00:00:17 Capitalism is a cancer. But I'm a smoker, anyway, so... 00:00:33 double cancer 00:00:33 `addquote Capitalism is a cancer. But I'm a smoker, anyway, so... 00:00:34 538) Capitalism is a cancer. But I'm a smoker, anyway, so... 00:01:07 what is socialism then? Lupus? 00:01:52 It's not lupus 00:02:02 oh. ok then. 00:03:29 "Everyone else seems to be able to cope with the people they spend most of their time with." That's what I tell myself at first. But then you hear about the divorces and custardy battles. The domestic violence. Cruel managers. Tyrannical dictators. "What would I do or how would I feel if I was one of them?" 00:03:42 no need to add every quote during my bizzaro rant though. 00:04:13 now see.. if they hadn't gotten to me i could just read my sentences normally 00:04:37 custardy battles are really gruesome things 00:04:43 but i also learned about this concept of embedded suggestions, hence: "add every quote during my bizzaro rant" and "read my sentences" 00:04:52 oerjan: Beat me to it 00:05:07 not intentionally though.. it occurs in normal grammar. 00:05:41 Wait, if you were the one on the receiving end, or the one on the giving end? 00:06:01 You'd probably feel pretty awesome in the latter case 00:06:30 #junethack is a boring channel 00:06:35 Hardly any puns at all 00:06:49 thack _is_ boring. also it's july already. 00:06:58 julythack 00:06:59 you should rename. 00:07:30 I got all my thacking done in May 00:08:18 so this crowd.. and believe me there is such a crowd, is saying that people filter out logic and just see imperatives 00:08:30 at least thats what they seem to be saying 00:08:48 itidus20: Those are what we call "voters" 00:11:08 there is an xkcd comic which reminds me of what im describing 00:11:40 http://xkcd.com/842/ 00:12:16 lol 00:14:59 Ju is not a month 00:15:33 >_> 00:15:34 in the street fighter back story there is a team of evil women named after months 00:15:36 portmantaeus don't exist 00:15:39 it was originally meant to be in June, but people decided July was better for them 00:15:55 ............................................______ __ 00:15:55 ....................................,.-‘”...................``~., 00:15:55 .............................,.-”...................................“-., 00:15:55 .........................,/...............................................”:, 00:15:58 .....................,?........................... ..........................., 00:16:01 .................../.................................................. .........,} 00:16:04 ................./.................................................. ....,:`^`..} 00:16:07 .............../.................................................. .,:”........./ 00:16:10 ..............?.....__............................ .............:`.........../ 00:16:13 ............./__.(.....“~-,_..............................,:`........../ 00:16:16 .........../(_....”~,_........“~,_....................,:`..... ..._/ 00:16:18 ..........{.._$;_......”=,_.......“-,_.......,.-~-,},.~”;/....} 00:16:21 ...........((.....*~_.......”=-._......“;,,./`..../”............../ 00:16:23 ...,,,___.`~,......“~.,....................`..... }............../ 00:16:26 ............(....`=-,,.......`........................(......;_,,-” 00:16:28 ............/.`~,......`-...................................../ 00:16:30 .............`~.*-,.....................................|,./.....,__ 00:16:33 ,,_..........}.>-._...................................|........... ...`=~-, 00:16:36 .....`=~-,__......`,................................. 00:16:38 ...................`=~-,,.,............................... 00:16:41 ................................`:,,.............. .............`..............__ 00:16:44 .....................................`=-,...................,%`>--==`` 00:16:46 ........................................_........ …_,-%…….` 00:16:49 Is my response 00:16:59 not a whale 00:17:06 Is that picture of something? Of what? 00:17:06 uhhh 00:17:06 very like a whale 00:17:07 NihilistDandy: that's not quite the same thing as accidentally pasting a 20-line struct into the channel 00:17:31 zzo38: /facepalm 00:17:38 NihilistDandy: yeah, don't do that. 00:17:49 Never again. I just realized a script I had did that 00:17:51 as an ardent pro-pasting activist, that's way too long 00:17:54 Is it facepalm? 00:18:05 Is it picture of facepalm? 00:18:06 elliott: would you abuse ops in #esoteric if you had them? 00:18:07 its not a whale 00:18:08 Yes 00:18:10 thats all i know 00:18:14 (not that I can grant them to you, I'm just curious as to the answer) 00:18:15 OK. 00:18:31 Juni and Juli are members of a special unit within Shadaloo called the "Dolls", also known as Bison Elite Guard, which is composed of twelve young women brainwashed to serve as Bison's personal assassins. The twelve members of the Dolls are named after the months of the Gregorian calendar in various languages, with Juni and Juli being German for June and July. 00:18:44 zzo38: thank you, i couldn't see it before you said it 00:18:46 Are there even any ops in #esoteric? 00:18:47 ais523: No. 00:18:49 NihilistDandy: Yes. 00:18:49 junithack, julithack 00:18:58 ais523: Well, I mean, maybe for a joke, but not beyond that 00:19:04 it can be hard to attract an op here sometimes 00:19:08 but not always 00:19:31 Op powers are only for trolling, as far as I can tell 00:19:39 As they should be 00:20:01 im bad with power i never know what to do with it 00:20:45 ais523: i was _this_ far from banning NihilistDandy 00:20:53 T_T 00:20:54 I have power in some places. Then left those places to rot 00:21:09 NihilistDandy: temporarily mind you 00:21:14 i am become Sgeo destroyer of virtual worlds 00:21:26 god that was a good one 00:21:26 elliott, since when are wikis virtual worlds? 00:21:30 stfu 00:21:40 hey oerjan gimme ops i need to find out for sure whether i'd abuse them or not 00:21:51 Sgeo: NetHackWiki? 00:21:57 funny guy 00:22:02 ais523: i figure i'll inevitably get ops if i just stick around here long enough 00:22:07 As compensation: http://www.smbc-comics.com/#comic 00:22:08 im going to be the only person here in thirty years 00:22:15 ais523, yes, and Creatures Wik, although I was a bit more involved there. And Superosity wiki 00:22:46 elliott: I tend to randomly pick up ops in all sorts of places when I don't expect it 00:23:08 elliott: You'll be dead in 30 years. HackEgo will become self-aware and destroy all meatbags in its quote file 00:23:14 hasn't happened here, yet, even though this is one of the channels people might most expect me to be an op in (given that I'm one of the two active wiki admins, and Keymaker rarely comes here) 00:24:02 -!- Nisstyre has quit (Remote host closed the connection). 00:24:34 ais523: you'll get ops when oerjan becomes a wiki admin 00:24:43 so you already have them, enjoy! 00:25:28 heh 00:26:08 -!- ChanServ has set channel mode: +q *!*@gateway/web/freenode/ip.76.99.21.147. 00:26:26 -!- ChanServ has set channel mode: -q *!*@gateway/web/freenode/ip.76.99.21.147. 00:27:28 If I do publish any of my computer program in books, I should also included the DVD of the program in the back of the book, too. (This is one idea about literate programming, although there are others, too.) 00:28:10 that, btw, is apparently the cloak of user "test" :P 00:28:20 oerjan: I was wondering 00:29:02 oerjan: who's test 00:29:09 oh 00:29:10 an person 00:29:16 * elliott says hi to them 00:29:19 elliott: the poor innocent victim of my chanserv testing 00:29:31 not victim, now friend!!! 00:30:05 Victim != friend? 00:30:09 I have been misinformed 00:30:12 now wtf would it segfault /there/ 00:30:12 you /msg'ed em? 00:30:13 I have some calls to make 00:30:27 oerjan: yes :D 00:30:32 how do you ascend a scope in gdb? ais523? 00:30:40 frame number 00:30:42 as in, frame 4 00:30:48 you can use bt to see a list of frame numbers 00:31:06 NihilistDandy: "Routine helps calm you down. Maybe you will talk to another friend. You talk to him every day for some reason. Though it's not exactly right to call him a friend, since you despise him. Your relationship with the fellow is difficult to describe. It should be noted that in troll language, the word for friend is exactly the same as the word for enemy." 00:31:13 Aww, it's "enemy", not "victim". 00:31:17 My tiny memory strikes again. 00:31:21 (Note: My memory is not quite tiny.) 00:31:22 ais523: thanks 00:31:26 ais523: isn't there "frame up" or something? 00:31:35 sigh, nope 00:31:37 Oh, well. The sentiment is on point, at least 00:32:02 In that case you use the word of troll language in case you need a new word since the normal English word not exactly right, maybe. 00:32:19 (gdb) print block_info[block] 00:32:19 $2 = {name = 0x0, type = AIR, trait = NO_TRAIT} 00:32:21 wat 00:32:21 oh 00:32:28 elliott: wat 00:33:01 olsner: it's a block that doesn't exist 00:33:11 elliott, derp. I was about to complain that routing calming the guy down makes no sense, but I was mistaken as to who that applied 00:33:17 elliott: Right. 00:33:37 Sgeo: good sgeo memory two thousand and eleven 00:33:44 01:32:44 [DIED] Failed to open file '/home/elliott/.mcmap/colors': No such file or directory 00:33:50 oh right, i probably shouldn't require it to exist 00:34:04 * olsner gets some burger-n-ds9 00:35:14 (the first part is watching unrar for 21 seconds, really boring) 00:35:38 G_FILE_ERROR_EXIST 00:35:38 Operation not permitted; only the owner of the file (or other resource) or processes with special privileges can perform the operation. 00:35:38 G_FILE_ERROR_NOENT 00:35:38 No such file or directory. This is a "file doesn't exist" error for ordinary files that are referenced in contexts where they are expected to already exist. 00:35:43 EEXIST is a good name for that 00:35:47 (it's EEXIST, right?) 00:36:05 olsner: then the second part is DS9 and it's even more boring :DDDDDDDdddddd oh snap 00:36:10 [asterisk]OH SNAP 00:36:18 G_FILE_ERROR_EXIST 00:36:18 Operation not permitted; only the owner of the file (or other resource) or processes with special privileges can perform the operation. 00:36:25 it's funny because DS9 has a plot, an attribute not shared with the entire rest of star trek 00:36:27 Why... call it G_FILE_ERROR_EXIST? 00:36:33 Sgeo: thus my lol 00:36:41 i think its from unix wow what the fuck happened to the map 00:36:44 elliott: both your statements are tainted by truth 00:36:50 01:36:36 MODE: surface긒澖缀è긒澖缀`跈Ā 00:36:52 what did i do 00:36:57 am i asian 00:37:02 i broke mcmap help 00:37:16 but yeah, I'm really looking forward for this "plot" you speak of 00:37:28 olsner: how many episodes in are you??? DS9 is great 00:37:36 elliott: 10 or something 00:37:41 i'm assuming you meant you haven't seen plot yet 00:37:43 season 1 ish 00:37:45 I liked TNG and Voyager, and DS9 is quite acceptable 00:37:49 I'd comment, but my opinions tend to be random relative to consensus 00:37:50 NihilistDandy: ...wow 00:37:54 NihilistDandy: You're the worst person 00:37:57 lol 00:37:57 I'm removing you from the whitelist now 00:38:01 NUUU 00:38:02 You think I'm joking 00:38:11 TNG really had no plot at all, and for the most part that was a good thing 00:38:13 * NihilistDandy cries and drinks 00:38:19 olsner: Exactly 00:38:38 NihilistDandy: I'm offended by your liking of Voyager, not your ambivalence to DS9 00:38:41 I didn't really watch Star Trek enough to be able to tell one series from another 00:38:42 elliott: You'll never get ops now. Your abusive nature is revealed 00:38:43 elliott very quickly abuses power 00:38:52 elliott: I only watched six episodes of Voyager 00:38:54 enterprise is currently my reference to startrek-with-plot 00:38:59 Maybe they were just six decent ones 00:39:02 NihilistDandy: OK yeah, stop saying that Voyager is an acceptable thing. 00:39:03 You have no idea. 00:39:16 Oh, all right, then 00:39:22 elliott, can we at least agree that the Voyager theme is good? 00:39:29 elliott: when I watched it I was too young to know good from bad, at least wrt Star Trek 00:39:31 Sgeo: lol 00:39:32 Sgeo: It's a bit boooooooooooooring. 00:39:33 so I have only your word to go on 00:39:41 So is DS9's, but it's the nicer of the two boring Star Trek themes 00:39:47 ais523: Literally everyone hates Voyager :P 00:39:48 olsner: I think Enterprise hurt me a little. I couldn't watch past the third episode 00:39:49 Apart from Sgeo. 00:39:59 NihilistDandy: Through a Mirror, Darkly is the only good Enterprise 00:40:07 ALSO ITS INTRO HAVE WE RAVED ABOUT ITS INTRO ENOUGH IN THIS CHANNEL YET 00:40:08 I'll have to look it up 00:40:43 NihilistDandy: http://www.youtube.com/watch?v=Dl5zw6fGjdA is the cool intro to In a Mirror, Darkly 00:41:02 They took all the exploration scenes from the normal intro and turned them into the far superior Earth-blowing-shit-up scenes 00:41:04 Neat 00:42:32 That was pretty cool. The music's kinda bland, though 00:42:35 Blandly epic 00:42:38 NEAT 00:42:55 OK, what the hell happened to this... 00:43:04 (mode == MAP_MODE_CROSS && map_flags & MAP_FLAG_FOLLOW_Y ? " (follow)" : ""), 00:43:04 (mode == MAP_MODE_SURFACE && map_flags & MAP_FLAG_CHOP ? " (chop)" : ""), 00:43:04 (map_flags & MAP_FLAG_LIGHTS ? " (lights)" : ""), 00:43:04 ((map_flags & MAP_FLAG_LIGHTS) && (map_flags & MAP_FLAG_NIGHT) ? " (night)" : "")); 00:43:09 How can any of that be 긒澖缀è긒澖缀`跈Ā? 00:43:20 NihilistDandy: Have you seen the normal Enterprise music. 00:43:22 Anything is better than it. 00:43:24 Literally anything. 00:43:30 Jesus. 00:43:34 I'll take your word for it 00:43:47 NihilistDandy: http://www.youtube.com/watch?v=ZPn-lTytfGo Enjoy terrible pop song 00:43:59 AHHHHHHHHHHHHH 00:44:05 Is this the actual intro? 00:44:07 elliott: have you tried encoding the garbage string as UTF-16, then decoding it as UTF-8? 00:44:18 WHAT THE FUCK 00:44:21 NihilistDandy: Yep. 00:44:25 I DON'T EVEN 00:44:28 I love how it looks like a bad redub by someone on YouTube 00:44:33 ais523: nope, I'm mostly trying to figure how I broke a piece of code I didn't even otuch 00:44:34 touch 00:44:51 Top comment on that video: Hey Im kinda new to the Star Trek franchise just bought the original series on dvd watching it tonight. Now which series has Wil Wheaton in it. 00:44:54 Rolfchoppa007 1 day ago 00:45:00 Trolololol 00:45:00 -!- Nisstyre has joined. 00:45:18 Maybe it has messy format strings? 00:45:25 I like how Wil Wheaton has about the same number of likers as Wesley Crusher has haters 00:45:35 Or, buffer overflows? 00:45:45 zzo38: but i didnt tucho that biteof code 00:46:14 Then maybe buffer overflows. 00:46:16 elliott: Wil Wheaton is a pretty nice guy, and Wesley Crusher is a terrible character. Simple. :) 00:46:45 Yup 00:47:29 I support the Wesley Crushers. 00:47:43 I'm watching Troll 2 00:47:54 (USER WAS BANNED FOR THIS MESSAGE) 00:48:17 I don't really get that though, wesley crusher was a likable geniusishy character, but all I've seen of wil wheaton is the big bang theory making him seem like a dick 00:48:28 ais523: how does use valgrind 00:48:33 olsner: wesley crusher was not likeable??? 00:48:33 elliott: for what? 00:48:35 olsner: Shh, you'll get us all killed 00:48:46 for checking for memory access errors, just write valgrind before the command name 00:48:48 elliott: I am saying he was 00:48:54 he was the accidental archetype of "smart but REALLY ANNOYING DICK" 00:49:06 and his smartness was just annoying because it made no sense at all 00:49:11 ais523: find memory corrupt 00:49:13 okays 00:49:16 yep, valgrind command 00:49:19 dunno if i'm smart enough for this but HERE WE GO 00:49:20 finding corruption is its default setting 00:49:26 also dunno if it'll like SDL 00:49:27 olsner: People get annoyed at Mary Sue wunderkind. 00:49:45 He has become Sheldon's nemesis. But we sympathize with Sheldon. 00:50:14 pikhq_: I think the actor who played Wesley Crusher is on record as saying that he hated Wesley Crusher too 00:50:18 i've never watched big bang theory but isn't sheldon meant to be a total jerk too 00:50:24 ais523: Yes, Wil Wheaton is on record saying that. 00:50:26 characters are not likeable just because they're intelligent 00:50:28 elliott: Yes. 00:50:37 YOU MEAN I'M DRUNK? 00:50:39 elliott: He's autistic taken up to 11. 00:50:47 wow what, bad packet id 00:50:49 I FEEL STRANGE, BUT ALSO GOOD! 00:50:51 minecraft why do you hate me 00:51:06 i wonder what i broke 00:51:08 NihilistDandy: No. You mean *I'm* drunk? I don't think I am drunk either. 00:51:11 You can't have your characters announce how they're feeling. THAT MAKES ME FEEL ANGRY! 00:51:51 elliott@katia:~/Code/mcmap$ valgrind hello let's be friends 00:51:51 > 00:51:54 ais523: help what do i tell valgrind 00:51:58 it wants to know something 00:52:01 it is > prompt... 00:52:11 elliott: valgrind doesn't take input 00:52:14 and just outputs to stderr 00:52:15 stupid elliott doesn't identify with intelligent unlikable characters on tv 00:52:18 so obviously hello is asking for inputs 00:52:19 then why is it asking me :((((((((((((( 00:52:19 *input 00:52:29 ais523: oh, should i say hello to hello??? 00:52:37 oh, it's the shell 00:52:40 you have an unmatched ' 00:52:41 > ==3150== Thread 3: 00:52:41 ==3150== Conditional jump or move depends on uninitialised value(s) 00:52:41 ==3150== at 0x57BBE40: inflateReset2 (in /lib/libz.so.1.2.3.4) 00:52:41 ==3150== by 0x57BBF2F: inflateInit2_ (in /lib/libz.so.1.2.3.4) 00:52:41 ==3150== by 0x40A47B: world_thread (in /home/elliott/Code/mcmap/build/mcmap) 00:52:42 wow, I don't even want to know 00:52:48 ais523: I know :P 00:53:10 bleh, I don't have hello installed 00:53:56 what the fuck? it looks like the only memory badness is in the lines i didn't touch 00:54:00 and that don't access any values i touched 00:54:41 (minecraft may be off-topic here, but mcmap isn't!) 00:58:42 THIS STRAW HAS NO BENDY-END 00:58:45 Remember when the Motorola Razr was a nice phone? 00:59:07 no 00:59:13 bendy is the best part of straws :( 00:59:24 what, the NY Times is running a story about /Dwarf Fortress/? 00:59:30 olsner: :( 00:59:31 is this for real 00:59:35 monqy: http://www.nytimes.com/2011/07/24/magazine/the-brilliance-of-dwarf-fortress.html?_r=1&pagewanted=all 00:59:36 monqy: well, this straw DOES NOT HAVE IT 00:59:37 elliott: Link? 00:59:39 Oh 00:59:51 Madness 00:59:56 did i die and end up in heaven 01:00:05 elliott: surely not 01:00:21 true, you're still here 01:00:25 the lack of bendiness is located to the straw 01:00:53 olsner: monqy: stop it. we don't want straw arguments here. 01:00:57 :D 01:01:01 :D 01:01:05 you're such a straw man, oerjan 01:01:17 oerjan: if my straw had bendies I could build a strawman 01:01:24 IF ONLY 01:01:55 ... 01:02:01 Where do.... Dwarf eat-- 01:02:02 In NyTimes!? 01:02:30 Lymee: your questions are irrelevant in this crisis 01:02:41 Now read the "level20" D&D recording file. 01:02:46 -!- ralc has joined. 01:02:48 the straw that broke olsner 01:02:52 "The only furniture in the small dining room is Scamps’s litter box." 01:02:55 Lymee: the straw, the bendy, it lacks it 01:03:01 good reporting 01:03:08 Lymee: hi, can you fix mcmap, 01:03:08 There may be some things I have missed and I will ask my brother and the DM. But other errors such as grammatical or whatever, you can tell me now. 01:03:15 elliott, NO! 01:03:17 "If much of Tarn’s apartment suggests a tenant who never fully moved in, his bedroom suggests a tenant who never sets a sock outdoors." 01:03:26 "Tarn wakes up around 3 p.m. every day, codes through the night and goes to bed around 6 a.m." 01:03:31 omg i am the tarn, its me 01:03:43 Tarn Hird 01:03:57 i think tarn's a republican though so :( 01:04:04 Hird Derp 01:04:10 also i would never make a game as boring as dwarf fortress 01:04:31 "Tarn consumed “maybe one glass” of water in the last three months, hydrating with soft drinks instead. “Water’s not doing it for me these days,” he said. “I know it’s bad, but the sugar goes right into programming the game. If I don’t drink soda now, I get a headache and can’t do any work.”" 01:04:36 i think they made this article just to laugh at him 01:04:37 Then don't make Dwarf Fortress, please. 01:05:22 elliott, nonsense! 01:05:29 That is the standard all programmers must aspire to. 01:05:45 -!- elliott has set topic: Tarn was pleased. “The hippos like the sewers!” he said. He took a celebratory swig of Dr. Pepper and rocked back and forth. | Logs: http://codu.org/logs/_esoteric/ and http://tunes.org/~nef/logs/esoteric/?C=M;O=D. 01:06:05 There is some property of soda which makes it less painful on sensitive gums. 01:06:35 it's probably all that acid 01:06:38 Is it the fizziness, or the acidity? :P 01:06:42 i know its ironic 01:06:43 Gregor: hifive 01:06:48 elliott: wooo 01:07:06 but water tends to irritate my gums 01:07:10 elliott: you know, this Tarn character sounds like someone from that story about the search for the Point 01:07:14 because i have some serious gum problems 01:07:22 olsner: search for the what now? 01:07:29 elliott: exactly 01:07:34 could be the salt 01:07:36 hahah 01:07:47 "The issue wasn’t aptitude so much as passion. He wanted to do math but also to make video games, a juggling act he managed as an undergraduate. This had become impossible. “They wanted 60 hours a week from you, giving you problems that would take 20 hours to solve,” he said. He grew depressed and, in his only encounter with drugs, snorted meth." 01:07:53 this is the silliest article 01:07:57 elliott: well, was it a search? I just recall something with "the Point" 01:08:05 you know... all the fizz and acids makes it possible for them to add salt to the soda 01:08:07 elliott: Agreed 01:08:18 Meth is so passé 01:08:33 all the real bedroom programmers snort pure liquified scheme 01:08:34 23mg of sodium per 250ml in my current soda 01:09:10 question is, how much sodium in water. and would i taste 23mg of sodium in 250ml of water 01:09:37 itidus20: never leave us ok 01:10:05 I played Dwarf Fortress for 20 minutes before becoming a mix of bored and disillusioned 01:10:11 I think I'm too tall for it 01:10:26 dwarf fortress is great, it's just not a game 01:10:41 more like a scientific measuring and probing tool for a universe that doesn't exist 01:10:41 people don't necessarily want games. 01:10:55 is super mario bros a game? 01:11:10 No, it is a video game. 01:11:12 itidus20: at least one yes 01:11:16 I spend most of my time modelling and probing things that don't exist. I don't need that in a game 01:11:30 is making an avatar jump onto a floating brick a game? 01:11:31 super mario brothers is a game, yes :P 01:11:33 i wish i could get into dwarf fortress 01:11:41 it sounds like the kind of game i would like 01:11:43 -!- cheater_ has quit (Ping timeout: 255 seconds). 01:11:43 monqy: get into elliottcraft instead 01:11:45 itidus20: No, that is just an idea. 01:11:47 the kind of more than a game i would like 01:11:49 elliottcraft???? 01:12:03 well.. in super mario bros, you do that just within the context of a game 01:12:04 monqy: yes i play minecraft, but it is a bad game, so i am making a totally different game and calling it a better minecraft 01:12:06 like.. uhh 01:12:07 It's minecraft, except that you're elliott 01:12:09 hahaha hahaha hahaa 01:12:13 It's a niche community 01:12:18 monqy: dwarf fortress-level detail and depth, but minecraft-style graphics and infinite world 01:12:27 plus you control a player rather than dwarves (although there are AIs in the world) 01:12:27 sounds neato 01:12:29 super mario bros is about jumping on blocks with a time limit 01:12:34 i'm not sure what the goal is yet :< 01:12:44 but hey, infinite in all directions unlike minecraft!! 01:12:48 (minecraft has a height limit) 01:12:50 itidus20: And running to the right 01:12:55 well yeah 01:12:55 itidus20: Well, it is about reaching the end of each level, I think 01:13:04 It is the goal, at least. 01:13:04 i study these things. 01:13:09 elliott: Lazily generated, one hopes 01:13:14 by study i mean i look at websites 01:13:15 Und jetzt: schalafen werden würden werden sein... Ja? 01:13:25 NihilistDandy: yes :P 01:13:40 originally shigeru said.. each level should take about 60 seconds 01:13:44 monqy: also competently coded multiplayer??? unlike minecraft's which is so bad you have no idea 01:13:49 elliott: Well, I mean if you have a real UTM you could do it strictly :D 01:13:59 NihilistDandy: utms can't do infinite computation in finite time, sorry no 01:14:03 the people said.. oh no. thats a lot of screens that have to be created per level if it is scrolling for a whole minute 01:14:03 :P 01:14:13 one time i played minecraft 01:14:15 i uh 01:14:17 then shigeru said.. but there will be things on the way to slow them down 01:14:18 it was a long time ago 01:14:24 elliott: http://www.youtube.com/watch?v=pJsPoI2w6-A 01:14:28 monqy: back when it was a java applet? 01:14:31 and you had infinite of all blocks 01:14:34 yeah 01:14:45 is it not a java applet anymore? 01:14:52 monqy: well it's java, but nobody plays it in the applet form any more 01:14:59 and it costs money, but not much 01:15:05 money :( 01:15:07 monqy: the beta is actually a fun game. the coding is so bad you have no idea, but it is actually fun. 01:15:12 Even *I* bought it, and I steal everything 01:15:23 monqy: 21 dollars, it seems 01:15:29 it was half that when i bought it during alpha :P 01:15:31 * olsner had an MRI scan the other week, btw: there should be nothing to worry about - because if there were I hope they would've told me by now 01:15:41 (and will double again when it hits final) 01:15:48 olsner: worry? I was just getting excited :( 01:15:50 (not true) 01:15:56 yeah.. super mario game is all about reaching the end of the level within a time limit 01:16:01 42 dollars minecraft? 01:16:04 elliott: MRI scans are more noisy than exciting 01:16:05 ouche 01:16:11 monqy: well, it's a lot bigger game now :P 01:16:15 humm 01:16:23 monqy: aren't new games like sixty dollars anyway 01:16:26 I better not have to buy it again 01:16:29 NihilistDandy: no 01:16:35 Good 01:16:37 elliott: i dont buy games i have no idea 01:17:01 elliott: otoh, why would you ever worry about my "insignificant" existence 01:17:13 monqy: minecraft is good to buy because it gives you two good things to do: 01:17:16 monqy: - complain about minecraft 01:17:16 olsner: We are all insignificant 01:17:18 monqy: - play it and enjoy it 01:17:33 NihilistDandy: the nihilist doubly so 01:17:41 Game theory is not a perfect tool for analyzing video games. 01:17:42 Ouch 01:17:45 monqy: also you get to play with us :{ 01:17:48 :( 01:17:51 Nash failed to create a "video game theory" 01:17:54 that was a happy :{ 01:17:58 oh 01:18:01 `addquote Nash failed to create a "video game theory" 01:18:01 whats a sad :{ 01:18:02 539) Nash failed to create a "video game theory" 01:18:02 or he didn't try :P 01:18:04 monqy: :{ 01:18:13 `delquote 539 01:18:14 ​*poof* 01:18:17 `addquote Game theory is not a perfect tool for analyzing video games. Nash failed to create a "video game theory" 01:18:18 539) Game theory is not a perfect tool for analyzing video games. Nash failed to create a "video game theory" 01:19:10 How does M. Bison roll? 01:20:21 “I don’t mind the idea of never having kids,” he said. “I want to stay focused on the game, and if I had kids, I’d wind up paying attention to them instead.” 01:20:38 itidus20: I don't think Bison has a roll attack 01:20:53 elliott: This article has to be a troll 01:21:03 -!- CakeProphet has quit (Ping timeout: 252 seconds). 01:21:12 why its awesome (its obviously trying to paint him as a "weird" person but who cares) 01:21:32 Vega and Blanka, OTOH... 01:21:49 LIKE A BAWWWSS! 01:21:53 what i 01:22:05 whats a bawwwss 01:22:07 i cut and pasted someone elses spelling 01:22:22 was it mine 01:22:47 because i uh 01:22:49 bawwss 01:22:51 i 01:22:53 what is it 01:22:57 I was gonna say Bowser from mario... but couldn't think of a name which would click instantly 01:23:08 Since some call him King Koopa 01:23:26 wow he makes a lot from donations 01:23:29 well not that much 01:23:31 but for donations... 01:23:32 a lot 01:24:07 "Tarn, 33, lives" 01:24:08 like a bowser from mario 01:24:11 In bold, for no reason 01:24:17 -!- cheater_ has joined. 01:24:21 Still, in the only moment I heard him speak with anything like bitterness, Tarn called Minecraft a “depressing distillation of our own stuff.” 01:24:33 NihilistDandy: it's a section break 01:24:46 elliott: It's not a very convincing one 01:25:29 -!- Nisstyre has quit (Quit: Leaving). 