter instead of = there as long as it's not in the verbatim-string.) 15:37:22 -!- zzo38 has left (?). 15:37:32 Well, I only want a superscript A and a subscript E. 15:38:08 -!- zzo38 has joined. 15:38:13 -!- zzo38 has left (?). 15:38:15 -!- zzo38 has joined. 15:38:17 -!- zzo38 has left (?). 15:38:42 There are super/subscript numbers, and a couple of other characters too. 15:39:05 Yes, I've looked. 15:39:28 zzo38: IRC is weird <-- I know the details about why name twice if you are interested 15:39:46 it isn't exactly what you suggested 15:41:31 What is there outside of the RFC's "If the parameter is specified, it sends the query to a specific server. It is useful if you want to know how long the user in question has been idle as only local server knows that information, while everything else is globally known" explanation? 15:41:45 http://filebin.ca/jrtvbo/first-test.pdf A most delightful X∃LaTₑX (see how hard I worked on that?) output. Bring Hoefler Text (or, wait, is it embedded in the PDF?). 15:42:16 fizzie, that is it yes. 15:42:21 Err, it's 16 March. 15:42:22 Not 17. 15:42:30 Please ignore that time-travelling document. 15:42:47 Yes, I think it embeds-by-default. 15:43:01 fizzie, using a non-server name means asking the server that first nick is on 15:43:03 Although I can't be sure, since I don't remember what Hoefler Text should look like. 15:43:06 so double name... 15:43:11 double nick* 15:43:33 Yes, I fail to see how that is different from my "nick nick means ask the server nick is on" explanation. 15:43:49 fizzie, well on freenode it is. Due to freenode's server hiding 15:43:53 fizzie: if you screenshot, I'll tell you if it's right :P 15:44:04 fizzie, you never get idle time with anything but repeating nick 15:44:15 security by obscurity 15:44:19 -_- 15:44:54 ehird: http://www.cis.hut.fi/htkallas/what_foolishness_is_this.png 15:45:10 LªTₑX! 15:45:17 fizzie: That's hideous, but I suppose that's Linux font rendering thar. 15:45:23 Asztal_: aha 15:45:28 X∃LªTₑX 15:45:31 Now we just need uppercase versions 15:45:49 fizzie: It has the right shapes, so, success. 15:46:08 By the way, that is a copyrighted image and I will sue you. 15:48:45 Asztal_: there should be a combining uppercase :P 15:49:04 There are COMBINING LATIN SMALL LETTER [AEIOUCDHMRTVX]; that's a very random-sounding set. 15:50:00 wth is up with http://www.cis.hut.fi/htkallas/what_foolishness_is_this.png 15:50:21 what do you mean what's up with it 15:50:26 it looks fine 15:50:32 well it is as you said hideous. 15:50:32 apart from the bad linux font rendering 15:50:39 AnMaster: try the pdf on your system 15:50:44 It will probably look nicer 15:50:44 ehird, pdftex generally renders better than that 15:50:49 It renders fine 15:50:51 on my system. 15:50:54 mhm 15:50:54 it's the font, Hoefler Text 15:50:59 It demands good rendering :P 15:51:08 ehird, well Apple has patents on the important rendering bits 15:51:09 ... 15:51:16 So you keep saying. 15:51:24 ehird, I even linked you some weeks or so ago 15:51:29 Yes. Yes you did. 15:51:55 That's viewed-with-xpdf, in case it matters. 15:52:08 -!- KingOfKarlsruhe has quit (Remote closed the connection). 15:52:48 * AnMaster wonders what Helvetica with serifs would look like 15:53:12 It does look rather different with, say, Evince. 15:53:20 fizzie, oh? 15:54:04 14:52 AnMaster wonders what Helvetica with serifs would look like <-- Unlike Helvetica. 15:54:14 http://www.cis.hut.fi/htkallas/evince_version.png 15:54:20 -!- sebbu2 has changed nick to sebbu. 15:54:24 fizzie: oh, that's significantly better 15:54:31 -!- BeholdMyGlory has joined. 15:54:35 as in I can actually read it and it looks similar to the rendering at my end 15:55:12 hm 15:55:29 fizzie, what about kpdf? 15:55:54 I don't have it on this thing. 15:56:00 hm ok 15:56:05 There's acroread, some version. 15:56:08 how comes pdf renders so differently? 15:56:35 I thought the point of pdf was to render the same 15:56:42 AnMaster: different font rendering 15:56:50 Here's how it looks on my end: http://imgur.com/74ZOM.png. It probably won't look very nice unless you have a high-DPI display with the right colour profile. 15:57:23 ehird, it looks hideous on this monitor 15:57:57 It looks reasonably nice on this. 15:58:28 Where "reasonably nice" means I like it more than Evince, I think. 15:59:39 -!- neldoret1 has changed nick to neldoreth. 15:59:47 Hoefler Text's serifs probably would fit better on print. 15:59:56 Heh, Firefox went and crashed when I opened gnome-control-center and twiddled with the font rendering settings. 16:00:06 Anyone here know GCC inline assembler? 16:00:15 AnMaster: gcc's manual does :P 16:00:21 Well, I've done it a little bit. 16:00:21 ehird, well not enough 16:00:27 I have read it and not found a solution 16:00:28 I think I missed the question. 16:00:29 It documents it all, afaik. 16:00:30 Ask #gcc. 16:00:32 I'm trying to something like this (but with more instructions to make it useful): 16:00:33 asm("leaq %[size]+%[var],%%rdx" : [var] "=m"(myvar) : [size] "i"(sizeof(myvar)) : "rdx"); 16:00:33 Where myvar is a static array of fixed size. Size is known at compile time, but may vary depending on compile time options. I would expect it to generate something like: 16:00:33 leaq 2097152+myvar(%rip),%rdx 16:00:33 But in fact it generates this invalid (at least gas thinks so) assembler: 16:00:34 leaq $2097152+myvar(%rip),%rdx 16:00:36 How can I get GCC to not include that first $ there? I have looked at the GCC documentation and found no way to work around it 16:00:44 Using the plain TEX notation $$ . . . $$ for displayed equations is not recom- 16:00:44 mended. Although it is not expressly forbidden in LATEX, it is not documented anywhere in the LATEX book 16:00:47 as being part of the LATEX command set, and it interferes with the proper operation of various features 16:00:50 such as the fleqn option. 16:00:52 Huh. 16:00:59 AnMaster: [size] instead of %[size]? 16:01:23 ehird, that doesn't substitute at all. I tried it. 16:01:34 leaq [size]+myvar(%rip),%rdx 16:01:40 Then you can't do it. 16:01:47 Or, maybe, 16:01:51 AnMaster: 16:01:53 I know! 16:01:54 ? 16:01:57 Well 16:01:59 something like 16:02:06 "leaq "#sizeof(foo)"..." 16:02:09 Or something 16:02:13 I was thinking like cpp stringification 16:02:15 and stuff 16:02:17 hm 16:02:31 anyone have a ttf of computer modern 16:02:35 ehird, I don't think there is a way to get the size of an object with CPP but hm... 16:02:58 Yes, you can't get sizeof() during the preprocessing; I was going to suggest stringizing too. 16:03:06 fizzie, really? how? 16:03:11 oh 16:03:14 "can't" 16:03:20 misread it as "can" 16:03:34 AnMaster: do you know the ranges of sizeof()? 16:03:40 Like, it's either 2 or 4 or 8 16:03:42 if so then 16:03:46 er, I dunno 16:03:47 just like 16:03:56 #if sizeof(foo)==1; #define foo "1" or whatever, 'cept, you can't do that in cpp 16:03:57 so I dunno 16:04:27 ehird, well it depends on compile time options. It could be 2097152 or 4194304 16:04:37 AnMaster: what options 16:04:43 Have you tried the 'n' constraint instead of 'i'? Although I really don't have a clue how they differ. 16:04:50 I have an idea 16:04:53 AnMaster: what option? 16:05:04 -DUSE32 -DUSE64 -DARRAY_SIZE_X -DARRAY_SIZE_Y 16:05:13 AnMaster: Then, just 16:05:23 #ifdef USE_32 16:05:24 the two latter toggle data type 16:05:27 err 16:05:29 two former 16:05:32 okay wait 16:05:33 AnMaster: 16:05:35 do this, for instance 16:05:36 the two latter toggle array size 16:05:49 #ifdef USE32; #define foo #ARRAY_SIZE_X; #endif 16:05:51 Or whatever 16:05:56 hm 16:05:59 Then just do "blah " foo " baz" 16:06:08 That sort of thing anyway 16:06:15 AnMaster: if that needs adding to 16:06:16 do 16:06:19 Then just do "blah " foo "+44 baz" 16:06:20 or whatever 16:06:34 size is sizeof(datatype) * ARRAY_SIZE_X * ARRAY_SIZE_Y 16:07:01 Right, you'll have to do that in parts then 16:07:03 #ifdef USE32 16:07:10 #define sizeofdatatype "4" 16:07:11 #endif 16:07:12 and 16:07:16 #define foo #ARRAY_SIZE_X 16:07:16 then 16:07:27 "sdfk " sizeofdatatype "*" foo 16:07:30 hm... 16:07:32 You get the idea 16:07:39 yeah 16:07:42 * AnMaster considers... 