00:05:22 -!- DHeadshot has quit (Read error: Connection reset by peer). 00:05:27 -!- DH____ has joined. 00:05:57 -!- nooga has quit (Ping timeout: 272 seconds). 00:06:57 -!- DHeadshot has joined. 00:07:02 -!- DH____ has quit (Read error: Connection reset by peer). 00:09:03 -!- copumpkin has quit (Ping timeout: 244 seconds). 00:10:55 -!- copumpkin has joined. 00:11:43 * kmc starts a new job tomorrow 00:11:51 where?? 00:12:58 i'll be employee #1 at a startup founded by a bunch of former ksplice people 00:13:06 ooh! 00:13:14 kentucky fried splice 00:13:19 doing what, and where geographically? still around boston area? 00:13:56 doing secret things :) (for now) 00:14:01 yeah, boston area 00:14:09 will have offices in central square most likely 00:14:12 zomg secret 00:15:11 -!- DHeadshot has quit (Read error: Connection reset by peer). 00:16:55 copumpkin: you should go by their offices and take some photos 00:17:11 will do 00:17:20 except I'm across the ocean right now 00:17:25 :O 00:17:25 but maybe next week 00:17:29 where what why 00:17:36 I'm in italy for family shit 00:17:40 offices are currently being procured 00:17:46 some walls need to be knocked down 00:17:58 let me know if you need to burn any bridges 00:17:59 I'm good at that 00:18:00 haha 00:18:09 i already offered to take down the walls myself, but noooooo 00:18:23 when i was at the hedge fund in NYC, LimeWire's offices were one floor below us 00:18:25 that sounds kinda fun actually 00:18:28 with a sledgehammer 00:18:32 and then they moved to new offices 00:18:49 and they had a moving out party which consisted of everyone getting piss drunk, spraypainting the walls, and then hitting them with sledgehammers 00:19:00 lol 00:19:02 o-o; 00:19:04 nice 00:19:05 is that.. what 00:19:06 i dont even 00:19:20 unfortunately in the heat of the moment, a distinction between interior walls and structural exterior walls was not observed with the degree of care one might hope for 00:19:29 hahhahaha 00:19:32 so there were Damages 00:20:09 did the building collapse 00:20:11 no 00:20:14 o 00:20:23 just cost someone a bunch of money 00:20:36 was the someone limewire 00:20:49 maybe 00:21:02 not sure if the expense was collectivized to the Lime Group 00:21:04 kmc had to pay out of pocket 00:21:11 because he was the one who did the damage 00:21:14 haha 00:21:15 he neglected to include that detail 00:21:17 no, i wasn't even at the party 00:21:23 i just walked through the wreckage the next day 00:21:37 if I were doing that, I'd have contributed a bit 00:21:39 who's gonna notice, right? 00:21:43 heh, true 00:21:53 we were actually there to loot the remaining non-damaged furniture 00:21:57 lol 00:21:59 wait 00:22:02 which provided most of the tables and chairs necessary for a 3-person apartment 00:22:03 through the wreckage? 00:22:13 oh inside you mean 00:22:15 my friend had the presence of mind to rent a U-Haul for this occasion 00:22:31 i thought you meant the wreckage of the exterior wall 00:22:58 they also left some computers behind, but we weren't sure whether they were technically evidence in an ongoing lawsuit 00:24:33 yeah, wouldn't want to do anything illegal 00:24:39 that's why you only stole the furniture 00:24:40 >_> 00:24:47 <_< 00:25:01 yeah well, petty theft > federal obstruction of justice 00:25:08 besides we had the tacit approval of the CEO to take the furniture 00:25:15 or so I was assured >_< 00:27:21 `addquote yeah well, petty theft > federal obstruction of justice 00:27:23 857) yeah well, petty theft > federal obstruction of justice 00:34:37 -!- DHeadshot has joined. 00:38:12 -!- augur has quit (Remote host closed the connection). 