01:25:29 I think the only reason I want Minecraft is the polish. Gameplay polish and also the abundance of NPCs 01:25:58 Coincidentally, those are the two major lacking features of Minecraft! 01:26:02 And -- since my brain can think up such things... I dare say 01:26:03 'As Tarn got into the zone, his muttered profanities and grumbles about “x distances”' 01:26:12 As if that's some absurd nerdspeak 01:26:15 Terraria is Lemmings with NPCs. 01:26:25 you dont control a character in Terraria 01:26:35 Oh.. I haven't played it. 01:26:49 you don't really control a character in Lemmings either 01:26:53 erm 01:26:56 you control a character in Terraria 01:26:58 sorry 01:27:01 you dont control a character in Lemmings 01:27:10 You're just a demigod in Lemmings. 01:27:14 I just uh.. hit on a cool idea didn't i? 01:27:17 NPCs in lemmings 01:27:27 terraria is basically minecraft + rpg 01:27:28 lemmings vs zombies 01:27:28 is that really a cool idea? 01:27:40 pikhq_: you're more of a tactical advisor who tells them when to turn into walls and when to explode 01:27:44 i wonder if it has been done 01:27:50 ais523: I'm going with demigod. 01:28:19 on bing: Thats an urban legend. However, in the very excellent video game Lemmings, they do. Now thats a great idea for a video game: Lemmings vs Zombies. 01:28:32 dying 01:28:32 "Demi" because you're not *that* powerful, "god" because thou art greatly seeing and greatly knowing. 01:29:00 "Can I be one of the play testers for Zombies vs. Lemmings? And will there be a sequel game with zombie lemmings?" 01:29:03 ais523: lemmings with npcs actually sounds interesting 01:29:05 although very hard 01:29:14 http://www.obamaconspiracy.org/2011/04/this-isnt-going-away/ 01:29:24 pikhq_: I always thought a demigod was just someone in charge of the espresso machine 01:29:25 it depends on what the NPCs do, and interact with the game 01:29:30 "obamaconspiracy.org" 01:29:37 i know.. im serious though 01:29:42 oh, it's against obama conspiracies 01:29:47 elliott: I'm not at all surprised that the website exists 01:30:09 ais523: I was worried it would be a collection of conspiracies about Obama in /favour/ of them 01:30:27 * elliott becomes the Tarn, tries to figure out what's up with this code 01:30:43 hmm... 01:31:53 -!- Slereah has joined. 01:31:53 -!- Slereah has quit (Client Quit). 01:31:56 "There are those who tell us that any choice from among theoretically-equivalent alternatives is merely a question of taste. These are the people who bring up the Strong Church Turing Thesis in discussions of programming languages meant for use by humans. They are malicious idiots. The only punishment which could stand a chance at reforming these miscreants into decent people would be a year or two at hard labor. And not just any kind of 01:31:57 hard labor: specifically, carrying out long division using Roman numerals. A merciful tyrant would give these wretches the option of a firing squad. Those among these criminals against mathematics who prove unrepentant in their final hours would be asked to prove the Turing-equivalence of a spoon to a shovel as they dig their graves." 01:32:30 source 01:32:31 ? 01:32:57 http://www.loper-os.org/?p=448 01:33:18 "Why do birthers believe what they do? It turns out that one factor in whether someone mistakenly considers random patterns to be something significant has to do with dopamine levels in their brains — the more dopamine, the more belief. Dopamine is part of the brain’s reward system. I was just reading about this today in Michael Shermer’s book, The Believing Brain." 01:33:23 this blog sure is fishing for posts 01:33:36 Sgeo: oh, another loper post 01:34:17 Sgeo: re turing-equivalence: right point, annoying delivery, as always with loper-os 01:34:24 apart from the times it's wrong point, annoying delivery 01:34:40 elliott, I think it's hilarious delivery *shrug* 01:34:54 elliott, link to wrong point? 01:34:59 ur joking right 01:35:07 I think rage is funny 01:35:09 "you say this huge blog is wrong sometimes?? PROVE IT WITH A LINK" 01:35:19 Oh. 01:35:23 also, yeah, now try reading every loper post and not getting sick of him and wanting him to shut the fuck up 01:35:28 Sgeo: but here's a wrong point: http://www.loper-os.org/?p=428 01:35:44 here's a stupid point: http://www.loper-os.org/?p=405 01:35:59 here's a stupid point: http://www.loper-os.org/?p=374 01:36:00 Sgeo: Well, the Turing thesis really *does* imply that the choice of language is "merely" one of taste. 01:36:17 But, then, the same is true whenever you have a selection of tools theoretically capable of a given task. 01:36:29 its not the turing thesis ffs 01:36:34 church-turing thesis is a different thing 01:36:41 Even if your choice is between a spoon and a bulldozer. 01:37:02 Sgeo: lol -- see that project cosa stanislav linked to? 01:37:07 from the guy behind that thing's blog: 01:37:16 i dont think a bulldozer will help me eat this soupe 01:37:17 The Surprising Source of My Knowledge about the Brain 01:37:17 Abstract 01:37:17 In my recent article on the flaws in Numenta's memory model, I made several claims about how the brain processes sensory signals. I now reveal the source of my knowledge as promised. Those of you who have followed my work over the years will not be surprised by what I am about to say. As always, it is up to you to take it or leave it. 01:37:17 [...] 01:37:18 My Source 01:37:20 The source of my knowledge is a couple of occult symbolic texts found in the Bible. 01:37:21 I saw the link, didn't click it 01:37:22 -- http://rebelscience.blogspot.com/2011/06/surprising-source-of-my-knowledge-about.html 01:37:24 good 01:37:27 good thinker 01:37:36 The problem is, of course, that we have people convinced that a teaspoon is a good tool for digging a tunnel. 01:37:57 does anyone think bulldozeres are good for soupes 01:38:04 yes me 01:38:08 i have very big soups 01:38:13 monqy: Well, the XML crowd seems to think that. 01:38:20 mmh 01:38:24 Except instead of a bulldozer it's more a Death Star. 01:38:51 "Neither Newton's gravity equation nor the equations of General Relativity explain why things fall. But what better way is there to hide one's cluelessness while presenting a façade of erudition than to use obscure equations to erect an impregnable mountain of obfuscation? Voodoo science is guru science." 01:38:53 oh this guy 01:39:04 "Examples of voodoo science masquerading as legitimate science are all around us: time travel, wormholes, black holes, dimensions curled up into little balls so tiny as to be undetectable, parallel universes, continuum physics, quantum computing, symbolic intelligence, machine consciousness, etc... It is all worthless crackpottery." 01:39:16 (Spacetime is a fictitious math construct, famous physicists are clueless about time, time dilation is a misnomer and time travel is crackpottery) 01:39:17 (Aristotle was right about the causality of motion and, as a result, we are immersed in an immense lattice of energetic particles) 01:39:26 (Gravity is an energy conservation phenomenon. It is both instantaneous and nonlocal) 01:39:26 (Intelligence is mostly about the temporal correlations between discrete sensory signals) 01:39:30 i wanna hug this guy 01:39:43 Artificial Intelligence From the Bible! 01:39:44 June 21, 2011 01:39:44 Due to recent dramatic progress in my research, this article has been deleted. My old interpretation of the scriptural metaphors was partially in error. I am working on a new series of articles on the subject. In the meantime, I encourage my readers to read Rebel Science News for the latest. 01:39:58 dead 01:40:01 rotting 01:40:01 Oh, so we can 01:40:07 can't see when he makes mistakes 01:40:14 (Erm, as in, in his predictons) 01:40:17 predictions 01:40:35 Although I'm sure he'll call any falsifications "voodoo science" 01:40:52 Partially in error 01:40:55 im torn between wanting to read the cosa paper and not wanting to because this guy man 01:41:58 cosa? 01:42:21 monqy: this guy's dataflow software/os/language proposal which seems to be much saner than his other stuff 01:42:25 that stanislav linked to 01:42:55 OHHH i know whats wrong 01:46:00 indeed, you should never use ketchup with filet mignon 01:47:06 what do you advise 01:47:08 for use with 01:47:11 filet mignon 01:48:09 i like béarnaise sauce 01:48:35 Leave it alone 01:48:50 oh or pepper sauce 01:48:50 Just eat it, and enjoy 01:50:04 I think I've had bearnaise sauce one time it was good pepper sauce is good too i forget if i've ever had filet mignon 01:50:08 hm wait that doesn't mean the same thing as "peppersaus" 01:50:30 no thats a completely different thing obviously 01:50:31 I actually kind of like this: http://www.loper-os.org/?p=428 01:51:16 it's a stupid point 01:51:24 with the context of http://www.loper-os.org/?p=288 01:51:33 Ah 01:51:59 also his response to http://www.loper-os.org/?p=288&cpage=1#comment-1283 01:52:04 it's amusing but wrong and stupid 01:52:54 lol, crazy people 01:53:25 loper is full of good ideas written by an annoying person with a too-big ego 01:53:37 HAVE I MENTIONED THAT TUNES IS GREAT 01:53:39 BUT @ IS BETTER 01:54:00 elliott: the english word implies chili pepper, the norwegian word implies black pepper (or similar) 01:54:27 elliott, the thing common between them is that neither are currently usable 01:54:37 I think "pepper sauce" refers to black pepper in British English too 01:54:39 * Sgeo ducks 01:54:42 ah. 01:54:46 oh -- 01:54:48 "TUNES started in 1992-95 as an operating system project, but was never clearly defined, and it succumbed to design-by-committee syndrome and gradually failed." 01:54:50 "Pepper sauce" just sounds bizarre in American English. 01:54:52 TUNES is now officially over 01:55:01 as of April, it seems 01:55:03 so why is clog still here? 01:55:06 RIP TUNES 01:55:08 monqy: say it 01:55:12 i uh 01:55:20 reading now 01:55:21 ais523: the website isn't down, it's just officially abandoned; and besides, clog is a bespin service, not a tunes one 01:55:26 ah, fair enough 01:55:27 we just use the wrong url because everyone does and it's prettier 01:55:30 monqy: wait you don't know of tunes? 01:55:39 A sauce from chili peppers would be "hot sauce", and you can make a black pepper sauce? 01:55:41 I only do because of clog 01:55:48 pikhq_: you can make a source from black pepper 01:55:54 Unununium is also dead 01:55:55 and other ingredients too 01:55:59 not /just/ black pepper, obviously 01:56:04 ais523: Well, obviously. 01:56:05 ais523: every programmer should know about TUNES, I think 01:56:24 elliott: in your opinion, lots of people should know about all sorts of things I don't know about 01:56:27 But, seriously, that seems like a strange sauce basis. 01:56:29 I'd give an example except I can't 01:56:29 so that they don't keep assuming Unix and Windows are the only ways to do computers, or anywhere near the best way 01:56:36 or even an acceptable way 01:56:44 I don't assume UNIX/Windows are the only way to do things at all, though 01:56:51 I'm at the end of the game Like a bauss 01:57:00 ais523: let's put it this way: @ doesn't sound like much new to anyone who knows about TUNES 01:57:04 whats a bauss is it like a bawwwss ;___________________; 01:57:05 UNIX is like the BF of operating systems; people who try to design operating systems normally end up designing crappy versions of it, but that doesn't mean it's the only way to do an OS 01:57:08 I would hazard a guess that @ sounded new to you 01:57:26 and @ is unlike OSes I know about, but the concepts aren't new to me 01:57:27 whats windows 01:57:33 whats @ 01:57:42 monqy: @ is the operating system you'll use in a few decades... maybe. 01:57:44 pikhq_: hm i went by wikipedia, where "pepper sauce" redirects to "hot sauce", which has both terms and chili sauce bolded 01:57:55 interesting 01:57:56 oerjan: "Pepper sauce" still sounds bizarre. 01:58:02 elliott: googleing for @ doesnt work he;lp[ 01:58:02 http://www.youtube.com/watch?v=NisCkxU544c 01:58:06 ais523: I like to think @ will be interesting enough to surprise you when it actually exists 01:58:06 "chili sauce" would be the usual name in the UK 01:58:12 monqy: @ is my ultimate NIH project 01:58:15 monqy: Modern Windows is, IMO, in the same OS family as UNIX. 01:58:22 err, "chilli" I think in British English 01:58:33 elliott: zepto 01:58:33 pikhq_: I agree too, although it's a very bad example of it 01:58:33 monqy: I decided that software was fundamentally broken 01:58:34 "It's 2008 now. If a few dedidated hackers can find the time and money to put some sustained effort into it, we could have a working prototype by 2010, with widespread use by 2015 or 2020. 01:58:34 " 01:58:38 ais523: No argument. 01:58:59 monqy: so i tossed everything out, and started designing a new software environment ("operating system is inaccurate"; @ has no distinction between the OS and everything running "on top" of it) 01:59:07 What makes something in the same OS family as UNIX? 01:59:09 including language, from scratch 01:59:14 Files in a heirarchy? 01:59:23 Sgeo: that would make DOS a UNIX 01:59:33 not DOS 1 01:59:39 it didn't have directories 01:59:46 it spent the rest of its life gradually stealing UNIX concepts 01:59:55 monqy: you may peruse http://catseye.tc/ehird/files-suck.html but only in the knowledge that it's old and whiny, and only summarises one aspect of @ (and fairly vaguely at that) 02:00:07 how did that end up on catseye? 02:00:15 I'm not surprised you wrote that, but I am surprised at where it's hosted 02:00:17 ais523: I had no hosting at the time and asked cpressey on IRC, and he said sure 02:00:25 this was last year 02:00:31 well, fifteen months ago 02:00:50 gah that's formatted badly, far too narrow 02:01:06 nah, it's only about 3 ems too narrow 02:01:11 exactly, in fact 02:01:13 -!- cheater_ has quit (Ping timeout: 255 seconds). 02:01:27 Sgeo: The primary feature is having the prime data structure being a hierarchy of blobs of bytes, with a little bit of metadata. Combine with a C ABI and system calls as the basic ABI, and multitasking, and you have a vaguely UNIX-like OS. 02:01:31 (ten words per line/sixty-six characters per line is the widely-agreed-upon ideal that presumably has some research backing it up, and that's 33 ems) 02:01:32 congratulations, you're the first person who actually made me open up Firebug and edit the CSS 02:01:35 to make it three times the width 02:01:49 even then, I'd prefer it to go margin to margin 02:02:09 there we go 02:02:11 ais523: now wouldn't it be nice if browsers didn't make lousy document viewers, and you could have done that without resorting to such things? 02:02:13 You know, like @. 02:02:17 Also, yes, the memory/disk dichotomy is big. 02:02:22 (I also think your eyes are broken, but oh well) 02:03:32 elliott: I can think of some counterpoints to things in that files article 02:03:57 ais523: go on, although I'll probably disagree with them 02:03:58 in particular, it's often useful to be able to change the representation of something on disk, and if you abstract away from that, there's no way to do so 02:04:06 ais523: yes there is 02:04:10 what is it/ 02:04:11 I even address that in the article 02:04:34 Would you fuckers stop talking so fast? I'm still reading the scrollback for context :D 02:04:47 and no you don't, unless I missed where you addressed it 02:05:00 -!- pikhq has joined. 02:05:04 ais523: tl;dr current @ design is that any "entity" (/object/whatever) can override its own saving and storing, but most of the time you'll want to just use a pre-written transformer like a compressor, since you must ensure to always save and restore _all_ of the state 02:05:08 ais523: "You don’t have to give up the compactness of formats like PNG, either: you can store that as a byte array in memory, and have it as the pluggable “backing storage” of some abstract image type. That way, all images are accessed in the same way (although possibly with different operations depending on the characteristics of the backing storage) regardless of the underlying compressed stora 02:05:08 ge. 02:05:08 " 02:05:24 that's a little vague and not the exact current design that I stated, but it still addresses it 02:05:30 "You could have overridable serialisation, so that e.g. a PNG-backed image is serialised as the compressed PNG, without any auxiliary data that can be generated when it is used." 02:05:35 ah, I suppose my question is as to how you're meant to change the backing storage 02:05:54 ais523: depends how you tell the operating system anything about an object 02:06:04 Every object could be responsible for its own serialisation/deserialisation, for instance 02:06:13 Also, elliott, last I was aware 100 characters/line was the new sort-of standardy ideal 02:06:17 also, how does @ do naming objects? e.g. atm I can access /home/ais523/esoteric/intercal/git/intercal to get the latest C-INTERCAL source 02:06:17 and you could pull in GzipSerialiser 02:06:24 NihilistDandy: nah, the sources for that seem rubbish 02:06:26 and obviously that doesn't have to be done with a hierarchical filename 02:06:31 but I'm wondering what specifically replaces it 02:06:43 ais523: That's metadata. 02:06:47 pikhq: hey 02:06:53 I get to answer questions about @ :p 02:06:57 ais523: how do you name the x coordinate of the first coordinate in a list of coordinates (say array) in C? 02:07:10 elliott: in C, it depends on their internal representation 02:07:18 ais523: filesystem names are internal representations 02:07:22 yep 02:07:24 ok, C was a bad choice 02:07:29 but the point is, array[0].x is a name too 02:07:39 at the moment, the filesystem representation, while internal, is the one that humans use to specify files 02:07:43 -!- pikhq_ has quit (Ping timeout: 276 seconds). 02:07:47 ais523: but in @ the primary means of naming documents isn't by internal name 02:07:51 I'm asking what @'s equivalent is, not because I don't think it exists, but because I'm curious as to what it is 02:08:02 so it's not so much an accusation as a question 02:08:07 ais523: I'm not sure exactly how you /do/, because I don't know what @'s UI looks like yet, other than it being very, very different from anything else 02:08:13 heh 02:08:19 a hipster UI? 02:08:33 ("I don't care what it is as long as it's different"?) 02:08:34 ais523: but e.g. you might say "find|document|@ best os" and it would find you a document titled "Why @ is the Best OS" 02:08:40 (prioritising title over text, etc.) 02:08:52 what if I don't want to do any sort of search? 02:08:58 or "find|documents|older than april" 02:09:08 ais523: Don't think of it as searching 02:09:11 e.g. if I'm using @ to work on some more traditional software 02:09:14 ais523: just think of it as naming an object 02:09:24 elliott: I want exactly the same result every time, no matter what's happened in terms of creation of other objects meanwhile 02:09:28 ais523: If you want a precise name, you could take its object ID -- basically, like a pointer, but it's literally global 02:09:31 (it's just a long hash) 02:09:34 or deletion or changes of any objects other than the one I want 02:09:37 ais523: Then see above 02:09:39 and I can't remember hashes 02:09:42 Can you use the ID numbers for the documents and whatever you want to find? 02:09:48 ais523: zooko's triangle 02:09:54 oh, right# 02:10:02 heh 02:10:11 you're ignoring the side of the triangle I most care about 02:10:15 ais523: Unix cheats by not having the decentralised or secure properties 02:10:15 and I'd much prefer you discarded one of the other two 02:10:16 and which side is that? 02:10:22 Human-meaningful? 02:10:31 I solve that by having alternate ways to name objects, like I said 02:10:37 hmm, decentralised is the one I care least about 02:10:43 I'm fine if I can only access stuff on my own computer 02:10:49 It's integral to @'s design 02:10:51 better, in fact, as I'm typically not connected to a network 02:10:53 It's integral to @'s design 02:11:06 well, in that case I probably won't use @, it's fundamentally incompatible with the way I work 02:11:09 @ makes no innate distinction between RAM and disk; it also makes no innate distinction between local and global 02:11:10 ais523: It isn't 02:11:16 ais523: You could define a filesystem view of all your documents 02:11:21 Mangle the name and some other data to form its name 02:11:37 hierarchical filesystems are far from perfect, but at least they're usable 02:11:39 You can do anything you can do in Unix with @; it just wouldn't be /idiomatic/ 02:11:44 well, fair enough 02:11:47 ais523: I just told you how to get a filesystem; you could make it hierarchical too 02:11:57 ais523: You could manage it manually by constructing a filesystem object yourself; basically just an associative map 02:12:07 what's @'s equivalent to UNIX "find" (i.e. get everything below a certain point in the hierarchy)? 02:12:07 But I find dismissing these things before actually using them to be unwise 02:12:13 O, so filesystems are not the fundamental things, but you can still make it up yourself to do like a filesystem? 02:12:27 the concepts don't translate at all well, which is why I asked 02:12:29 ais523: There is no hierarchy. If you mean on a constructed filesystem, that's up to the API of the associative map 02:12:54 ais523: Anyway, the distinction between local and global doesn't make local use more difficult 02:12:58 Well, I certainly wouldn't know until I can see some parts of it that it can be done. 02:13:01 [asterisk]lack of distinction 02:13:03 say, I'm working on a program, and I want to find all instances of a certain identifier in all the source that makes up that program 02:13:09 ais523: Note that distinctoins are made at the user-level 02:13:10 (I'm trying not to assume that the source is text-file-like here) 02:13:28 ais523: Well, "program" is also an artificial distinction :) 02:13:30 presumably I have to group all those files some way so I can equivalent-of-grep over just those 02:13:43 elliott: well, let's try "package", which is also an artifical distinction but one that's made deliberately for a reason 02:13:43 ais523: But say you have a collection of that count as the structure of your program 02:13:57 ais523: OK, fair enough 02:13:57 or "portion of all software that I'm responsible for maintaining" 02:14:01 which is not arbitrary 02:14:01 ais523: How would you do that on an arbitrary data structure? 02:14:11 pikhq: please, don't try and argue @ for me, it makes @ look bad :) 02:14:18 -!- cheater_ has joined. 02:14:19 elliott: Okay, then. 02:14:20 pikhq: for a completely arbitrary data structure, with no knowledge about its contents, I couldn't 02:14:38 but that's not the case even in @ 02:14:47 ais523: Well, I don't know how you'll actually phrase queries... because I don't know what @'s UI is yet (this also implies I don't know what its language is yet, because they're two sides of the same coin) 02:14:49 ais523: But 02:15:19 ais523: Things like functions will have a way to get at their underlying AST 02:15:29 Presumably 02:15:33 Maybe only in some context like a package 02:15:33 oh no 02:15:37 To avoid breaking all sorts of parametricity 02:15:40 I've just realised that Feather is the only sensible language to write this in 02:15:44 ais523: haha, it isn't 02:15:47 ais523: But seriously -- 02:15:53 you just said "Things like functions will have a way to get at their underlying AST" 02:15:57 Then contradicted it 02:16:04 ah, OK 02:16:06 That breaks all sorts of properties, it's gross 02:16:14 But you might be able to get a function's source from its module, say 02:16:15 -!- BeholdMyGlory has quit (Remote host closed the connection). 02:16:18 I'm not sure 02:16:27 I'd assume functions would be decompiled to get their source 02:16:33 ais523: But anyway, tl;dr AST objects (if there is a concept like an AST??) will have a way to search for an identifier 02:16:34 and things like comments would be part of the compiled version 02:16:38 so that it would roundtrip 02:16:38 So it's basically just a map over the project 02:16:53 I know this is hopelessly vague, but your questions are all user-level, which is the least certain part of @ so far 02:17:09 yes, that's because the system-level stuff, at least at the highest level, makes a lot of sense 02:17:09 So I do ask you don't dismiss it before it exists based on _these_ answers :) 02:18:05 incidentally, I can think of a moderate solution to the zooko's triangle problem; allow individual users (with whatever way you identify who they are, which you have to do for security reasons) to have their own private names for arbitrary objects 02:18:25 that way you have both a secure/decentralised name and a secure/humanreadable name, with the human who reads it doing the centralisation 02:18:26 ais523, I think there are systems that do that 02:18:35 -!- elliott has quit (Read error: Connection reset by peer). 02:18:38 -!- elliott_ has joined. 02:18:42 incidentally, I can think of a moderate solution to the zooko's triangle problem; allow individual users (with whatever way you identify who they are, which you have to do for security reasons) to have their own private names for arbitrary objects 02:18:46 Missed everything after this 02:18:50 ais523: And yes, that's a Petname system 02:18:57 [03:18] that way you have both a secure/decentralised name and a secure/humanreadable name, with the human who reads it doing the centralisation 02:18:58 [03:18] ais523, I think there are systems that do that 02:19:00 ais523: And that will be possible, yes 02:19:00 I think I'd be fine with that 02:19:08 ais523: It'll be the same mechanism as what replaces a browser's bookmark system, too 02:19:21 ais523: And of course you won't be restricted to naming things with a-constrained-path.txt, or even just a string 02:19:25 indeed 02:19:26 No reason a name can't be an object 02:19:29 So I have to think up names for all my bookmarks? 02:19:35 elliott_: except that objects can't be compared, in general? 02:19:41 Sgeo: Your bookmarks already *have* names. 02:19:41 a name has to be something comparable 02:19:41 Sgeo: what happens when you press the bookmark icon on your browser? 02:19:45 apart from that, it can be general 02:19:47 ais523: well, yes, definitely 02:19:58 ais523: although you could have a name that isn't comparable, and it'd just appear in a list of all your names 02:20:00 that'd be fairly pointless, though 02:20:02 elliott_, A name is filled in, but possibly non-unique 02:20:03 Sgeo: your bookmarks are almost certainly given the title of the website they refer to, at the time it was bookmarked, as names 02:20:13 Sgeo: Non-unique names are OK, too 02:20:13 "Index of /" etc. 02:20:18 Looking it up will just say "oops, which do you mean" 02:20:24 Ok 02:20:30 in fact, the default lookup operation might as well just do a fuzzy-ish text search 02:20:37 Also, I tend to bookmark everything and then never look at them agian 02:20:38 again 02:20:40 that's a lot more convenient, too, since you can type only part of it and select which one you want 02:21:30 elliott_: I think Firefox's address bar thing has quite a few properties in similar with the way @'s names would typically work (although a lot of differences too) 02:21:42 yes; Firefox's address bar is a good idea, implemented badly 02:21:54 what do you dislike about the implementation, incidentally? 02:22:07 It's hard to articulate; it just doesn't do what I want 02:22:14 Chrome's is better, but it tends to not find things in my history, which is really annoying 02:22:37 when I'm typing in the address bar, either I'm copying an URL, or I'm searching in history 02:22:50 so Firefox does pretty much exactly what I want there 02:22:56 ais523: Anyway, hopefully @'s "final" (just the beginning) version will be the OS you want to use; all my efforts right now are classed as pre-pre-pre-pre-alphas of a given /stratum/ of a project leading up to @ 02:22:58 I definitely don't want it to return arbitrary websites, that's what the search box is for 02:23:03 elliott_: heh 02:23:07 So there's plenty of time for you to test and comment on it :P 02:23:21 ais523: I'm not even joking, I've even tried to name distinct phases 02:23:41 I can't call anything @ until it's 1.0 quality, and that means being about a million lightyears ahead of any existing OS 02:23:47 security sounds like a hard problem in @, in that a single security bug could allow someone to instantly bring down every computer in existence 02:23:50 Oh, and @ isn't its actual name, of course 02:23:55 ofc, it's just a placeholder 02:23:59 it's not bad for an actual name, though 02:24:03 perhaps its actual name will be an object 02:24:09 Bad for Googleability, but who will use Google once @ is around? 