16:07:44 http://www.nopaste.com/p/aVqjYoeUbb <- the source to that LaTeX document; please excuse any noobishness 16:07:59 Also excuse the wrapping; TeXShop doesn't seem to do that automagically. 16:08:06 Er, lack of wrapping, rather. 16:09:22 ehird: Judging from some other latex, you can indeed escape \ with the verbatim mode, so you could write \verb=\chapter= instead of the bulkier \textbackslash{}chapter. 16:10:52 Say, does anyone have a HIGHLY ILLEGAL copy of the Univers font? Well, the copy doesn't have to be highly illegal. 16:13:22 huh wth 16:13:27 what 16:13:28 error: stray ‘#’ in program 16:13:36 * AnMaster goes read C99 spec 16:13:51 Ah. 16:13:53 Ah. 16:13:56 AnMaster: I think you must do 16:13:59 foo(x) #x 16:14:03 oh ok 16:14:06 and foo(ARRAY_SIZE_X) would give "ARRAY_SIZE_X". 16:14:09 So this is perhaps a slight dead end 16:14:13 Hmm hmmm. 16:14:27 AnMaster: are the X and Y bounded? 16:14:30 You need the double-macro thing. 16:14:34 fizzie: Oh? 16:14:34 ehird, hm? 16:14:35 Do tell 16:14:44 #define foo(x) #x -- #define bar(x) foo(x) -- bar(ARRAY_SIZE) 16:14:45 double macro? 16:14:51 That will evaluate ARRAY_SIZE before stringizing it. 16:14:52 fizzie: Aha. 16:14:55 Yes, AnMaster, do that. 16:14:56 AAAAAAAAAAAAAAA 16:14:57 ok 16:15:02 It's very tricky, and I always get it wrong. 16:15:03 I love^Whate cpp :-) 16:15:11 But the comp.lang.c faq has some examples, anyway. 16:15:13 mhm 16:15:37 Incidentally, on the topic of I'm Talking About How OS X Is Awesome To Annoy AnMaster (just kidding, AnMaster, kay?): I like how Ctrl-A and Ctrl-E from emacs are available in every text input field. 16:16:06 #define CPP_SILLY_STRINGIFY(x) # x 16:16:06 #define CPP_SILLY_EVAL(x) CPP_SILLY_STRINGIFY(x) 16:16:09 something like that? 16:16:15 Yep. 16:16:20 Although, well, I'd call it 16:16:37 #define CPP_STRINGIFY_ARGH(x) #x 16:16:44 #define CPP_STRINGIFY(x) CPP_STRINGIFY_ARGH(x) 16:16:53 To more accurately convey the correct emotion. 16:17:22 Sometimes people use the same name with a trailing _, but something like that anyway. 16:17:28 ehird, ok 16:17:32 makes sense 16:17:35 fizzie: That is evil. 16:17:42 It does not convey feminine emotion of human vitality. 16:17:44 Or something. 16:17:51 -!- MizardX has quit ("011000 100110 000101 110011 011001 010010 000000 110010"). 16:18:00 AnMaster: Anyway, this even more simplificates it: 16:18:17 http://c-faq.com/ansi/stringize.html 16:18:21 #if USE32 16:18:27 #define CPP_SIZE 4 16:18:31 #elsif USE64 16:18:33 #define CPP_SIZE 8 16:18:34 #endif 16:18:35 then 16:18:53 CPP_STRINGIFY(CPP_SIZE) "*" CPP_STRINGIFY(X) "+" CPP_STRINGIFY(Y) 16:18:54 or whatever 16:18:58 no need for extra definitions, I mean. 16:18:59 hm 16:18:59 -!- FireyFly has joined. 16:19:05 Just CPP_STRINGIFY* and CPP_SIZE (as an int). 16:19:25 ehird, that would make it pass 80 columns, which look silly 16:19:33 So wrap it. 16:19:40 CPP_STRINGIFY(CPP_SIZE) "*" 16:19:43 CPP_STRINGIFY(X) "+" 16:19:44 etc 16:19:56 ehird, looks silly to have multiline inline asm expand to single line asm -_- 16:20:02 No. It really doesn't. 16:20:07 These are the only times you use those stringifications, so assigning them a name is ridiculous. 16:21:39 hm 16:26:09 http://thanksants.com/ <- <3 16:27:47 :D 16:42:13 -!- MigoMipo has joined. 16:47:28 -!- FireyFly has changed nick to FireFly. 16:53:22 ehird, uh what? 16:53:30 AnMaster: Look Around You reference 16:53:43 oh 16:53:45 enable javascript 16:54:04 ehird, I did enable javascript 16:54:07 and I still don't get it 16:54:09 Ah. 16:54:13 Well, it's a Look Around You reference. 16:54:22 * AnMaster googles 16:54:37 Get some gary gum 16:54:52 It's garry 16:55:08 I hate you :( 17:00:01 -!- ais523 has joined. 17:01:11 hi ais523 17:01:19 hi ehird 17:01:36 can i have my note credits now? :P 17:01:52 ehird: wrong channel, and let me read email first to figure out what you're talking about 17:01:53 hello ais523. 17:01:56 hi AnMaster 17:02:08 1) there is no right channel that I am currently present in; I was just passing on a one-line note 17:02:13 2) your cron job fired 17:02:18 although murphy beat you to it 17:02:45 ais523: also, goethe was in _another_ scam secrecy contract with another group of players, plotting the same scam. 17:02:47 as far as I can tell 17:02:56 so if you entered an agreement with him, he tricked you. 17:03:00 I didn't 17:03:05 ah 17:03:08 i read wrong then 17:03:12 but again, wrong channel, your refusal to join the right channel does not make this the right channel 17:03:27 nobody else is talking, so. 17:03:33 it was just a little note 17:06:51 * AnMaster wonders why the hell gcc generated this code: 17:06:53 sub $0xffffffffffffff80,%rax 17:07:04 what is wrong with adding a bit instead? 17:07:15 less omg optimized 17:07:16 claerl 17:07:17 y 17:07:26 AnMaster: did the stringification work out? 17:07:30 ehird, yes it did 17:07:35 hoorah 17:11:06 ais523, any idea about that sub? 17:11:19 it was in a loop gcc generated: 17:11:24 sub $0xffffffffffffff80,%rax 17:11:25 cmp $0xa42620,%rax 17:11:29 AnMaster: it might change the processor flags differently 17:11:29 jne 0x41c1f0 17:11:31 to addition 17:11:37 ais523, hm... 17:11:40 there are lots of that sort of thing in asm 17:11:59 or it may be subtracting %rax /from/ that large number, rather than subtracting the large number from %rax 17:12:16 ais523, rax is a pointer to an array 17:12:38 -!- Hiato has joined. 17:12:40 also it only uses sub if it is unrolling the loop. It uses add otherwise 17:12:53 like: add $0x10,%rax 17:13:12 yes the sub jump is larger, but that is because it was unrolled 17:16:20 -!- MizardX has joined. 17:26:48 * ehird does some more logics in haskell typeth system 17:26:53 hmm, I forgot, I'm a ghost 17:26:56 -!- ehird has changed nick to ehirdghost. 17:26:58 whooooooo 17:27:14 Ahah, now I recommandeth my speakings of the ghostular enhanced communicatoungh. 17:29:10 you were supposed to wait until Easter to resurrect :( 17:29:32 yeah, yeah, sorry, wait, I'll remove that previous shit from the timestream 17:29:39 done, if you still see it you're hallucinamating 17:36:47 ehirdghost, did you finish your bef93 in qbasic? 17:37:12 no, it was too trivial that I fell asleep 17:37:24 that was before I died... 17:37:28 good times, good times 17:38:00 ehirdghost, why not write a befunge93 in SQL (probably with some procedural extensions) 17:38:14 I ask you s/not // 17:38:30 is ehirdghost writing befunge in SQL? 17:38:37 No. 17:38:41 if so, which extensions? SQL isn't actually Turing-complete without extensions 17:38:43 ais523, no it was a suggestion for something to do 17:38:44 AnMaster wants me to, I don't see why it's interesting. 17:38:46 but then, neither is befunge-93 17:39:01 ais523, I would suggest Pg/SQL 17:39:15 (or whatever it is called) 17:39:42 ehirdghost, no I didn't "want you to", just a suggestion for something to do 17:39:45 you seemed bored. 17:40:01 dodecahedron 17:40:06 dodecahedronasaurus 17:40:14 PL/pgSQL, if you mean the PostgreSQL thing. 17:40:22 fizzie, ah yes indeed 17:40:26 didn't remember the name 17:40:29 Or just PL/SQL for the Oracle thing. 17:41:54 Google Image Search doesn't find any dodecahedronasaurii. :/ 17:42:17 Try "dodecadicks" 17:42:41 http://www.google.com/search?q=dodecadicks&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:fr:official&client=firefox-a 17:42:48 What do you know, it actually exists! 17:42:58 hm 17:43:12 how many bits is needed to represent a 2 MB address space? 17:43:18 31 17:43:23 oh, MB? 17:43:24 21 17:43:28 ah 17:44:23 ais523, sure it isn't 20? 17:44:26 Since 2^10 is a kilobyte, 2^20 is a megabyte and 2^30 is a gigabyte. 