00:04:22 -!- MDream has joined. 00:07:04 -!- MDude has quit (Ping timeout: 246 seconds). 00:10:59 good eeing 00:15:36 Hellu 00:15:52 gragh. my aes key expansion still isn't working 00:27:56 Are you... reimplementing AES? 00:28:06 yes. In visual basic 00:28:55 http://hastebin.com/anopomiyay.vbs 00:29:23 thats what I have so far but the key expansion is not giving the righyt output 00:30:12 I thought you were going to use existing crypto. 00:30:30 bah. 00:30:51 I'll do that if this gets too frustrating 00:31:31 what's weird is only a few of the bytes of the expanded key are wrong 00:33:00 (in the particular unit test of the key 00000000000000000000000000000000) 00:33:13 Have you checked the iteration-by-iteration key expansion examples in the standard? 00:35:50 -!- ProofTechnique has joined. 00:36:00 that's true, I should just log the same things they did and compare... 00:36:05 Helloren 00:37:56 ok never mind I had an off by one error 00:38:18 hi ProofTechnique 00:38:37 I'm making AES in Visual Basic 00:38:49 I'm sorry 00:38:51 http://arin.ga/CpcwH6 00:41:53 It working! I fix it! 00:42:18 ...on one example. I should add more unit tests 00:42:51 before counting it as "done" 00:43:35 however this is a major step towards getting the ssh to work 00:45:46 oren: Um... 00:45:58 Did you just use two different links to the same thing? 00:46:37 I want to go on the record here that despite the suggestion, I'm in no way promoting homebrew cryptography. 00:46:37 oh. whoops 00:47:00 -!- impomatic_ has joined. 00:47:02 -!- darkl0ck_ has quit (Ping timeout: 260 seconds). 00:47:15 -!- impomatic_ has quit (Client Quit). 00:48:39 -!- darkl0ck has joined. 00:57:32 -!- clog has joined. 01:01:53 -!- mauris_ has quit (Ping timeout: 256 seconds). 01:28:03 [wiki] [[Pikalang]] http://esolangs.org/w/index.php?diff=44015&oldid=41282 * 68.50.42.113 * (+26) /* Commands */ 01:29:50 [wiki] [[Pikalang]] http://esolangs.org/w/index.php?diff=44016&oldid=44015 * Grotr * (+3) /* External resources */ 01:37:35 -!- mauris_ has joined. 01:37:35 -!- mauris_ has quit (Changing host). 01:37:35 -!- mauris_ has joined. 01:45:01 I want to go on the record here that despite the suggestion, I'm in no way promoting homebrew cryptography. <-- eh, someone's gotta brew it, right? 01:45:31 -!- darkl0ck_ has joined. 01:46:59 i mean, there are presumably existing, well-tested implementations of AES for visual basic, so i think it's a bit silly to write your own, but like, if oren writes their own and tests it thoroughly as well, i see no reason to trust any existing implementation more than oren's 01:48:38 -!- darkl0ck has quit (Ping timeout: 272 seconds). 01:50:02 -!- fowl has quit (Disconnected by services). 01:50:31 -!- fowl has joined. 01:50:57 (imo i'm not a professional cryptozoologist though so what do i know! i'm definitely interested to find out if there's more to it than "this produced the same output as an existing trustworthy implementation for a couple thousand random inputs, so it's trustworthy") 01:51:30 -!- fowl has quit (Disconnected by services). 01:52:01 -!- fowl has joined. 01:56:16 -!- fowl has quit (Disconnected by services). 01:56:31 -!- notfowl has joined. 01:59:06 -!- hppavilion[1] has quit (Ping timeout: 240 seconds). 02:02:45 -!- grotewold has joined. 02:04:36 -!- notfowl has changed nick to Guest19768. 02:06:17 -!- atslash has quit (Quit: This computer has gone to sleep). 02:08:05 -!- mauris__ has joined. 02:09:51 -!- hppavilion[1] has joined. 02:11:19 -!- mauris_ has quit (Ping timeout: 246 seconds). 02:12:22 mauris__: there are other problems, relating to the trustworthiness of VB itelf. typically crypto is written in C or even assembly so the compiler hopefully can't introduce weakness 02:13:01 -!- grotewold has quit (Quit: My Mac has gone to sleep. ZZZzzz…). 02:14:23 in other words just because my code looks right and produces the right output doesn't mean Microsoft's compiler or garabage collector or whatever isn't screwing you 02:21:05 -!- grotewold has joined. 