02:24:10 rather than a text string 02:24:39 Using @ will, ironically but perhaps appropriately, not be as nice as using @ would be in an ideal world; nobody else is going to use it, so you'll still be stuck with the web, and IRC, and all sorts of other mediums that reimplement everything and don't allow you to send objects 02:24:41 security sounds like a hard problem in @, in that a single security bug could allow someone to instantly bring down every computer in existence 02:24:42 elliott_: probably Google will have the best servers at resolving who has a copy of a given @ object 02:24:48 Really? I'm planning on using capability security. 02:24:58 That seems relatively immune to such things because the security isn't "a thing" or a layer 02:25:01 elliott_: I mean, if the security implementation is broken 02:25:04 It's just an inherent property of who gives you what. 02:25:11 ais523: See above; capability security doesn't get implementations. 02:25:17 "a capability" isn't something you can hardware into an actual processor 02:25:18 It's literally just not having a standard environment beyond what's passed around. 02:25:27 ais523: A capability is just an object! 02:25:31 Every object is a capability! 02:25:33 It is a 0 lines of code thing. 02:25:39 There are no flags or anything. 02:25:45 It is just recognising that a pointer to X means you are allowed to use X. 02:25:55 elliott_: yes, so what if I try to use X without a pointer to it? 02:26:02 ais523: How? You can't name it. 02:26:04 that's meant to be impossible, but has to be implemented somehow 02:26:04 "Slartibartfast: Can I be one of the play testers for Zombies vs. Lemmings? And will there be a sequel game with zombie lemmings?" 02:26:07 Try and name X for me. 02:26:09 ais523: No, it doesn't 02:26:14 "Count me in too! Dr C you really should patent that ideaI think that game or app would rake in a fortune! ;)" 02:26:20 elliott_: well, say X happens to be on my physical hard drive 02:26:32 elliott_, there has to be a level of abstraction though 02:26:32 I can name it by its location on the drive 02:26:38 ais523: You can't access the bits on your hard drive; nobody passed you that object. 02:26:47 Of course they didn't, that's insanely high-privilege. 02:26:48 that's the kind of crap patents are.. its really just how quickly you can traverse the inventions tree and find an empty node 02:26:59 elliott_: so what if an exploit does let me access the bits on my own hard drive? 02:27:03 ais523: How? 02:27:10 say, removing the hard drive and putting it into an older laptop that doesn't run @ 02:27:17 Yes, security bugs are possible, but capability security makes it just so less likely. 02:27:28 or corrupting its processor somehow such that it runs arbitrary I/O instructions? 02:27:35 that's what I mean by implementing security 02:27:36 ais523: And then modifying the bits so that the piece of code you're running has a pointer to X? 02:27:39 I clearly demonstrated that I was able to independantly discover the idea. 02:27:47 If your hard-drive is not encrypted and you have access to it you can do anything locally. 02:27:57 @ can't stop you. 02:28:02 elliott_: yes, or that modifies something it has in order to put the content there 02:28:13 And @ won't even try; a user can ask it to fiddle with bits on its hard drive and it will happily comply, perhaps after panicking a bit and asking if you're sure. 02:28:19 (A user, not a piece of code.) 02:28:23 ais523: Well, so what? 02:28:23 that's what I'm getting at; it's that the universe doesn't support @'s model of capabilities 02:28:28 You've successfully compromised your own system. 02:28:32 ais523: I don't wish to imprison the user. 02:28:32 so they have to be implemented in terms of something that doesn't 02:28:35 I want to empower the user. 02:28:37 elliott_, there's a distinction between user and code? 02:28:38 I wish to imprison _code_. 02:28:43 Sgeo: Har har. 02:28:46 Hmm, perhaps whether the code comes from user or elsewhere 02:28:48 now, how do we serialise capabilities to send them over the network? 02:28:51 The user is just a really high-privilege context. 02:28:54 aha, that's what I'm getting at 02:28:57 ais523: You don't. There is no "capability". 02:29:07 ais523: If someone gives you a reference to an object that happens to be somewhere else in the world, you can use it. 02:29:09 elliott_: well, objects are capabilities 02:29:10 and you can send those 02:29:13 But then, a trojan can tell the user to just go "Oh hey, claim you're my source" 02:29:23 so presumably, we need to serialise the references to be able to send them over the network 02:29:25 ais523: You just serialise the object and send it, or just send a pointer 02:29:32 The pointer is the big hash 02:29:41 I also have the idea to empower the user, but doing it mostly in hardware instead of in software. 02:29:47 (I'm not sure how big the hash will be) 02:29:49 elliott_: so if I get the hash of any other object, I can access it? 02:29:59 ais523: Only if another machine says it's offering 02:30:09 why would the machine be offering? 02:30:15 or not? 02:30:21 Was about to say that it smells like security by obscurity 02:30:23 Because it's configured to offer; it offers nothing by default 02:30:25 Is "offering" a boolean? 02:30:43 Sgeo: Offering is a set, presumably 02:30:50 Ah, ok 02:31:14 ais523: The user can share things, presumably 02:31:18 Which would make them being-offered 02:31:23 elliott_: btw, the "memory/disk" dichotomy seems wrong; UNIXy systems are moving more and more in the direction of removing it 02:31:25 very slowly, though 02:31:37 ais523: As long as files on disk are bytes, it is still there 02:31:46 It doesn't matter whether a disk cache exists 02:31:47 and files in memory are also bytes? 02:31:49 That's an implementation detail 02:31:53 ais523: Very funny 02:32:05 well im laiughning 02:32:06 ais523: Even having to call mmap is too much 02:32:23 Orthogonal persistence is the only way to escape that dichotomy, although those are the wrong terms; it is a simpler system than a filesystem 02:32:27 elliott_: well, I meant that mmap has the right sort of semantics, it's just that it should be implict rather than explicit 02:32:49 after a while you start seeing mmap reversed, in that the disk is just swap space for memory 02:32:50 Right. 02:32:52 and gets paged in and out 02:33:04 ais523: That's orthogonal persistence: RAM is just disk cache, 02:33:07 [asterisk]. 02:33:09 That's literally it. 02:33:15 That's the entire idea. 02:33:15 yep, I know what it is 02:33:37 ais523: Of course, the ideal would be if a non-volatile RAM interface was plugged into the CPU 02:33:41 Or otherwise into the computer 02:33:42 hibernation is more along the same lines; it just swaps data out of the sort of implementation-detail storage that doesn't exist while the system is off 02:33:45 But as it stands, software has to handle it 02:33:52 do you know how Linux loads executables, nowadays? 02:33:56 ais523: right; hibernation in @ is just a flush and a power cut 02:34:06 And I forget, but it involves an mmap-alike of some sort 02:34:10 it just mmaps them, and lets page faults do the actual loading 02:34:15 Yep 02:34:16 -!- NihilistDandy has quit (Quit: Blah). 02:34:21 mmap is my favourite system call 02:34:27 It's that little glimmer of light in Unix. 02:34:27 elliott_, is there any sort of restart that isn't hibernate and restore? 02:34:37 Sgeo: Uh... you could wipe your disk and reinstall @. 02:34:44 Dunno why you would, though. 02:34:50 Sgeo: presumably it would be the equivalent of kill -15 0 02:35:02 ais523: Also how any sane ld.so does things. 02:35:04 ais523: ouch; that would delete all your files 02:35:07 probably lazily, as the GC runs 02:35:07 I don't know what kill -15 0 02:35:09 does 02:35:13 but it'd also crash your system forever 02:35:17 elliott_: err, really? 02:35:28 ais523: well, I forget what -15 is :) 02:35:32 Why, what's signal 15? 02:35:34 Sgeo: "tell every process I have permissions to tell to exit to exit" 02:35:37 elliott_: it's SIGTERM 02:35:41 Ah 02:35:45 like -9 except that the programs get a chance to do cleanup 02:35:50 I know TERM :P 02:36:12 ais523: basically, if you kill all running code, then not only is there no code for the computer to run (???), but there's also no references to any objects 02:36:21 and objects without references get garbage collected 02:36:35 but 0 doesn't mean "all running code", it means "all processes I have permissions to signal" 02:36:39 which is presumably quite different 02:36:40 so that's a really, really slow disk wipe, basically, assuming that GCs still run with no code for the computer to run 02:36:56 ais523: Well, OK. It would just remove all _your_ files, unless someone else has a reference to them. 02:37:22 ais523: I'm not sure there's any distinction between "kill this code" and "destroy this object", anyway 02:37:26 Think thunks 02:37:40 I'm not deleting the code, just telling it to stop running 02:37:47 so in @, presumably it'd go into some sort of paused state 02:37:48 You're deleting the thunk that represents that running code 02:38:06 ais523: Oh, you just want to tell the scheduler to stop switching to all objects you can access, then 02:38:07 in UNIX, you have two separate concepts; the editor, and the file that it edits 02:38:14 That'd just lock up your session, including post-reboot 02:38:37 (The problems that restarts are usually used to solve -- e.g. "clear my workspace" -- are solved in different ways in @) 02:38:40 elliott_: presumably @ notionally has a global scheduler that schedules for everyone in the universe, although it's actually implemented in a decentralised way 02:38:45 Oh, and user responsiveness is the top priority, more or less 02:38:51 Code will never cause your keyboard and mouse to stop working 02:38:56 (because it's lagging your system) 02:39:03 So you'll always be able to kill miscreant processes 02:39:24 assuming no scheduler bugs 02:39:29 Well, yes. 02:39:40 But @ is basically designed so that every component is too simple to fail. 02:39:50 so what's responsible for scheduling the scheduler? or is it somehow more "special" than all the other functions? 02:39:53 e.g. capability security in itself cannot fail, because it's more an observation than code at all 02:40:04 ais523: Well, objects could be responsible for their own scheduling, although that has a whole bag of issues; I'm not sure what my scheduler model looks like because I'm not sure what my concurrency model looks like because I'm not sure what my language looks like. 02:40:06 (See a pattern yet?) 02:40:12 indeed 02:40:19 @ may seem underdesigned, but it's really just that there's so _much_ to think about 02:40:24 although I'd think the language would be caused by the concurrency model, not cause it 02:40:31 ais523: interdependency 02:40:33 Just like the UI 02:40:45 ais523: there's this big knot at the core of @, but if you look closely, it's actually a point 02:41:00 It looks like a bunch of concepts tangled up that you have to work out at the same time, but they're actually all the same concept 02:41:01 I still dislike the statement that capability security is an observation 02:41:14 ais523: it is; the only time an actual bit started existing was at the network boundary 02:41:36 well, what prevents you just guessing at references to objects? 02:41:51 or is the /internal/ storage for them kilobyte-long hashes? 02:42:05 The internal storage for them is a pointer. But you don't have the ability to cast a random integer to a pointer. 02:42:12 elliott_: aha 02:42:14 Giving you that capability would be insane; it's insanely privileged. 02:42:23 ais523: Note that @ does not rely on the hardware's really coarse, bad mechanisms for security. 02:42:28 but if a bug does give me the ability to cast a random integer to a pointer, it completely blows open the security 02:42:31 ais523: This is why it can run in ring 0, and do things like syscall inlining. 02:42:41 ais523: Well, it blows open the local security. 02:42:43 I think that's what I was getting at 02:42:46 ais523: That's like a root escalation exploit in Linux. 02:42:51 yep 02:42:53 It still doesn't give you the power to demand objects from other computers. 02:43:02 and presumably, a similar bug would be a remote arbitrary code execution bug 02:43:10 which could also occur 02:43:18 Right; but you don't see that much in high-level languages at all. 02:43:21 it's just that there's only the one layer of security 02:43:25 Without eval, or whatever. 02:43:33 ais523: Nah; there's as many layers as there are objects 02:43:35 elliott_: we discovered one in Rodney (#nethack's bot) recently 02:43:39 Language? 02:43:41 Perl 02:43:45 Any eval calls? 02:43:47 but it was using eval, without taint checks 02:43:50 See. :) 02:43:59 ais523: BTW, in @, eval would be insanely rare, because who stores code as a string? 02:44:05 it was trying to regex the eval, but there was a misplaced ^ 02:44:07 OK, you can have eval : AST → a 02:44:09 but it's still insane 02:44:14 ais523: ouch 02:44:19 ais523: all I see is typing bugs 02:44:30 (taint checks are basically a really rudimentary type system) 02:44:34 indeed 02:44:42 ais523: BTW, casting integer-to-object-leaking bugs seem really contrived, because that's another low-level-ism 02:44:44 actually, Perl's taint checking is identical to the Identity monad 02:44:57 Perl taint checking == phantom types 02:44:59 data OK 02:45:00 data NotOK 02:45:04 data Data okornot = ... 02:45:09 unsafe :: Data OK -> ... 02:45:16 thisiscool :: Data NotOK -> Data OK 02:45:17 etc. 02:45:17 elliott_: here's a fun fact: Linux has a system call that puts a process into a "secure" mode where it only accepts 4 system calls 02:45:20 ais523: there's this big knot at the core of @, but if you look closely, it's actually a point <-- so, basically, just the same as feather? >:) 02:45:24 ais523: haha 02:45:30 ais523: Which 4 calls? 02:45:35 ais523: I take it you know this because you're using it? 02:45:36 which are just enough to get data in and send it out, doing anything else gets you kill -9ed 02:45:39 no, I'm not 02:45:40 heh 02:45:43 I think you've told me about this before 02:45:48 there's no way to get out of the mode, because the syscall that would take you out of it is not one of those 02:45:54 Ah, so. Read, write, exit, and...? 02:46:07 I can't remember precisely what they are 02:46:11 oerjan: :) My current plan to tackle it is to try a bunch of "mini-@s" that don't try to do nearly as much, and have a specific, inadequate instantiation of the inner point. 02:46:11 I doubt arbitrary read/write are allowed 02:46:12 Actually, probably mmap. 02:46:15 That should give me more insight on what the point has to be. 02:46:22 the funny thing is, that there's also a getter to see whether you're in this mode or not, but no way to run it 02:46:22 "what the point has to be" -- heh. 02:46:26 ais523: You could only read/write to things where you already have the file descriptor. 02:46:31 well, if you aren't in the mode, it returns 0, and if you are in the mode, it crashes 02:46:32 And also help me figure out a lot of details, especially relating to its implementation on existing hardware. 02:46:40 (Yes, @ stops at the hardware layer!) 02:46:43 so the getter always returns 0 even though the value is settable 02:46:47 (I know, it surprises me too! Let's hope I don't forget that!) 02:47:03 ais523: haha 02:47:12 ais523: you could set up a signal handler before you get put in the box 02:47:15 and catch your crash 02:47:16 oh 02:47:19 you can't catch kill -9 02:47:21 It's read, write, _exit, and sigreturn. 02:47:25 pikhq: aha 02:47:36 also, sigreturn's allowed? that means that signal handlers are 02:47:40 haha 02:47:40 ais523: I've always really wanted an exploit in kill -9; it replaces your code with a suicider, and sets your priority really high 02:47:42 but I assume they can't actually /do/ much but read/write/exit 02:47:48 ais523: So if you could somehow override that code before the kernel jumps to you... 02:47:50 elliott_: heh, is that how it works? 02:47:54 Try to kill me?? I WILL COME BACK STRONGER THAN EVER 02:47:56 ais523: yep 02:48:00 ais523: or is that OOM kill? 02:48:02 I forget, but anyway 02:48:05 I assumed it just marked the process as zombie, or something like that 02:48:19 Maybe sigreturn is in case you set signals before entering restricted mode 02:48:29 in fact, it should just do the same code as _exit, which already has the right semantics 02:48:38 ais523: it's probably OOM kill I'm thinking of 02:48:41 since it would need to free up its memory 02:48:45 pikhq: what's the syscall to get into that mode? I thought it might be prctl, but I was probably wrong 02:48:48 It probably just turns it into an _exit call 02:48:51 The suicide code, I mean 02:48:51 well, I was wrong 02:48:54 prctl. 02:49:06 ah yes, found it 02:49:07 With PR_SET_SECCOMP 02:49:13 I was reading the man page for prctl, but missed it 02:49:16 ais523: anyway, I hope you have a bit more insight to @ now 02:49:20 elliott_: yes 02:49:27 And hopefully like it a bit more, or at least it hopefully surprises you a little 02:49:39 I sort-of had the fundamentals down (they're pretty similar to something I was thinking of myself), but the details are interesting 02:49:55 I note that @ sort of changes a bit whenever I talk about it to respond to criticism; does Feather do that too? :p 02:50:00 Well, criticism or questions 02:50:05 elliott_: no, Feather changes whenever I try to implement it 02:50:09 haha 02:50:33 the main problem with Feather is not so much design flaws, as inability to start without going into an infinite loop 02:50:45 an implementation is badly needed to prove that that doesn't happen 02:51:09 I've already had to add a boolean to every object that simply queries if it's # or not 02:51:13 in order to get off the ground 02:51:16 * elliott_ wonders how long until someone stops an @ program from accessing more than a certain few types of object to stop it leaking memory (e.g. including no arrays), but gives it integers 02:51:21 which is a really specific thing to have, and I'm not sure what to name it 02:51:32 and then I'll run a program that leaves them a nice message and gets to work on constructing Graham's number 02:51:35 (note: I won't) 02:52:02 ais523: oh, wait, you didn't know that gnome's file opener let you type in a location? 02:52:23 elliott_: no 02:52:25 ais523: that's the top-left icon in the dialogue; click it and it'll focus by default 02:52:26 you're welcome 02:52:32 thanks, that's blown my mind to 02:52:33 you can also type directories in there, and their names complete 02:52:34 *too 02:52:39 and will change to that directory 02:52:45 I know it does that, it used to do that before I couldn't find the location bar at all 02:52:56 I was fine with that dialog, other than its insane slowness at loading /usr/bin 02:53:05 but pressing / makes the location bar magically appear 02:53:09 (other characters often don't) 02:53:19 (although I suspect ~ probably does) 02:53:30 another pet hate of mine: alt-f2 doesn't expand ~ but interprets it literally 02:53:48 if I write ~/esoteric/intercal, it interprets it as /home/ais523/~/esoteric/intercal 02:53:51 which is obviously not what I wanted 02:53:53 yeah, that sucks 02:54:05 I really wish I knew what @'s point looked like :( 02:54:31 when I try to envision what a screenshot of @ would look like, I honestly just see a black screen 02:55:23 wear a special hat and think about what you want user interface 02:55:31 I don't think it would look too dissimilar from a modern OS 02:55:46 because the UI and internal representation are not necessarily linked at all, nor should they be 02:55:46 ais523: your conception of that being a bunch of terminal windows and Emacs? 02:55:59 elliott_: that's not actually what my desktop normally looks like 02:56:02 ais523: The UI is just how you look at the internal representation 02:56:12 atm, it has IRC, email, two browsers, terminal 02:56:14 I mean, yes, things can have a nice veneer over that 02:56:21 But you can always look at an object directlry 02:56:22 directly 02:56:23 inspection-style 02:56:30 @'s interface also doubles as the best programming interface ever 02:56:35 and inspector/debugger, too 02:56:37 now, in @, I'd have things available for doing the same activities 02:56:48 and I imagine their UI would be similar 02:56:59 ais523: yep, but it won't have a traditional window manager 02:57:03 I'm not sure what it /will/ have 02:57:03 e.g. whatever I was using to look at a document full of hyperlinks probably wouldn't be Firefox, or even a Web browser 02:57:07 but it would control in a similar way 02:57:12 I used to think it'd look like Emacs, but with objects instead of text 02:57:14 now I have no idea 02:57:15 and look similar, as that's what people want hyperlinked documents to look like 02:57:22 Oberon has a good interface, I can take inspiration from that 02:57:27 but it's too text-oriented 02:57:46 one big advantage of being text-oriented is that you can edit everything with the same operations 02:57:56 You can do that with objects too, the operations are just more mind-bending 02:58:04 I've converted images into whatever that text-based X format is before, and edited them with Emacs 02:58:15 xbm, I think 02:58:16 I'm not sure what the equivalent for objects would be 02:58:29 ais523: an ASCII-ish version of the serialised object? 02:58:31 I've been meaning to try Oberon 02:58:35 elliott_: fair enough 02:58:36 But really, "text" is a bad word. 02:58:45 I tried to try it, but I remember having problems 02:58:47 Programmers think: text, string, and they think of... well, strings. 02:58:52 editing the serialised version of something directly is a hack 02:58:59 ais523: so is editing xbm in Emacs 02:59:00 It's arbitrary to say that text is a bunch of Unicode codepoints 02:59:05 Why those, and not bolding? 02:59:06 elliott_: no, that's what I mean 02:59:11 what I was doing was a hack, in a way 02:59:13 ais523: right 02:59:24 ais523: Well, the more idiomatic version would be looking at it in an object inspector 02:59:26 also, bold is ESC [ 1 m 02:59:29 And poking around changing things 02:59:32 ais523: Very funny 02:59:35 That's hardly semantic :) 02:59:55 Basically I think strings as we see them in modern systems don't appear naturally in @ much at all 03:00:04 because "document segment" works just as well 03:00:05 hmm... I suppose the problem is that objects never really have the interfaces you want them to have 03:00:10 Why can't you name a bookmark with a bolded wrod in it? 03:00:11 word 03:00:13 With a link? 03:00:27 With a video? (ok, not lookupable -- this isn't SpectateSwamp Desktop Search -- but the point remains) 03:00:27 because the name would be a pain to type 03:00:34 elliott_: gah you mentioned SSDS 03:00:36 ais523: you don't have to type names in full 03:00:40 in an appropriate context 03:00:45 :D 03:00:54 But basically, the point is that strings are only common because our I/O interfaces do strings. 03:01:17 Most of the time, you just want to print/show something, or to read something, and in @ you don't need to coerce things into Unicode codepoints to do that 03:01:18 how do you access things in SSDS, anyway? I know that everything is named with videos, but it seems a little unwieldy 03:01:25 ais523: you just give them tags, I think 03:01:27 And it searches those tags 03:01:29 (originally filmed with camcorder pointed at screen) 03:01:33 in a text file 03:01:37 elliott_: that seems altogether too sane 03:01:52 ais523: SSDS = grep with automatic video playing, I think 03:01:57 ah, right 03:02:07 "SS claims SSDS is a "desktop search" application. However, SSDS qualifies as "desktop search" only in SS's own definition and when using SS's own computer usage patterns; it does not qualify as a "desktop search" application in any other definition of the term. In most definitions, a "desktop search" system indexes the user's personal files, extracts metadata from the files, and allows the user to perform searches on this metadata; for example, "songs 03:02:07 by a composer called John" or "instant messaging conversations that mention pizza" or "email about Project X". SSDS, on the other hand, requires the user to convert all data to plain text format, thus losing all formatting and metadata, and merge everything in one file. For certain operations, the user is required to maintain such index themselves. Further, ordinary desktop search systems allow the user to see all metadata, such as "this matching song 03:02:07 is 4:33 long" or "the last pizza conversation took place yesterday" or "the boss mailed to you about Project X using Microsoft LookOut, the bastard"; SSDS only allows you to see matching likes in SSDS index file. For example, if you search for "Project X", it is difficult to see when, by who, and (for example) with what program the e-mail was sent with." 03:02:12 -- http://www.thestupidestmanonearth.com/DesktopSearch.aspx 03:02:14 long paste, but oh well 03:02:16 "At best, SSDS is a simplistic linear search application that attempts to match search terms against a text file - and with extra features that allows the user to show images and videos, also in random order." 03:02:23 heh, so it doesn't even play the videos you mention 03:02:28 it's {grep, random mplayer} 03:02:34 ah 03:02:39 also, /me /clears 03:02:44 because that much SSDS is too much for me 03:02:50 Well, yes, the UNIX stuff with pipes and grep and so on works much better. But, of course, that is UNIX. SSDS is a similar (but not as well designed) things for Windows. 03:03:05 And anyways you can do that stuff on Windows by using Cygwin or whatever, too. 03:03:11 pipes are a good idea 03:03:17 unfortunately, they're less effective on bytes 03:03:27 elliott_: have you seen PowerShell? 03:03:38 ais523: yes; good ideas, but it's not usable 03:03:39 There shouldn't be just stdin and stdout if you're putting in pipes 03:03:45 ais523, I haven't 03:03:48 You can have one command that shuffles the lines of stdin and that way, you can play videos in random order. 03:03:56 I have the feeling that someone at Microsoft thought of @, realised they'd never be able to implement it, and wrote PowerShell instead 03:04:05 * elliott_ notes http://okmij.org/ftp/Computation/monadic-shell.html 03:04:06 which is unusable for most operations because it doesn't really mesh with things well 03:04:11 | is (>>=) 03:04:18 zzo38: it's called shuf(1) 03:04:36 ais523: I have a feeling @ is the kind of thing many people start thinking about, but then realise what they're doing and quickly stop 03:04:39 ais523: Yes, then use that. 03:04:39 elliott_: PowerShell is apparently very useful for Windows administration merely because it can access things like the registry easily that other tools can't 03:04:59 Unfortunately, I wasn't clever enough to stop, and now I have to implement it 03:05:09 With barely no knowledge of standard PC hardware and OS design to start with 03:05:31 Incidentally, oklopol's probably-long-abandoned OS that he talked about ages ago has similarities to @ 03:05:40 elliott_: anyway, I think I've figured out what a capability is, at the hardware level; it's simply a bit pattern stored /in a memory location that lets it be used as a pointer to an object/ 03:06:09 gah, I'm tired, my eyes read there as being an unmatched opening paren in my last line, and there aren't any parens at all 03:06:16 I just keep seeing one for some reason 03:06:16 heh 03:06:18 and I can't even work out where 03:06:24 mediaplayer `ls *.ogv | shuf` 03:06:30 ais523: A capability is a pointer, pretty much; which ends up as bits in memory that are interpreted in a certain way 03:06:34 yep 03:06:39 zzo38: s/ls/echo/ 03:06:50 and it's distinguished by an integer with the same value by the type system 03:06:52 elliott_: Yes, or echo will do too 03:07:01 ais523: Well, integers have tag bits in @. 03:07:03 (Maybe.) 03:07:03 as this is standard hardware, the type system is being checked at "compile time" (actually "convert to x86 machine instrucitons time") 03:07:06 Small integers, that is. 03:07:23 whether that's upon editing, or much later 03:07:48 No... echo won't do 03:07:50 it'll be agonising when/if @ is "done"-ish, but has insufficient compatibility layers to actually use day-to-day 03:08:01 zzo38: oh right; you need them separated by \n 03:08:01 Because echo put everyone on one line 03:08:15 (fun fact: TUNES wanted a DOS emulator to start with) 03:08:28 did you know that Linux has two syscalls that only exist for dosemu? 