17:44:32 hm 17:44:33 http://filebin.ca/jrtvbo/first-test.pdf <- Relinking this since everyone must see it. 17:44:34 AnMaster: 1 MiB is 20 17:44:37 because 1 KiB is 10 17:44:37 ah 17:44:42 2 MiB is therefore 21 17:44:48 and 2 MB is slightly smaller, therefore still 21 17:44:48 * AnMaster wonders what he is miscalculating then 17:45:22 wait, I see that I made an error, but why is it only off by half... 17:45:23 ehirdghost: but you never linked it before... we were just hallucinamating that, right? 17:45:40 Also there's the whole A20 line stuff in the legacy-x86 world. 17:45:40 I didn't calculated in 16 bit numbers... 17:45:43 Asztal_: Do not question me. 17:45:43 calculate* 17:46:03 Lets see. How many bits do you need to represent 1024*512 ? 17:46:24 10+9. 17:46:51 If you mean "represent all numbers in the range [0, 1024*512-1]". 17:47:34 Or "represent 1024*512 different entities", more generically. 17:47:55 well the latter 17:48:32 fizzie, I'm trying to work out how I would do bit interleaving for using a z-order space filling curve to index the static funge space 17:48:41 and I just can't get it straight 17:49:52 ais523, btw since you know gcc quite well. How do you make gcc expand asm("leaq %[size]+%[var],%%rdx" : [var] "=m"(myvar) : [size] "i"(sizeof(myvar)) : "rdx"); to "leaq 2097152+myvar(%rip),%rdx" rather than "leaq $2097152+myvar(%rip),%rdx" 17:50:01 AnMaster: wtf 17:50:03 we just told you how 17:50:06 for now I worked around the issue with some ugly macros 17:50:07 we spent ages explaining it with cpp 17:50:16 ehirdghost, yes but I was wondering "is there no better solution" 17:50:21 it's not a bad solution 17:50:26 it's just stringifying some expressions 17:50:29 ehirdghost, it is an ugly one 17:50:33 C is ugly 17:50:37 well yes 17:52:58 heh, it's funny how well suited haskell is to logic in the type system 17:53:45 Here's a funny bit of x86 trivia: the A20 gate (which controls whether the A20 line is enabled or not; if it's not enabled, the 21th bit in memory addresses is forced to be 0, wrapping the [1MB,2MB) range on top of [0,1MB) and same for 3-4, 5-6 etc.) used to be connected to the *keyboard controller*. 17:54:34 fizzie, heh... 17:54:48 fizzie, keyboard DMA? 17:55:31 It's just that their keyboard controller had a spare I/O pin they could use. The keyboard controller can also reset the CPU. 18:00:00 mhm 18:00:19 so I need to bit interleave a 9 bit and a 10 bit integers in the fastest way possible... 18:00:28 * AnMaster looks at the bithacks page fizzie linked 18:01:18 [0,1) should be valid haskell pintax 18:01:22 -!- Deewiant has quit (Read error: 60 (Operation timed out)). 18:01:39 ehirdghost, pintax? 18:02:00 Syntax but with safety pins and income tax. 18:02:15 heh 18:02:30 ehirdghost: still a ghost? 18:02:33 I always thought the [) notation looks silly 18:02:36 * ehirdghost walks right through Judofyr 18:02:40 Any questions? 18:03:16 AnMaster: The correct range semantics for (N..M) is including N and excluding M, anyway. See: Djikstra. They compose better. 18:03:33 ehirdghost, yes I know what it is for 18:03:33 And (0..N) gets you N items; fits in with array-type stuff 18:03:50 So [X,Y) is actually useless as there's only One True Solution :P 18:04:07 Unfortunately, haskell includes M. 18:04:09 in the range 18:04:09 I just thinks it looks silly with [X,Y) Typographically silly I mean 18:04:15 Mm. 18:04:31 s/thinks/think/ 18:04:43 -!- Deewiant has joined. 18:06:39 -!- ineiros has quit (Remote closed the connection). 18:08:36 Sq, Q prqpqsq thqt qll vqwqls qrq rqplqcqd by "q". (So, I propose that all vowels are replaced by q.) 18:08:53 (It actually works fine for everything but i, pretty much. So let's try it.) 18:11:47 -!- ineiros has joined. 18:12:52 Nq? 18:13:06 -!- Mony has joined. 18:13:33 plop 18:13:37 hi Mony 18:13:43 Yqq mqqn plqp. 18:14:05 yeah, true 18:14:29 Qxqctly. 18:16:09 qqs523: Wqll yqq jqqn my pqlgrqmqgq frqm vqwqls? 18:16:55 no 18:17:00 :( 18:18:09 Tqlkqng wqthqqt vqwqls qs sq pqssq́. 18:18:26 Sq's yqqr fqcq. 18:25:17 wow you can do fast bit interleaving with SSSE3. But not with SSE3 18:25:22 so useless to me 18:25:33 what, you mean the intercal operation? 18:25:53 ais523, almost. I'm talking about a space filling Morton curve here. 18:25:57 oh 18:26:05 ais523, which can be done with bit interleaving 18:26:09 so yes kind of 18:28:08 ais523, hm a hillbert curve would provide better locality of reference than. Wonder how you can calculate it. 18:28:12 Might be worth comparing 18:28:32 AnMaster: err... you're using space-filling curves for the memory of your Befunge interp to avoid cache misses? 18:28:36 Yes. 18:28:40 ais523, that is the plan yes 18:28:43 are you /sure/ that doesn't waste more time calculating than it does reading from cache? 18:28:51 Yes, and it's pretty much exactly the intercal mingle. 18:28:55 ais523, it is possible. That is why I want to profile 18:29:17 I can't be sure if I haven't looked at it at all 18:29:48 if I jump over a bridge, will I die? I can't be sure if I haven't tried it at all 18:29:50 ais523, so I can't say I'm sure until I even tested with space filling curves. 18:29:56 ais523: I'm repeating myself a bit here, but: 18:29:56 [2009-03-15 19:59:38] < fizzie> Notably, calculating the z-order coordinate from x, y is just a single application of the INTERCAL mingle operator. 18:29:58 [2009-03-15 19:59:54] < fizzie> Of course your silly C might lack the always-useful $ operator. 18:32:11 ais523, anyway it is quite possible this could differ a lot between different CPUs, if you have a very small cache you could possibly gain from it. 18:32:20 and the reverse 18:32:27 yes 18:32:28 anyway it is worth trying 18:32:45 ais523, and my sempron has a 128 kb L2 cache, and no L3 cache 18:33:02 128kb? o_O 18:33:10 ehirdghost, yes it is very small. 18:33:32 I have a Pentium 3 with twice as big L2 cache 18:33:34 Why dqn't yqq mqcrq-qptqmqsq qn q dqcqnt mqchqnq?! 18:33:49 what? 18:34:01 * ehirdghost rqllqyqs. 18:34:12 Asztal_: Yqq trqnslqtq, mmkqy? 18:34:14 tell me when you decide to make sense. 18:34:47 Why don't you micro-optimise on a decent machine‽ 18:34:50 Incidentally, are you doing "thqs q thqng" manually or automagically? 18:34:55 because I don't have one? 18:35:04 Asztal_: Thqnks. 18:35:25 maybe ehirdghost will provide the money? 18:35:27 you mean Qsztql_, or are nicknames excluded? 18:35:42 fizzie: Mqnqqlly, bqt nqw I'm nqt: tr qqqqq qqqqq 18:35:57 Asztal_: /nqck Qsztql_ qnd jqqn qqr qrdqr. 18:36:04 -!- ehirdghost has changed nick to qhqrdghqst. 18:36:16 -_- 18:36:32 The-artist-formerly-known-as-ehirdghost: your new name looks like a MMX opcode. 18:36:36 Hmm, tr qqqq q wqrks tqq. Nqcq. 18:36:40 fizzie: Qt qs. 18:36:44 fizzie, :DDD 18:36:52 Sq qnywqy. 18:37:49 I wandar haw at gaas wath a anstaad af q. I wender hew et gees weth e ensteed ef q. I wondor how ot goos woth o onstood of q. 18:37:57 fizzie, my favourite MMX opcode is CVTTPD2PI 18:38:13 wait that one is SSE I think 18:38:16 anyway I like it 18:38:36 I wndr hw t gs wth nstd f q. 18:38:38 it is SSE but operates on mmx registers, instead of xmm registers 18:39:39 qhqrdghqst, why not replace non-vowels instead of vowels? 18:39:49 Qn Q-spqqk, CQBQL = CQBQL. Cqqncqdqncq? 18:39:58 err 18:40:02 QnMqstqr: Tq mqrq qccqrqtqly glqqk mqqnqng frqm cqntqxt. 18:40:20 I suggest all consonants -> i 18:40:32 no idea if it will work out well 18:41:08 ieiiiii, ieiiiii oie iio iiiee 18:41:14 ah not very well 18:41:20 sounds ghostly though 18:41:27 CQBQL = CQBQL <-- was that COBOL = C.B.L? 18:41:56 I read it as COBOL = CABAL 18:42:00 Cqmmqn Qrqqntqd Bqsqnqss Lqngqqgq = Thqrq Qs Nq Cqbql 18:42:05 Asztal, oh that could work 18:42:10 ah 18:42:17 Qsztql gqts q cqqkqq. 18:42:19 CYBYL looked silly 18:42:32 Qsztql <-- looks like a monster in nethack? 18:42:40 some A iirc 18:42:44 forgot the name for it 18:42:50 something like that anyway 18:43:01 Hqy, fqrst pqrsqn thqt mqkqs q scrqpt thqt grqps /qsr/shqrq/dqct/wqrds tq qdd thq vqwqls bqck qn gqts q cqqkqq. 18:43:27 cqqkqq? 18:43:43 also I'm too lazy to make such a script 18:43:59 Why nqt grqp /qsr/shqrq/dqct/wqrds tq fqnd qqt whqt thqt wqrd cqqld bq? Jqst s/q/./ wqqld wqrk fqr qnq wqrd. 18:44:24 well true. But I think it was cookie now 18:44:49 Yqq gqt q cqqkqq. 