02:22:02 How can I make a language which's programs are harder to hack and are more secure? 02:22:03 Well 02:22:31 1) Provide well-tested standard libraries for security (no hacky hashing for you!) 02:22:38 oren: oh, like the plaintext lingering around in memory? 02:23:07 mauris__: Yes, like that xD 02:23:33 2) Provide easy mechanisms to catch crashes so that you don't leave something open to kill it 02:24:05 -!- grotewold has quit (Client Quit). 02:24:53 3) provide a nocache storage modifier that disables disk cahing, and a zero() function that takes such a variable and zeroes it 02:25:18 -!- grotewold has joined. 02:25:20 Don't C compilers optimize stuff? 02:25:27 Which is not what you need with crypto 02:25:39 (modern compilers are known to optimize out the part of an algorithm that deletes that plaintext) 02:26:00 Sgeo: yes, which is why it needs to move to assumply 02:28:38 [wiki] [[Ingredients-Oriented Paradigm]] http://esolangs.org/w/index.php?diff=44017&oldid=44000 * Hppavilion1 * (+675) Added some stuff and rearranged 02:31:04 [wiki] [[Ingredients-Oriented Paradigm]] http://esolangs.org/w/index.php?diff=44018&oldid=44017 * Hppavilion1 * (+313) Exception Handling 02:42:27 Oooh 02:42:34 Constraint Programming looks interesting 02:42:46 -!- grotewold has quit (Quit: My Mac has gone to sleep. ZZZzzz…). 02:44:13 -!- idris-bot has quit (Ping timeout: 246 seconds). 02:45:24 -!- Melvar has quit (Ping timeout: 264 seconds). 02:46:08 !tell ais523 WUUI question: "Whenever (to be precise, just after) any expression is evaluated," <-- how many expressions is, say, "x[0]/3"? 02:46:10 ais523: WUUI: question:: Whenever (to be precise, just after) any expression is evaluated,: <--: how: many: expressions: is,: say,: x[0]/3?: 02:46:21 @tell ais523 WUUI question: "Whenever (to be precise, just after) any expression is evaluated," <-- how many expressions is, say, "x[0]/3"? 02:46:21 Consider it noted. 02:46:47 `thanks EgoBot 02:47:02 Thanks, EgoBot. ThEgoBot. 02:48:18 `thanks walrusman 02:48:19 Thanks, walrusman. Thalrusman. 02:49:42 Ok, how is it even possible that I did the shift rows step wrong? 02:50:59 well that's solved, bu the output is still worng. back to the drawing beard 02:54:23 -!- Melvar has joined. 03:00:45 `thanks lllllleeeeroooooy_jenkins 03:00:45 Thanks, lllllleeeeroooooy_jenkins. Theeeeroooooy_jenkins. 03:00:53 Interesting 03:00:56 -!- Phantom__Hoover has quit (Read error: Connection reset by peer). 03:01:47 I'd like to see a Programming Language which is centred around defining abstract machines (turing, kolmogorov, FSMs, etc) and letting those evaluate for you 03:02:01 Declarative programming in a nutshell, I suppose xD 03:02:51 Anyone want to define an EsoFSM? 03:04:01 @tell hppavilion[1] you are a walrus. Fear your wrath 03:04:01 You can tell yourself! 03:04:10 @tell hppavilion[1] you are a walrus. Fear your wrath 03:04:10 You can tell yourself! 03:04:24 @clear 03:04:24 Maybe you meant: clear-auto-reply clear-messages clear-topic learn 03:06:06 -!- idris-bot has joined. 03:06:42 @help 03:06:42 help . Ask for help for . Try 'list' for all commands 03:06:49 @list 03:06:49 What module? Try @listmodules for some ideas. 03:06:55 @help list 03:06:55 list [module|command]. Show commands for [module] or the module providing [command]. 03:07:06 @list command 03:07:06 No module "command" loaded 03:07:09 @list help 03:07:09 help provides: help 03:07:18 @list list 03:07:18 system provides: listchans listmodules listservers list echo uptime 03:07:30 @list echo 03:07:30 system provides: listchans listmodules listservers list echo uptime 03:07:36 @echo "Walrus Man" 03:07:36 echo; msg:IrcMessage {ircMsgServer = "freenode", ircMsgLBName = "lambdabot", ircMsgPrefix = "hppavilion[1]!~Lordofthe@133-171-58-66.gci.net", ircMsgCommand = "PRIVMSG", ircMsgParams = ["#esoteric",":@echo \"Walrus Man\""]} target:#esoteric rest:"\"Walrus Man\"" 03:07:48 @@ 03:08:18 @tell walrus GET OUT OF MY BODY 03:08:18 Consider it noted. 03:09:09 @hppavilion[1] 03:09:09 Unknown command, try @list 03:09:17 @tell tell tell 03:09:17 Consider it noted. 