00:38:21 :) 00:44:28 kmc: Ah, tomorrow? 00:45:06 (Please disregard previous message.) 00:50:20 what was previous message? 00:50:52 Are you asking because you've already disregarded it? 00:53:26 no 00:53:30 'm confuseod 00:55:02 I asked if you were starting your new job tomorrow. Then I realized it was a silly question because you just said you did, and I didn't have anything to add. 00:55:07 ok then 00:55:21 Exactly. 01:02:08 elliott: GUESS WHAT 01:02:19 You didn't read the post? 01:02:23 WRONG 01:03:30 You read the post? 01:03:38 Yes. 01:03:44 *gasp* 01:03:53 It was a grueling, arduous task. 01:04:05 i feel such responsible behavior sets a bad precedent for the channel. 01:04:15 But I grit my teeth and and did it! 01:04:17 shachaf: I have another post for you to read. 01:04:25 I did it! I did it! You said I couldn't do it but indeed I did! 01:04:50 @tell monqy what's the another post? 01:04:50 Consider it noted. 01:05:48 @messages 01:05:48 shachaf said 58s ago: what's the another post? 01:05:57 @ask elliott 01:05:57 Consider it noted. 01:06:09 shachaf: OK so, read http://www.impredicative.com/ur/tutorial/tlc.html first. Then read http://blog.ezyang.com/2012/07/polymorphic-variants-in-urweb/. 01:06:09 elliott: You have 1 new message. '/msg lambdabot @messages' to read it. 01:06:18 @ask monqy 01:06:18 Consider it noted. 01:06:26 Or just skip to the second one. But the other one is good too. 01:06:28 elliott: Can I other order ;( 01:06:31 Oh. 01:14:17 shachaf: Are you reading??? 01:14:31 @ask elliott no :"( 01:14:31 Consider it noted. 01:15:07 @clear-messages 01:15:07 Messages cleared. 01:15:16 @clessages 01:15:16 You don't have any new messages. 01:15:50 monqy: Did you ever read that one? 01:16:03 @clear-messages 01:16:03 Messages cleared. 01:16:12 the blog thing? 01:16:20 ye 01:16:21 i read at least the first paragraph 01:16:27 good job monqy 01:16:29 but i wasnt in a reading mood 01:16:41 did mspell get any furthre btw .. 01:16:43 *further 01:16:48 oerjan: btw op me 01:16:59 ye 01:17:14 elliott: btw op me :'( 01:17:15 monqy: how frtuher 01:17:26 um 01:17:35 i removed both kinds of special abilities! 01:18:06 grats 01:18:16 -!- derdon has quit (Ping timeout: 246 seconds). 01:18:19 I like the symmetry between product and sum types. 01:18:40 YOU MEAN PRODUCTS AND COPRODUCTS 01:18:58 elliott: Did you know all monads were covariant??!?! 01:19:05 "amazing" 01:19:06 switch :: (a -> r, b -> r) -> (a | b) -> r; swotch :: (a -> r | b -> r) -> (a * b) -> r 01:19:09 shachaf: yes 01:19:37 switch swotch 01:20:13 switch : (a -> r * b -> r) -> (a | b) -> r 01:20:32 switch : (r^a * r^b) -> (a + b) -> r 01:20:40 swotch : (r^a + r^b) -> (a * b) -> r 01:22:28 r^a * r^b = r^(a+b) 01:23:00 (a -> r) + (b -> r) = (a*b)^r 01:23:07 Um, r^(a*b) 01:23:09 What? 01:23:23 What? 01:23:24 r^a + r^b = r^(a*b)? 01:23:27 That's nonsense. 01:23:39 It's not =. 01:23:45 a -> b isn't a = b. 01:23:56 -!- derdon has joined. 01:24:00 Oh, it's not nonsenes because r^a + r^b has a bit more information. 01:24:15 (A BIT get it) 01:24:20 :') 01:24:29 s/not // 01:24:33 It *is* nonsense because ... 01:24:47 elliott: You know the way foldr's type should mirror unfoldr's? 01:24:57 switch and swotch let you get all the information out, right? 01:25:05 swotch (Left id), swotch (Right id) are fst and snd... 01:25:11 So I guess so. 01:25:27 Well, it's not polymorphic. 01:25:33 Huh? 01:25:43 Oh, no. 01:25:44 shachaf: Anyway, if you read that post I linked you'd understand why the sum-product relation is even more cool! 01:25:53 It's a continuation of that other post. Except it does sum types instead. 