18:45:20 AnMaster: I'm not quite sure what "Qsztql" could refer to. Quetzalcoatl is the lawful archeologist god, though. 18:45:59 fizzie, you tend to run into them when you are high level and near the top of the dungeon... Maybe it was Quetzalcoatl 18:46:14 But that's a god, not a monster. 18:46:34 There's an A called couatl, maybe that. 18:47:06 fizzie, in ASCII it is A, in QT mode it is some brown/pink blurry tile that looks vaguely like a brown snake with pink wings 18:47:10 iirc 18:47:45 Quetzalcoatl is the feathered snake. And couatl is a D&D monster that refers to that, and has the A symbol. So it's probably that. 18:48:04 fizzie, what about "couatl" a[4]: Monster: 'A' angelic beings: couatl, Aleax, Angel, ki-rin, Archon 18:48:17 "There's an A called couatl, maybe that." 18:48:20 interesting, changing vowels to q's isn't really even noticeable for short words, but i have no idea what ieiiiii, ieiiiii oie iio iiiee is 18:48:22 oh yes just saw that 18:48:23 That *is* what I'm talking about. 18:48:38 fizzie, I did /msg #esoteric from in the privmsg with Rodney 18:48:41 so I didn't see that 18:49:24 Hmm. My scrqpt fqqls qn yqq gqt q cqqkqq, fqr thqrq qrq mqny pqssqbqlqtqqs. 18:49:43 Oh. And incidentally, why does irssi prefix a + or - to all incoming messages now that I have an irssi-proxy thing going on? 18:49:51 fizzie, still I think CVTTPD2PI is the MMX/SSE instruction with the nicest name 18:49:55 don't you agree? 18:50:02 fizzie: Ah, right. 18:50:03 Is it miau? 18:50:07 qhqrdghqst: well why not use a markov chain, don't you just love those? 18:50:15 fizzie: http://miau.sourceforge.net/faq.html 18:50:16 oklofok: :D 18:50:42 fizzie, freenode adds +/- if you request it when you connect 18:50:48 Yes, and miau does. 18:50:50 Thus, see http://miau.sourceforge.net/faq.html. 18:50:52 it means identified to nickserv or not 18:50:53 I haven't seen anything else do i 18:50:54 t 18:51:10 qhqrdghqst, xchat enables it if available and uses it 18:51:26 Oh. Heh, yes, I did connect with xchat to the irssi-proxy. 18:51:32 Is it a toggleable setting somewhere? 18:51:54 what is the proxy 18:51:59 fizzie, yes, by reconnecting. You could also make the proxy filter this so the client never sees that the server supports it 18:52:05 I made another Enigma level, by the way, just for fun; it's pretty easy 18:52:09 ais523: oo 18:52:10 and again not the sort AnMaster likes 18:52:12 ais523, what type? 18:52:14 oh ok 18:52:17 AnMaster: it's a memory level, mostly 18:52:21 and a bit of forward planning 18:52:27 ais523, so what is your opinion on CVTTPD2PI? 18:52:40 AnMaster: I generally don't have opinions on particular asm opcodes I don't know much about 18:52:48 the name I mean! 18:52:54 ehird: irssi-proxy's a module of sorts for irssi which makes it act a bit like a bouncer. I wanted to try a non-monospaced font in IRC, but that's not very viable in a terminal. 18:52:59 looks typical for bloated x86 asm opcodes 18:53:01 "Convert Packed Single-Precision Floating-Point to Packed Doubleword Integers, Truncated" 18:53:04 is what it means 18:53:19 according to the AMD reference docs 18:53:25 fizzie: ah. 18:53:30 Why not just use miau if you want a bouncer? :P 18:53:46 CVTTPS2PI mmx, xmm/mem64 18:53:53 Because I already had irssi running, and didn't want to disconnect for this experiment. 18:54:14 Sounds like a very exciting experiment. 18:54:16 http://filebin.ca/ufepe/ais52304_1.xml 18:54:18 Welcome to 2000 :P 18:54:33 I'm not sure I like it here in 2000. 18:54:56 Wimp 18:55:25 ais523: enigma time 18:55:36 AnMaster: Anyway, did you say I can tell this X-Chat to not enable that identify-msg thing? I'm not sure I want to do any filtering in irssi-proxy. 18:56:07 fizzie, no 18:56:17 I suggested you would filter this in the bouncer 18:56:26 so xchat would never see that it was enabled 18:56:33 also filter any requests to enable it 18:56:34 About instructions, I think I like the name of PUNPCKHBW. It's got, you know, punch. 18:56:57 Unpack and Interleave High Bytes... 18:57:00 interesting 18:57:21 http://khjeron.de/index.php?ELEMENT=300 wat 18:57:23 Aw. I'm not quite sure how to do it. Irssi-proxy is not a very configurable bouncer, it's rather rudimentary. 18:57:27 PUNPCKHQDQ just sounds lame 18:57:51 fizzie: just greenify 18:57:54 also, xchat sux :| 18:58:01 fizzie, xchat is not a very configurable client, it is rather advanced but single minded. That is if you don't like the defaults you don't have a lot of options to change it 18:58:10 qhqrdghqst: how are you getting on with my level? 18:58:15 Awfully. 18:58:26 I'm wondering if I should make it harder, probably not if you're finding it hard 18:58:30 qhqrdghqst, that should be Qwfqllq 18:58:41 Yes, it seems that way. But this xchat is several magnitudes better than I remember it being back in, you know, 2000 or so. 18:58:46 ais523: I suck at enigma. 18:58:50 I'm almost convinced oklofok would do it first time and think "that was boring", but then he's oklofok 18:59:05 my back hurts, can't really concentrate 18:59:10 AnMaster: I will comply if and only if fizzie and you and ais523 start filtering all messages through (tr aeiou q | tr AEIOU Q). 18:59:11 Deal? 18:59:21 Just for a bit. :P 18:59:24 comply with what? 18:59:36 actually been working on this puzzle for ages now, even though i solved it pretty fast last night (my points weren't registered so i had to do it again) 18:59:38 = continue qing. 18:59:39 I'm not sure what would be a nice graphical IRC client. Colloquy sure seems nice-looking, but it's just OS X. 18:59:48 oklofok: which puzzle? 18:59:49 qhqrdghqst, nah, you are free to stop it 18:59:57 3d logic 2: stronghold of sage 18:59:57 fizzie: colloquy is pretty awful; LimeChat is nice. 19:00:08 Colloquy is buggy and crashy and underfeatured 19:00:27 but, err, graphical IRC, hm. 19:00:29 fizzie, there is an xchat fork called conspire. Haven't tried it. May be worth checking it out 19:00:30 Dunno. 19:00:56 you have a grid on three faces of a cube, and you need to connect dots of same color. 19:01:07 I'm going strictly based on screenshots-shown-on-the-software's-web-site here. LimeChat seems like an OS X thing too. Of course one would assume that anything nice-looking is. 19:01:27 LimeChat is OS X only, yes. 19:01:41 started playing kongregate since all the cool kids seem to be doing it 19:01:54 fizzie: Just use plan9. 19:01:59 Problem solved. 19:03:39 you can use plan9 userspace on Linux btw 19:03:46 ported tools 19:03:48 yes 19:03:50 not the same 19:03:56 that's true 19:03:58 you don't get any of the device magic that actually makes it worthwhile 19:03:59 or /proc 19:04:08 qhqrdghqst, but not everyone can switch to plan9 19:04:12 sure they can. :P 19:04:28 oklofok: I played the original version of that game 19:04:36 well iirc fizzie worked on some workstation owned by the university 19:04:38 so that could be hard 19:04:47 qhqrdghqst: I just got a time of 1:53 on that level, btw, that's faster than the record written in the file 19:04:52 ais523: cool 19:05:04 qhqrdghqst, also I don't remember how the /proc of plan9 was but Linux have a /proc. Maybe different. 19:05:15 yes 19:05:18 plan9's is far more extensiv e 19:05:20 Yes, well, I'm currently at home; I think I'll stick to rxvt-unicode and irssi for chatting at work. 19:05:22 anyway: http://swtch.com/plan9port/ 19:05:40 OS X too 19:05:46 qhqrdghqst: harder, easier, or you didn't pass it? 19:05:54 AnMaster: I already have it. 19:05:59 wait 19:06:00 qhqrdghqst, ok :) 19:06:01 oklofok: dunnos 19:06:04 it had no stupid grass shit 19:06:06 how could you know whether it's harder or easier 19:06:07 :D 19:06:17 (that was a question, answer) 19:06:20 what is kongregate? 19:06:26 I can guess it's a KDE program from the spelling 19:06:28 a site with flash games 19:06:31 but don't konw anything beyond that 19:06:31 basically you get points out of playing flash games. 