03:08:55 they both put an x86 system into virtual 8086 mode (with different parameters), and fail with ENOSYS on any other sort of system 03:09:00 yep 03:09:09 * elliott_ has configured a kernel before, shockingly enough 03:09:19 heh, I didn't realise they were involved in kernel config 03:09:39 I think it mentions vm86 or something 03:09:41 elliott_: Most of the traditional knowledge on OS design you've already thrown out the window. 03:09:50 pikhq: Well, Genera exists. 03:09:53 So does Smalltalk. 03:09:59 I did say "traditional". 03:10:01 The low-level details are what matter to me, anyway. :p 03:10:07 And orthogonal persistence does have quite the bit of literature on it. 03:10:11 it's similar to Smalltalk 03:10:17 So does Smalltalk. 03:10:18 more so than typical OSes 03:10:27 elliott_: I know, I was agreeing with you 03:10:28 Yes, you're ignoring the low-level details that are really painful to deal with. 03:10:32 ais523: incidentally, one of my main worries about @ is that non-programmers won't be able to learn it, but... 03:10:34 What would be the algorithm to find the best way to use registers in a DVI file? 03:10:39 Most obviously, dealing with hardware's task switching. 03:10:40 I still like programs to be serialisable as text, incidentally 03:10:52 Still will need to handle paging, though. 03:10:54 ais523: I think @ just has a learning curve, maybe a larger one than most common OSes, but it'll confuse you less once you learn it and you'll be able to accomplish so much more 03:11:02 ais523: Also, OK, but why? 03:11:10 Sharing over IRC? Yes, desirable, but this is a limitation of IRC. 03:11:14 elliott_: fun fact: GNU Hurd was originally aimed at 32-bit processors, because 16-bit was common then and they wanted to future-proof 03:11:17 (@-chat would just share objects) 03:11:21 (usually text documents) 03:11:24 ais523: heh 03:11:28 isn't it possible to make a non-porgrammers friendly shell over @ 03:11:37 ais523: The same is true of the rest of GNU. 03:11:37 or am i loon 03:11:45 monqy: Sure, but that wouldn't be @... what's the difference between programmer and user? 03:11:58 ais523: Their idea was "by the time this is usable, 16-bit will be an old piece of shit, so why care now?" 03:12:03 elliott_: I think it's because it makes them easier to store on a wide range of devices (including things that aren't computer-related at all), and a wide range of implementations 03:12:05 monqy: If you want I can paste you an explanation of @ I wrote semi-recently into /msg (it's long) 03:12:07 pikhq: yes 03:12:13 pity they didn't start with 64 bits instead 03:12:19 ais523: What non-computer-related devices is it easier for? 03:12:24 And you mean codepoints, not text. 03:12:25 It's a bit of an easier jump, though. 03:12:27 elliott_: writing down on paper 03:12:35 Especially if you consider x86. 03:12:41 ais523: Um, you've used paper before, right? 03:12:43 and I mean what exists on my keyboard 03:12:44 You can draw boxes and things there. 03:12:48 Segmentation to a flat memory model was one *hell* of a jump. 03:12:49 elliott_: not reliably 03:13:04 ais523: Well, that's different; keyboard-inputs are one ... view of @ code. 03:13:04 you can be off by a millimetre or too really easily 03:13:17 the set of characters in printable ASCII has better error correction 03:13:20 Basically it would be reifying the intentions. (elliott buzzword!) 03:13:34 elliott_: sure. as-is i only have a really vague idea of @ that i can't really describe 03:13:34 ais523: basically you want to be able to get the actions that are required to recreate some code 03:13:39 monqy: me too 03:13:41 elliott_: yes, I think so 03:13:49 ais523: Right 03:13:51 including in an implementation that nobody currently envisages 03:13:54 ais523: Well, that's not necessarily text, but. 03:13:58 ais523: oh, well that's a lot harder 03:14:12 ais523: what you want is, I think, something that's not actually possible, but not something that /should/ necessarily be possible 03:14:16 in particular, I don't want it to be just a serialised form of the internals of something 03:14:19 But you can achieve what you want in _specific_ situations 03:14:21 Oh, well right 03:14:27 But it'd depend on /interfaces/ 03:14:29 You can't avoid that 03:15:10 well, you can, in that I think that even, say, 100 years into the future, no matter what formats and interfaces people are actually using, they'll have some way to represent sequences of octets 03:15:13 ♪ 03:15:15 to store old files "unchanged" 03:15:33 and you'll probably be able to get C89 compilers, even if they're seen as something antiquated and only of academic interest 03:15:43 does anyone know how to convert BBCode or HTML to Markdown that uses footer links? 03:15:45 I want to paste this to monqy :P 03:15:48 and it has a bunch of links 03:15:49 even @ could plausibly have a C89 compiler, although there wouldn't be much reason to use it 03:16:07 ais523: I plan some kind of "weakly-integrated" POSIX layer out of necessity] 03:16:14 ais523: Maybe in case you want to run C programs 03:16:35 why would anyone ever want to do that 03:16:39 zzo38: I would say you'd missed the point, but then I'd have to describe what the point is, and I'm not sure I can 03:18:15 this elusive point, so hard to describe 03:18:54 -!- ralc has quit (Quit: Leaving). 03:18:59 speaking of crazy possible future things, does anyone know what logs have stuff relating to feather? I'm suddenly really curious 03:19:08 monqy: hmm, I'm really tempted to just put this up as HTML somewhere 03:19:08 -!- azaq23 has quit (Read error: Operation timed out). 03:19:12 also, grep -r Feather . 03:19:38 monqy: did you find that files-suck.html too narrow as well as ais 03:19:40 -!- azaq23 has joined. 03:21:18 I'm not really sure what it's getting at as a solution 03:21:52 monqy: i was talking about the view 03:21:55 oh 03:21:56 as in visually 03:22:34 oh 03:22:43 I can see it just fine 03:22:47 also, saying that existing abstractions are bad is OK even if you don't have something better to suggest 03:24:03 right but i also didn't understand how saying something is bad could possibly be "too narrow" so I pulled some miserable conclusion jumpery 03:27:39 does anyone have a unicode return arrow 03:27:40 like 03:27:44 ↵ 03:27:44 caret 03:27:46 - 03:27:46 --- 03:27:47 monqy: ah thanks 03:30:13 and on greping for feather in logs, is this with the expectation that i have every log file :( 03:31:49 yes 03:31:54 exclamation mark logs to get them all 03:32:00 ok does anyone have a place for me to put this html file 03:32:08 monqy: also, install Bitstream Chartered it is good font 03:33:08 !logs 03:33:13 like that 03:33:34 yes 03:33:37 see your notice 03:33:38 use rsync 03:33:40 yep 03:38:04 monqy: also do you have bitstream charter this is totes important 03:38:26 i searched fro it but then it wanted me to buy it for $99 i think i did something wrong 03:38:33 !logs 03:38:33 -!- MDude has changed nick to MSleep. 03:38:55 monqy: its in your package manager 03:38:57 I should put the Cygwin installer on my desktop or something 03:39:26 monqy: xfonts-scalable or something 03:39:32 or hmm 03:39:40 monqy: oh, texlive-fonts-recommended should get you it 03:39:51 that give syou postscript fonts 03:39:53 which is better 03:40:58 help i dont use debia n o rhwat ever that is help 03:41:15 maybe i secretly already have it but never knew 03:41:23 monqy: what os do you use 03:41:29 is it a, bad os 03:41:31 arhclinux is that bad 03:41:36 am i ab ad person 03:41:40 :'( 03:41:43 yes :) 03:41:45 :'( 03:42:07 lemme find charter for arch 03:42:15 monqy: pacman -S xorg-fonts-type1 03:42:19 pacman -S arch-sucks 03:42:43 :'( 03:43:14 at least i dont use (worse distributuieon here) 03:43:44 does dcc work for you 03:43:50 i think so maybe 03:43:58 do you see that 03:44:18 that's literally just my forum post reformatted and with one bit crossed out 03:44:19 woo it failed 03:45:01 monqy: meh, do you have a gpg key 03:45:10 i forget if i ever bothered 03:45:15 probably not 03:45:15 psht 03:45:19 I'll just sprunge it 03:45:31 are there any good distributions i am honestly completely dumb at knowing things about them 03:45:41 Opa's do statement seems a bit magical :( 03:45:42 only @ 03:45:48 I use Debian because it works unlike Arch 03:45:53 -!- lambdabot has joined. 03:45:59 yaaaaaaaaaaaay 03:46:02 me hug lambdabot 03:46:09 * Lymee hugs elliott_ <3 03:46:12 -!- ghoulmaster has quit (Remote host closed the connection). 03:46:22 i glanced at opa then decided it is not for me 03:46:41 oh shit wait monqy 03:46:44 learning ur and a bunch of other unrelated languages first at the very least 03:46:46 don't download that i havent stled the hr yet 03:46:48 :(999 03:46:54 its UNEBARBEL 03:46:56 > fix$var.("yay! "++).show.(:[]) 03:46:57 yay! [yay! [yay! [yay! [yay! [yay! [yay! [yay! [yay! [yay! [yay! [yay! [yay... 03:47:00 too late but i havent started viewing it yet 03:47:03 i will remove it 03:47:08 rmeoved 03:47:10 > "\^AACTION hugs elliott_\^A" 03:47:10 "\SOHACTION hugs elliott_\SOH" 03:47:13 :( 03:47:57 monqy: ok see /msg 03:48:03 my mistake will, never be known, 03:48:13 downeloded 03:48:54 > text "\^AACTION hugs elliott_\^A" 03:48:54 ACTION hugs elliott_ 03:49:43 I think the control characters are stripped 03:50:33 ^ul (ACTION may or may not still work.)S 03:50:33 * fungot may or may not still work. 03:50:51 heh 03:50:53 !hug 03:50:56 This one works. 03:51:12 !python print "test" 03:51:14 test 03:51:30 !show python 03:51:30 sh python 03:51:37 monqy: btw this is from March, so of course @ has changed almost entirely since then :) 03:51:41 (not really) 03:51:42 EgoBot definitely doesn't work with ACTION any more. 03:51:44 :( :) 03:51:49 :( 03:51:51 monqy: also that is from my editing of it right now 03:51:52 the strikeout 03:51:57 !python print "ACTION test" 03:51:57 ​.ACTION test. 03:52:01 oop 03:52:09 rip 03:52:21 elliott_, this is WTFy on Homestuck 03:52:23 oop was even a bad way to describe what i was thinking about in the time 03:52:28 Sgeo: oh thanks for the reminder 03:52:32 There's no way to access current update from previous update 03:52:37 monqy: sorry i cannot answer your questions for about ten minutes I must READ HOMESTUCK 03:53:07 OH GOD HE;S JDGING ME 03:53:21 oh my god the top-left panel Sgeo click it 03:53:33 good thing i hate homestuck 03:54:02 monqy: when did your opinion develop from nonplussedness to hatred 03:54:06 is nonplusedness a word 03:54:19 elliott_: when i tried reading it after expressing nonplusedness 03:54:30 how far did you get 03:54:34 `addquot its UNEBARBEL 03:54:36 No output. 03:54:39 `addquote its UNEBARBEL 03:54:40 540) its UNEBARBEL 03:54:40 * elliott_ prepares for answer he's already guessed 03:54:55 i think i started roughly where i left off i.e. somewhere in the middle maybe 03:55:01 monqy: lol which act 03:55:05 i cant recall 03:55:12 monqy: which page number 03:55:14 which anything 03:55:17 no recolaection 03:55:25 what was happening 03:55:27 then i skipped around a bit because i am an awful person 03:55:35 (plenty of people think they're about half-way through but are actually in the first act) 03:55:41 it wasn'ta ct 1 03:56:24 you should start it again and go until at least the game that starts act four. it starts out boring but it really does develop. and skipping around is just going to confuse you. 03:56:44 is nonplusedness a word <-- no, it needs another s, but then it's a doubleplusgood word. 03:57:21 i dont hate things for being confused by them 03:57:51 i never said that's why you hated it but obviously it is completely inconceivable that someone could actually hate homestuck 03:58:06 03:58:09 (ok probably not. i doubt there's many people who actually hate homestuck while liking sbahj, though.) 03:58:09 What does that do? 03:58:13 Lymee: thx for the CTCP "" 03:59:10 i cant fully remember why i hate it but i remember finding the troll gimmicks insufferable and having to pay attention and care to understand things didn't help and it always felt really awkward to me like it's trying too hard or something idk 04:00:04 it is only a mild hatered 04:00:39 the gimmicks are meant to be insufferable, there's a bookmarklet thing that normalifies it though... but really, 99 percent of things that would class as "trying too hard" I'd say are jokes that only make sense with the context, assuming that was during skipping around 04:00:43 oh well I'll just conclude you're subhuman 04:00:52 totaly 04:01:13 !@#$ broken update 04:01:15 there's a lot of trying-too-hard near the start too but that's mostly intentional to set the scene 04:01:33 !@#$ crashed MSPA site 04:01:59 one thing i classify as trying too hard that is maybe not in the typical normal nonsubhuman definition: i mean when it tries to take itself too seriously 04:02:12 n/m 04:02:36 Chrome: Stop having troubles connecting 04:02:37 in addition to the other stuff at the beginning yeah 04:02:37 monqy: considering that almost all the major plot developments are sandwiched by jokes, I dunno if that's true 04:02:46 but really you should start again and just go through it all in order, it really picks up in act three, act four, and act five act two (in that order) 04:02:53 MAYBE YOU DON'T HAVE A SPARE WEEK THOUGH 04:02:55 that just makes you subhuman 04:03:20 clases ends next week 04:03:38 subhumans like me learn school during the sumer 04:04:26 CHROME: STOP HAVING TROUBLE WITH MSPA 04:04:26 -!- elliott_ has quit (Read error: Connection reset by peer). 04:04:29 -!- elliott has joined. 04:04:53 but really, homestuck has issues, and they're mostly pacing issues. and the first three acts are kind of a drag, although pretty funny. 04:04:53 elliott, how many working panels were there? Just one (Jade)? 04:05:07 monqy: but personally i'd be surprised if you didn't like it if you gave it a chance from the very start to the end 04:05:09 Sgeo: yes 04:07:12 Sgeo: ok im waiting for the site to stabilise before reading 04:07:23 * Sgeo is too impatient for that 04:07:32 monqy: ANYWAY THAT @ EXPLANATION HUH 04:07:37 * Sgeo may be part of the problem 04:07:46 sorry i was too busy trying to rationalize my subhuman hatred of homestuck 04:07:59 anyway i';ll finsih reading it after brushing my teeth i have to do that now 04:08:13 monqy: you should start from the beginning >:| 04:08:14 ....finish reading it? In less than 6 hours? 04:08:16 (i realise you weren't serious) 04:08:26 re-start, i mean 04:08:43 elliott, what's a reasonable amount of time to read all of Homestuck? 04:08:50 I think I did it in less than a week 04:09:07 Sgeo: um well it's about twenty hours of consecutive effort 04:09:19 maybe more 04:09:32 let's say twenty four hours of consecutive effort 04:09:36 so if you put in five hours a day you can read it in five dyas 04:09:37 days 04:11:09 http://i.imgur.com/oh8Fl.png <-- meanwhile, reddit discovers the least tasteful way to do a remembrance comic ever 04:11:12 I know this instability is Hussie trolling us 04:11:37 ha, looks like a fake 04:12:35 1) i meant finish reading at.html (2) haha just kidding about doing it after brushing my teeth i have other stuff to do also woops 04:12:53 elliott, link to thread? 04:13:05 monqy: but it's...short... 04:13:08 Sgeo: top of main page 04:17:42 how do you mix an rgb with an rgba using the alpha im dumb 04:19:07 -!- bsmntbombdood has joined. 04:19:18 at.html? 04:19:47 Sgeo: yes 04:20:00 "old.r * (1-alpha) + new.r * alpha" (plus same for g/b) is very traditional. 04:20:09 fizzie: thx 04:20:17 fizzie: (I've pushed configurable colours and blockinfo to mcmap) 04:20:18 elliott, I expect you to be psychic and realize that I want a link 04:20:22 Sgeo: no link 04:20:33 fizzie: Now I'm working on respecting the alpha values in a dumb way by only looking one down 04:20:38 But it's a start. 04:20:49 It's a fart. 04:21:10 fizzie: Also: new is the one with the alpha, right? 04:21:13 i.e. the RGBA. 04:21:25 Well, it's the one "on top". 04:21:33 Right. 04:21:42 And that was assuming the 1 == full opacity scheme, sometimes it's flipped. 04:22:01 fizzie: Oh, erm, it's two-five-five = full opacity. 04:22:19 Well, scale as necessary. 04:22:37 (old.r * (255-alpha) + new.r * alpha)/255 or some-such. 04:22:51 Right. 04:25:03 fizzie: Next thing I work on will be the status bar, to make that blockinfo actually useful beyond the configuration. 04:25:12 Which I decided to just put at ~/.mcmap/colors because XDG seriously sucks. 04:25:16 Hope that's okay. 04:25:25 But my ~/.config! (Yeah, that's okay.) 04:25:31 Ugh, I still have that horrible bug that shouldn't even be happening. 04:25:33 I broke code I didn't even touch. 04:25:42 tell("MODE: %s%s%s%s%s", 04:25:42 modenames[map_mode], 04:25:42 (mode == MAP_MODE_CROSS && map_flags & MAP_FLAG_FOLLOW_Y ? " (follow)" : ""), 04:25:42 (mode == MAP_MODE_SURFACE && map_flags & MAP_FLAG_CHOP ? " (chop)" : ""), 04:25:42 (map_flags & MAP_FLAG_LIGHTS ? " (lights)" : ""), 04:25:42 ((map_flags & MAP_FLAG_LIGHTS) && (map_flags & MAP_FLAG_NIGHT) ? " (night)" : "")); 04:25:44 How could this send things like 04:25:51 05:25:11 MODE: surface[GARBLED KANJI] 04:25:56 java.io.IOException: Received string length longer than maximum allowed (23040 > 16) 04:26:12 I do not even understand it. 04:26:53 That is a bit weird. 04:27:07 Okay, gone for ~12 hours or so. -> 04:27:19 fizzie: BUT YOU NEED TO FIX MY BBUGEEE 04:28:51 Friend post on a tech help forum 04:28:55 Gives a model number 04:29:02 Helper asks which model 04:29:18 05:29:05 §bMODE: surface 04:29:19 05:29:05 MODE: surface끾 04:29:20 OK, that's insane. 04:29:30 The string is correct, but inject_to_client is fucking it up. 04:29:33 Somehow. 04:30:02 fizzie: Ah. 04:30:05 fizzie: I think you broke it, man. 04:30:09 jshort_write(lenb, conv_len); 04:30:10 With this, somehow. 04:30:18 Or at least, it just changed and it's the FIELD_STRING code. 04:30:30 How do you nicely tell someone who's trying to help you that they asked a really stupid question? 04:30:41 Could be; I didn't test it too much. 04:31:02 Sgeo: Ignore them. 04:31:07 Sgeo: Or answer it. 04:31:27 fizzie: OK BUT ALSO what's the formula for how many iterations something with alpha N can mix with things below it before it will have no effect. 04:31:34 (With byte-sized colour and alpha components.) 04:31:39 I want to ~optomize~ alpha blending. 04:33:23 Depends on the original color, I suppose. 04:33:34 "Until the target color equals the new color" is the easy test, but due to rounding-down that's probably a bit too conservative. 04:33:50 fizzie: Meh, good enough. 04:35:16 Oh yes, I did in fact break it. 04:35:16 Well, my friend answered it 04:35:28 It needs to be conv_len/2. 04:35:40 "Really... I give you a model number and I'm asked which one is it?" 04:35:54 Then he answered it... again 04:36:13 fizzie: Committing a fix or I? 04:36:17 Why am I teling you this 04:36:27 You do it, I'm about to miss my bus. 04:36:37 Feel free to put in a disparaging commit message. 04:36:57 Something about everyone else always having to clean up after me or something. 04:37:30 fizzie: I'll just include your address and tell everyone to send the bombs that-away. 04:40:23 I think I realized what is wrong with my optimization algorithm, although I do not know how to make a proper algorithm. I can explain by example. 04:41:02 Say that there is a command - which does the next letter, = which does the next letter and sets a register to its value, and . which does the value of the register. 04:41:13 -a-a-a-a-a-b-b-b-b-b-b-b-a-a-a-a-a-a-a-a-a-a-a-a 04:41:27 To optimize using my current algorithm might do like: 04:42:01 =a....-b-b-b-b-b-b-b............ 04:42:09 Due to the way it computes the scoring. But better would be: 04:42:21 =a....=b......=a........... 04:43:12 That is a simple circumstance, and actually there are other complications such as stack and multiple registers, and a lot more ways of interleaving. So, how to make up the proper algorithm? 04:44:27 hmm 04:44:44 zzo38: I think you just have to do reordering 04:45:06 elliott: What do you mean? Can you give examples? 04:45:07 hmm 04:45:13 ah wait hmm 04:45:16 well looking at -a-a-a. vs. =a..=b it seems like you should always do = if you have a run of 3 or more? 04:45:19 zzo38: You just have to iterate it, don't you? 04:45:54 There are other complications. Although I will try oerjan's suggestion. 04:46:12 -a-a. =a.=b seems to indicate it's harmless to do so with 2 as well 04:46:15 It would make sense. However the other complications might cause it to fail; I don't know for sure. 04:46:23 ah. 04:46:30 But I can try. 04:47:27 if you have something like -a-a-b-c-a-d-a-a it might get trickier... 04:47:34 im returned from not being here. at.html time?? 04:47:37 you probably want to save only a there 04:47:45 monqy: possibly!!! 04:48:12 For example, let's say there might be two or three or four letters for some commands. And that, you might have two registers. And a stack that saves the values of the registers but has other effects too. And, yes, also your other example -a-a-b-c-a-d-a-a 04:48:40 (Or even remove the first -a and the problem still holds) 04:48:53 elliott: write it in zepto... 04:49:23 monqy: no :P 04:49:30 yocto 04:49:43 It is alarming that my computer keeps running low on memory 04:49:57 zzo38: well if you have _enough_ complications you probably end up with something NP-complete. 04:50:39 Chrome, Y U EAT SO MUCH RAM? 04:51:00 Sgeo: no 04:52:00 Sgeo, Y U NO PUT NO IN Y U? 04:52:41 PORQUE NO TE CALLAS? 04:54:33 elliott: finsihed. unrelated to design, but are any languages even suited to implementing it? 04:54:50 monqy: Existing languages? No. 04:55:11 Although this is an analogy, it is close enough to the problem I am having. Now let's extend this system. Each name can be from one to four letters long, and these values are stored in the register. Say there are two registers, where "=" sets the "." register and "+" sets the "," register. Now add "[" and "]" which push the values of the registers onto the stack and pop. However, "[" and "]" have other side effects and you *cannot* change or mo 04:55:23 monqy: There are two bootstrap options: Write a minimal interpreter for @lang in assembly (C is not an option at any stage), write a @lang compiler with that, switch to it; 04:55:41 monqy: Or implement @lang somehow in existing OSes (is this even possible? Maybe just a facade of @lang), implement a @lang compiler in that, port it over. 04:55:42 assembly? 04:55:53 monqy: That's what the lower layer is being written in, yes. 04:56:00 Which won't be much, once @lang is up and running. 04:56:04 which flavour 04:56:12 (This is equivalent to the problem as long as you assume that there can possibly be up to 256 "letters" of the "alphabet") 04:56:21 monqy: Probably nasm, but it'll be replaced by an assembler written for and in @ when all is said and done. 04:56:30 * Sgeo wants to learn more about @lang 04:56:35 Sgeo: Me too. 04:56:48 Sgeo: If you find out, let me know. 04:56:48 Hmm 04:57:02 making a compiler for a high-level language in assembly sounds like a pain 04:57:07 monqy: Interpreter 04:57:13 -!- pikhq_ has joined. 04:57:13 even so 04:57:13 This makes me wonder if one could make a language that requires solving an unsolved problem just to see the complete specs 04:57:17 monqy: Well, yeah. 04:57:26 -!- pikhq has quit (Ping timeout: 258 seconds). 04:57:32 monqy: But the other bootstrapping route sounds like a pain, too :) 04:57:41 monqy: This is why I'm picking a simpler language than @lang for one of the initial prototypes. 04:57:46 A Lisp, say; not that hard to do in assembly. 04:57:53 Wildly inadequate, but I have to start somewhere. 04:59:18 hard? maybe not. painful? I'd imagine so. maybe assembly just isn't my thing. I'd rather make myself a simple compiler than write it directly. 04:59:31 monqy: Sure, if it's Lisp I can go tha troute. 05:00:33 this reminds me, I really have to remember my ideas for the serious prgoramming langauge i hope to make sometime to solve all of my problems forever. 05:00:34 Would you know any ideas about the proper algorithm to solve my problem now? It is more complicated now that I added that stuff. 05:00:40 monqy: just use @ 05:01:24 zzo38: i doubt i know, but you also got cut off after "and you *cannot* change or mo" 05:02:04 ... you *cannot* change or move them! 05:02:53 zzo38: ok. that means the only the only thing you could adjust with [ and ] are which values are pushed and popped 05:04:15 oerjan: Yes. You cannot move, add, or remove, any [ and ] commands yourself. 05:04:51 hm the value of the registers after the [ will be precisely the same as after the matching ] 05:04:55 But changing other commands can change what values are saved with the [ and restored with ] 05:05:10 oerjan: Yes that is correct. 05:05:17 so you can think of it as a kind of three branching instead of stack manipulation 05:05:23 *tree 05:06:20 OK. 05:07:02 A [ B ] C becomes a tree with commands A at the root and two branches with B and C respectively 05:07:44 Yes, that would work, I think. 05:07:44 where B and C might contain further branching 05:08:27 well also A, in which case [B] C belongs to A's last branch 05:09:07 Yes. 05:12:18 hm if names have different lengths that also is a complication - it may be cheaper to save (and so reuse) a longer name 05:12:41 Yes I have realized that, too, already. 05:13:06 (I didn't realize the tree stuff at first but now that you explain it makes sense and I agree) 05:14:33 i still have a suspicion this _might_ be NP-complete 05:16:25 everything is NP-complete 05:16:29 everything interesting, at least 05:16:45 perhaps 05:17:17 monqy: btw will you use @ i need committed users for funding 05:18:55 i will use @ 05:20:08 at least i'll play around with it and maybe implement stuff if not use it regularly (it would be hard to use it regularly without stuff implemented) 05:20:23 same with @lang 05:23:47 monqy: stuff implemented but but it'll have a web browser, an irc client, and a posix layer WHAT MORE DO YOU NEED WHAT DO YOU WANT FROM MEEEEEEE 05:25:32 oerjan: You might be correct about NP-complete. 05:26:04 elliott: it depends on if it's usable for me (if not i'd like it to be usable for me) 05:26:09 intentionally vague 05:26:11 However it probably can still halt. 05:26:17 because i do not know exactly what makes things usable for me 05:26:27 it's going to be so homestuck you have no idea 05:26:29 :( 05:27:19 Would you know if there are other channels that they can also help with these kind of algorithms? 05:27:50 zzo38: well yes. one obviously could use the algorithm of iterating through all possibilities of =, + and - at each spot where the name isn't already in the register, and that would halt but probably be exponential. 05:29:13 i only read the first 2 or 3 days of feather chat but i think i know enough now to say it's insane 05:29:22 monqy: it's beautiful 05:29:26 That is why I would want to figure out if there is better things 05:29:27 and amazing and perfect 05:29:28 that too 05:29:35 monqy: keep reading btw, you'll see ais gradually go completely insane 05:29:40 in the everyday sense 05:30:19 zzo38: i'm not frequenting any optimization channels, alas. 05:30:33 i like the part where the guy skeptics all over feather and then likes php 05:30:50 this is because i am a horrible person 05:31:17 monqy: you like to see movies where people fall into sewers and die, i take. 05:31:27 the best 05:31:32 elliott: btw whats the everyday sense of insane 05:31:49 () 05:32:30 monqy: the sense that isn't used in the idiotic sense of "LOL WE'RE ALL CRAZY HERE BECAUSE WE AM LIKE ESOLANGS" 05:32:38 * elliott STARES AT OERJAN FOR NO REASON 05:32:46 -!- Nisstyre has joined. 05:33:21 i like the part where the guy skeptics all over feather and then likes php 05:33:23 elliott: so the good sense 05:33:23 oh god which day 05:33:29 uhh 05:34:59 * oerjan hits elliott with the saucepan ===\__/ 05:35:11 the one where ais explains how feather has no primitives or syntax and RodgerTheGreat claims it impossible and ehird accuses him of hating most non-Java non-BASIC languages 05:36:04 no syntax either? oh dear. 05:37:05 monqy: rodgerthegreat is a bad person 05:37:12 true facts :{ 05:37:27 ./combined:20:49:49: * tusho watches RodgerTheGreat come and say "THAT'S IMPOSSIBLE! JUST LIKE FEATHER!" 