03:09:38 I should stop before I make someone very powerful* angry 03:09:49 *(Not at all powerful) 03:10:59 Someone love me 03:11:43 oren b_jonas rdococ shachaf there are lonely people here 03:12:00 -!- shachaf has left. 03:13:24 So all the sub steps of AES are working perfectly, and yet the ouput is wrong?!! wtf is happenink here 03:13:43 You missed a substep perhaps? 03:13:55 Perhaps fungot does not will that you have encryption 03:13:55 hppavilion[1]: ( currycpseid a b). that means there's at the very least, they did. it's in all the shitty languages out there 03:14:19 fungot fungot fungot <-- I wonder if this prints more than one 03:14:19 hppavilion[1]: http://www.schemers.org/ documents/ standards/ fnord/ hallways/ etc? 03:14:22 Nope 03:16:42 I don't think so. In fact my main cipher() function is virtually identical to the pseudocode in the standard 03:21:20 @tell ais523 "An entire program is a single command." <-- I think you either forgot to put {} around the example program, or meant to write "list of commands" 03:21:20 Consider it noted. 03:25:06 -!- hppavilion[1] has quit (Ping timeout: 240 seconds). 03:29:51 -!- ProofTechnique has quit (Quit: Textual IRC Client: www.textualapp.com). 03:51:14 FUUUUUUUUUUUUUUU 03:51:40 -!- hppavilion[1] has joined. 03:52:11 those... assholes who wrote the standards document store their arrays with columns in sequence! 03:52:35 What kind of fucking monster does that! 03:56:36 oren: What?? 03:56:49 Which standards document? 03:57:11 fungot: You. Fungot. Fungotify. 03:57:11 hppavilion[1]: i recognize that the idea of doing them like that, it means 03:57:13 http://csrc.nist.gov/publications/fips/fips197/fips-197.pdf 03:57:40 look at page 33 how the keyis loaded into the matrix 03:58:00 er, the input 03:58:28 4 bytes of the input goes into each column 03:58:55 So I realized that the snub cube has a problem, considered as a die. 03:59:01 (Besides the fact that it's not fair.) 03:59:24 tswett: What's the problem? 03:59:33 Now, there are three classes of faces: squares; triangles adjacent to squares ("edge triangles"); and triangles not adjacent to squares ("corner triangles"). 03:59:33 (I love how the universe has continuity) 03:59:51 Every square has an opposite face. Every corner triangle has an opposite face. Edge triangles do not have opposite faces. 04:00:05 Oh... oh god 04:00:26 Each *pair* of edge triangles has an opposite pair of edge triangles, but the two pairs are at a 90 degree angle to each other. 04:01:05 So if the die lands on an edge triangle, there's no top face; there's instead a top pair of faces, where one of the middle vertices is higher up than the other one. 04:01:22 Wouldn't it be a quartet of edge triangles? 04:01:24 Wait 04:01:24 So apparently those assholes should have put a Transpose(State) at the start of their pseudocode but couldn't be bothered 04:01:26 No it wouldn't 04:02:04 So you could simply mark those "edge triangle pair middle vertices" with numbers instead of marking the edge triangles themselves with numbers. 04:02:16 That's kind of inelegant. 04:02:38 But does it have the number of sides you wanted? 04:03:05 OOOOOOOOOOh 04:03:18 I missed the "So if..." part 04:03:36 "So if the die lands on an edge triangle, there's no top face; there's instead a top pair of faces, where one of the middle vertices is higher up than the other one." 04:03:41 It does have the number of sides I want. 04:03:50 I figured that out 04:04:04 So are you going to hunt down and kill the person who suggested a snub cube? 04:04:04 I pretty much can't imagine going with a shape that's not either a snub cube or a variant of the snub cube. 04:04:23 No, I think the snub cube was a great suggestion, and I'm really grateful to whoever it was. 04:04:33 I'm just going to have to alter it a little. 04:05:48 Well I should get around to finding out what a snub cube 04:05:49 is 04:05:57 dropped a word 04:05:58 there 04:07:34 I want 3D Animated SVG 04:07:54 SVOPENGLIF if you will 04:15:44 -!- haavard has quit (Read error: Connection reset by peer). 04:16:18 -!