19:06:35 yeah. 19:07:13 * ais523 wonders if Three Times Through is always possible 19:07:16 ais523, this one ends in 04? I saw 01 too. What about 02 and 03 19:07:17 I suspect it is, but haven't proved it 19:07:19 never saw them 19:07:24 AnMaster: well, no 19:07:27 they aren't finished yet 19:07:28 ah 19:07:33 are you having a go at it? 19:07:34 ais523, any in the style I like? 19:07:39 well, not yet 19:07:42 more incentive to actually finish games, and try all kinds of stuff out; which of course is good only if you consider flash games educational, which i do 19:07:59 ais523, I don't know if I mentioned it. But I rather like that level "robin's wood" 19:08:01 _02 and _03 are almost finished, they just need AIs 19:08:03 forgot what pack 19:08:10 AnMaster: I don't, it just looks big and tiresome 19:08:18 ok 19:08:51 oklofok: http://www.kongregate.com/games/AlexMatveev/3d-logic 19:08:52 3d logic wun 19:10:04 ais523, is there anything in your last level preventing the first two stones matching each other? 19:10:14 AnMaster: no, it's completely random the arrangement 19:10:18 normally you get a mix of luck 19:10:25 -!- Hiato1 has joined. 19:10:27 because there are so many oxyds to place 19:12:08 ugh that level is irritating 19:12:12 oklofok: i can't beat original level 7 <.< 19:13:36 wait I just did. 19:14:27 ais523, I liked "floppy floors" too 19:14:37 on easy, never tried it on hard 19:14:53 AnMaster: what, my level? 19:14:54 however it was another level I was looking for 19:14:58 ais523, ? 19:15:00 i did the first 12 levels in about 2 minutes, then made a mistake, and closed it 19:15:16 ais523, in engima 1.0 new 19:15:21 (i mean i don't actually want to play atm) 19:15:32 AnMaster: oh, I was referring to the one I just pasted 19:15:50 and nothing's forcing oklofok to play, especially as the level I just pasted is likely far too easy for em 19:15:53 I just mentioned some levels I liked 19:16:11 ais523: are you trying to force me to try it :P 19:16:17 no, there wouldn't be any point 19:16:20 ais523, there was some level that was a "who did it" iirc 19:16:24 that was rather interesting 19:16:42 AnMaster: I've done that one, it isn't really because it's just elimination and luck 19:16:57 cluenigma 19:16:59 hm 19:17:25 ais523, there was one with lots of hidden tools. Split up in four screens 19:17:32 I was doing that one recently 19:17:38 ais523, don't remember name 19:17:44 I remember first room was very white 19:17:45 I think I tried twice and failed 19:17:51 I like that one 19:17:57 just don't remember where or name 19:18:09 both times because coffee wasn't implemented, so getting the last pair of oxyds depends entirely on luck and fast mouse movement, you need both 19:18:16 and I was unlucky and not fast enough anyway both times 19:18:25 hmm, i should probably implement some coffee 19:18:40 ais523, indeed 19:18:52 ais523, do you remember what level pack? 19:19:19 * ais523 looks 19:19:26 it won't be one of the enigma ones because it had coffee in 19:19:42 ais523, also there are some levels in "enigma 0.92" called "Pentimino", any clue what they are about? 19:19:48 yep 19:19:51 what kind of coffee are we talking 19:19:51 oh? 19:19:54 they're about pentominos, pretty obviously 19:19:57 qhqrdghqst: the item in Enigma 19:20:03 that does nothing because they haven't programmed it yet 19:20:04 ais523, and what the *** is that? 19:20:17 AnMaster: look it up on Wikipedia or Google or somewhere 19:20:23 mhm 19:20:29 Tetris shapes are tetrominoes, pentominoes are like that but with one more square 19:20:32 ais523: what's it meant to do. 19:20:39 18:20 AnMaster: ais523, and what the *** is that? <-- what the ass? 19:20:42 found it. "tool time" in 0.92 new 19:20:55 qhqrdghqst: it's described vaguely as "pause the game" 19:20:58 qhqrdghqst, "hel" 19:20:59 in the docs 19:21:16 but from looking at the levels that use it, I suspect it multiplies durations by infinity 19:21:21 so your umbrellas last forever, etc 19:21:33 maybe it's limited-duration itself, or only when it's the first item on your list, or something, though 19:22:01 AnMaster: it's actually "heel" 19:22:08 wow, Slalom Skiing in 0.92-1 is hard (#17) 19:22:08 heh 19:22:58 -!- Hiato has quit (Connection timed out). 19:23:54 qhqrdghqst: yep, that one took me quite a while 19:23:58 there's a trick to it, though 19:24:23 even better, it's a trick you can work out entirely on visible information, it's not like there's a hidden thing you have to find or something like that 19:24:49 In Oxyd®, you could take a break with this item. During the break, you could analyse the whole level stresslessly. In Enigma, the cup does not have any special properties yet. 19:24:53 the definition of the coffee 19:25:14 I managed the slalom one 19:25:37 on both easy and hard 19:25:50 I did it on hard, and doing it on easy can be done the same way as doing it on hard 19:25:53 below par for easy 19:26:05 2 seconds above par for hard 19:26:33 #58 light barriers, how do you get that block?! 19:26:47 I solved it, above par 19:27:04 AnMaster: did you do salom, above par, on hard, you solved it? 19:27:06 qhqrdghqst, also the mirrors duh 19:27:12 qhqrdghqst: 58 in which pack? 19:27:15 if so, did you solve slalmon, on hard, above par? 19:27:18 ais523: 0.92-1 19:27:25 oh wait 19:27:27 qhqrdghqst, I solved it on both, on easy below par, on hard above par 19:27:28 you can move the lasers 19:27:31 as I said 19:27:38 AnMaster: I do not like them, Sam I am 19:27:46 qhqrdghqst, what? 19:27:51 YOU HAVE NO CULTURE> 19:28:02 qhqrdghqst, what the hell are you referring to 19:28:05 >__________< 19:28:25 AnMaster: a Dr. Seuss book 19:28:31 they're books of nonsense intended for children 19:28:31 ehird: Your culture is not universal, you know. Although I've heard enough by cultural osmosis to understand that much. 19:28:37 and are great fun to read out loud 19:28:39 * AnMaster googles 19:28:45 I'd say Green Eggs and Ham is fairly universal 19:28:54 i've heard that thing, don't know what it's about tho 19:28:56 I did google "sam" but that returned Seattle Art Museum 19:28:57 and such 19:28:58 that sam i am thing 19:29:01 anything in English is universal 19:29:05 AnMaster: google "green eggs and ham" 19:29:06 lament: :P 19:29:08 with the quotes 19:29:09 actually i think it was just the name of an episode of some series 19:29:10 ais523, ...? 19:29:13 Incidentally, we watched some sort of green-eggs-and-ham cartoonification just the-day-before-yesterday. 19:29:20 "I do not like green eggs and ham. I do not like them, Sam-I-Am." 19:29:40 what about "automaton magic" 19:29:48 err 19:29:49 what about it. 19:29:54 solved it below par here. and just two seconds above world record 19:30:11 so 29 seconds instead of 27 19:30:20 I like that level 19:30:25 http://www.youtube.com/watch?v=RdR0LXOiEB8 I think was the clip. 19:30:31 engima 0.92-1 #83 19:33:39 A dark house on them am do I or there and eat if rain they anywhere eggs in Sam train are fox let asy tree be goat like see try boat good may so will box green me thank with car ham mouse that would could here not the you. 19:33:45 s/asy/say/ 19:33:48 The most zen sentence ever. 19:34:30 qhqrdghqst, did you make that up now? 19:34:41 No, it's the complete set of words used in Green Eggs and Ham. 19:34:46 http://en.wikipedia.org/wiki/Green_Eggs_and_Ham#Lexicon 19:34:57 All 50 of them, of which 49 are monosyllabic. 19:35:11 ah 19:36:39 ok. 19:37:01 qhqrdghqst, I don't think they are intended to be written together like that 19:37:10 Oh really? 19:37:10 ;P 19:37:28 eggs in Sam train or fox... I think Dr Seuss is warning us from the grave 19:37:29 *are 19:37:34 qhqrdghqst, no, it is "for beginning readers" after all 19:37:37 The eggs in any train owned by Sam are inevitably foxes! 19:37:39 50 words in one sentence? 19:37:44 oh true. 19:38:05 qhqrdghqst, even an experienced reader have trouble keeping the context in such a long sentence 19:38:23 Someone make a huge sentence that ends with 10 proposition :-P 19:38:24 s 19:38:42 qhqrdghqst, and someone made a sentence out of a single word: buffalo 19:38:44 your point? 