05:37:32 not helpful grep... 05:37:43 17:24:06: ehird: utterly untrue. I like BF, ///, LOGO, PHP, Postscript, LUA and some of my own languages, in addition to Java and BASIC 05:37:46 grep for that 05:37:59 00:14:41: I love JS too :D 05:38:00 WHY DID I SAY THIS i must have been joking 05:38:07 monqy: im ehird btw 05:38:10 i know 05:38:11 i guess you knew that 05:38:15 SORRY IF YOU THINK LESS OF ME 05:38:45 what year was this i think i was an idiot back then 05:39:18 monqy: two thousand and eight 05:39:36 yeah i was probably most certainly an idiot. i don't want to think about it. 05:39:40 hmm, something is reminding me to play nethack again. prolly ais talking about acehack 05:39:55 monqy: DO YOU PLAY ANY ROGUELIKES i bet your taste in roguelikes is as bad as your taste in webcomics >:( 05:40:20 i think you talked about roguelikes once 05:40:20 sort of 05:40:22 or was that cpressey 05:40:23 dunno 05:40:24 i did? 05:40:32 anyway i don't know of any roguelikes i like 05:40:50 which did you plau 05:40:51 y 05:41:05 its complicated 05:41:12 youre avoiding answering because you have bad taste 05:41:16 yes 05:41:35 google "monqy nethack" -> oh god results for crawl 05:41:38 i dont really play crawl anymore but i hang around in ##crawl 05:41:41 FROM THIS YEAR 05:41:42 NOOOOOOOOOOOOOOOOOOOOOOOO 05:41:47 i cant talk to you any mor 05:41:47 e 05:41:49 and ##crawl-dev 05:41:52 die forever :{ 05:41:55 8) 05:42:03 i actualyl dislike crawl 05:42:07 does that make you feel better 05:42:16 one of the results from that google highlighted this part of crawl learndb 05:42:16 bhaak 05:42:16 Still believes that Nethack is the most popular roguelike. 05:42:16 Also believes that the earth is flat. 05:42:16 Happens to maintain a Nethack fork, UnNethack. If you're curious how a Nethack with some thoughts spent on balance could look like, give it a try: http://sourceforge.net/apps/trac/unnethack/ 05:42:17 Is on the record as stating that he is not a shark. 05:42:19 Likes ke$ha 05:42:21 good isnults A+ 05:42:24 monqy: maybe a bit........... 05:42:41 i know ais has been like kicked from crawl-dev several times for "trolling" so i have to MILDLY DISLIKE YOU... 05:43:09 I think I've been kicked once and muted once. does that help. 05:43:13 17:24:06: ehird: utterly untrue. I like BF, ///, LOGO, PHP, Postscript, LUA and some of my own languages, in addition to Java and BASIC 05:43:16 this must be from the day before 05:43:19 erm, after 05:43:30 oh wait no 05:43:31 i just grepped wrong 05:43:46 monqy: yes ok. 05:43:46 are there any good roguleikkes 05:43:49 monqy: how anal are those guys even 05:43:54 monqy: yes. vagrant 05:44:05 it's so beautiful you have no idea. 05:44:08 will check it out 05:44:22 monqy: good luck with that 05:44:27 monqy: its mine its in python but it's GOLFED PYTHON 05:44:29 it's like 05:44:31 a kilobyte big 05:44:34 and so amazing 05:44:38 i should work on that sometime 05:44:47 maybe make it use perlin noise instead of just... random() to generate the map 05:44:49 as for those guys do you mean the players or the developers 05:44:59 whoever ops crawl-dev 05:45:03 oh 05:45:44 varies between okay and insufferable?? i dunno 05:46:40 where is vagrant i am bad at finding it 05:47:20 monqy: it is not online :( 05:47:21 i addded ehird but google thought i meant third i will put quotes around it 05:47:23 oh :( 05:47:28 well it is 05:47:30 but mostly not. 05:47:40 i think i coerced cheater into pasting the latest version a while back after i lost it by pretending i liked him. 05:47:41 felt dirty. 05:52:02 17:22:28: If the first thing a language does is define it's own syntax, the syntax for defining syntax IS THE SYNTAX. 05:52:08 monqy: you should be so happy you weren't around when this guy was 05:52:13 im hapy 05:52:53 -!- Lymee has quit (Ping timeout: 260 seconds). 05:53:29 I actually still talk to him fairly regularly. 05:53:38 how bad of a person 05:53:47 Usually, not *too* bad, actually. 05:53:52 usually :D 05:53:55 He's developed a strong fondness for Forth. 05:54:29 isn't he like only four years older than me or something, i seem to recall him ranting at me for ages for being immature and blaming it all on my age once 05:54:54 He's a bit older than that. 05:54:57 But not much. 05:55:08 Wanna say, uh, 22? 05:55:40 Also, I'm pretty sure that https://secure.wikimedia.org/wikipedia/en/wiki/Fushigi_no_Dungeon is the most popular roguelike... 06:00:12 the difficult question is at what point does it become opcodes 06:00:32 hi 06:00:35 Argh, fizzie is gone for twelve hours. 06:01:09 the non-esoteric languages are usually used for bootstrapping right? 06:01:33 itidus20: Very few languages actually bootstrap, TBH. 06:01:41 Well, language implementations. 06:01:44 like unix was used to bootstrap gnu if i understand correctly 06:02:40 i don't "really" understand bootstrapping though 06:02:43 Clearly. 06:03:04 isn't it a physical impossibility to bootstrap something? 06:03:15 rather a systemic impossibility 06:03:17 Not really. 06:03:25 there was this thing i read somewheer 06:03:35 To bootstrap something is to build something with itself. 06:03:42 about how 06:03:45 There is no requirement that your initial build was bootstrapped. 06:04:06 nevermind ^^; 06:04:14 hmm 06:04:24 For instance, you could bootstrap a C compiler written in C, using a C compiler not written in C. (I don't know if such a thing exists. :P) 06:04:28 Argh, fizzie, you am done confound me. 06:04:34 Or, you could compile it *by hand*. 06:04:39 pikhq_: zeta c was written in lisp, I think 06:05:02 so one meaning of bootstrap is to have a compiler for a language written in itself then right? 06:05:06 The WEB program is itself written in WEB, although if you have the file TANGLE.PAS then you can start from there and then you can modify TANGLE.WEB and compile (with or without modification) WEAVE.WEB too. 06:05:07 itidus20: Yes. 06:05:19 Are there C compilers written in assembly languages? 06:05:26 zzo38: probably 06:05:32 zzo38: None notable, at least. 06:06:06 interpreter is probably a better word 06:06:25 Oh, yeah, you could also have the bootstrapping compiler run on an interpreter. 06:06:44 C interpreter written in assembly language would also work since the C compiler could then compile itself 06:06:53 I think that's how GHC started off. 06:07:29 itidus20: basically every compiler written in its own language must have been bootstrapped at one point. 06:08:11 -!- Taneb has joined. 06:08:18 hi Taneb 06:08:21 Hello 06:08:27 Frankly, bootstrapping compilers annoy me. I dislike circular dependencies. 06:08:30 although there are several options for how that could happen: a compiler written in another language, an interpreter written in another language, or (probably the very first languages) by hand. 06:08:34 pikhq_: but but @ 06:08:45 anyway it's not circular if it's portable 06:08:47 oro. 06:08:51 i am very much confused 06:08:53 elliott: For Linux package management. 06:09:04 hahahahhaha 06:09:05 Isn't GHC circular? 06:09:07 elliott: Namely, I dislike GCC having a build dependency on GCC. 06:09:11 Taneb: Yes. 06:09:18 pikhq_: a package manager that can't handle circular dependencies is broken 06:09:23 itidus20: for example erlang was originally interpreted in prolog (and still has a very similar syntax) 06:09:28 elliott: Build dependencies. 06:09:34 pikhq_: yep 06:09:48 is the bootstrapped section just the minimum necessary to build the rest? 06:10:27 oh i see now.. 06:10:35 the thing is when you update a language 06:10:45 and the updates to the language are coded in the language 06:10:48 itidus20: you can't use a feature you just implemented in the compiler without bootstrapping 06:10:56 obviously 06:11:17 so you begin in say, C/asm, for example.. 06:11:26 elliott: The only way to solve circular build dependencies is manual intervention so that the package depending on itself actually exists before you build it. Which is *really annoying*. 06:11:40 oh i guess it doesn't matter what you begin in 06:11:44 pikhq_: or it could install a binary first 06:11:47 same shit really 06:12:15 elliott: Yesss, that's manual intervention so that the package depending on itself actually exists before you build it. 06:12:20 Which is really annoying. 06:12:37 pikhq_: no, that's doing it in the package manager 06:12:47 It's still *really annoying*. 06:13:33 So the fundamental idea though is about initially writing a minimal compiler/interpreter in say: C, C++, Lisp, Haskell, what-have-you.. and using that to compile/interpret a new version of your language 06:13:52 and using that to build your language 06:14:06 yeah 06:14:57 I guess that it very much depends on many things. 06:15:23 Like, you wouldn't want to rely on an interpreter running underneath your language 06:15:47 indeed interpreters written in the language itself aren't useful for bootstrapping :P 06:16:11 although there are still uses for those 06:16:36 I can see that you could tell lisp to output asm code for an interpreter. 06:17:05 uhh... or something 06:17:06 lisp may have both an interpreter and a compiler for itself, where the interpreter is used for quick running and the compiler is used for optimizing 06:17:46 inside the same implementation 06:17:50 oh i mean.. suppose you wanted to implement a brainfuck interpreter. you wouldn't want to have the interpreter running on top of an interpreter. 06:17:52 Actually, Lisp almost always has a self-interpreter. That's the "eval" part of REPL. :P 06:18:36 sometimes you might 06:18:38 pikhq_: well it _might_ but it could also compile on the fly immediately, like ghc does 06:18:51 oerjan: Yeah, true... 06:18:51 *ghci 06:19:10 although that's still a slightly simpler compilation to bytecode 06:20:28 in any case, i figure you only really want 1 layer of interpreters running most of the time 06:20:46 itidus20: it's ok to use an interpreter to run your initial compiler, though, since that's a one time thing 06:20:47 and 2 for cases wheer it doesn't matter 06:21:22 oerjan: sbcl compiles to native code always, fwiw 06:21:30 ah 06:21:43 including in the repl 06:21:56 but you wouldn't really want to load haskell, and then load brainfuck on top of it, and then supply a brainfuck interpreter's source to that 06:22:50 or maybe you would 06:22:51 :) 06:23:02 im just not really moving things forward 06:23:18 itidus20: well for esolangs you might want to just because it can be done 06:23:28 yeah, its not such a bad idea 06:23:28 (see: my unlambda interpreter in unlambda) 06:23:49 im just in the habit of attaching a value judgement to everything i say 06:27:00 otherwise if you value efficiency you'd want to get a compiler pretty soon, or at least an interpreter that is itself compiled 06:28:15 heavy.. 07:12:11 You know what's odd? 07:12:28 What my brother calls an awkward turtle is different to what I do 07:12:40 what's an awkward turtle 07:12:52 a personality type i think 07:12:59 a clumsy tortoise 07:12:59 a turtle goes into their shell 07:13:01 A hand signal made during awkward silences 07:13:06 oh. lol 07:13:23 The point of it is to change the subject to the awkward turtle 07:13:58 those things feel artificial to me. subject changes to how artificial they feel. nobodys happy. 07:14:02 I prefer to make things more awkward and then sink deper into depression 07:14:14 classic itidus20 07:15:05 -!- monqy has quit (Quit: hello). 07:15:21 To quote David Morgan Mar, "If you find yourself in an awkward silence... Fill it! Say something! Anything! You will never have a better chance to make an awkward situation even more awkward!" 07:15:22 Everything seems to go wrong 07:16:23 does he ever come here? 07:16:51 not to my knowledge 07:16:56 No, I don't think he does 07:17:01 has he ever? :P 07:17:17 same answers? 07:17:20 I'm not even sure he's seen the wiki 07:17:22 thats fascinating 07:17:24 hahahahahahaha 07:18:03 wow, the bloglike section on Mezzacotta hasn't been updated for over a year 07:18:32 of course he has scary time planning powers, so he probably wouldn't fit in something that useless 07:18:36 it's like.. which friggen esolang club does he hang with then? 07:18:48 whereas Square Root of Minus Garfield has an update yesterday, so presumably that's still going 07:18:52 i thought this was the only one 07:18:59 as does Lightning Made of Owls 07:19:10 itidus20: well afaik all his esolangs are old... 07:19:21 ah ok 07:19:39 All the webcomics, bar Awkward Fumbles, are active 07:19:49 And Infinity on 30 Credits a Day 07:20:04 ais523: actually LMoO is skipping many updates nowadays 07:20:17 its funny how that works out 07:20:18 it is 07:20:26 something like that, it would astonish me if it came out daily 07:20:34 but it's still being updated when there are submisisons 07:20:35 *submissions 07:22:27 and SRoMG shows no signs of abating 07:24:42 btw the mezzacotta comic is rather dead if you look at the hall of fame section, not enough votes. although just this week some started trickling through. 07:25:08 Well, sqrt(-garfield) is brilliant. 07:26:03 hm... 07:26:22 i actually read some of it before coming here 07:26:52 it really tore apart the comic medium in fun ways 07:26:54 -!- Vorpal has joined. 07:27:14 itidus20: Also, it's producing something intelligent out of Garfield. 07:27:33 Which is quite a feat. 07:28:52 I am not sure if I approve of Scott McCloud though. HE should try his hand at esolang though 07:29:08 Given that Garfield is a comic written for one purpose: making money... 07:29:25 Scott McCloud is the most likely person to publish a book about esolangs :D 07:29:33 He has the mindset 07:29:41 He just doesn't know it yet. 07:31:20 I mean Mr. McCloud will never be Jack Kirby or Akira Toriyama. 07:31:47 ok, as far as can see there is no evidence that DMM has ever edited our wiki. 07:31:51 *i can 07:32:03 oerjan: im sobbing loudly 07:32:12 is he somewhat a celeb in esolang terms? 07:32:28 itidus20: it's separate, he's an esolang designer /and/ a celebrity 07:32:31 he's a celeb in every term. well, apart from the one normal people use. 07:32:32 but not famous for esolang design 07:32:33 ahhh 07:32:40 he writes this webcomic :P 07:32:49 he writes a somewhat popular webcomic 07:32:57 "Irregular Webcomic!" 07:33:03 and has a heavy hand in several others 07:33:04 ais523: Vorpal isn't appreciating this feature I coded into mcmap, probably because he's away; please tell me about how great it is 07:33:13 i actually discovered DMM when I started asking someone a question "theres not really any popular australian websites like google, yahoo etc" 07:33:17 elliott: I, umm, don't know what it does and why it's useful 07:33:22 but you managed to do something nontrivial in Java 07:33:26 that's worth an accolade of itself 07:33:29 itidus20: im lol 07:33:32 ais523: who said it was java??? 07:33:37 oh, good point 07:33:39 ais523: mcmap is a proxy, it's pure retrostyle c99 07:33:40 I assumed it was a Minecraft plugin 07:33:44 ais523: nah, it's fizzie's project 07:33:46 also, "retrostyle c99"? 07:33:48 that I also write some stuff for 07:33:48 and i went on wiki and found DMM 07:33:57 ais523: yes, fizzie thinks this is the nineties, and does things like use bitshifts instead of division 07:34:03 ais523: and mark functions "inline" 07:34:04 it's adorable 07:34:05 oh, right 07:34:12 I use bitshifts instead of division when it's clearer 07:34:33 say, if I'm dividing by a large power of 2, it's clearer to write x >> 20 than x / (1 << 20) 07:34:39 or worse, x / 1048576 07:35:04 x >> 1 vs. x / 2, though, they're pretty much equal in terms of legibility 07:35:55 http://en.wikipedia.org/wiki/Category:Australian_websites "Irregular Webcomic!" 07:36:00 ais523: Yes, but you're using bitshift when, logically, you are shifting bits. 07:36:05 i found him by that means 07:36:09 ais523: Which is only proper. 07:36:21 What if the shift is zero-extend? 07:36:34 ais523: oh, and he tries to make things thread-safe by marking them volatile 07:36:46 elliott: *wince* 07:36:50 pikhq_: dont wnice its amazing 07:37:01 DMM is the most famous esoteric programming language inventor with a PhD in Astrophysics 07:37:09 That's the "make the compiler stupid" button. 07:37:09 i can respect a man who uses macros because he doesn't trust the compiler enough to optimise an inline function 07:37:22 pikhq_: well it _is_ a mutex 07:37:35 australia kicks ass 07:37:44 elliott: that, umm, can actualy work sometimes 07:37:50 I lived in Australia for 11 months 07:37:50 (making things volatile to make them thread-safe) 07:37:52 itidus20: hmm, the most popular australian website may be Whirlpool (it's the only one I can think of off the top of my head) 07:38:00 which I'm not sure /why/ it's famous, because it's all about Australian ISPs 07:38:00 we have kangaroos and DMM and isn't afraid of anything 07:38:02 but I know it anyway 07:38:09 ais523: And in some cases, it just makes the compiler stupid. 07:38:12 but expecting it to happen without knowing why is probably going to end in failure 07:38:14 `addquote australia kicks ass we have kangaroos and DMM and isn't afraid of anything 07:38:17 541) australia kicks ass we have kangaroos and DMM and isn't afraid of anything 07:38:23 elliott: I suspect the most popular Australian website is google.au 07:38:26 but am not certain 07:38:29 It depends quite heavily on what you're doing. 07:38:30 ais523: that's not really an _Australian_ website 07:38:32 And I meant most famous 07:38:56 * elliott wonders how many Australians are proud of kangaroos despite having never seen one 07:39:07 * ais523 resists urge to say "Google's pretty famous" 07:39:24 Interestingly, there are some feral wallabees in Scotland 07:39:26 elliott: they're fairly common in the Australian countryside, aren't they? and even citybound Australians probably see them in zoos 07:39:34 even I've seen a kangaroo, in a zoo 07:39:56 I've seen a kangaroo in a zoo... in AUSTRALIA 07:40:01 you don't have to travel far to see one.. you just have to leave the cities 07:40:13 ais523: that was a nice rhyme 07:40:18 they're everywhere once you actually get away from suburbia 07:40:26 elliott: coincidental, and it doesn't scan properly 07:40:28 kangaroo / in a zoo / there's pretty much no way this rhyme isn't going to end up talking about kangaroo poo, so I'm stopping it here 07:40:40 which means it's a good rhyme in a bad context to put a rhyme in 07:40:46 everytime you go camping in melbourne you're bound to see one 07:41:17 hrm 07:41:22 they're everywhere 07:41:30 its just they're not in suburbia 07:41:30 * coppro sees a reasonable path to getting an Erdos-Bacon number of 65 07:41:31 *6 07:41:59 i don't get out much.. but i have seen them in the wild several times 07:42:16 coppro: Do tell. 07:42:29 Also, s/Erdos/Erdős/ 07:42:34 every encounter with a kangaroo tends to be memorable too 07:43:03 i guess they would get boring quickly enough 07:43:48 coppro: 65??? WOOOOOOOOOOW 07:44:13 i suspect 65 might actually be _harder_ than 6 07:44:49 pikhq_: My father has a Bacon number of 3; there are multiple profs at UW with Erd(I don't know how to compose the accent with my keyboard layout)s numbers of 1 07:45:11 Here, it's Compose = o 07:45:27 pikhq_: Although my father's Bacon number of 3 doesn't apply if you only count credited actors 07:45:32 as he was an extra 07:45:38 pikhq_: like in SML iirc 07:45:42 (and I haven't run the cast list of the film to see if he actually has a 2) 07:45:48 Crediting doesn't typically matter for Erdős-Bacon. 07:46:00 is it Erdős * Bacon that you're using to do the calculation? or appending digits? 07:46:07 Erdős + Bacon 07:46:12 aha 07:46:17 so 65 is really quite difficult? 07:46:28 Unless you have a time machine. 07:46:41 Bye, everyone 07:46:47 -!- Taneb has quit (Quit: Page closed). 07:46:49 Then, you can be enough generations removed to pull it off. 07:47:03 I still don't know what my Erdős number is 07:47:20 although I'm pretty sure I have one, given that I've coauthored papers with people who have coauthored lots of papers 07:47:21 * pikhq_ really would like for Bacon to appear in a film using stock footage of Erdős. 07:47:33 pikhq_: to give the lowest possible value of 1? 07:47:41 Yes. 07:47:42 it'd be funnier if they coauthored a paper 07:47:50 ais523: grep https://files.oakland.edu/users/grossman/enp/Erdos2.html 07:47:53 ais523: well i assume you'd have to get a long chain of people to cooperate with each other to get it that high 07:48:05 oerjan: Or time. 07:48:10 -!- cheater_ has quit (Ping timeout: 255 seconds). 07:48:15 * pikhq_ really would like for Bacon to appear in a film using stock footage of Erdős. 07:48:19 i think something like that actually happened 07:48:41 I thought it was just someone with a Bacon number doing that, though. 07:48:41 ais523: Vorpal isn't appreciating this feature I coded into mcmap, probably because he's away; please tell me about how great it is <-- what feature? 07:48:47 hmm, or not 07:48:52 Vorpal: see -minecraft 07:49:01 just after you joined 07:49:02 Yup. 07:49:06 it's currently broken btw :P 07:49:10 oh wait i didn't commit the broken 07:49:11 Erdős has an Erdős-Bacon number of 3. 07:49:11 it's not broken 07:49:40 ais523: oh wait you filter URLs 07:49:51 no he doesn't 07:49:55 not any more 07:50:01 I delink them 07:50:07 but can still see what they say 07:50:09 Which is the lowest known Erdős-Bacon number, though a few people have that. 07:50:10 What counts as a "paper" for Erdos number? 07:50:23 an academic paper 07:50:27 I did a simple proof with ais523. Does this mean I might have a .. blah 07:50:31 Sgeo: The *stock* definition is an academic, published paper. 07:50:51 well, my Erdős number is at least 4, it seems 07:50:55 Sgeo: no. 07:51:19 ais523: clearly we should collaborate and I should get a research term with one of the two profs at UW I know of with 1 07:51:24 that will give you a firm 3 07:51:28 Now I will sleep soon. Tell me now or later, if you have more ideas about my optimization problem or any other IRC channels you know that they could discuss these things. 07:51:41 -!- zzo38 has quit (Quit: >). 07:51:57 ais523, clearly, we should write a paper >.> 07:52:03 (I actually have an eye on one in particular since he both works in an area I'm interested in and taught me last term) 07:52:03 Oh, sorry, it's any research collaboration resulting in a published work. 07:52:11 07:52:20 A collaborative, unique proof would probably count. 07:52:23 bleh, does that mean that elliott has an Erdős number because I mentioned what I was doing to him? 07:52:26 ais523: :D 07:52:28 and he replied? 07:52:34 ais523: everyone you have EVER LOOKED AT has an erdos number now 07:52:41 ais523, is our GoE proof unique? 07:52:58 Sgeo: I don't know what you're referring to 07:53:04 well, I mean, my coding style is slightly influenced by elliott 07:53:10 from being exposed to his programs 07:53:18 and I did a lot of programming towards one of the papers 07:53:35 (I'm still trying to figure out why I switched from Allman-style braces to K&R-style, though) 07:53:39 ais523, that thing about any sufficiently small pattern in a bounded Game of Life universe being a result of some Garden of Eden 07:53:46 (for when I'm not trolling with EOL braces) 07:53:54 Sgeo: it probably isn't new 07:54:00 Oh, darn 07:54:01 someone else has probably wondered that at some point 07:54:11 well, I mean, my coding style is slightly influenced by elliott 07:54:12 really? 07:54:15 or is this just in a butterfly effect sense 07:54:32 Sgeo: what is the result, precisely? 07:54:39 elliott: well, I've had style arguments with you, and they help to sharpen my opinions of my own position 07:54:49 and I've started using K&R/OTBS style for no obvious reason at all 07:55:02 mcmap puts { on the line after branch constructs and it makes me unhappy 07:55:08 *OTB style 07:55:19 elliott: I used to do that, because the books I learned C from did 07:55:41 I really dislike {-on-its-own-line 07:55:42 coppro, every sufficiently small (5 width and height smaller than the universe) pattern in a bounded Game of Life universe has at least one Garden of Eden that results in it. 07:55:47 iirc 07:55:53 coppro: do you really dislike }-on-its-own-line? 07:56:00 OTBS is like Perl, it's inconsistent in the name of looking nice 07:56:01 ais523: I do 07:56:06 (not) 07:56:23 Sgeo: interesting. What's the proof? 07:56:29 also bounded -> toroidal? 07:56:52 hmm, another plausible indentation style is { lined up with if(x), on the same line as the line after, and } at the end of the line, next to the margin (i.e. column 78) 07:57:01 I wonder why nobody uses it? 07:57:10 ais523: ewwww 07:57:17 coppro, yes, although it works with other topologies. And probably also with always-off beyond a point, but not sure, and ais523 wasn't paying attention. 07:57:30 coppro, let's see if I can remember it offhand. It's somewhere in logs 07:57:33 Sgeo: "yes" is not a proof :P 07:57:52 Yes was to bounded -> toroidal 07:58:04 ah ok 07:58:16 coppro: now I'm trying to think of something that's proved by infinitely many copies of the letter 'y' 07:58:18 hey ais523, here's a screenshot you won't understand: http://i.imgur.com/Oygim.png 07:58:18 but I can't 07:58:28 * itidus20 notes that there is no coding-style applicator editor that i have heard of 07:58:28 it could make a decent counterexample to various banal and pointless statemetns 07:58:39 coppro: now I'm trying to think of something that's proved by infinitely many copies of the letter 'y' 07:58:45 "there exists an infinite stream of 'y's" 07:58:51 (constructive-style proof) 07:58:53 elliott: something vaguely interesting 07:58:58 itidus20: you mean a reformatting editor? 07:59:07 ais523: The only thing I dislike more than putting opening braces on their own line is putting them on their own line AND indenting them halfway 07:59:09 ais523: that there exists infinite anything would be considered interesting to some people 07:59:11 (see ultrafinitists :P) 07:59:11 itidus20: several programs, like Emacs and Kate, will shuffle lines left and right to comply with a coding style you give them 07:59:20 NetBeans, at least, will completely re-pretty-print your program if you tell it to 07:59:27 and vim 07:59:45 elliott: perhaps you should give ultrafinitists a copy of yes 07:59:49 and watch their minds explode 07:59:49 ais: I can independantly discover any feature given enough time 07:59:58 coppro, in a bounded universe, every pattern is either a oscillator or a precursor to one. 08:00:18 (counting things like empty universe as an oscillator) 08:00:33 -!- cheater_ has joined. 08:00:34 elliott: I assume it's Minecraft-related; the first thing that came into mind when I saw it was that it was a Dwarf Fortress map dump 08:00:42 and Minecraft map dumps are likely to look similar 08:00:45 There are no infinite growth patterns. Eventually, you're going to start repeating. 08:00:47 as people do similar things to the map in those games 08:00:55 Sgeo: Right. I know that result 08:01:14 y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y 08:01:16 y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y 08:01:17 y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y 08:01:38 elliott: I assume it's Minecraft-related; the first thing that came into mind when I saw it was that it was a Dwarf Fortress map dump 08:01:44 It's not really a dump, more a live view 08:02:14 elliott: perhaps you should give ultrafinitists a copy of yes 08:02:14 and watch their minds explode 08:02:25 ais523: they'd, unfortunately correctly, probably argue that it was still bound by C's restrictions 08:02:28 at least, if they were IN THE KNOW 08:02:33 because they know the OS is written in C 08:02:35 yes doesn't count its iterations, does it? 08:02:40 If you take a pattern that's just a precursor, and work "backwards" (for all possibilities of backwards), you can't end up with the same pattern again, otherwise, it wouldn't be a precursor 08:02:53 ais523: No, but the OS has to print it out 08:02:57 and depending on what you're outputting, the OS doesn't have to either 08:03:02 *outputting to 08:03:06 say you're outputting to a parallel printer 08:03:06 ais523: OK, but the hardware will still degrade :P 08:03:12 it's going to run out of paper eventually, I suppose 08:03:18 but the ys will still be there conceptually 08:03:19 ais523: You could print to screen and wipe anything that trails off 08:03:21 You can't go "backwards" infinitely, it has to stop sometime before 2^area 08:03:56 ais523: How come Junethack has muted goals in AceHack? Isn't AceHack basically vanilla for the purposes of gameplay? 