01:25:56 What I'm saying is that you can't ((a * b) -> r) -> (a -> r | b -> r) 01:26:04 It's a one-way thing. 01:26:20 Just like you can't (a -> b) -> Either (Not a) b. Or something. 01:26:34 i guess i'll read the post....................................... 01:26:41 monqy: NO WAIT 01:26:47 monqy: THINK ABOUT WHAT YOU'RE DOING 01:27:01 monqy: ur gr8 01:27:09 monqy: ur web 01:27:24 -!- augur has joined. 01:27:35 shachaf: The post also shows you how you can do scrap-your-boilerplate-type things with the system elegantly! 01:27:40 As in "without any ugly deriving (Data) stuff". 01:27:59 elliott: Logitext is "vnev more ocool!!" 01:28:07 What? 01:28:23 http://logitext.mit.edu/main 01:28:26 "the future" 01:30:08 This looks boring. 01:30:23 DID YOU PROVE ALL THE PROOFS 01:31:32 no 01:32:00 elliott: i am a terrible person 01:32:08 -!- oerjan has quit (Quit: Good night). 01:32:17 I have a ... -> x that I want to turn into a ... -> Type x 01:32:26 where ... can be any arity 01:32:30 and clearly TH is the way to go 01:32:43 it's ok, everyone does shit like that when they start out 01:32:57 you realise why it's a bad idea later 01:33:23 Unless you're Oleg. 01:33:46 oleg doesn't really do things like that 01:34:18 kmc: should i make a perfect language 01:35:34 elliott: I know why it's a bad idea now 01:35:38 and it's not stopping me 01:36:28 are you what they call a "bad person" 01:36:40 monqy: oh no 01:36:46 monqy: are you a "bad person" too 01:36:54 because "it takes one to know one" 01:36:59 "as they say" 01:37:01 ye 01:37:05 -!- azaq23 has joined. 01:37:07 coppro: people think they know why it's a bad idea to start with, too 01:37:13 does that make me "a bad person" ! 01:37:18 ye 01:37:39 it takes one to know that it takes one to know one 01:37:41 as they say 01:38:38 elliott this is a good blog post 01:38:38 elliott: well then you tell me why i am wrong 01:38:49 coppro: but being old is so much easier 01:38:56 monqy: i can't tell if you're being serious or not........................... 01:39:03 im serious it's good !! 01:39:06 yaey 01:39:07 is that wrong of me..... 01:39:23 nop 01:39:30 people are sarcastic often and it is had to tell!! 01:39:40 -!- derdon has quit (Remote host closed the connection). 01:40:04 elliott: When you say foo.bar in Rust, can bar just be a regular function? 01:40:26 Or is it a "method of the class of foo" "as they say" 01:40:27 coppro: but in summary (a) TH in general is bleh for many reasons (some theoretical, most practical) and self-preservation dictates to avoid it wherever possible; (b) you will run into pain whenever you end up writing higher-order functions of some kind with this method; (c) it does not really save you much at all blah blah 01:40:37 shachaf: I don't think Rust has "classes" proper? 01:40:44 elliott: Oh. 01:40:48 * shachaf doesn't know Rust. 01:40:56 elliott: How does "for 5.times" work? 01:40:59 shachaf: It has ADTs and traits (which are like OOP interfaces). 01:41:02 elliott: yes, those are the reasons 01:41:06 And you can implement traits for ADTs. 01:41:09 i am choosing to ignore them because i am stupid 01:41:31 coppro: It's OK, you'll do it the proper way when it slows down your compilation times 2x. 01:41:45 shachaf: I think that's the implementation of a trait for the integer types. 01:42:00 So 5.times is actually implementation_of_times_for_the_relevant_integer_type.times(5). 01:42:08 OK. 01:42:12 "but still" 01:42:48 What do you mean but still? 01:44:29 shachaf: I don't understand what your question is. 01:44:44 Noq uestion. 