19:38:47 ais523: "I do not like green eggs and ham. I do not like them, Sam-I-Am." <<< wait green eggs? :D something started gnawing me about that sentence, but i could not quite put my finger on it until now 19:38:53 oh make not made 19:38:54 misread 19:38:57 Buffalo buffalo buffalo buffalo buffalo buffalo buffalo with. 19:39:07 (ofc green eggs might actually mean something other than rotten eggs, i just don't know what) 19:39:21 in the book, they're just like ordinary eggs, except they're green 19:39:33 oklofok, odd bird? spilled paint? 19:39:42 In the cartoon, the ham is also green, IIRC. 19:40:34 hmph, i've been trying to leave irc for like 20 minutes now, now seriously 19:40:35 -> 19:42:41 -!- MigoMipo has quit ("QuitIRCServerException: MigoMipo disconnected from IRC Server"). 19:43:54 hmmmm 19:45:23 http://swtch.com/plan9port/screenshots/opensolaris.png <-- is that window manager gnome? 19:45:36 I thought Solaris had some custom one 19:45:43 Not nowadays. 19:46:00 I see 19:46:09 qhqrdghqst, I still think OpenWindows was a nice one ;P 19:46:17 AnMaster: it's certainly possible to theme Gnome to look like that 19:46:24 err, that is gnome 19:46:25 but you could do that with other window managers too 19:46:26 very obviously 19:46:32 see: window decoration, the style of taskbar buttons 19:46:34 the show desktop button 19:46:38 and the desktop selector 19:46:40 and the icons next to it 19:46:45 and the icons on the desktop & their shadow 19:46:49 and the text rendering 19:46:51 ah, I don't use default Gnome icons anyway 19:46:52 and the terminal's menu bar and icon 19:47:03 you're right, the taskbar buttons look like unthemed Gnome 19:47:15 it's themed, just the default theme :P 19:48:24 Solaris' CDE wasn't what I'd call nice. Glrbh. 19:48:40 the default theme is bluer than that 19:49:00 err 19:49:02 is it? 19:49:04 no it's not 19:49:10 ais523, well of course you can theme it, But the icons looked gnome style. So did the applets. 19:49:14 I just wasn't sure 19:49:23 since I remembered solaris using something else 19:49:29 well it's debian's default theme at least 19:49:42 and I haven't used gnome for years 19:49:57 http://mail.gnome.org/archives/desktop-devel-list/2005-July/msg00269.html I think I understand why gnome is shit now 19:50:33 qhqrdghqst, the linked images are 404 19:50:43 Yes. Because it is from 2005 and linkrot. 19:50:49 Congratulations for noticing. 19:51:10 qhqrdghqst, did it get accepted? 19:51:15 what 19:51:20 that patch 19:51:22 or change 19:51:23 or whatecer 19:51:25 whatever* 19:51:27 /facepalm 19:51:33 I'm dropping this conversation thread 19:51:37 err 19:51:39 I read the mail 19:51:42 _exit(1); 19:51:43 just not the responses yet 19:51:50 so what the hell are you talking about 19:52:00 excuse me, what are we talking about? 19:52:05 http://mail.gnome.org/archives/desktop-devel-list/2005-July/msg00269.html 19:52:07 about that 19:52:09 _exit(1); 19:52:12 it seems to be a patch to change the theme 19:52:17 the default that is 19:52:18 what, they're making it-sensor visible in Enigma 1.01? 19:52:29 they can't do that, or at least they should give an option for invisible sensors 19:52:35 it-sensor? 19:52:35 ais523, ?? 19:52:48 qhqrdghqst: it's an invisible item which makes commands run when you go over it 19:52:58 ais523, also "making" implies engima 1.01 isn't released yet? I'm pretty sure it is 19:52:58 like an invisible trigger, but you can't hear it and it doesn't care about stones 19:53:02 *1.10 19:53:05 ah 19:53:32 ais523, can you place it below some other tile? 19:53:45 you can have one floor, one stone, one item, and any number of actors on a square 19:53:51 mhm 19:54:05 ais523, what about those levels where you go under something that looks like a floor? 19:54:07 how do they work 19:54:16 those are hollow stones which look the same as the floor beneath them 19:54:22 ah 19:54:27 so they can have an item under them 19:54:34 and a floor under that (you wouldn't want to fall, would you?) 20:03:18 Sq, Qs, qnyqnq? 20:03:33 ais523, mhm 20:04:07 -!- Judofyr has quit (Read error: 54 (Connection reset by peer)). 20:04:08 ais523, what about the things you gets with explosives. and remove with spades 20:04:14 don't know the name 20:04:18 hollows are items 20:04:21 ah 20:04:35 you know, because you can't drop an item on their square, but you can have different sorts of floor under them, and push stones over them 20:06:11 -!- Judofyr has joined. 20:07:06 -!- Judofyr_ has joined. 20:22:15 -!- Judofyr has quit (Read error: 110 (Connection timed out)). 20:25:09 -!- MigoMipo has joined. 20:25:33 ais523, heh 20:25:49 ais523, what are actors? 20:26:10 in engima I mean 20:26:23 AnMaster: marbles, tops, etc 20:26:28 ah 20:26:31 things that aren't restricted to integer coordinates 20:27:06 well I would assume they are, just integer coords == pixels instead of == tiles 20:27:15 or do they really use float? 20:28:00 arbitrary reals 20:33:58 not likely 20:33:59 AnMaster: it's floats 20:34:06 ais523, why on earth? 20:34:22 AnMaster: You were an X-Chatter, right? Do you happen to know how the "Colored nick names" thing picks colors for nicks? 20:34:23 so their physics simulations work better 20:34:33 oh ok 20:35:00 fizzie: it picks the most appropriate color for the personality 20:35:52 * FireFly is randomly happy 20:36:07 lament: Do you feel purple, then? 20:36:49 lament's a sort of mauve on my client, the same colour as fizzie 20:36:51 I do like purple prose. 20:36:54 and FireFly, for that matter 20:37:06 AnMaster's green, and qhqrdghqst's cyan 20:37:21 oklofok is a slightly redder purple than lament 20:37:34 I've succeded in displaying a pic at my DS 20:37:36 and Asztal_'s grey 20:37:37 I'm just wondering, because this has decided that ais523 and AnMaster have the same color, which is non-optimal as you people so often coincide temporally. 20:38:15 fizzie, I used to use xchat once upon a time 20:38:18 nowdays I use ERC 20:38:25 and I never used coloured nicks in xchat 20:38:31 I don't like that feature 20:38:34 Right, right. 20:38:44 it just hashes the nickname 20:38:49 I tend to use three colours: "normal, highlighted, own message" 20:39:26 ais523, and I'm more a dark blue person 20:39:30 than green 20:39:34 though green is ok 20:39:35 hey, I'm a dark blue person too 20:39:37 but also cyan 20:39:46 * qhqrdghqst transparent, like ninja. 20:40:08 ais523, http://en.wikipedia.org/wiki/Ultramarine is nice 20:40:31 * qhqrdghqst octarine 20:40:52 qhqrdghqst, doesn't go with transparent I'm afraid 20:41:09 Well, the text part already has the normal-highlighted-own split, so I don't mind nicks being rather colorful. 20:41:40 fizzie, I do that for all of the lines, and no nick column rainbow 20:42:12 -!- Hiato1 has quit (Read error: 60 (Operation timed out)). 20:42:55 qhqrdghqst, what do you think of the look of plan9 20:43:08 It's usable but very ... well, 1992s. 20:43:11 Seems that color_of(char *name) is just sum of all the bytes in name, modulo amount of colors in the fixed set used for nicknames. 20:43:12 well ok 20:43:13 s/s././ 20:43:23 qhqrdghqst, you think GTK is worse? 20:43:29 Absolutely. 20:43:39 qhqrdghqst, what about QT? 20:43:48 fizzie: that's so boring I could take a shit on it. 20:43:53 if you get my analogy. 20:43:59 AnMaster: Qt is acceptable. Sometimes. 20:44:06 qhqrdghqst, mhm. Motif? 20:44:20 AnMaster: Motif I can help it 20:44:26 (For guidance on the above sentence, see oerjan.) 20:44:31 Also the particular piece of code divides by sizeof (char), which is a rather silly way of saying 1. 20:44:53 qhqrdghqst, well it is a "meta" style pun I think 20:45:01 that is it is supposed to sound like something else 20:45:03 AnMaster: Yes. 20:45:04 but I don't know what 20:45:07 Pronounce it out aloud 20:45:10 "motif i can help it" 20:45:25 just did. But I'm not sure how to pronounce motif in English 20:45:27 It's "not if" when your nose is stuffed. 20:45:38 fizzie: You spoiled it :P 20:45:40 ah 20:45:57 Yes, I'm a spoiler. 