08:04:09 So every precursor has a Garden of Eden that results in it. 08:04:15 coppro: it's because I didn't want to break save compatibility to add a patch that was implemented really badly 08:04:21 Sgeo: right 08:04:23 so I restricted myself to stuff that the game tracked already 08:04:28 apparently (according to the source of translations of the tipitaka) buddha said that the first being in a realm(not sure if i have this right) assumes himself to be god 08:04:30 ais523: ah 08:04:40 planning to play in junethack, btw? 08:04:49 what are muted goals? 08:04:55 (I'm still astounded that Slashdot accepted the story) 08:05:00 Now, for every pattern of a certain size or smaller, you can trivially make a precursor by putting a single live cell outside of causal contact with the pattern. Therefore, the pattern has a precursor, and therefore, there's a Garden of Eden that results in it. 08:05:02 elliott: they're not really muted, just slightly different than the other variants 08:05:18 elliott: muted meaning less difficult 08:05:30 ooh, the submitter of the slashdot submission is the one with that insulting crawl infodb entry 08:05:37 ais523, did I forget anything/make any mistakes? 08:05:38 well insulting-ish :P 08:05:40 I thought of that in particular because "beat sokoban" is replaced with "consult the oracle" 08:05:47 the others are close to isomorphic 08:05:51 (long story involving monqy as to why I know/recall this so recently) 08:05:51 Or be too long-winded for something simple? 08:06:02 Sgeo: I wasn't playing attention 08:06:07 oh, the luckstone from mine's end isn't 08:06:09 elliott: bhaak is the developer of UnNetHack 08:06:14 it's replaced with getting into the quest 08:06:18 coppro, ping 08:06:20 "bhaak 08:06:20 Still believes that Nethack is the most popular roguelike. 08:06:20 Also believes that the earth is flat. 08:06:20 Happens to maintain a Nethack fork, UnNethack. If you're curious how a Nethack with some thoughts spent on balance could look like, give it a try: http://sourceforge.net/apps/trac/unnethack/ 08:06:20 Is on the record as stating that he is not a shark. 08:06:21 Likes ke$ha" 08:06:30 and possibly the person with the most (although not the largest volume) of AceHack patches but me 08:06:39 Sgeo: makes sense 08:06:47 goodplayers 08:06:47 10 wins. Add yourself to the list with !nick goodplayers yourname. Don't be a monqy and delete the list. 08:06:47 The number of wins required to be a goodplayer increases over time. It is defined to be one more win than monqy currently has, except when it gets fixed at nice numbers like 10 or 27 for a while. 08:06:55 monqy needs to extricate himself from this learndb 08:07:23 http://twitpic.com/monqy <-- google search result for monqy, COINCIDENC???? 08:07:32 a picture of a twit? 08:08:03 also, I'd be very surprised if Crawl were more popular than NetHack 08:08:09 but Crawl's playerbase, I expect to be easier to count 08:08:12 ais523: I wouldn't, at all 08:08:12 Sgeo: Now, for the killer question: can you work out whether this applies in the totally general case? 08:08:22 NetHack's a part of UNIX culture, Crawl isn't 08:08:29 ais523: Nobody on Windows plays NetHack, lots of people on Windows play Crawl (tiles, universally) 08:08:36 (Yes I know not "nobody" etc. etc. etc.) 08:08:42 elliott: most of the YouTube results for NetHack are played on Windows 08:08:44 coppro, general as in generalizes to other rules, or to infinite sized universes? 08:08:45 Also, Fushigi no Dungeon. 08:08:50 Also see sheer volume of DCSS on http://www.reddit.com/r/roguelikes, although I realise the sample might be biased 08:08:53 ais523: bad logic 08:08:58 Sgeo: To all finite patterns 08:08:58 elliott: indeed 08:08:59 X=>Y =/= Y=>X 08:09:10 (a series of Japanese console rogue-likes) 08:09:11 I'd be very surprised if NetHack had more active players than DCSS 08:09:14 but it's still relevant, you just have to plug the numbers into Bayes' Theorem 08:09:20 pikhq_: you don't have to tell me what the Mystery Dungeon series is 08:09:24 I'm too lazy to be a Bayesian 08:09:28 some of the games in it are released in the UK too 08:09:33 "monqy 08:09:34 killed good players, is a horrible person" 08:09:34 ais523: Not everyone would know. 08:09:36 yep 08:09:59 coppro, good question, I'd like to try, but I have no reason to believe I can. I am currently mentally playing with more GoE stuff though 08:10:21 I'm actually quite a fan of Pokémon Mystery Dungeon; its bonus levels (Zero Island) have a balance unlike anything I've seen in any other roguelike, and are sufficiently fun that I'm considering writing my own roguelike based on similar principles 08:10:37 it's like what I think Crawl ought to be like, rather than what it actually is like 08:10:39 ais523: interesting; your recommendation might make me pick one up 08:11:00 coppro: be careful; the first 85% or so of Pokémon Mystery Dungeon is insultingly easy, and you have to play through it before you reach the fun part 08:11:11 most people have given up before they reach it 08:11:16 squarelos totally looks like spanish ors omething 08:11:35 ais523 is going to reply to this with an opinion on squarelos 08:11:48 elliott: nah, I'm not good enough at Crawl to have a serious opinion on it 08:11:59 but apparently, the developers mostly dislike it, and the top players mostly like it 08:12:15 presumably the wider playerbase mostly dislikes it because it looks worse than circlelos, but I'm not sure 08:12:21 ais523: monqy seems to be behind it 08:12:45 clearly, we need to inundate him with these mixed opinions 08:13:13 ais523: I was planning to go for the discount bin anyways 08:15:34 ais523: completely unscientific comparison based on seeing a link to an SA forums thread in the crawl learndb: Crawl's thread has amassed 141 pages in three months; I can't even /find/ a thread for NetHack, but Minecraft's has attained 574 in six months 08:15:52 I think it's fair to say that Crawl is the most popular roguelike THAT PEOPLE ACTUALLY THINK IS A ROGUELIKE WHEN THEY'RE PLAYING IT OK OK PIKHQ 08:17:36 Now back to mcmap 08:18:05 another completely unscientific comparison: NetHack's been slashdotted more often than Crawl 08:19:04 ais523: oh, /come on/ 08:19:11 that's about seven hundred times more biased than mine 08:19:54 also, the SA thread for Crawl was, for ages, the main forum to discuss the game 08:19:59 which is why it's mentioned there 08:20:04 a? 08:20:07 *sa? 08:20:12 coppro: something awful 08:20:22 they moved away to the Tavern because they didn't like a pay site having most of the discussion about the game 08:20:24 one of the largest forums on the internet, with some site that nobody reads attached to it 08:20:41 is it the largest paid forum, I wonder? 08:20:41 elliott: it is indeed something awful 08:20:53 things like many of the 4chan boards are presumably larger, but you don't have to pay to post there 08:20:55 ais523: according to http://www.big-boards.com/, no; offtopic.com is larger 08:21:09 coppro: what an original joke 08:21:09 heh, I'm not surprised there are forum rankings 08:21:18 but you only mentioned one other forum, so I'm guessing that they're second 08:21:23 ais523: unfortunately, that ranking omits the largest forum on the internet because it doesn't have the data they want 08:21:30 and no, they're just the only one I remember being pay-for 08:21:32 aha 08:21:34 I checked the biggest ones out a while back for some reason 08:21:42 (the largest forum on the internet is 2channel, by a large margin) 08:21:48 oh, that makes sense 08:21:50 (but they don't have e.g. membership information as they don't have members) 08:22:00 oh wait "RegistrationOptional, USD33.00/year" 08:22:02 maybe they do :-P 08:22:21 does a paid forum end up being better? 08:23:05 itidus20: depends on what you mean by "better" 08:23:17 it's going to discourage a huge number of potential posters, and probably many readers too 08:23:30 but that's not necessarily going to be a bad thing 08:23:53 * elliott has ended up coding mcmap in gedit, I wonder why? 08:24:01 its a weird idea 08:24:12 Well, I know why, it's because vi displays tabs as spaces, so I can't copy and paste properly with my terminal, and because emacs just feels wrong here 08:24:12 http://www.mezzacotta.net/archive.php?date=2012-12-22 My goodness that is wonderful. 08:24:32 ais523: SA seems to do well enough with those: 08:24:32 5,187 Users Logged In 08:24:33 3,550 registered users logged in. 08:24:33 155,024 users total. 08:24:41 (OK, it's probably one of the only successful for-pay forums.) 08:24:45 elliott: if you're a paid forum, being big is definitely to your advantage 08:24:53 just like if you're a social network, being big is definitely to your advantage 08:24:54 Obvious statements woo :P 08:25:04 pikhq_: what is so special about that date? 08:25:15 coppro: you can figure that one ou 08:25:15 t 08:25:20 pikhq_: haha, that must have been specialcased 08:25:21 coppro: Day after the Mayan calendar ends. 08:25:26 ais523: Undoubtedly. 08:25:27 it's not 08:25:29 http://www.mezzacotta.net/archive.php?date=2012-12-20 is the same 08:25:40 as is http://www.mezzacotta.net/archive.php?date=2012-11-01 08:25:44 you're doing something wrong 08:25:47 oh wait, duh 08:25:48 oh right, those comics haven't been written yet 08:25:49 you can't see future comics 08:25:59 Aaaah, it only peeks into the past. 08:26:08 yeah, I know the date 08:26:14 just nothing seemed special about that site at that date 08:26:24 coppro: well, that the comic didn't exist 08:26:32 on account of the universe not existing, just like mezzacotta extends back to the beginning of time 08:26:41 it's only amusing if you know what mezzacotta is 08:27:20 I do not 08:27:40 http://www.barnesandnoble.com/w/the-assemblies-of-god-holy-bible-god/1030075545?ean=2940012671967&itm=2&usri=assembly%2bof%2bgod oh, this pattern of "helpful" votes isn't suspicious 08:28:01 coppro: It is a webcomic which has an update for every day. 08:28:06 it has the largest archives of any webcomic ever 08:28:25 ais523: well, MSPA will surpass it in a few years, I'm sure 08:28:35 presumably it was going as a newspaper comic before it first came online, and as a pamphlet comic before that, and as a stone carving comic before that 08:28:50 and as a plsama painting comic before that 08:28:51 plasma 08:28:55 Specifically, every day on the proleptic Gregorian calendar. 08:29:34 And then some. 08:29:37 I GUESS MY JOKE TRULY WAS THAT BAD 08:29:48 http://www.mezzacotta.net/archive.php?date=-9999999999999-01-01 Here's the earliest comic. 08:29:49 pikhq_: DMM could tell that the Gregorian calendar was going to be introduced millions of years before it was, so he worked on that assumption 08:29:58 also, my fingers tried to tab-complete Gregorian, and it half-worked 08:30:06 hahaha 08:30:10 -!- azaq23 has quit (Quit: Leaving.). 08:30:19 hmm, "I think this comic is going downhill." must have been special-cased as the first line 08:30:55 coppro, you should read all of them. 08:30:57 elliott: it could just be chance 08:31:16 There's a comic worth archive-binging. 08:31:25 Especially since you'll need to invent immortality to do so. 08:31:35 ais523: yeah, but /come on/ :P 08:31:36 elliott, ais523, or, perhaps the function determining the comic was designed with that result in mind 08:31:37 Well, not necessarily immortality. 08:31:46 Just being highly prolonged. 08:32:16 ahh.. is mezzacotta a procedural comic? 08:32:23 Yes. 08:32:27 itidus20: no, it is hand-written by the comic irregulars 08:32:28 pikhq_: fuck you 08:32:31 Either that or DMM is amazing. 08:32:39 IT IS NOT JUST DMM DO NOT DISCOUNT THE CONTRIBUTIONS OF OTHERS 08:32:55 today's mezzacotta is funny 08:33:11 although the first and second panels should be swapped 08:33:51 -!- Sgeo_ has joined. 08:33:57 "The rod of alertness is planted (set) to prevent each instance of a beetle." 08:34:17 * elliott really wants to play frictionless tennis now 08:36:19 -!- Sgeo has quit (Ping timeout: 255 seconds). 08:37:14 the D&D markov chain person strikes again? 08:37:21 ? 08:37:53 elliott: one of the mezzacotta people just does markov chains of the d20 SRD (basically, D&D minus trademarks) 08:38:10 and itidus20's quote sounds like it was generated like that 08:38:40 there is a tetris dude 08:38:44 goes by holdnext 08:38:47 ohhhh 08:38:55 i swear to god every post he makes sounds like a markov chain text generator 08:38:58 so theres actually characters 08:38:58 but apparently he's a real person 08:39:22 um fizzie 08:39:27 Uint32 v = *b; 08:39:27 if (v < 64) 08:39:27 rgba = RGBA_OPAQUE(4*v, 4*v, 0); 08:39:27 else 08:39:27 rgba = RGBA_OPAQUE(255, 255-4*(v-64), 0); 08:39:29 b is the array of blocks 08:39:34 don't you want to use y, not v?... 08:39:57 oh wiat no 08:40:00 b is different for that 08:40:01 argh 08:43:41 ais523: 08:43:41 int lv_block = c->light_blocks[bx*(CHUNK_ZSIZE*CHUNK_YSIZE/2) + bz*(CHUNK_YSIZE/2) + ly/2]; 08:43:42 int lv_day = c->light_sky[bx*(CHUNK_ZSIZE*CHUNK_YSIZE/2) + bz*(CHUNK_YSIZE/2) + ly/2]; 08:43:42 if (ly & 1) 08:43:42 lv_block >>= 4, lv_day >>= 4; 08:43:42 else 08:43:42 "Not being one-to-one is not considered sufficient of a function for it to be called one-way (see Theoretical Definition, below)."? 08:43:44 lv_block &= 0xf, lv_day &= 0xf; 08:43:44 Huh? 08:43:46 typical mcmap code 08:44:04 gah, stupid italics indentation 08:44:10 (my client interprets tab as "toggle italics") 08:44:23 Sgeo_: yes, those are different concepts. 08:44:50 one-to-one == injective (probably what you want for GOL stuff) 08:45:00 I'd say the bitshifts/bitmasks are correct there, as it looks like it's trying to unpack a format that packs multiple 4-bit chunks into an octet 08:45:03 oerjan, I'm on this page for different reasons 08:45:04 one-way = cryptographical 08:45:06 and the mathematical behaviour is irrelevant 08:45:29 oerjan, and I'm kind of just wondering about hashes not necessarily being one-way 08:46:00 Sgeo_: well no one has proved that one-way functions definitely _exist_ 08:46:22 (requires P != NP) 08:46:33 I should go to sleep now 08:47:20 oerjan: does not require P != NP; if checking is O(n^2) and breaking is O(n^999), it's going to be safe for all practical purposes with high enough n 08:47:39 well depends on your precise definition then 08:48:00 P != NP iff P != 0 and N != 1 08:48:08 08:48:15 it's been made before 08:48:41 also, what if P = +Inf, N = 2? 08:48:44 +Inf is not a number 08:48:56 no, that's NaN 08:49:03 > (1/0) == (0/0) 08:49:04 False 08:49:07 elliott: +Inf * 2 = +Inf 08:49:14 > (1/0) * 2 08:49:15 Infinity 08:49:28 it was a joke 08:49:37 ais523: I would just like you to know that I edit mcmap with four-wide tabs 08:49:37 oh, I see 08:49:40 in response to coppro 08:49:49 YOU ARE LIVING A LIE 08:50:14 Who doesn't use four-wide tabs for stuff? 08:50:19 elliott: why not edit it with half-width spaces too? 08:50:26 Sgeo_: FREAKS 08:50:28 ais523: SOOO ANGRYYYYYYYYY 08:50:33 Sgeo_: oh god don't 08:50:37 you're not me, so he'll actually respond 08:50:41 and we'll be here for hours 08:51:28 I actually posted that Linux coding style guide quote to someone in a different channel earlier 08:51:35 when we were about to get into a tabs vs. spaces vs. tabs=8 war 08:51:39 I need to go to sleep 08:51:46 but it killed the conversation, obviously they were unable to argue with ti 08:51:47 when we were about to get into a tabs vs. spaces vs. tabs=8 war 08:51:48 *it 08:51:49 uh oh ais523 08:51:58 you implicitly acknowledged that there MAY exist tabs that are not equal to eight psaces there 08:52:05 better be careful!!!! 08:52:19 elliott: no, I acknowledge that some people use tabs for purposes that most programs are incapable of reading 08:52:26 Erm, I meant "4 spaces for each indent level" when someone said tabs 08:52:31 * Sgeo_ hides 08:52:40 the only programs I can think of where tab width is variable, are some programming editors, and word processors 08:52:43 everything else uses tab=8 08:52:50 (well, QBasic uses tab=14, but it's insane) 08:52:51 I think my Python is showing 08:52:53 ais523: hehehehhehehe WOW LOOKS LIKE I'M NOT GETTING INTO THIS AGAIN WITH YOU how strange 08:52:57 Wow, that sounsd wrong 08:55:39 hey, ahtewa's using ä in Eodermdrome programs 08:55:41 that's cheating! 08:55:48 although, I suppose I didn't specify the alphabet 08:55:55 and it does make it more pronounceable 08:56:56 -!- ais523 has set topic: Tarn was pleased. “The hippos like the sewers!” he said. He took a celebratory swig of Dr. Pepper, and rocked back and forth. | Logs: http://codu.org/logs/_esoteric/ and http://tunes.org/~nef/logs/esoteric/?C=M;O=D. 08:57:07 clearly we need to get someone vietnamese to write the eodermdrome C interpreter 08:57:15 (all I added was a comma; it's correct either way but means something slightly different now) 08:58:46 ais523: it looks to mean the exact same thing to me 08:58:49 since I know the context 08:59:01 how has the meaning changed? 08:59:14 oh, I think with the comma it implies that the swig of Dr. Pepper and the rocking happen one after the other 08:59:19 and without, it leaves it ambiguous as to whether they overlap 08:59:23 haha 08:59:25 OH THE MEANING-CHANGE 08:59:33 I said "slightly different" 09:00:15 slightldyifferent 09:00:30 ais523: the source of that is an amusing read, by the way 09:00:47 ais523: It's an NY Times article that purports to be about Dwarf Fortress, but is mostly a list of all the ways in which the NY Times writer thinks Tarn is weird 09:01:11 oh, I, err, looked at that article 09:01:24 but it was paywalled 09:01:37 that paywall is bypassable with a single line of JS, IIRC 09:01:41 it's literally just an overlay 09:02:08 not in my case, it wouldn't scroll, and I assume the article was longer than a screenful 09:02:16 it's possibly because I declined cookies and didn't use JS 09:03:03 ais523: umm, you can disable scrolling, I think 09:03:05 but fair enough 09:03:16 but you could see the article greyed-out with the one I'm talking about 09:03:20 I couldn't 09:04:32 -!- oerjan has quit (Quit: Good night). 09:04:58 ais523: do you know how to check whether the window is focused in SDL? 09:05:05 I suppose I have to handle the event... 09:05:15 elliott: I don't, I've hardly used SDL 09:05:23 I can't even remember if DNA Maze is written in it or not 09:05:34 I know I straight-ported it to something, but can't remember if it was SDL, although it seems plausible 09:06:08 it is :P 09:06:32 straight ports generally don't require much thought 09:06:46 sufficiently so, that you can forget what libraries you're working with 09:07:16 I've learned very little SDL working on mcmap, although I've mostly left the actual map parts alone until now 09:08:33 -!- ais523 has quit (Remote host closed the connection). 09:19:02 Ok so an idea came to me. 09:19:46 The degree to which a randomly generated world is enjoyable is probably correlated to the amount of cooperative multiplayer conducted within that world 09:20:22 in other words, always looking for a human meaning 09:20:28 a great many people enjoy single-player minecraft but don't enjoy it online 09:20:36 even cooperatively 09:20:52 but minecraft at least has NPCs I guess thats the diffeernce 09:21:05 maybe you don't need humans 09:21:10 maybe NPCs are enough 09:21:46 lots of people play in peaceful, and then there are no NPCs inside any real structure 09:21:54 and sheep and pigs are generally more of an ignored annoyance outside 09:21:58 cows too 09:22:02 chickens are too tiny to notice 09:22:52 ok so.. perhaps it is relative to how busy the persons social life is 09:23:10 like, the more social the more enjoyable? 09:23:19 oh.. i mean.. 09:23:22 If anything it's the opposite, although I doubt even that; plenty of MC players seem to be sociable 09:24:36 i dont have much of a social life .. so gradually I don't enjoy being too alone 09:25:06 Well, with the size and speed a lot of builds are constructed, I doubt a lot of the big ones involved too much social interaction during the build period... 09:26:14 I mean.. I'm indoors at home all day every day. At first it seemed like the perfect chance to be alone with computers and books. But gradually I found dependance on chatrooms etc 09:28:34 looking for human meaning? mc's randomly generated worlds are way better than those constructed by hand for more expensive games, exactly because they don't look like they've been drawn by a human 09:29:33 well i was thinking about this while reading mezzacotta 09:29:41 oh well that explains it then 09:30:26 you should not share ideas before they are properly cooked 09:30:44 lol 09:31:04 oklopol: so ais should never have talked about feather? :p 09:32:25 " Well, with the size and speed a lot of builds are constructed, I doubt a lot of the big ones involved too much social interaction during the build period..." <<< doing stuff together is just a more primitive form of social interaction 09:33:27 i referred to solo builds 09:33:40 oh sorry 09:33:56 years ago playing final fantasy 3/6 on a SNES I started to imagine "This is great. It's a shame the NPC is 'stuck' to a set of scripted events." not in those exact words 09:34:03 i thought you meant they were multiplayer builds but so big and quickly done that the players must've just been building like crazy 09:35:08 ugh, scripted events... is there a worse 09:36:18 But as I get older, I start to realize you can't replace the storyteller with a computer. 09:36:34 Sometimes the story is well-written 09:36:42 well read a fucking book 09:36:58 the point of games is you can be creative 09:37:08 and the game does a good job of expanding the story in the game 09:37:14 no it doesn't 09:37:16 it never does 09:37:18 yes it does 09:37:21 no it doesn't 09:37:22 it never does 09:37:27 I have to disagree 09:37:31 well i have to agre 09:37:32 e 09:37:35 ...with myself 09:37:36 k then 09:37:43 lol 09:37:45 k indeed 09:37:49 obviously this disagremeent can be settled in only one fashion 09:37:54 obviously 09:37:55 what is that 09:38:01 fight to elliott's death 09:38:22 coppro> I have to disagree (with oklopol) well i have to agree (with oklopol) 09:38:39 elliott has the Magic Spirit of Rightness in him 09:38:52 the person who kills him and takes the Magic Spirit of Rightness will, thereafter, be right 09:39:28 coppro: i say oklopol is right because i like him more 09:39:32 issue resolved 09:39:38 " and the game does a good job of expanding the story in the game" <<< afaik, in every game the story can only be expanded locally, or you can explore a few main branches. to me it just feels like they couldn't come up with a long enough story so they added some parts where i have to press x repeatedly. 09:40:23 with slightly (very slightly) varying values of pressing x 09:40:43 looks like pixelcomic is over 09:40:52 are you mocking me? 09:41:08 no 09:41:10 "THE FACT THAT IT IS NAMED HURRICANE PIXEL IS WHAT MAKES IT INTERESTING 09:41:10 IMAGINE A WEATHERMAN IN REALITY TALKING ABOUT "HURRICANE HUMAN" 09:41:11 TO WARRANT SUCH A NAME THIS HURRICANE WOULD HAVE TO BE CAPABLE OF DESTROYING THE EARTH 09:41:11 (WHICH WOULD HAVE A DELICIOUSLY IRONIC VALUE IN NAMING IT "HUMAN" MWAHAHA AHEM)" 09:41:15 are you sure 09:41:18 the annotations of this were ten times better than the comics itself 09:41:19 oklopol: yes 09:41:31 because that would've been a rather subtle form of mocking 09:41:45 ...:| synchronicity 09:41:47 http://pixelcomic.net/016.shtml 09:41:50 "ARE YOU MOCKING ME OR SOMETHING" 09:42:19 PIXELCOMIC UPDaTeD? 09:42:22 Damn shift key 09:42:36 *pixelcomic updAtEd 09:42:38 Sgeo_: nope 09:42:42 but its index page is now a list of all strips 09:42:45 which implies overness to me 09:43:54 :( 09:44:51 I AM NOT ALONE WHEN I SAY THAT 09:44:51 I ALWAYS FOUND IT ODD HOW LOIS LANE WAS RENDERED INCAPABLE OF RECOGNIZING SUPERMAN 09:44:51 SIMPLY FROM A PAIR OF GLASSES 09:44:51 I WONDER IF ANYONE HAS EVER BEEN INSPIRED TO ROB A BANK WEARING GLASSES 09:44:51 AND THEN SUCCESSFULLY ESCAPE BY THROWING AWAY THE GLASSES AND WALKING CASUALLY 09:45:11 http://pixelcomic.net/020.shtml 09:45:14 these are awesome 09:46:11 oklopol: i love how it'd be a really shitty comic without the annotations making fun of it 09:47:45 i would love it anyway 09:47:49 elliott, btw, MSPA is more stable now 09:47:55 Sgeo_: i read it ages ago 09:47:57 not as good as quimbox but anyway 09:48:04 oklopol: less than three hundred comics for yr enjoyment 09:48:27 the last strip is pretty much the best 09:48:49 http://pixelcomic.net/044.shtml <-- oerjan 09:52:03 oklopol: also what's quimbox 09:52:21 well you read the comic right? 09:52:54 which comic 09:53:01 quimbox 09:53:03 you must have 09:53:10 because you said it was the worst comic in the world 09:53:20 is that the vjn thing 09:53:35 yes 09:53:43 istr it being amazing 09:53:49 anyway quim = pussy, assuming you live a few hundred years ago 09:54:25 elliott: The Magic Spirit of rightness is not transitive 09:54:33 the punchline is going to be amazing. 09:55:22 -!- nooga has quit (Remote host closed the connection). 09:55:51 "I LOVE TO IMAGINE THAT WHEN A PIXEL STARTS ROLLING SOME SNOW 09:55:51 IT AUTOMATICALLY REMAINS SQUARE SHAPED" 09:56:30 -!- Somelauw has joined. 09:56:37 elliott: okay pretty quickly the annotations turn out crucial for the jokeness 09:56:50 okay so my ridiculous gmail plan might actually work 09:56:56 oklopol: it's great :D 09:57:09 -!- Somelauw has left. 09:57:16 oklopol: he's like "oops this is totally unfunny, let me say funny things as an annotation" 09:57:53 i love how even a comic about pixels gets significant art improvement 09:58:30 I GUESS SOME PIXELS BECOME CELEBRITIES FOR BEING IN SO MANY MOVIES 09:59:33 there is some controversy over whether pixels are infact rectangular 10:00:10 but thats to be expected i guess 10:00:44 there's that thing about them ... not being 10:00:50 for algorgorgotihms 10:01:10 HEY IDENTICAL PANELS THAT ARE DIFFERENTIATED SOLELY BY THE DIALOGUE 10:01:10 I AM A TRUE COMIC ARTIST NOW 10:01:10 contravarsial dinosaur comics dis 10:02:37 I ENCOUNTERED THIS FIERCELY ORANGE CREATURE IN MY BACK YARD 10:02:37 AND WHEN FETCHING MY DIGITAL CAMERA I OF COURSE FELT THE NEED TO RUN 10:02:38 IN FEAR THAT THIS THING WOULD ESCAPE IN 8 WHOLE SECONDS 10:02:54 Bloody useless and trivial result: In an m x n section of space, there are no oscillators with period larger than 2^(m*n) 10:03:28 are you sure? 10:03:40 :D 10:04:12 * Sgeo_ is referring to two-state CA with an I-forget-the-name neighborhood 10:04:17 if state is binary 10:05:00 conway's game of life neighborhood? 10:05:04 Yes 10:05:22 -!- Taneb has joined. 10:05:33 Hello 10:05:35 Although the neighborhood is irrelevant, come to think of it. well, hmm. What exotic neighborhood could break it? 10:05:38 have you proved it or merely sampled it? 10:05:38 092 - EXERTION OF POLAR OPPOSITION - 020922 10:05:38 I AM ENJOYING THE FACT THAT I AM RAPIDLY BECOMING THE MASTER OF NOT MAKING SENSE 10:05:38 IF ANYONE IS INTERESTED IN SUBSCRIBING TO AN APPRENTICESHIP FOR NOT MAKING SENSE 10:05:38 PLEASE CONTACT ME VIA EMAIL OR AIM 10:05:42 but yeah if you have a configuration in {0, 1}^{Z^2} that has period vectors (0, m) and (n, 0), then necessarily the orbit of that configuration in your CA is eventually periodic with period less than or equal to 2^{mn}. 10:05:45 it links to both :D 10:06:01 trivial to prove but it's actually used all the time 10:06:08 itidus20, I'm too lazy to sample 10:06:28 What are we talking about? 10:06:43 " Although the neighborhood is irrelevant, come to think of it. well, hmm. What exotic neighborhood could break it?" <<< even if the neighborhood and rule are different at every cell, the result follows, just as trivially 10:07:22 your state set has size 2^{mn} so a deterministic system will be come periodic with period less than or equal to that 10:07:42 Taneb: cellular automata 10:07:47 isn't that just "finite state machines cycle" 10:07:50 Okay 10:07:53 ZOMG WOWZ 10:08:43 oklopol, actually, my statement is _slightly_ different from that. I mean a finite sized pattern on an infinite mostly empty grid. Same idea though, but applies only to patterns that don't grow past m*n 10:08:47 Same thing though 10:08:59 Um, did I state "doesn't grow past m*n"? 10:09:02 I might not have 10:09:07 Sgeo_: but wrapping the rule around? 