- haavard has joined. 04:17:14 -!- hppavilion[1] has quit (Ping timeout: 244 seconds). 04:34:58 -!- mauris__ has changed nick to mauris. 04:35:11 -!- mauris has quit (Changing host). 04:35:11 -!- mauris has joined. 04:59:25 It work! 05:00:25 http://arin.ga/Gd7e6L 05:00:54 (also has my very incomplete SSH class) 05:00:58 -!- Walpurgisnacht has joined. 05:02:33 Explain what monads are in full depth from memory but very simply 05:04:10 Iv'e heard they're used as a way to encapsulate statefulness away from the pure functional code 05:04:40 Mmm 05:04:45 Sounds useful 05:08:38 However in addition Haskell represents basically all data structures as monads, for uh... reasons. 05:15:31 Walpurgisnacht: the programming definitions: they're a way to chain computations with a certain context together 05:15:55 Mmm 05:15:56 here "context" is a very vague thing, of course! 05:18:25 it can mean, "optionally not returning a value" (a -> Maybe b), or "optionally returning many values representing possibilities for some non-deterministic computation" (a -> List b), or "computations using IO" (a -> IO b) 05:19:42 in each of these, a monad is essentially some fluff that lets you chain (a -> m b), (b -> m c), (c -> m d), ... together in a way that makes sense 05:20:37 (and also, supplies an (a -> m a) computation -- one that "does nothing", but in the context in question) 05:21:08 here's a very simple computation that can fail and not return a value: 05:21:29 @let halve x = if even x then Just (x `div` 2) else Nothing 05:21:31 Defined. 05:21:41 > [halve 2, halve 3, halve 4, halve 5, halve 6] 05:21:43 [Just 1,Nothing,Just 2,Nothing,Just 3] 05:21:58 -!- hppavilion[1] has joined. 05:22:04 I see 05:22:08 How useful 05:22:10 its type is: 05:22:13 :t halve 05:22:14 Integral a => a -> Maybe a 05:22:37 okay, haskell made that confusing. but it's (Int -> Maybe Int), essentially. 05:23:13 now, suppose we want to halve a thing three times, but if we fail along the way there, return Nothing 05:23:48 we can't do: halve (halve (halve 24)) 05:24:04 because halve wants an Int, not a Maybe Int! 05:24:16 so we use the confusingly-typed: 05:24:19 :t (>>=) 05:24:21 Monad m => m a -> (a -> m b) -> m b 05:24:46 it takes the result of some computation, and the next step to pipe that value through 05:25:17 so this works: 05:25:34 > halve 24 >>= halve >>= halve 05:25:37 Just 3 05:25:47 -!- nisstyre has quit (Read error: Connection reset by peer). 05:27:00 this is much nicer than what we'd have to write otherwise: https://bpaste.net/show/898533b7f3f7 05:27:29 grfmbl 05:28:55 Walpurgisnacht: do you know what the type IO a means, in Haskell? 05:29:24 Not really I'm still pretty new with haskell 05:30:35 ok. it's roughly something like... a recipe for how to get a value of the type 'a' from the outside world 05:30:45 :t getChar -- for example 05:30:47 IO Char 05:31:29 the crucial thing to understand about these is that you can't "call" them! 05:32:14 (oh, i meant to write: "get a value of the type 'a' from the outside world, optionally doing some other stuff on the side") 05:32:35 in fact, the only one that ever gets ran is main :: IO (). 05:34:32 ah 05:34:38 when you write a Haskell program, all you're doing is describing how the steps for "main" fit together 05:35:37 you can never execute a sub-step (there is no function "IO a -> a"), only combine a bunch of them and call it "main". this is why Haskell is pure! 05:36:14 but then how do you interleave inputs and outputs to write an interactive program? 05:37:04 and as you have probably read somewhere, this "chaining steps together" stuff happens using a monad 05:38:00 oren: i'll demonstrate that in a bit 05:38:34 the type of (>>=), the "bind" combinator we used before, in the IO monad, is: IO a -> (a -> IO b) -> IO b 05:40:13 which means: 'given an IO computation that gets us an "a", and another IO computation that WOULD return a "b" if only we knew which "a" to use, combine them into an IO computation returning a "b"'. 