20:46:01 if (! 20:46:08 it sound more like "note if" than "not if" 20:46:20 or am I mispronouncing motif? 20:46:28 qhqrdghqst, ? 20:46:38 Well yeah. 20:46:40 It's a bad pun 20:46:59 fizzie: Here's a nice colourerer: abs (foldl xor 255 nick). 20:47:00 * IPA: /məʊ'tif/; you do need a bit of imagination there. 20:47:07 qhqrdghqst, so is it a long or a short t in motif? 20:47:11 er 20:47:13 fizzie, I can't really read that 20:47:19 foldl xor 255 (map ord nick). 20:47:21 qhqrdghqst, err o not t 20:47:50 colourNick nick = foldl xor 255 (map ord nick) 20:47:54 "Moo-tif", the interface of choice for cows. 20:48:03 fizzie, hah 20:48:30 Prelude Data.Char Data.Bits Test.QuickCheck> colourNick "AnMaster" 20:48:30 236 20:48:31 Prelude Data.Char Data.Bits Test.QuickCheck> colourNick "ais523" 20:48:33 176 20:48:35 Perfect distinguishotron. 20:48:37 fizzie, err so what is the difference between "note" and "not" then. It isn't "long/short" o 20:48:51 qhqrdghqst: what is the number? selection between 256 possible colours? 20:48:56 ais523: yep 20:49:01 since motif ends up a bit like note if the way I say it 20:49:08 it may be hard to find 256 different-looking colours 20:49:12 yeah 20:49:17 Just use the standard palette 20:49:22 You don't get too many similar colours 20:49:37 let colourNick nick = foldl xor 16 (map ord nick) `mod` 16 20:49:40 also seems to work acceptably 20:50:47 AnMaster: "note" is /nəʊt/, "not" is just /nɒt/. So there's an "ou"-style diphthong in "note". 20:50:55 FireFly, ah 20:50:57 err 20:50:58 fizzie, ^ 20:51:15 AnMaster: But yes, "motif" is closer to "note if" than "not if". 20:51:28 indeed 20:51:32 -!- kar8nga has joined. 20:51:59 hm 20:52:05 I just got an idea for the perfect OS 20:52:22 probably quite different from ehird's/my 20:52:29 That is what I was thinking, ais523... 20:52:30 It would be a combination of Genera, Plan 9 and QNX 20:52:31 which are somewhat different from each other 20:52:49 what do you think? 20:52:49 AnMaster: well, first, those are so completely different that you couldn't combine them reasonably 20:52:57 qhqrdghqst, yeah probably 20:52:57 secondly, QNX isn't very interesting apart from being embedded 20:53:10 qhqrdghqst, it is very stable though 20:53:18 QNX didn't invent stability 20:53:19 what about just combining the first two? 20:53:23 qhqrdghqst, that is true 20:53:27 "AnMaster: well, first, those are so completely different that you couldn't combine them reasonably" 20:53:33 qhqrdghqst, hm 20:53:37 Adding random good stuff together isn't a recipe for success. 20:53:43 well true 20:54:04 QNX does have interesting aspects. The distributedness stuff is fancy. 20:54:24 and having almost everything in userspace 20:54:28 Plan 9 has a distributed CPU system 20:54:30 which is excellent 20:54:37 I mean, IPC and scheduling are in kernel, that's about it 20:54:48 iirc 20:54:51 my OS is so microkernel, it even has its userspace in userspace! 20:54:58 ais523, -_- 20:55:27 all your features are belong to userspace 20:55:45 There was also something funny related to the file systemics, but I've forgotten what it was. My only QNX experiments were several years ago. 20:55:54 fizzie, same 20:56:41 http://plan9.bell-labs.com/plan9/img/mirtchov/interpolate.gif <-- that's pretty 20:57:01 it's like every ELER comic ever! 20:57:23 Did you mean: ELLE comic 20:57:24 ?? 20:57:31 http://geekz.co.uk/lovesraymond/ 20:57:35 Everybody Loves Eric Raymond 20:58:17 "And GIMP now supports CMYK" <-- ? Really? *looks* 21:02:45 looks like a future version will have it 21:17:03 -!- fizzie has left (?). 21:17:21 -!- fizzie has joined. 21:17:25 -!- kar8nga has quit (Read error: 110 (Connection timed out)). 21:17:28 Whoopsie. 21:18:05 I am definitely not ready for this third-millennium gooey-IRC thing. 21:18:16 fizzie, what? 21:19:02 -!- MigoMipo_ has joined. 21:19:23 I think I should make a realtime javascript raytracer. (At this point qhqrdghqst dies of unbelievable stupidity.) 21:19:43 fizzie: did you press Ctrl-W? 21:19:45 qhqrdghqst: it'll take you something like 20 years for the hardware to catch up 21:19:46 I somehow accidentally closed this tab-or-whatever-it-is, maybe with ^w or something. I've never accidentally typed "/part #esoteric" or something. 21:20:09 accidental mouse clicks are what most commonly close tabs by mistake for me 21:20:10 ais523, ctrl-m in xchat is "move marker of last line read in channel" iirc 21:20:16 or was it clear window? 21:20:18 Hey, with TraceMonkey it'll be native-code-speed. 21:20:24 It's "mvoe marker line". 21:20:28 AnMaster: err... why are you nickpinging me with that details? 21:20:29 s/mvoe/move/ 21:20:37 ais523, what? 21:20:45 fizzie: unfortunately native speed isn't fast enough for realtime raytracing either 21:20:58 ais523, I must have misread 21:21:00 ais523: oh, it's been done... 21:21:02 two things 21:21:12 -!- BeholdMyGlory has quit (Excess Flood). 21:21:14 " fizzie: did you press Ctrl-W?" turned out as " fizzie: did you press Ctrl-M?" 21:21:16 no idea how 21:21:16 ... 21:21:32 -!- BeholdMyGlory has joined. 21:22:16 There's been things that you could consider real-time ray-tracing, if you want to be polite about it, in demoscene prods a long time. 21:22:52 On June 12, 2008 Intel demonstrated Enemy Territory: Quake Wars using ray tracing for rendering, running in basic HD (720p) resolution. ETQW operated at 14-29 frames per second. The demonstration ran on a 16-core (4 socket, 4 core) Tigerton system running at 2.93 GHz.[10] 21:22:59 I want a 16-core system, me. 21:23:06 That would be pleasurable. 21:23:19 AnMaster: http://slashdot.org/article.pl?sid=09/03/16/1839231 21:23:22 qhqrdghqst: Just get a couple of microwaves, then you can do another core-counting experiment. 21:23:31 it seems that Intel and AMD are rowing over x86 21:23:43 if they end up revoking each other's licences, all sorts of ridiculous things could happen 21:23:57 -!- MigoMipo has quit (Nick collision from services.). 21:24:01 ais523: probably they'll just find a loophole and change a minor bug to make it "not x86" 21:24:04 -!- MigoMipo_ has changed nick to MigoMipo. 21:24:22 AnMaster: http://slashdot.org/article.pl?sid=09/03/16/1839231 <-- huh 21:24:30 why is that a huh? 21:24:35 because only ehird links people to things randomly? 21:24:43 http://goatse.ca/ 21:24:50 ais523, yes 21:24:53 ... which is now squatted 21:25:10 qhqrdghqst, what about the cx original? 21:25:23 AnMaster: someone else owns it; it has a picture of a LEGO thing that looks like a goatse but it's SWF. 21:25:28 Safe for work that is. 21:25:29 Not flahs 21:25:31 *flash 21:25:31 heh 21:25:39 and they're trying to sell it 21:25:58 "Buy a piece of Internet history." 21:26:12 qhqrdghqst: what happened to the picture of Bill O'Reilly? 21:26:16 if it ever existed? 21:26:23 -!- kar8nga has joined. 21:26:24 it was replaced 21:30:10 Oh wow. 21:30:12 Horrible idea. 21:31:05 brb -> 21:31:05 what's the horrible idea? 21:31:15 Maybe he went to implement it immediately. 21:34:09 hah 21:45:39 -!- ais523 has quit (Remote closed the connection). 21:45:53 It seems to have been a really horrible one. 22:10:11 heh 22:10:26 qhqrdghqst, btw do you know any good breakout game for OS X that is free? 22:13:30 -!- Judofyr_ has quit (Read error: 110 (Connection timed out)). 22:17:52 -!- MigoMipo has quit ("QuitIRCServerException: MigoMipo disconnected from IRC Server"). 22:21:57 -!- Mony has quit ("Quit"). 22:22:43 -!- kar8nga has quit (Read error: 110 (Connection timed out)). 22:31:58 -!- qhqrdghqst has changed nick to ehird. 22:33:53 -!- Judofyr has joined. 22:35:16 * comex wonders how to implement motion blur in canvas 22:35:42 where by "motion blur" I just mean blend in some previous frames 22:35:50 I guess you could create invisible canvases 22:39:10 You could just always draw N frames, but that doesn't sound very fast. 22:40:17 by invisible canvases do you mean translucent canvases? 22:41:07 have 5 canvases, and draw to each one in turn, changing the Z-order so that the most-recently-drawn-to canvas is at the top 22:41:31 that would work :| 22:42:44 -!