01:47:19 shachaf: https://github.com/mozilla/rust/blob/master/src/libcore/iter.rs#L19, https://github.com/mozilla/rust/blob/master/src/libcore/int-template.rs#L91, https://github.com/mozilla/rust/blob/master/src/libcore/uint-template.rs#L83, which are then instantiated like this: https://github.com/mozilla/rust/blob/master/src/libcore/int-template/i64.rs. 01:47:36 elliott: I thought there was no question. :-( 01:47:58 Also, please mark your URLs in an easy-to-copy way. 01:48:04 Don't mix valid punctuation into them. :-( 01:52:46 shachaf: Does that answer your non-question? 01:53:06 (Also apparently traits are like typeclasses in some way.) 01:53:13 (So I guess they're more powerful than Java interfaces.) 02:20:39 -!- DHeadshot has quit (Read error: Connection reset by peer). 02:20:44 -!- DH____ has joined. 02:37:16 -!- DH____ has quit (Read error: Connection reset by peer). 02:37:21 -!- DHeadshot has joined. 02:37:45 -!- kinoSi0 has quit (Read error: Connection reset by peer). 02:38:13 -!- kinoSi has joined. 02:48:32 cvttsd2si 02:48:35 What a great instruction name. 02:55:35 -!- azaq23 has quit (Quit: Leaving.). 03:00:05 -!- DHeadshot has quit (Read error: Connection reset by peer). 03:00:28 -!- DHeadshot has joined. 03:38:44 So, there's this library that's supposed to tell you the distance between a point you give it and the nearest point in the database. 03:39:01 The thing is, the library breaks. You can only use it about ten times before it starts giving you numbers that are obviously wrong. 03:39:10 Trial period expired. 03:39:16 Once, it gave a number that was too high by over a hundred orders of magnitude. 03:39:19 INSERT COIN 03:39:33 Anyway, that doesn't necessarily mean it's wrong. 03:39:53 It was wrong by over a hundred orders of magnitude. 03:47:18 a hundred orders of magnitude isn't cool 03:47:43 you know what's cool? a hundred and three orders of magnitude 04:03:14 https://www.google.com/search?q=goats+in+trees&tbm=isch 04:06:15 -!- DHeadshot has quit (Read error: Connection reset by peer). 04:18:38 -!- MoALTz has joined. 04:36:08 -!- heroux has quit (Ping timeout: 244 seconds). 04:41:30 -!- MoALTz has quit (Ping timeout: 264 seconds). 04:55:25 There is a open source FPGA, called ZUMA. However, no implementations exist, as far as I know. 04:56:04 I thought ZUMA was a game with marbles. 04:56:37 http://www.computer.org/portal/web/csdl/doi/10.1109/FCCM.2012.25 04:56:51 it's an FPGA implemented on another FPGA °_° 04:57:02 "Previous attempts to map an FPGA architecture into a commercial FPGA have had an area penalty of 100x at best. Through careful architectural and implementation choices to exploit low-level elements of the host architecture, ZUMA reduces this penalty to as low as 40x." 05:03:02 -!- impomatic has quit (Ping timeout: 244 seconds). 05:04:09 good 05:08:12 progress 05:13:21 hey kmc did mosh get any releases in the past ever 05:13:23 maybe i should update my shit 05:13:53 also does anyone know if that ghc version with type-level strings came out yet 05:13:54 shachaf? 05:14:04 YAWG 05:14:18 the latest mosh release is in the topic in #mosh 05:14:20 and on mosh.mit.edu 05:14:39 shachaf: Yawg? 05:14:43 kmc: :( 05:14:46 My server is running an old version of mosh. :-( 05:14:49 kmc: I need a personal social web 2.0 connection with my software, man! 05:14:51 That means asking you. 05:14:52 So I can't file bug reports until I upgrade. 05:14:55 But upgrading is a hassle! 05:14:58 HLEP 05:15:11 Also I don't even know what version I have, so ha. 05:15:35 $ mosh --version 05:15:38 1.2.3 05:15:44 POINT 4 05:15:52 I'm on 1.1.94d. 05:16:00 I bet that's old. 05:16:16 Apparently 1.2.2 is the latest release. 05:18:13 shachaf: Is there a way I can tell g++ and/or clang++ that a macro I have is an assertion macro? 