05:40:36 the second thing (a -> IO b) might be confusing, but there's a really simple example: 05:40:38 :t readFile 05:40:39 FilePath -> IO String 05:41:03 this WOULD be an IO String, but it needs to be told which file path to read from first 05:41:07 :t readFile "foo.txt" 05:41:08 IO String 05:41:18 ^ and now it's an IO String. 05:41:45 so now we can combine two IO steps together where the second depends on the result of the first 05:41:57 :t getLine -- from standard input 05:41:59 IO String 05:42:36 :t putStrLn -- to standard output, with a dummy result () (think "void" in C-like langs) 05:42:37 String -> IO () 05:43:21 now we can write a program that reads a line and prints it by piping them together with (>>=)! 05:43:48 :t getLine >>= putStrLn -- has the right type to be "main"! 05:43:49 IO () 05:44:20 we still haven't done any line-getting or string-putting, here. we *only* described a program that does those steps in order -- the second depending on the result of the first 05:45:16 :t getLine >>= readFile >>= putStrLn 05:45:18 IO () 05:45:49 another valid program: "get a line; with the result read a file; with the result write a string." 05:47:01 :t (getLine >>= readFile) >>= putStrLn -- for clarity 05:47:03 IO () 05:48:22 a common case, of course, is that the second step doesn't depend on the first -- we simply want to put them in order 05:48:32 there's an operator for that, too: 05:48:34 :t (>>) 05:48:36 Monad m => m a -> m b -> m b 05:49:42 -!- hppavilion[1] has quit (Ping timeout: 244 seconds). 05:52:07 -!- JesseH has quit (Read error: Connection reset by peer). 05:52:42 Walpurgisnacht: oren: https://bpaste.net/show/5525a8366547 <-- here is a more complicated example 05:53:42 "\name -> blah" is a lambda. a very common pattern is: 05:54:27 someAction >>= (\foo -> ...) -- where ... is some other action that needs the result of the previous one -- which we've called "foo" 05:55:01 (in the type "IO a -> (a -> IO b) -> IO b", foo is our "a") 05:55:16 (and the whole lambda is the "(a -> IO b)") 05:57:06 oren: the program is interactive! we chain the steps "get line; print; get line; print" together to one big "main", but when it gets executed, those steps aren't necessarily done all at once; we can perform the first step without worrying about the third one. 05:58:16 now you're thinking: "that looks freaking awful. i don't want to write those pesky >>= and >> and lambdas everywhere! i just want to write my steps in order and be done with it." 05:59:57 and the Haskell people agree! there's some syntactic sugar called "do-notation" that lets you do just that: https://bpaste.net/show/7df7e88f4f6e 06:00:12 Eh. >> is ; and >>= is function composition where unlike C, the functions are written in the order they are executed 06:01:00 not so awful, just uses symbols in unorthodox ways 06:01:46 "p >> q" became "p; q" (you can put semicolons instead of newlines), and "p >>= (x -> q)" became "x <- p; q" 06:02:43 Jeeze I have a lot to learn 06:02:54 do-notation is a bit confusing, because writing "do { line <- getLine; putStrLn line }" looks a *lot* like a bit of imperative code that you can execute. but in reality it's *still* just a pretty way to describe how they fit together! 06:03:59 anyway, i'm derailing a bit. understanding do-notation is something you can worry about separately :) 06:04:07 that seems like a meaningless distiction to me. it executes when you run main, in the order you wrote it, just like in C 06:05:13 oren: That's the goal. The trick is, *not everything in Haskell works like that*. 06:05:52 The parts that aren't IO don't have to be in any order, but that is ALSO true in C with modern compilers 06:06:00 The goal is, you write imperative code, pretty much as imperative code, and then functional code as functional code. 06:07:29 Also worth noting -- IO is not the only monad. 06:07:33 You can write your own. 06:07:41 So, >> is ;, but it's a programmable ; 06:07:48 still, very nice of the Haskell makers to include an algol-derivative-mode 06:08:49 if haskell tutorials started with main = do that would make haskell much more popular 06:08:56 the goal also is: there's no way to *call* the imperative parts from the functional parts. you can only click functional parts together, or click them onto imperative parts to get bigger imperative parts 06:09:31 -!