- BeholdMyGlory has quit (Read error: 104 (Connection reset by peer)). 22:42:54 I just crashed firefox 22:42:57 or hung 22:43:37 You can also use the toDataURL("image/png") or something to get the current frame as an image, then render those. But that would mean one extra PNG creation operation and N-1 extra PNG decoding operations per frame. The "pile of canvasii" approach sounds most sensible. 22:44:18 I was going to do a translucent draw under the assumption that it's probably faster than css 22:44:57 but since that means copying data and the other approach doesn't 22:44:58 :p 22:46:32 There seems to be some sort of getImageData functions you maybe could use; it still involves copying, but at least there's no PNG creation stuff. 22:46:39 the canvas3D canvas context would make this a lot faster :) 22:47:02 yeah, I'm using that 22:47:04 BUT 22:47:09 looks like I can't putimagedata between canvases 22:47:10 lame 22:47:37 OH THANK GOD 22:47:43 SOMEONE is finally havnig hardware-accelerated 3d 22:48:28 scratch that 22:48:31 I was just setting it up wrong 22:51:31 WELL HELLO THERE 22:51:38 firefox is leaking memory like crazy 22:51:54 it's not freeing the image datas 22:51:55 ew 22:52:16 well, I'll use the overlay-canvases approach 22:53:34 It does sound simpler, at the very least. 22:53:45 oi! 22:58:23 http://qoid.us/cv.html <-- don't run in a slow browser 23:00:51 ok.. 23:01:04 you should add friction. 23:01:55 * comex adds more balls 23:02:02 where by "ball" I mean "square" 23:02:17 add friction! 23:02:21 and drag! 23:03:51 I wonder how slow collision detection will be 23:03:52 probably not very 23:04:00 maybe slow if I have 100 balls 23:04:17 if you do quad tree searching it should be efficient 23:04:23 what the fuck is that 23:04:26 :p 23:05:12 well, for your thing it wouldnt be an issue since you're using squares 23:05:40 actually, screw collision detection 23:05:42 instead, 23:07:32 -!- neldoreth has quit ("leaving"). 23:07:44 -!- neldoreth has joined. 23:07:49 xor http://qoid.us/cv.html 23:11:37 :o 23:11:38 magic! 23:11:45 collision is better. 23:12:41 too lazy 23:12:46 because what if they're colliding by multiple pixels 23:14:06 then you better get a time resolution good enough for that to happen :) 23:14:10 or not happen 23:17:49 comex: quad trees are one of the data structures that split R^n into hypercubes so you can check whether containing hypercubes intersect before doing the actual collision check. 23:17:54 (glad i could help) 23:18:38 psygnisfive: it's not fast enough :p 23:18:42 err actually quad tree is not one of them. 23:18:46 I guess I can do physics faster than actually drawing 23:18:50 but, 23:18:51 * comex tries to get canvas 3d to work on beta 3.1 23:18:55 but guess it can be used as such 23:18:57 oklofok! 23:19:02 let me tell you about quantifiers :D 23:19:05 comex: that's shit 23:19:10 ehird: what 23:19:10 :) 23:19:24 proof: 23:19:28 http://www.blahbleh.com/whyiesucks.htm 23:19:36 a 3d cube, with motion blur, in canvas, getting ~50fps 23:19:41 ur doing it wrong, evidently :P 23:19:46 ehird: um, so? 23:19:49 I saw that 23:19:57 yeah, and... motion blur isn't that hard? 23:19:58 the boxes are going perfectly fast (not measuring fps) 23:20:06 oklofok, can i can i huh huh huh 23:20:10 AnMaster: lbreakout2 23:20:12 I believe that's drawing the cubes repeatedly for the motion blur 23:20:21 but they do collide by more than one pixel I think 23:20:23 anyway, my delay was me scraping most of the skin off the back of my foot ^_^ 23:20:33 By mistake, that is. 23:20:44 50 FPS *with an encoding run going*. 23:21:03 * comex stabs pikhq with rapier 23:21:05 psygnisfive: you can tell me seven sentences 23:21:16 anyway, awful idea time 23:21:22 all learnable natural language quantifiers are conservative. 23:21:31 HOWS SEVEN WORDS 23:21:49 And this on 4 year old hardware. 23:22:02 psygnisfive: interesting! what does that mean?` 23:22:04 Canvas is still shit :P 23:22:08 you can have another 7 words. 23:22:08 orly? 23:22:12 it's quite fast and a lot better than flash 23:22:15 *mean? 23:22:21 ehird: Less shit than everything else. 23:22:21 comex: compare that to SDL 23:22:31 Well, everything else on the web. 23:22:31 heck you could script SDL with spidermonkey 23:22:34 it'd be 39487539457345 times faster 23:22:48 Compared to a proper programming environment, well, yeah. Canvas sucks. 23:22:57 suppose Q is a quantifier, relating two sets, e.g. Q(X,Y) = |X intersect Y| > |X-Y| (== "most X are Y") 23:22:58 ehird: I doubt it 23:23:15 no, really. 23:23:18 then Q is conservative if and only if: Q(X,Y) iff Q(X, X intersect Y) 23:23:30 really 300 trillion times faster? 23:23:33 well, no. 23:23:57 Now. 23:23:58 Evil time 23:24:29 http://hg.mozilla.org/users/vladimir_mozilla.com/canvas3d/ 23:24:33 how the fuck am I supposed to compile that 23:24:37 with butts 23:24:42 * comex looks up 23:24:49 http://hg.mozilla.org/users/vladimir_mozilla.com/canvas3d/file/f050229f6011/Makefile.in 23:24:51 Makefile.in; happy? 23:24:59 not very 23:25:50 actually, my evil idea is kinda related to this 23:26:00 -!- FireFly has quit ("Later"). 23:27:10 autoreconf&&./configure&&make&&make install? 23:30:07 psygnisfive: awesome 23:31:00 its even cooler 23:31:01 because 23:31:14 all conservative quantifiers can be built up in very simple was 23:31:16 ways* 23:31:17 namely: 23:31:24 seeing psygnisfive get excited about linguistics amuses me 23:32:21 any boolean operation over conservative quantifiers gives a conservative quantifier 23:33:02 and: Q(X intersect C, Y) is conservative, for any set C 23:33:18 and these two together produce ALL and ONLY the conservative quantifiers 23:33:33 if you start with a single conservative quantifier all(X,Y) 23:33:34 :) 23:33:43 or some(X,Y). 23:34:46 cool 23:35:01 which is /very/ interesting indeed 23:35:52 -!- olsner has quit ("Leaving"). 23:35:57 well i think i don't actually know what you mean by quantifier when it comes to language 23:36:10 words like all, some, most, many, few 23:36:12 all the numbers 23:36:20 hmm 23:36:30 more-than 23:36:52 -!- tombom has quit ("Peace and Protection 4.22.2"). 23:36:59 what quantifier would 5 be? 23:37:14 5 cats are small 23:37:21 hmm righ 23:37:22 t 23:37:30 5(X,Y) = |X intersect Y| = 5 23:37:36 == 5* 23:37:38 yeawh 23:37:40 *yeah 23:37:45 well actually 23:37:46 technically 23:37:53 5(X,Y) = |X intersect Y| >= 5 23:38:02 ....yeah :) 23:38:43 so how would you construct most out of all? 23:40:00 no clue :D 23:40:06 most is 23:40:18 most(X,Y) = |X intersect Y| > |X - Y| 23:40:21 you already defined most 23:40:28 we can try to reword it 23:40:34 hmm. 23:40:39 what is legal in the transformation? 23:40:52 boolean combinations of other conservative quantifiers 23:40:57 just all(all(X),all(all(Y),all(Z))) kinda stuff? :| 23:41:08 so Q(X) -> !Q(X) 23:41:09 and 23:41:17 ohh. 23:41:33 Q(X,Y), R(X,Y) -> Q(X,Y) op R(X,Y) 23:41:38 for op some boolean operator 23:41:43 right right. 23:41:48 and Q(X,Y) -> Q(X intersect C, Y) 23:41:52 for any set C 23:42:41 all(X,Y) = X subset Y = X - Y == 0 23:43:30 yeas 23:43:31 ill look at my references and see if they mention how to construct most from all 23:43:46 do look, i'm not really in a thinking mood, kind of a math overdose 23:45:02 oh my god i want to learn chemistry 23:45:28 lol 23:45:53 well you know molecules and stuff they're very pretty. 23:46:56 "A semantic characterization of natural language determiners" is one of the papers that discusses this 23:47:21 well why don't you go look then :-) 23:47:38 i cant get it 23:48:16 i have another article i can give you a copy of 23:48:30 nooooo exam next monday and i forgot to begin my reading journey today. 23:52:18 i have two articles for you. 23:52:31 buttt... i need to start my readings! 23:54:07 wellnowwhat.net/linguistics/quant1.pdf 23:54:09 wellnowwhat.net/linguistics/quant2.pdf 23:54:28 Star Trek: Some hydrogen stars go trekking. It's a gas! 23:55:13 *rimshot*