10:09:13 err 10:09:17 wait what 10:09:29 oh oscillator 10:10:21 THE MORAL OF THIS STORY IS THAT 10:10:21 EVOLUTION REQUIRES A GIANT DRIP OF SWEAT 10:10:44 110 - AN END - 030124 10:10:44 GIVE ME ONE GOOD REASON WHY I SHOULD CONTINUE WITH MY LIFE 10:10:44 AND BY LIFE I MEAN COMIC STRIP 10:11:09 this comic is way too smart for me 10:11:31 i cannot follow the conversations fast enough for this to even count as a comic 10:11:39 oklopol: it's study 10:11:43 yeah 10:12:52 -!- augur has quit (Remote host closed the connection). 10:13:01 -!- augur has joined. 10:13:13 elliott, i'll give you no good reasons! 10:13:57 THIS IS WHAT HAPPENS WHEN YOU GO TO COLLEGE 10:13:57 WHAT YOU LEARN IN COLLEGE IS 10:13:58 WHAT YOU TRULY VALUE IN LIFE 10:14:06 oklopol: it's like poetry for the soul 10:14:32 I want to treat areas of the grid as usefully finite for X time but it is probably imposible, at least usefully 10:15:07 oklopol: 123 is like oerjan but better 10:15:12 rip oerjan++ 10:15:57 "I WOULD HAVE BEEN BETTER OFF IF I WAS A HAMSTER RUNNING INSIDE MY HAMSTER WHEEL OF HAMSTER FORTUNE AND HAMSTER VIRTUE" 10:16:39 THE TOP OF THE LINE IN FLAT PANEL MONITORS USES PLASMA PIXELS 10:16:51 Derp misread that as it's used for the top row 10:17:03 ONLY $19.95 10:17:03 BUT WAIT 10:17:04 CALL NOW AND GET AN ADDITIONAL 10:17:04 MINUTE OF YOUR LIFE STOLEN 10:17:24 THE QUESTION IS 10:17:25 SUPERFLUOUS WEALTH SOMEHOW GIVES ONE 10:17:25 THE ABILITY TO GROW AN ERRATIC MOUSTACHE 10:17:25 a question apparently 10:17:35 -!- FireFly has joined. 10:18:38 Get out of here as soon as you can 10:19:32 wat 10:19:40 FireFly is an idler 10:19:51 Yeah 10:19:59 Why, is it illegal to idle? 10:20:08 " I want to treat areas of the grid as usefully finite for X time but it is probably imposible, at least usefully" <<< if this was still about CA, idgi, please explain 10:20:47 FireFly: nope, Taneb seemed to think you were new 10:20:56 areas of the grid would seem to mean (m x n section of space) 10:21:09 Nah, I was trying to stop more people be drawn into this single conversation 10:21:15 Rather than the channel as a whole 10:21:19 itidus20: yeah i got that 10:21:43 usefully finite sounds like the pipe dream part 10:22:36 so once again let me mention that injective on finite configurations <==> surjective 10:23:27 valid in any dimension and on many many many other groups 10:26:51 I can never remember jectives 10:27:25 surjective = onto, injective = one-to-one; onto = all have preimage, one-to-one = different go to different in the mapping 10:27:57 sur is just french for onto 10:27:59 There's another one 10:28:10 well bijective but that's easy to remember 10:31:57 -!- elliott has quit (Remote host closed the connection). 10:32:41 -!- elliott has joined. 10:35:45 well yeah but elliott doesn't know that 10:36:21 ofc i don't 10:38:08 really you only know what you've been told, or otherwise gathered from your perceptions, and also the stuff you've independently thought up in your head 10:38:40 and perhaps some other stuff that was there in the first place 10:38:57 plus i dunno some bonus maybe 10:39:00 pizza 10:39:04 -> 10:45:00 or is that just what they want you to thin 10:45:05 ^think 10:52:33 Derp. 10:53:04 I just proved to myself that there are countably infinite finite-sized oscillators, then realized that there are countably infinite finite-sized patterns 10:53:13 Ok, I think everyone's sick of my failures 10:56:07 nah 10:56:09 those aren't failures 10:56:29 they are just unremarkable successes 10:57:27 -!- MigoMipo has joined. 10:58:07 I assume the actual algorithm for enumerating through all finite-sized oscillators is also unremarkable? 10:58:31 As remarkable as the one for bit strings with a condition 11:00:49 I _just_ realized how to determine if a finite-sized pattern is a GoE or not... and I was talking about related stuff just hours ago. 11:00:52 * Sgeo_ slaps self 11:01:11 (And yes, that's what I was looking for for quite a long period of time here) 11:06:28 all of math is unremarkable 11:06:42 most of math is unremarkable 11:06:44 not all 11:07:24 Sgeo_: This sort of stuff is good. Keep at it, and tell me if you can prove or disprove the existence of oscillators with no entry points in finite grids of games of life 11:08:06 coppro, are you saying you know the answer? 11:08:48 Any such oscillators with no entry points would have to be a certain size, and have no internal gaps of a certain size. 11:09:20 all of math is unremarkable 11:09:23 so i take it you can't do any math? 11:09:45 (joking, but with a point.) 11:10:09 i mean, the answers are only remarkable until some other rule comes along and generalizes them 11:11:53 Let's restrict myself to an n*n grid. 11:11:55 Sgeo_: No, I don't 11:12:08 however, surely there are reasons for math beyond our understanding 11:12:32 the basic foundations of abstract algebra are utterly boring 11:12:38 but the wondrous things you can do with it are not 11:13:54 n*n grid, (n-5) by (n-5) patterns all have GoEs. But some of those patterns may oscillate into each other, so that there is 1 GoE for each... wait, no! 11:14:23 haha 11:14:48 Even in the case of oscillators, there are multiple GoEs for each phrase of the oscillator (as long as each phrase is at the particular size or smaller) 11:14:57 http://www.youtube.com/watch?v=svXeDHE6tBU 11:15:09 Because you could put that spare dot in several locations 11:15:19 You can even do it for n by n-5 11:15:55 most of math is unremarked, but quite markable. 11:16:15 Here is what I have so far: Any pattern with a 5x5 space admits an eden to its next generation 11:16:18 This puts a rather high lower bound on number of GoEs: 2^[n(n-5)] 11:16:18 We've got a sword 11:16:50 Since you can stick a dot in the center of that space without affecting the result (which is, in CA parlance, to say that they are twins) 11:17:48 Well, that greatly increases the lower bound 11:17:50 what's an entry point? 11:18:19 Say that | | are symbols for repeat 11:18:22 A |B C| 11:18:26 I've been holding a sword 11:18:27 A would be an entry point 11:18:33 I think 11:18:34 " the basic foundations of abstract algebra are utterly boring" <<< no they are not! 11:18:40 they are utterly awesome 11:18:58 oklopol: Axioms are boring 11:19:02 things that you do with axioms are cool 11:19:17 oh, i thought you meant things like the homomorphism theorem and stuff 11:19:32 Sgeo_: Thus to prove the existence of an oscillator with no eden predecessors, you must have an oscillator that never contains a five-by-five hole 11:19:44 this is necessary although not sufficient 11:19:49 it's "nothing new under the sun" but with numbers. 11:19:52 hehehe 11:20:24 coppro, ooh, I think, on a torus, "no 5x5 or larger hole" also forces it to be the necessary size 11:20:44 Sgeo_: yeah 11:20:44 obviously there cannot be an oscillator that's a GoE 11:20:50 or what do you mean by oscillator? 11:21:08 and so i take entry point = preimage 11:21:31 oklopol: We're trying to reason about the existence of an oscillator with no eden that will result in the oscillator through repeated evolution 11:21:52 On a finite grid such as a torus 11:21:58 well you do know you will never find a GoE for GoL? 11:22:04 we're working on toroids right now 11:22:07 oklopol: uh what 11:22:13 they are big. 11:22:54 well dunno maybe you are really smart or have a smart program or can read wp 11:22:57 There's an 11x11 known orphan 11:23:09 right, if you start from that, then you'll find an orphan easily 11:23:12 by doing nothing 11:23:24 but you would never find that yourself 11:23:24 oklopol, even in infinite grids, they exist. And ais523 and I proved that for any pattern on a finite grid that has a 5x5 or larger hole (coppro, I'm using this formulation now), there's a GoE that results in it 11:23:46 oklopol: no, but someone else found it for me 11:23:48 so whee 11:24:22 results in it? you mean evolves into it? no that's obviously false 11:24:44 s/finite/toroidal/ 11:25:02 well unless 0 iterations of the rule are allowed as well 11:25:19 yeah, this is accepting 0 iterations 11:25:25 actually that one holds in any finite space 11:25:42 no 11:25:46 oh for gol again 11:25:51 but the one about oscillators might be different between finite limited space, and finite toroidal space 11:25:55 yes this is all gol 11:26:01 CAs exist without edens 11:26:07 obviously 11:26:26 anyway, back to the subject at hand 11:26:48 " oklopol, even in infinite grids, they exist. And ais523 and I proved that for any pattern on a finite grid that has a 5x5 or larger hole (coppro, I'm using this formulation now), there's a GoE that results in it" <<< how did you prove this? 11:27:01 oklopol, read logs 11:27:01 and this is for periodic configs, that is, toroidal? 11:27:20 okay i have a counterexample 11:27:22 oklopol, yes. Or, might work for other finite as well. Certainly other finite that wraps 11:27:26 Oh? 11:27:39 no wait, maybe not, hmmhmm 11:27:42 where in the logs, today? 11:27:49 Either today or yesterday 11:27:55 I was explaining to coppro 11:28:01 oklopol: the proof is fairly straightforwad 11:28:04 Timezones scare me 11:28:07 *forward 11:28:15 oklopol: Just add a dot in the 5x5 hole 11:28:46 and what then 11:29:26 Work backwards. You're not in an oscillator, so all possible patterns going backwards must eventually be a GoE. 11:29:54 oklopol: More lucidly this time 11:29:56 so it was for periodic ok 11:30:41 and why aren't you in an oscillator? 11:31:06 If you aren't, you already must have a finite number of predecessors since there are only finite states to work from and repeated application will never get you back to an earlier one 11:32:05 well obviously, if you restrict to configs with periods m and n; why aren't you in an oscillator? 11:32:39 Let me restate the proof better 11:32:51 There's a really simple answer for that that is in my head that just won't come out 11:32:54 if you restate the trivial part, i will slap you 11:32:58 Unless my thinking is muddled 11:34:36 If you keep going backwards, and you're in an oscillator somehow, then you get...now I'm wondering if this proof only works for starting from oscillators... but if it's not an oscillator then it's a preimage to one 11:34:37 really i'd like to hear your definition of hole 11:34:54 like a 5x5 empty space you cannot move out of, using only white cells? 11:35:18 5x5 dead cells surrounded by anything 11:35:28 i mean obviously otherwise you can just take a 5x5 empty space and surround it with small oscillators to find something with infinitely many preimages 11:35:30 oh 11:35:36 Given some state, an infinite number of applications must eventually oscillate due to finite state space. If the repeated applications do not arrive at the original, you are not in an oscillator and must have finitely many predecessors. If you are in an oscillator, make a cell live in the center of a 5x5 square of dead cells. This pattern is a twin to yours, so they have the same successor. But the oscillator cannot reach this twin through re 11:35:40 well then it's obviously false 11:36:13 since the state wiht the added live cell is a predecessor as its successor is a part of the same oscillator 11:36:23 ah that's a good point 11:38:13 either the original thing with a 3x3 hole or the one with an added dot must not be an oscillator, since they have the same image, and therefore one of them has an n'th preimage that's a goe 11:38:22 yeah 11:38:33 does the size 5x5 give you that in fact it's the original one? 11:38:54 5x5 is required to ensure that the added cell doesn't interact with anything 11:39:14 oh indeed it is 11:39:27 forgot what the rule was 11:39:35 so did I at first :) 11:40:07 and now I want to tackle the harder problem :D 11:40:39 Can we count how many patterns there are with 5x5 or larger holes? 11:40:54 Because that's the lower bound on GoEs 11:40:58 anyway it's usually the one with the dot that has a goe nth preimage, since you can, as i said, just have small oscillators and a 5x5 hole in between them 11:42:31 anyway if that 11x11 thingie can be made to collapse into the all-0 configuration, then all finite configs have a goe preimage (although it might not fit in your favorite torus) 11:42:39 coppro: what's the harder problem? 11:43:08 oklopol: Do all finite configs have a geo preimage? 11:43:30 a finite goe preimage, on an infinite grid? or torus again 11:43:35 torus 11:43:50 wrapping around? 11:43:52 If not, which sizes of torii admit states with no goe preimage 11:43:57 that's what torii do 11:44:09 usually yeah 11:44:26 *tori 11:45:12 goe is too complicated for this to be at all interesting really 11:45:34 too complicated to be interesting? what? 11:45:48 well mathematically interesting 11:45:51 i dunno where you get your kicks 11:47:31 now I'm pretty sure that anything smaller than 3x3 admits nothing other than the trivial oscillator 11:47:36 " oklopol: Do all finite configs have a geo preimage?" <<< well for this particular question, actually the answer is trivially "yes", but questions that actually depend on gol, kind of hard 11:47:52 hmm 11:47:59 wait what the fuck am i saying :D 11:48:00 oklopol, wait, it is? 11:48:08 oklopol, describe please 11:48:23 it's no trivially true for all non-surjectives 11:48:25 *not 11:49:16 yeah :D 11:49:47 how far have you checked? 11:50:08 From experimetnation, it appears that 3x3 admits only period-1 oscillators and patterns that produce the trivial oscillator in one or two generations, and if by two, the intermediate state is always the all-on state 11:50:18 not very far, I'm playing around manually 11:50:29 I may write a script tomorrow to do serious checking 11:50:59 We're actually testing stuff? That's boring :/ 11:51:13 well there's no way to do this any other way 11:51:24 for small patterns 11:51:52 yeah 11:51:59 experiment for the small case 11:52:00 you can probably prove it for all large enough patterns by looking at the images of the goe on the wiki 11:52:01 *cases 11:52:03 If we can prove that the lower bound on GoEs = the number of patterns without 5x5 holes 11:52:11 pedia 11:52:14 oklopol: That's hardly sufficient 11:52:29 How many patterns are there without 5x5 holes? 11:52:35 coppro: whoosh 11:52:57 I suspect that the answer is constant beyond a certain size 11:53:06 coppro: it might be. if it's not, what else can you do really. 11:53:15 oklopol: See if there's a pattern 11:53:32 what if it's only prime by prime tori? 11:53:32 to what? 11:53:36 Is there an upper bound on number of GoEs? 11:53:44 Sgeo_: yes. The number of states 11:53:47 :P 11:53:58 if you take a random problem, you will find a random answer. it's not the prime by prime tori. 11:54:25 Actually, number of patterns without 5x5 holes is the upper bound 11:54:48 interesting 11:54:55 The question is equivalent to: Is the number of GoEs = to the upper bound? If not, then there's a pattern without a 5x5 hole that is not a GoE 11:55:02 oklopol: prove it :P 11:55:09 Sgeo_: Dude 11:55:15 Sgeo_: that makes no sense 11:55:36 an upper bound is an estimation; you need some criterion for the upper bound 11:55:42 making conjectures is a great way to progress 11:55:52 but that particular one is obviously false 11:55:53 Ok, replace "upper bound" with "number of patterns without 5x5 holes" 11:55:54 Sorry 11:56:00 ah ok 11:56:05 in which case I go with oklopol 11:56:15 so fun fact: any 3-cell pattern on a 3x3 torus immediately fills the entire torus and dies 11:56:38 every cell sees every cell so obviously 11:57:06 only depends on the amount of live cells 11:57:17 Come on, someone tell me a way to count patterns without 5x5 holes 11:57:21 whether it's instantly everyone alive, or instantly everyone dead 11:57:21 There has to be an easy way 11:57:37 oklopol: right, thanks for beating sense into me :D 11:58:09 Sgeo_: I could do it in a few minutes, but it doesn't seem worth it 11:58:37 it's just a counting problem 11:59:10 Yes, and if the lower bound on GoEs ever equals it for some sized tori, we'll have some answers 11:59:23 that seems quite unlikely 11:59:41 except on small tori 12:00:12 once you're above 16x11 then you definitely won't hit that bound 12:00:17 since there is a known 11x11 orhpan 12:01:19 ooh 3x3 torus. nifty 12:01:21 also i still don't get how you know 5x5 hole => goe nth preimage, all we've established is either that pattern or the one with a dot in the middle has one, how did you do the final step? 12:01:38 i mean there are patterns with a 5x5 hole that are in an oscillator 12:01:45 as i said 12:02:16 coppro: will every 3x3 torus eventually die? 12:02:25 itidus20: in at most 2 steps 12:02:27 itidus20: or be stable 12:02:28 homework: why? 12:02:44 oklopol, if it's in an oscillator, then that cell will die, and the oscillator will eventually reach the same position except without the dot 12:02:47 oh hmm 12:02:52 Or am I misunderstanding your question? 12:02:58 * oklopol tries to remember the rule 12:03:24 oklopol: Suppose that one of the two has no goe nth premiage. Then it must be in an oscillator as the state space is finite. But that means its successor is a predecessor, and its successor has the other twin as an nth preimage and hence has a geo nth preimage; contradiction 12:03:26 right the rule is not symmetric w.r.t. live and dead cells, in that case 3x3 is not quite as trivial as i said 12:03:38 still pretty trivial though 12:03:47 rule is 2 or 3 to stay alive, 3 to birth 12:04:22 it's not a contradiction that the config with a 5x5 hole is in an oscillator 12:04:40 because that can happent 12:04:42 *happen 12:04:44 oklopol: sure 12:04:47 oklopol, did I misunderstand your question? 12:04:56 Suppose it's a p2 oscillator 12:05:01 But it has to have a goe nth preimage because its successor does, and its successor is itself an nth preimage 12:05:06 since it's in n oscillator 12:05:08 *in an 12:05:09 Pattern in phase 1 + dot has goe primage. 12:05:11 preimage. 12:05:23 Then, phase 2 + no dot has goe preimage 12:05:31 Then phase 1 + no dot has goe preimage 12:07:09 " But it has to have a goe nth preimage because its successor does" <<< are you claiming this is a general truth for CA? 12:07:34 oh hmm 12:07:37 oklopol: No 12:07:53 oklopol, what CA is it not true for? Nondeterministic? 12:08:03 Are there nondeterministic CAs? 12:08:30 well i don't see why that's true 12:09:04 Oh wait, I misread the quote as "predecessor" 12:09:09 oklopol: Okay, let me try this again from the top. Given a state S with a 5x5 hole, put a dot in that hole and call it H. 12:09:19 why don't you just explain the part i don't get 12:09:25 oklopol, I'm trying to 12:09:31 Sgeo_> Pattern in phase 1 + dot has goe primage. 12:09:31 preimage. 12:09:31 Then, phase 2 + no dot has goe preimage 12:09:31 Then phase 1 + no dot has goe preimage 12:09:32 because I'm not quite sure what part that is 12:09:45 Maybe I'm not getting what part you're not getting? 12:09:48 Both S and H cannot be in the same oscillator as the generation function is bijective within an oscillator and G(S) = G(H) 12:10:06 and they cannot be in different oscillators for the same reason 12:10:13 So one is not in an oscillator 12:10:24 oh lol 12:10:28 yeah okay igi :D 12:10:38 hey that's a neat trick 12:11:38 this code is a shit :( 12:12:14 elliott, do Game of Life stuff with us 12:12:14 Now I want to write a script which will enumerate every state of a finite torus GoL modulo symmetry and create a graph of the results 12:12:23 but I think sleep must come first 12:12:35 Sgeo_: no 12:13:26 so to generalize 12:13:34 if you have two preimages, then you have a goe nth preimage 12:13:38 yeah 12:13:53 in finite state space, anyway 12:13:56 yes 12:14:28 why do I suddenly feel like this is going to become category theory 12:16:05 a good analogy can apply to anything 12:16:40 also if one of your images has a goe nth preimage, then you have one as well 12:16:49 in finite space 12:17:29 I'm going to need to go out soon 12:17:40 Chicken sandwich for breakfast :D 12:17:52 -!- Taneb has quit (Ping timeout: 252 seconds). 12:18:07 oklopol: right 12:18:20 Sgeo_: omg how amazing D::DDDD 12:18:32 that's an interesting lemma for functions acting on a finite space, certainly, nothing to do with ca tho 12:18:50 it's all a function of IRC 12:18:52 3x3 admits oscillators with no goe nth preimage, as does 1xn with n>=3 12:19:35 actually wait 12:19:39 With no goe nth preimage, or no goe nth preimage provable with a 5x5 hole? 12:19:39 cool 12:19:43 has to be even n 12:20:06 Or, if you can prove there's only one preimage, that would do it 12:20:18 sheesh, stop obsessing about 5x5 holes pervert 12:20:39 hmm... actually, odd may admit a seriously wacky oscillator 12:21:11 so I am looking at a video with the heading "morphism category theory" and I'm getting basic knot theory explained 12:21:53 ok so 1xn with odd n admits oscillators 12:21:55 neat ones too 12:22:12 would be good for the patterns of random lights that repeat in weird cycles in scifi schows 12:22:33 oh there we go 12:22:42 they each represent a symmetric 2-morphism 12:22:45 NO WONDER 12:22:58 well the only way not to admit an oscillator is to have everything go to zero 12:23:09 err sorry 12:23:23 non-goe predecessor oscillator 12:23:45 need a good adjective 12:23:47 stranded? 12:23:52 i don't think we have a name for this concept in math 12:23:58 so come up with your own 12:24:07 k I'm calling them stranded oscillators 12:24:12 oklopol, I consider everything at 0 to be a period 1 oscillator 12:24:33 there's a name for ones which have arbitrarily long preimage chains, such points are said to belong to the limit set of the CA 12:24:33 also the 1x11 torus is weird 12:24:33 Is that an abuse of terminology? 12:25:01 oklopol: ah ok, that works then 12:25:12 Sgeo_: that's the terminology i'd use, i just assumed coppro meant there has to be a 1 since he said it admits one 12:25:14 actually wait, no it doesn't 12:25:31 I agree with that terminology 12:25:40 if every point goes to the all-0 config for some state 0, the CA is said to be nilpotent 12:25:49 (then there exists a k such that all points to go all-0 in k steps) 12:25:49 symmetric 2-morphism? 12:26:51 1x5 is nilpotent 12:26:55 Sgeo_: talking about the video 12:27:39 1x7 has a weird spaceship oscillator 12:27:44 :o 12:28:30 actually, how about a xenophobic oscillator 12:28:38 ...wha 12:28:39 ? 12:28:42 I like that name 12:28:51 does it not like other kinds of oscillators 12:28:56 nope 12:29:04 it doesn't like any state except its own members 12:29:10 exclusive might work too 12:29:27 -!- Taneb has joined. 12:29:29 What do you mean by that? 12:29:46 -!- Taneb has left. 12:29:57 -!- Taneb has joined. 12:30:15 Sgeo_: Just trying to name these oscillators with no predecessors 12:30:23 Ah 12:30:45 I suppose a closed oscillator would probably be best 12:31:08 well you have a set that cannot be entered, kind of like a fortress. so call each config in there a *drumroll* brick 12:31:20 How did you prove they exist for the 1xn and ... the other thing you proved? 12:31:59 I mean, experimentally, but how do you demonstrate they have no other preimages? 12:32:10 closed is good, in fact it's a closed set in the topology given by one-way orbits 12:32:23 (clopen in fact) 12:32:39 1xn for even n I'm not 100% sure about but could probably sketch up a proof 12:32:55 topology given by one-way orbits being that a set is open if it's closed under the map 12:33:02 any zebra stripe pattern is stable and I'm pretty sure also has no other predecessors 12:33:32 if you can go through all the configs, it's easy enough to enumerate stranded ones 12:33:37 yeah 12:33:48 not sure if I prefer stranded or closed 12:33:50 rendaer a q 12:34:07 closed is more conventional but stranded sounds more like the concept 12:34:15 Sgeo_: 3x3 is straightforward 12:34:47 Sgeo_: in 3x3 all configurations with exactly 3 cells alive are stable and all others die out in at most 2 generations 12:35:28 oklopol: yeah, at some point I will write a script to enumerate these 12:35:34 Hmm, ok. 12:35:59 enumerate ones that belong to a closed oscillator? 12:36:49 yeah, at least for small tori 12:37:15 remember to use the fact you only need to look at the first preimages, and then enumerate the whole orbit if you have multiple 12:37:22 otherwise that's gonna take some time 12:37:49 I suppose 12:37:52 If you have multiple preimages for a state, then it's not a closed/stranded oscillator 12:37:54 iiuc 12:38:30 no it's not, point is you never need to look further than that because a point is in a closed oscillator iff it has a preimage not in a closed oscillator 12:38:48 Sgeo_: sketch of prrof for 1xn in my head 12:38:56 (for even n, anyway) 12:40:18 1x9 appears to really really like a single oscillator 12:40:39 have yet to find a pattern other than all-off and all-on that doesn't result in i 12:40:43 *in it 12:41:22 (another problem: Is there a general solution to the number of distinct oscillators on an mxn torus? 12:41:57 anyway, bed 12:42:00 nope 12:42:20 oklopol: nope there isn't one or nope there can't be one? 12:42:37 nope there can't be one, assuming the way you can implement a tm in gol is in any way sensible 12:42:50 ah 12:42:55 -!- cheater_ has quit (Ping timeout: 255 seconds). 12:42:59 night 12:43:10 Huh? 12:43:21 First: You can't impement a TM on a torus 12:43:22 -!- cheater_ has joined. 12:43:28 you can't? 12:43:44 it grows forever even if its tape doesn't? 12:43:49 oklopol, TM implies infinite storage 12:43:54 A torus is not infinite 12:44:19 but does the gol implementation need expanding space for constant storage 12:44:21 Erm, that's what I usually mean when I say torus, anyway 12:45:00 if it doesn't, there can't be a general solution to the number of distinct oscillators 12:45:04 at least not in the form of a tm 12:45:04 oklopol, what? 12:45:21 what? 12:45:25 You need expanding space for expanding storage, I'd assume 12:45:37 obviously, that's not what i asked 12:45:41 And you can't forbid expanding storage without forbidding TM 12:46:14 ... 12:46:16 Surely, a counterexample, then: Why can't you simply enumerate every mxn pattern, and count the oscillators? 12:46:42 err sure you can do that :D 12:46:45 *surely 12:46:49 what i meant originally 12:46:57 was that the fact you can implement a tm 12:47:04 means there can't be any sort of nice formula 12:47:11 but not in any rigorous sense 12:47:16 What is a "nice formula"? 12:47:21 "Nice formula" has no meaning 12:47:31 that's why i said "not in any rigorous sense" 12:47:42 i then forgot that this was all i meant and claimed you can't even calculate them 12:48:18 which is not true since given m and n, you can certainly just count the oscillators 12:48:48 but for instance this is certainly p-complete 12:48:54 because you can implement tm's 12:48:58 So, hm. I'm still not quite sure I get what finite state automa have to do with this 12:49:03 And I'm refusing to call it TM 12:51:52 well the point is you can draw an initial part of the infinite tape with a tm using logspace, this is all you usually need for p-hardness results 12:54:48 and what i mean is that whether a single config is in an oscillator should be p-hard, since you can reverse your computation and start over with yes-instances, and go to all-0 or something for no-instances 12:55:34 Sorry, I'm a bit confused 12:56:06 well to, hopefully, unconfuse you, you are certainly right in that you can just count oscillators 12:56:07 Um, is doing several GoL generations and comparing generally considered p-hard? I'm a bit lost with complexity classes. 12:56:23 Oh, wait 12:56:32 p-hard 12:56:49 what i'm saying is there can't, intuitively, be a nice formula for this, because there never is for this kind of things. 12:56:55 elliott: (sic) 12:58:06 basically just means you can draw a finite part of a turing machine config in the state 12:58:12 or a circuit 12:58:14 Is running GoL for 2^mn generations not intuitively a nice formula? 12:58:33 it's a nice algo, i wouldn't call it a formula 12:58:39 Ok 12:59:21 oklopol: is the algo "for i in ... setp gol" 12:59:26 step 12:59:44 i wouldn't call it a formula 12:59:53 i would call m^n + n^m a formula 13:00:23 Hmm. 13:00:28 something that's actually some kind of solution to the problem 13:00:42 that you surely cannot have, because you can implement a tm. 13:00:51 I guess I know what nice means, kind of. We would like a nice means of determining the existence of stranded oscillators, without brute-forcing 13:01:20 oklopol, but there must be a formal definition of what it is that you cannot have? 13:01:38 i just studied the 2x2 toroidal by hand.. because formulas for these things are simply over my head. 13:01:39 Sgeo_: i'm not aware of one, i suppose that would be complexity classes 13:01:43 and it was fun to see 13:01:50 itidus20, get Golly 13:01:59 but all i'm saying is tm ==> all hope is gone, mathematically 13:02:05 Although I don't even have Golly open right now 13:02:11 -!