05:18:28 What should it do differently? 05:18:59 shachaf: I don't know. Not complain about me assuming what I asserted later on? 05:19:23 Oh, be smart like that? Can it even do that? 05:19:32 Does it happen with #include 's assert()? 05:19:55 shachaf: I think so? 05:20:03 I want to be as fancy as assert.h. 05:20:14 Alternatively, is there a way to hook into when an assert fails or something? 05:20:22 I want my fancy stuff *and* the platform's fancy stuff. :( 05:21:09 assert() raises SIGABRT doesn't it? 05:21:31 it will still kill the process even if you handle SIGABRT 05:21:50 All the big C++ codebases I've used defined their own assert, but I've never noticed anything smart like "telling the compiler about your assumptions". 05:22:07 Some asserts raise SIGTRAP in debug mode. Or do other things. 05:22:34 Hmm, I wonder whether SIGTRAP has much of an advantage over SIGABRT. 05:22:35 There's some inline asm in the current assert() implementation. :( 05:22:50 Is it asm("int3")? 05:22:56 Snippet of the fun: http://sprunge.us/RQcf 05:23:05 shachaf: Close enough: #define MyDebugBreak() _asm {int 3} 05:23:13 You're working with Win32 code? 05:23:23 Oh, whoops. I actually removed IsDebuggerPresent95. 05:23:41 BUT before I did, 05:23:47 shachaf: (how do I see a file at a revision in git) 05:23:50 (Without checking it out.) 05:23:57 git show? 05:23:57 if you mark your custom abort function as attribute((noreturn)), and/or insert a __builtin_unreachable() 05:24:02 then gcc can do clever things 05:24:19 I think unreachable is better than noreturn? 05:24:23 [elliott@dinky source]$ git show 5e814759930bf282ba3b5741643bebe1587557f1 debug.cc 05:24:23 [elliott@dinky source]$ 05:24:24 Help. 05:24:34 git show commitish:filename 05:24:46 thx 05:24:50 yw 05:25:21 shachaf: http://sprunge.us/FaBg 05:25:28 kmc may also click that link. Everyone else is forbidden. 05:25:37 elliott: thx 05:25:42 that's some code 05:25:43 on kmcs bhlf 05:26:02 By the way, this isn't some 90s nonsense, this code is from ~2005-ish. 05:26:18 (Okay, maybe this bit is older. But it's newer than, like, 2000.) 05:26:25 what is it 05:26:28 Crawl? 05:26:56 It's monqy. 05:27:16 monqy: ruopensouce?!! 05:28:00 monqy's codebase has some great comments. 05:28:03 monqypatching 05:28:04 // XXX: must fix species abils to not use duration 15 05:28:04 // -- ummm ... who wrote this? {dlb} 05:28:23 The unforgivable faux pas of not signing your todo comments. 05:28:35 elliott: has this been in git since 2005?? 05:28:58 If by "2005" you mean "2011ish", then yes. 05:29:11 2011 is the year of git on the desktop 05:30:37 kmc: Did you know Windows still exists? :( 05:30:53 http://gcc.gnu.org/onlinedocs/gcc-4.7.0/gcc/Function-Attributes.html 05:30:58 gcc has a lot of fancy attributes 05:32:18 shachaf: Do you want to hear a joke? I know you love jokes. 05:33:19 elliott: Oh boy, a joke! 05:33:50 shachaf: The joke is const correctness. 05:35:20 shachaf: Do you want to hear another joke? 05:35:50 Time's up. 05:35:54 monqy: What about you? 05:36:11 ye 05:36:14 monqy: Tough. 05:36:35 probably for the best 05:36:40 im eating so if i laughed bad things might happen 05:36:46 -!- williannys has joined. 05:36:52 like that 05:37:12 `welcome williannys 05:37:15 williannys: Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: http://esolangs.org/wiki/Main_Page. (For the other kind of esoterica, try #esoteric on irc.dal.net.) 05:37:21 Wait, are you the sneeze person? 05:38:44 going to sleep, 'night all 05:38:53 kmc sleeps? 05:39:11 yes i sleep quite a lot 05:40:10 Weird. 05:40:53 -!