- MigoMipo has quit (Ping timeout: 240 seconds). 13:02:32 oklopol, tbh, that sounds more like a rule of thumb. And again, you can't have TM on a finite torus 13:02:50 yes, a rule of thumb that's always true 13:02:57 kind of like the church-turing thesis 13:03:04 12 out of 16 died... and the other 4 turned into fascinating "up" "down" "left" "right" shapes 13:03:23 you can have a tm on a finite torus, it will just explode if it tries to use too much space 13:03:51 my thumb still applies 13:03:52 by turned into i should say "remained as" 13:06:03 Hmm, there's no "nice" way to see the result of a TM without running it... but all that really equates into is there's no way to see the result of a TM without the thing computing the result, whether by simulating the TM or not, definitely not halting 13:06:34 Not really sure how to apply that here. 13:06:34 -!- BeholdMyGlory has joined. 13:07:46 you apply it by noting you cannot know whether a tm drawn on a config is in an oscillator without running it, which means the only possible formula is brute-force. 13:08:11 this is not a formal proof, but only an ass would not accept it 13:09:16 I'm just going to mentally s/tm/fsm/ what you said and say ok 13:09:20 point is, you cannot take a random problem and expect it to be mathematically interesting, the solution will be just as random 13:09:54 well if you want to be a fucking outstanding retard, go ahead 13:10:39 * oklopol prepares more insults 13:11:00 anyway if you substitute fsm where i say tm, that really means nothing 13:11:10 I can solve the halting problem for your "TM" on the torus 13:11:13 the point is it's a certain kind of fsm, one that has a tape. 13:11:21 I always try to direct everything i think about onto gaming. I figure that to give a game depth at its core then some mathematics is required. 13:11:50 Then the trick becomes how to turn a mathematical problem into a fun game. 13:12:02 hasn't been done yet 13:12:13 surprise us all and be the first 13:12:28 oklopol, hmm? Are you talking to me or itidus20? 13:13:01 I can solve the halting problem for your "TM" on the torus 13:13:03 tm =/= jutm 13:13:06 utm 13:13:16 Oh 13:13:21 hmm 13:13:30 elliott: is the joke that you're missing the point as badly as Sgeo_? 13:13:56 oklopol: the joke is all the words 13:14:03 i only read that line anyway 13:14:10 i just kind of assumed anything Sgeo_ said would be something like that 13:14:15 oklopol, here's the solution to the halting problem: Run it. If it crashes, or otherwise doesn't return to the initial state after 2^mn, it halts, otherwise, it doesn't. 13:14:34 well he was mostly missing the point and expressing it in random ways 13:15:54 well work time 13:16:30 since last night i realized all lattices with continuous shift-commuting operations over S^Z can be recoded into pointwise lattices 13:17:12 that is, you have a conjugate subshift and the lattice operations only look at the ith symbols of their arguments to determine the ith symbol of the image 13:18:05 elliott, oklopol seems to think that turing-machines can exist in finite space. And just to be clear on the idea of finite tape possibly being TM vs UTM, he mindboggled when I said that I could solve the halting problem for his TM (although maybe he thought I meant with stuff _on_ the machine, which is impossible) 13:18:47 that wacky oklopol with his strange uncommon mathematical beliefs 13:19:14 elliott, if I'm missing something, please tell me 13:19:27 mathematics is my religion 13:19:42 i don't think i ever actually proved anything 13:20:00 for actual games one thing that tends to be necessary is interfacing the math models with non-math things. 13:20:56 whether the interfaced things is actually non mathematic is questionable though 13:21:21 Sgeo_: sorry i was talking to itidus20, i have you on ignore 13:21:31 i did not mindboggle at you claiming that trivial thing 13:21:58 sorry to everyone that i am always offtopic 13:22:55 well so, its a bit like what the chinese did with yin and yang, and trigrams. applying meanings. 13:23:00 -!- MigoMipo has joined. 13:23:59 and so 00=north, 01=east, 10 = west, 11=south 13:24:24 congratulations, you discovered binary 13:25:03 but the pure mathematicians never actually apply it to anything.. the "interfacing" 13:25:10 i guess its horses for courses 13:25:31 they make it all possible etc 13:26:09 oklopol, is our disagreement here just a matter of terminology? 13:26:17 i am sure that if i did that better i could say north + east = northeast 13:28:00 itidus20, easy way for the future is to assign 1 bit to 1 ... bit 13:28:14 So north/south gets 1s place, east/west gets 2s place, or something like that 13:28:26 Although that doesn't... quite... hm 13:28:33 Kind of assumes defaults 13:29:42 ah.. it was right how i had it.. no doubt 13:30:40 ahh fuck it. time for me to take a break. 13:30:40 -!- MSleep has changed nick to MDude. 13:30:44 Oh, it trivially doesn't work 13:30:49 My scheme, I mean 13:31:13 8 possible states, so you need 3 bits 13:31:28 00 + 01 = 01 .. that creates a problem 13:31:38 just do a bitmask duh 13:31:39 powers of two 13:32:53 ais523: yes, fizzie thinks this is the nineties, and does things like use bitshifts instead of division <-- I think the most bitshifts are where there are negative numbers, where bitshift != round-towards-zero division. (Though it admittely uglily assumes sign-extending arithmetic shift for signed quantities.) 13:32:54 I'm trying to think of a non-ugly way to do it in 3 bits now. 13:33:49 A NORTH flag, a WEST flag, and one more flag 13:33:52 applying meanings to numbers is approximately as fun as processing the numbers 13:34:39 Oh, duh. If I only had two flags, so NORTH or not NORTH (south), and WEST or not WEST (east), then you only get diagonal directions 13:35:41 perhaps it doesn't work out neatly if using addition 13:35:47 addition is kind of a luxury 13:36:24 Trivial ordering, probably ugly though, let's see what happens: 000 = N, 001 = NE, 010 = E, 011 = SE, 100 = S, 101 = SW, 110 = W, 111 = NW 13:36:38 I'm honestly thinking ternary at this point 13:36:57 You can trivially do it in two trits 13:37:30 Um, with space left over. Oh, for neutral 13:37:45 just do a bitmask duh 13:37:45 powers of two 13:37:57 oh wait, three bits 13:38:03 meh 13:38:11 seems space left over is the price for being able to add them 13:38:16 elliott, you'd need NORTH, NOT NORTH, and SOUTH 13:38:26 itidus20, look 13:38:35 0+ is North, 0- is South 13:38:41 00 is Neutral 13:38:47 ah.. so its a 2s compliment thing? 13:39:01 +0 is West, -0 is East 13:39:16 itidus20, no, it's base 3 instead of base 2, but instead of 0, 1, and 2, we go -1, 0, and 1 13:39:19 ah ok this is the trits 13:39:41 So, +0 + 0+ is North + West = NorthWest 13:39:45 = ++ 13:39:53 probably the netral is a necessary placeholder for the addition 13:39:58 ^neutral 13:40:20 itidus20, it's because naturally there isn't on or off, there's forward, backwards, and neither 13:40:22 -!- Taneb has quit (Ping timeout: 252 seconds). 13:40:35 hmm ok 13:40:55 If you're going north, than what would east/west be? Neither. 13:41:06 If you're going west, north/south must be Neither 13:41:15 so it's to account for the uh.. center point 13:41:42 ohhh 13:41:45 right ok i see now 13:41:54 thats deep 13:42:44 And these balanced ternary are just numbers 13:43:18 -- = -1*3^1 + -1*3^0 = -3 + -1 = -4 13:43:56 -0 = -3, -+ = -2, 0- = -1, 00 = 0, 0+ = 1, +- = 2, +0 = 3, ++ = 4 13:44:50 -!- oklopollen has joined. 13:45:15 And with bitmasks, usually you use bitwise or instead of addition, although addition works as long as you're careful that + and - only combine with 0s. Not sure what you use with trits 13:45:33 quickly, make a guess! 13:45:39 Erm, sorry, that "as long as" is for trits, not bits 13:46:05 Sgeo_: to your last question: probably. 13:46:36 -!- itidus20 has changed nick to itidus20|afk. 13:46:53 Ok, so no more arguments about terminology? 13:47:12 no never ever. 13:47:36 but no makeup sex since i have to work 13:48:28 actually i just have to copypaste some stuff and generalize it :-DSASD 13:48:34 -!- Lymee has joined. 14:09:48 -!- pikhq_ has quit (Ping timeout: 240 seconds). 14:09:51 -!- pikhq has joined. 14:12:14 -!- CakeProphet has joined. 14:40:17 -!- oklopollen has quit (Quit: Page closed). 14:41:07 -!- Taneb has joined. 14:41:12 Hello 14:48:22 Believe it or not, etc. etc. etc. 14:48:25 Bye 15:01:44 Wow, that's pretty etc, etc, etc. 15:02:28 I'd rather not think about the details so, etc, etc. 15:05:05 sup CakeProphet 15:07:20 nothing much... 15:07:28 chilling after work. tired as shit. 15:07:41 I don't know if you knew this or not, but shit gets pretty tired. 15:07:46 not a lot of energy in those motherfuckers. 15:24:36 i'm also pretty tired, worked for almost an hour 15:25:36 -!- Taneb has quit (Ping timeout: 252 seconds). 16:09:50 hai fizzie 16:09:57 fizzie, I'm Elizacat, and I've taken interest in mcmap :3 16:10:12 fizzie, I've never gotten around to poking you because I'm shy as hell. But hello! 16:11:06 Well, um, hello, I guess. (I never really know the correct responses.) 16:11:06 fizzie, I'm interested in helping! I'm told you guys have a private channel for this, if you want to give me an inv I'd be much obliged, but you don't have to if you don't want 16:11:09 hahahahaha 16:11:10 it's ok :P 16:11:22 I'm not an extrovert even online, but one of us had to initiate it 16:11:30 and well, you don't know who I am 16:11:31 so :p 16:11:44 the onus would be on me for that wouldn't it 16:12:02 I am extrovert online 16:12:06 because I don't give a shit. 16:12:18 fizzie, I'm normally pretty nice if not rather unserious :p 16:12:26 fizzie, just be aware :P 16:12:35 it's always serious business around here. 16:12:39 * Elizacat is a good friend of Vorpal's 16:12:44 internet is serious business CakeProphet 16:12:46 you wanna go 16:13:01 dude I'm so good I don't even have to I've already won. 16:13:07 Well, there's a "private" channel for Minecraft-related stuff so that it doesn't clobber things here, but I don't think it's really all that private, it's just a "-minecraft" suffix and I'd be surprised if it hasn't been mentioned publicly here. 16:13:10 fuck punctuation. 16:13:11 CakeProphet, come at me bro 16:13:11 :p 16:13:40 yeah I've known of it, but I don't like minecraft. 16:13:58 * Elizacat encases CakeProphet in a bedrock house and pours lava 16:14:09 "this is why Id on't like it..." 16:14:15 my RL friend made a castle that rebuilds itself. 16:14:20 I did that 16:14:21 :p 16:14:30 Vorpal and I built a self-regenerating house <3 16:14:33 well 16:14:36 he did the logic 16:14:38 I just had the idea 16:15:19 please tell me Vorpal isn't your boyfriend 16:15:27 i might have to kill myself 16:15:41 but they build self-regenerating houses in minecraft together it would be so adorable 16:16:33 i once asked a girl to build stuff with me in mc but she said we were moving too fast 16:18:10 I tried to teach my ex Python. 16:18:17 I think there is probably a reason she is my ex. 16:18:54 Because she has taste in languages? 16:19:06 not quite. 16:19:09 i tried to teach my ex to program a few times but she just got really mad because she didn't get it 16:19:19 or didn't get it quickly enough 16:19:23 right. 16:19:45 Elizacat: could you please tell me that 16:19:53 i'm scared by your silence 16:20:12 please tell me Vorpal isn't your boyfriend <-- lol, no I'm not. 16:20:19 thank god 16:20:33 oklopol, just happen to play on the same minecraft server. 16:30:04 -!- oklopol has quit (Read error: Connection reset by peer). 16:34:24 -!- monqy has joined. 16:42:11 -!- elliott_ has joined. 16:45:41 -!- elliott has quit (Ping timeout: 258 seconds). 16:56:06 haha 16:56:11 oh god 16:56:13 me dating Vorpal 16:56:17 I'm sorry 16:56:18 that is just 16:56:19 no :P 16:56:24 no offense Vorpal 16:56:26 but I wouldn't date you :P 16:56:28 and you wouldn't date me 16:56:30 I'm too crazy 17:05:44 -!- Taneb has joined. 17:10:27 Blood is thicker than water, but corn syrup is thicker than blood. (An American national pride slogan by me :P ) 17:11:25 Lava is thicker than corn syrup 17:11:57 I don't think that lava is anyone's (de facto) national beverage. But blood probably is? 17:13:00 Might be Icelands? 17:13:39 But I was talking about threats of violence 17:15:50 You're ruining my slogan buzz :P 17:33:12 Right. Lasagne went in at 6:32 17:33:24 uh oh sgeo 17:33:45 Uh ohhhhhhhhh, sgettio 17:34:07 Y'know where I'm off to? Lancaster! 17:34:19 Lancaster? 17:34:25 But that's in Lancashire! 17:34:32 Dun dun DUNNN 17:34:38 I'm going to Durham tomorrow 17:34:55 And, after tomorrow morning, I won't be online 'till Tuesday 17:40:11 that was crazy... X broke... By refusing the believe in the mouse cursor. For example, xkill said it couldn't grab mouse. And mouse did nothing... 17:40:21 well it moved around just fine 17:40:30 but click didn't work 17:43:43 -!- GuestIceKovu has changed nick to Slereah. 18:07:05 iceland;s national beverage is alcoholic pine tree 18:21:33 I wonder if you can ferment or distill or whatever coffee beans 18:22:31 If it won' 18:22:38 t kill yeast, you can ferment it. 18:22:46 And if it's liquid, you can distill it. 18:25:12 -!- itidus21 has joined. 18:25:12 -!- itidus20|afk has quit (Read error: Connection reset by peer). 18:30:11 -!- zzo38 has joined. 18:34:06 I'll bet mercury would kill yeast ... 18:38:42 Do you know what the test service numbers are for my telephone service? Including, telling you your own telephone number, disconnecting your line for a few minutes, etc 18:49:51 -!- pikhq_ has joined. 18:50:29 -!- pikhq has quit (Ping timeout: 276 seconds). 18:51:07 http://therighttool.hammerprinciple.com/statements/i-often-get-angry-when-writing-code-in-this-langua 18:51:10 *giggles* 18:53:59 everyone likes haskell on this site 18:55:03 /win 1 18:58:42 In spite of the wording, that's clearly a popularity contest. 18:58:56 Few people get angry while programming in Haskell because few people program in Haskell. 18:59:25 O, yes, probably that would be why. 19:02:41 Strange that that site doesn't let you rank things equal. 19:07:55 -!- ais523 has joined. 19:10:35 hmm, weird observation: within the first 3 or 4 minutes of my computer loading, the touchpad won't move the mouse cursor above the bottom half of the screen, it keeps bouncing down 19:10:39 after a while, it works fine 19:25:24 -!- Taneb has quit (Ping timeout: 252 seconds). 19:32:44 I have now regopherized the Space Weather Prediction Center. 19:33:37 (Some of the files don't work but that is not my fault.) 19:34:25 ais523: whoa 19:41:27 Did you know that? I don't know why some files don't work, but the problem is on their end, not on my end. 19:42:12 -!- Taneb has joined. 19:42:19 (The same files fail to work when using other protocols too, including HTTP and FTP.) 19:45:47 -!- oerjan has joined. 19:55:37 oklopol: 123 is like oerjan but better 19:55:38 wat 19:58:02 is this still referring to that pixelcomic 20:00:56 yes 20:01:37 those kind of puns are too square for me 20:22:32 -!- Taneb_ has joined. 20:24:05 -!- Taneb has quit (Ping timeout: 252 seconds). 20:25:53 -!- Taneb_ has changed nick to Taneb. 20:41:18 -!- Taneb has quit (Ping timeout: 252 seconds). 20:42:41 In the D&D game my character current has two items: a 25-foot rope and a navy guest uniform. 20:42:58 Now make up a computer game text adventure game 20:44:14 -!- Lymee has quit (Ping timeout: 260 seconds). 20:49:50 -!- Lymee has joined. 20:49:51 -!- Lymee has quit (Changing host). 20:49:51 -!- Lymee has joined. 20:50:45 -!- ais523 has quit (Remote host closed the connection). 20:52:30 I didn't end up posting my password in here, did I? The graphics driver died, and I flailed around blindly trying to "sudo /etc/init.d/gdm restart" 20:53:33 You sent nothing in the past few minutes look in the log files to make sure 20:54:19 -!- pikhq has joined. 20:54:19 -!- pikhq_ has quit (Ping timeout: 240 seconds). 21:01:53 Do you know anything else about my optimization problem now? 21:12:46 zzo38: what was it again? 21:13:14 -!- TeruFSX has joined. 21:16:01 quintopia: Start searching the recent IRC log for "Say that there is a command - which does the next letter" 21:16:47 (The starting timestamp is 1311396023 although there are at least three different formats and from different computers, so just search the text instead) 21:19:58 okay 21:20:05 doesnt quite make sense 21:20:10 i'll read it better soon 21:20:22 tunes' time zone is so broken you cannot even search on the minutes. 21:22:56 quintopia: use codu 21:23:04 advicewithoutcontext.com 22:12:54 elliott_: there's a thread on haskell-cafe about that reallyUnsafePointerEq# you used http://www.haskell.org/pipermail/haskell-cafe/2011-July/094103.html 22:14:17 heh 22:18:43 seems my use was safe with forseeable ghc 22:18:57 -!- Sgeo_ has quit (Read error: Connection reset by peer). 22:19:38 -!- SgeoN1 has joined. 22:22:43 I am alarmed by the fact that I now feel the need to run memtest 22:33:30 quintopia: What stuff doesn't quite make sense? 22:42:44 Try eat your chopsticks right now! 23:03:45 -!- itidus21 has changed nick to itidus20. 23:04:54 but i have no chopsticks! 23:05:05 or do i... 23:08:50 Look in your cupboard or drawer or closet or cabinet or wherever you keep it. 23:10:09 wouldn't you know, i do have a couple! 23:10:22 sadly, they're a bit too nicely decorated to eat. 23:10:32 follow the white rabbit 23:11:00 nah i don't want to be chasing the clock 23:11:04 O_O FFFFFUUUUUUUUUUU it's a marsh hare! a marsh hare! 23:11:14 wat 23:11:47 we planted the chopsticks for you to eat 23:12:04 ...i don't believe you. 23:12:21 oerjan: we are HERE 23:12:24 good. i apologize 23:12:25 -!- elliott_ has quit (Quit: disconnection error). 23:12:51 elliott_: no you are not. 23:13:07 -!- elliott_ has joined. 23:13:11 haha, great timing itidus20. 23:13:24 now oerjan will not BELIEVE THAT THING that WAS DEFINITELY not TRUE at all. 23:13:46 oerjan: LETJOISE HAVE A PERARTY 23:14:08 QANTUM BEER NEEDSITY? 23:15:23 what if oerjan was that guy in norway who did that bad stuff, would we even know. :/ 23:15:31 i am just saying that there are a lot of people in norway 23:15:34 and you cannot trust them all 23:15:44 yes. i doubt they would let him chat from prison. 23:16:16 are you saying you escaped :| 23:16:26 oh this is probably in bad taste :DDDDD 23:16:30 oerjan: im sorry ive deeply offended u 23:16:39 i have not escaped. but not from the same place, either. 23:17:29 oh 23:17:34 are you trying to escape 23:17:35 of course there are similarities. i am blond, somewhat tall, and not a muslim. 23:17:35 THIS 23:17:36 MORTAL 23:17:37 COIL 23:18:21 hi oerjan 23:18:36 i'll ignore that question. 23:18:42 -!- Sgeo has joined. 23:19:11 oh. sory. lo oerjan. 23:19:34 oh riead as suggestion 23:19:39 oerjan: hows a ijf 23:19:49 very rieadbale 23:20:09 and thatw as when oerjan turned into a cat 23:20:12 ~Thend/ 23:20:29 miawat 23:21:05 i mcofnused 23:21:20 monqy: itns eryvoene 23:21:55 My memory passed 23:22:05 rest in peace sgeos memory 23:23:34 `addquote My memory passed rest in peace sgeos memory 23:23:35 542) My memory passed rest in peace sgeos memory 23:23:43 did you ever finish writing at.html 23:23:45 reading 23:23:51 elliott_, /me wants to read it 23:23:55 what ae re verbs actulay 23:24:01 elliott_: me yes i did 23:24:03 Sgeo: iduno if anyone else can stan,d,,,,,the @,, 23:24:16 elliott_, you're acting drunk 23:24:19 it has, maximal @ level contents. 23:24:20 I doubt you're drunk, but 23:24:34 Sgeo: so's your face OHOHHHHHHHHHHH BURNSJAP 23:24:47 no if i was intoxicated i would be sure to be much, much more amusing than this 23:24:56 so basically if i'm ever really funny........... 23:25:42 anyway drunk people don't act like anything, they just act like irritable, annoying, stupid versions of themselves 23:25:49 which is a shame since it's rly boring 23:25:55 elliott_: so are you old enough to drink yet? 23:26:23 i think you can drink indoors from like six years old here??????? 23:26:26 but no 23:26:41 three and a bit months before that, but it seems pretty boring so i probably won't bother 23:28:14 alcohol education taught me never to drink to get drunk 23:28:23 an important life lesson 23:28:48 monqy: did you follow it 23:28:58 im not of legal drinking age......... 23:29:07 monqy: so yes then 23:29:26 i'm not actually sure at all why people would get drunk alcohol honestly just seems like the most boring drug, i'm utterly sincere 23:29:50 -!- MigoMipo has quit (Read error: Connection reset by peer). 23:30:19 cigaretes and drunking "lame highs for lame people" 23:30:39 yeah if it doesn't make you go crazy and shoot at people, it's not worth it. take it from a norwegian. 23:30:55 monqy: so is squarelos the best,,, it is a long story involving me reading the crawl learndb and talking to ais 23:31:04 `addquote cigaretes and drunking "lame highs for lame people" yeah if it doesn't make you go crazy and shoot at people, it's not worth it. take it from a norwegian. 23:31:05 well stab people, in the case of weapon-deprived british 23:31:06 543) cigaretes and drunking "lame highs for lame people" yeah if it doesn't make you go crazy and shoot at people, it's not worth it. take it from a norwegian. 23:31:18 elliott_: public humiliation 23:31:19 oerjan: stop making me into a person i don't want to be with laughter 23:31:24 elliott_: is not cool????? 23:31:32 monqy: i am asking,,,sincere questions:((( 23:31:35 :( 23:31:36 is squarelos,,cool 23:31:41 i do not play, crawl, at all, ever, so, 23:31:44 oh 23:31:47 its sort of a joke 23:31:47 it is good to improve things even if they are bad things 23:32:07 ais says that devs dislike it experienced players like it and everyone else dislikes it 23:32:10 is this true and also is this redundant 23:32:13 see crawl has circular line of sight which is ridiculous 23:32:23 and 23:32:25 i would like to have eyes that have circular line of sight 23:32:28 that would be great 23:32:30 well spherical 23:32:37 well um do i 23:32:40 i dont know much about eyes 23:32:44 :( 23:33:20 square los was tried but then got taken down so i made an account called squarelos to lament its demise 23:33:26 line of sight tracing a mandelbrot fractal 23:33:29 rip 23:33:51 anyway pretty much the only way i ever play crawl anymore is doing stupid things as squarelos true story 23:33:55 oerjan: stop youre just making the universe seem relaly inadequate 23:34:02 monqy: thats a good way to play games 23:34:23 -!- TeruFSX has quit (Ping timeout: 250 seconds). 23:34:26 What's wrong with circlelos? 23:34:52 was ist los 23:35:34 line of sight 23:35:48 Sgeo: distance is such that an actual circle in crawl would look like a square this is how crawl space works....yet there's a bucnh of inconsistency like line of sight is based on what looks like a circle to unenlightened humans 23:35:54 Sgeo actually likes crawl though because he's a dumb 23:36:00 -!- SgeoN1 has quit (Quit: Bye). 23:36:09 "distance is such that an actual circle in crawl would look like a square this is how crawl space works" 23:36:11 this is beautiful 23:36:15 i want to see crawl space 23:36:27 wait 23:36:30 isn't that just manhattan distance 23:36:31 its just chebyshev space i think but it pretends it isnt 23:36:32 I think I understand 23:36:45 manhattan is taxicab right? crawl has diagonals. 23:36:50 or am i confusing things 23:36:51 I think manhattan is orthogonal moves only 23:36:53 taxicab yeah. and diagonals suck 23:37:29 hmm, sometimes I get urges to write things in C that relaly shouldn't be written in C 23:37:33 help;p 23:37:49 * Sgeo wants to try Chicken again for some reason 23:37:51 What kind of things do you mean? 23:38:01 the best reason to write anything in anything is that it shouldn't be done 23:38:02 And what programming language do you think those things should be written in? 23:38:04 zzo38: i cannot say, it is too abhorrent 23:38:27 elliott_, web toolkit's been done (for either C or C++, don't remember) 23:38:33 Sgeo: no shit 23:38:44 http://www.webtoolkit.eu/wt 23:38:44 olsner: No, the best reason to write anything in anything is that it *cannot* be done (or at least, it seems like it, until it works) 23:39:01 olsner: i was trying to replicate java-style generics except without type erasure because i don't think you can even do that in C. why? well because it shouldn't be done. also because it meant i could implement a custom allocator which, like, knew all about your objects and could inspect them? which would be cool. but. basically i want to implement half a jvm at c level just because it sounds hard/fun 23:39:04 that's 23:39:06 probably a bad sign 23:39:26 @pl \x = (magnitude x) (line 1, column 4): 23:39:26 unexpected "=" 23:39:26 expecting operator, pattern or "->" 23:39:31 @pl \x -> (magnitude x) (< iterations) . magnitude 23:39:40 No, it is good thing learn, try, see what happened in case you can make something like that. 23:40:25 Wt sounds fun but why C++? 23:40:32 olsner: unfortunately my first start ended up with a member of type TypeInfo##T for an array with element type T. and the problem is that the type of an array with element type T is "struct { TypeInfo##T ...; ... }" which doesn't really paste on to TypeInfo to create a valid token. 23:40:37 TeX also has its own memory allocation algorithms. In fact there are a lot of features of Pascal it doesn't use, although it does use some nonstandard features (but those ones are controlled by macros so you can change it a bit) 23:40:40 so I kind of need generic typeinfos. this is kind of gross. 23:40:49 if i remove those, i'm not sure how to typecheck it 23:40:50 hmmmmmmmmmmmmmmmmmm 23:40:51 this is hard 23:43:07 this is like c++ sudoku: hard edition :( 23:43:51 one time i treid searching for c++ soduko but i only got sudocu solvers i tw as awcufl ;_; 23:44:07 You can create static type checking (usable in ? : operators and such) in C89, even. 23:44:24 monqy: c++ sudoku is my invention and mine alone, also im the only known player, you should try though, you just need a copy of the C++0x features list, a recent g++ compiler, and an ability to forget that things aren't jokes 23:44:36 zzo38: do you have an example? 23:44:39 You just need to use unions and arrays and stuff and a few things 23:44:46 a sprunge or? 23:45:15 And then you can use the sizeof operator which is a constant and can make a macro that checks types at compile time to decide what to do! 23:46:05 ok but 23:46:06 ok 23:46:08 * Sgeo is considering getting a Nook, but the DRM is worrying 23:46:11 I suppose I can try to make up an example, just a few lines I can type directly on here to describe 23:47:24 union { struct { int something; int something_else; int who_cares; } main; char typecheck1[2]; short typecheck2[1]; }; 23:47:30 On the other hand, apparently B&N DRM is easy to remove... 23:47:40 Stuff may have changed though 23:48:06 #define figure_out(x) (sizeof(x->typecheck1[0])==1?abc(x):xyz(x)) 23:48:33 This is my thoughts ideas, at least. 23:49:09 okay, thank you 23:49:18 Yeah, all this stuff looks old 23:49:41 zzo38: But how does that fail at type-time if it'st he wrong size? 23:49:49 the size is obviously always 1 23:50:23 Well, if (x) is of a different type, then sizeof(x->typecheck1[0]) is not 1 23:50:39 Or you can omit the [0] for another number 23:50:52 ah 23:52:04 Of course this is using completely standard C89. GNU C has its own things. And what is probably acceptable in many compilers (or at least, should be) is make a member of the structure with zero elements array, and then you can measure the size of the array element it can be any number you want without messing up the rest of the program! 23:53:32 Although I don't know for sure, this idea might also be possible with LLVM to force each different structure type to be considered different by having different sizes of data of zero length array at the end. 23:55:31 http://www.barnesandnoble.com/w/jesus-deepak-chopra/1012718717?ean=9780061980404&itm=7&usri=jesus for the life of me, I can't figure out who the target market is 23:58:18 Did you ever read about the Uncarrot Tarot? I did have a idea of a four-player trick-taking game using those cards, which I called "Rulers". (It uses all the cards, including the metas.) 23:59:28 If a Hitchhiker is played to a trick, then you must Hitchhike.