- williannys has left. 05:45:58 Ok, so I heard JVM startup time is slow, but this is insane 06:19:20 Disadvantages of existing FPGA architectures includes: * Proprietary nature of hardware. * Requires you to use their software, specific to certain operating systems and certain computers. * Vendor locked. * No dynamic reconfiguration (some have this feature, but again, you need to use their software so you cannot do it with your own software). 06:19:40 Therefore you cannot use it to make your own computer system since it can only be used with the computer systems they say it can be used! 06:20:08 Another disadvantage: * They don't sell them for a long enough time. After they sell it, they will be obsolete and you have to buy another one. 06:39:07 elliott: Yes! 06:39:23 elliott: What's the another joke? 06:39:41 elliott: He has, like, a job and stuff now, man. 06:39:44 He sold out on us! 06:40:06 He's probably going to sleep early because of that! 06:44:32 -!- zzo38 has quit (Remote host closed the connection). 06:46:34 -!- nys has quit (Quit: quit). 06:47:35 `? zzo38 06:47:40 zzo38 is not actually the next version of fungot, much as it may seem. 06:47:46 `? fungot 06:47:46 shachaf: he looks gay. 06:47:49 fungot cannot be stopped by that sword alone. 06:47:52 fungot: thx 06:47:52 shachaf: what the hell would anyone srvive that the computer, it was dat. oh and apologies in advance for the most beautiful girl i've ever heard of nick pitera on this 06:48:30 ^style 06:48:30 Available: agora alice c64 ct darwin discworld europarl ff7 fisher fungot homestuck ic irc iwcs jargon lovecraft nethack pa qwantz sms speeches ss wp youtube* 06:48:34 Ah. 06:49:07 In retrospect, "of course". 06:49:26 ^style darwin 06:49:26 Selected style: darwin (Books by Charles Darwin -- you know, that evilution guy) 06:49:31 fungot 06:49:33 shachaf: :( i was too glad to accept :) this, in fact, take them away from any spot which itches: monkeys search each other for aid; and as they flit from tree to tree, and what i lay much stress on it, and really great, if we take/ fairest standard :) comparison because, :)/ corals just mentioned, and :)/ other photographs given by dr. o. regnelli,/ illegitimate unions :) dimorphic and trimorphic plants, as described by himsel 06:49:49 Charles Darwin sure used a lot of smileys. 06:50:00 He was just such a fun guy. 06:50:24 fungot: are you a fun guy??????? 06:50:25 shachaf: a species may depart from/ type :) both its parents were common peacocks. lastly, i have had some strange proceedings on board/ :( beagle" anchored a few miles, i could see this argument answered. 07:08:43 shachaf: it seems like noreturn is the same as normal function call + unreachable 07:08:55 http://gcc.gnu.org/onlinedocs/gcc/Other-Builtins.html#index-g_t_005f_005fbuiltin_005funreachable-3286 07:09:12 elliott: Oh. 07:09:17 elliott: Is exit() noreturn? 07:09:43 If I write if (x == 0) exit(1); will the compiler assume that x is != 0 or something? 07:09:50 I think it ought to be allowed to assume that for unreachable. 07:11:41 I think exit is noreturn, yes. 07:11:41 C11 standard prototype for exit is "_Noreturn void exit(int status);" 07:11:49 At least in glibc. 07:12:13 shachaf: I suspect GHC might insert a __builtin_unreachable() after every call to a noreturn function, implicitly. 07:12:42 That's very clever of GHC. 07:12:46 gcc. 07:13:00 Anyway, that makes sense. 07:13:17 Oh, you mean the noreturn comes *after* the return from the function call. 07:13:30 Makes sense. 07:13:56 The unreachable, you mean? 07:14:03 Yes. 07:16:41 http://sprunge.us/hXUh -- assumptions: it has those. 07:17:25 Oh, you mean the noreturn comes *after* the return from the function call. 07:17:26 Huh? 07:17:33 00:13