- oren has quit (Remote host closed the connection). 06:09:46 -!- oren has joined. 06:12:17 Walpurgisnacht: if you got lost at any point, feel free to point that out! 06:12:40 > let foo a = do { x <- a; return (x+1) } in foo [1,2,3] 06:12:42 [2,3,4] 06:12:54 ok 06:13:12 the "real" way to monad enlightenment is: write a lot of code for each individual monad and then see the overarching pattern, because it's very vague 06:13:47 the IO and Maybe monads do very different kinds of things in one way, but they're very similar in another! 06:14:39 (yes, a list is also a monad) 06:14:49 :t Just 1 >> Just 3 06:14:51 Num b => Maybe b 06:16:47 > Just 1 >> Just 3 06:16:49 Just 3 06:17:23 :t >>= 06:17:25 parse error on input ‘>>=’ 06:17:35 :t `>>=` 06:17:36 parse error on input ‘`’ 06:17:36 > Just 1 >>= (\valueInsideJust -> Just 3) -- equivalent 06:17:38 Just 3 06:18:13 > Just 1 >>= (\valueInsideJust -> valueInsideTheJust + 3) -- also possible, and if we don't supply a Just to start with: 06:18:15 Not in scope: ‘valueInsideTheJust’ 06:18:15 Perhaps you meant ‘valueInsideJust’ (line 1) 06:18:22 > Just 1 >>= (\valueInsideJust -> valueInsideJust + 3) -- also possible, and if we don't supply a Just to start with: 06:18:24 No instance for (Show b0) 06:18:24 arising from a use of ‘show_M889533674962114763625185’ 06:18:24 The type variable ‘b0’ is ambiguous 06:18:38 ugh. lambdabottttt 06:19:01 oh, duh. 06:19:16 > Just 1 >>= (\valueInsideJust -> Just (valueInsideJust + 3)) 06:19:19 Just 4 06:19:29 > Nothing >>= (\valueInsideJust -> Just (valueInsideJust + 3)) 06:19:30 Nothing 06:20:06 for >>= in Maybe, if we start with a failed computation, that failure gets propagated; else we continue with the result. 06:21:19 > Just "heya" >>= Just >>= Just >>= Just >>= (\x -> Nothing) >>= Just >>= Just 06:21:22 Nothing 06:21:39 if *any* of the steps go wrong, the whole chain fails 06:22:13 (Just is simply (\x -> Just x), so it's a valid right-hand operand to >>=) 06:26:06 yay, a monad tutorial 06:26:42 the thing to remember is that monads are a lot like burritos 06:39:33 here's something depressing: 06:39:38 "When the Java 8 library team was designing Optional there was some opposition to the idea that it should contain some useful methods (essentially Optional.map and Optional.flatMap) on the somewhat spurious grounds that they didn't want their Optional to be a Monad." 06:42:38 The Java team is always trying to restrict what sort of code can be written 06:45:03 they in particular loathe functions 06:53:29 why did they add lambdas then 07:01:15 how do i test if $myprogram is faster than $otherprogram if the disk io is the bottleneck? 07:01:39 -!- Walpurgisnacht has quit (Quit: Fuck me). 07:01:49 uhm, i need a fast way to generate pseudorandom crap 07:02:30 like a few gb 07:02:34 any ideas? <.< 07:03:07 -!- shachaf has joined. 07:04:08 another pleasant interjection from our clear wire friend 07:09:16 -!- oerjan has joined. 07:10:46 umtry dd if=/dev/urandom of=temp bs=1048576 count=%nunmber of megabytes% 07:11:01 -!- AnotherTest has joined. 07:12:15 that's saving it to disk 07:12:32 how do i feed it to my programs without disk being the bottleneck? 07:14:18 -!- zadock has joined. 07:15:05 prog Great repository names are short and memorable. Need inspiration? How about strident-rutabaga. <-- sorry, but my heart long since decided on "reactive banana" hth 07:15:36 -!- zadock has quit (Remote host closed the connection). 07:16:41 wait yours doesn't exist, cheater! 07:17:01 -!- PinealGlandOptic has quit (Ping timeout: 244 seconds). 07:17:39 i must admit that reactive banana is memorable 07:23:32 `` cc $'#include \n#include \nmain() { srand(time(0)); for(;;) putchar(rand()); } | dd of=/dev/null bs=1M count=1K 07:23:33 bash: -c: line 0: unexpected EOF while looking for matching `'' \ bash: -c: line 1: syntax error: unexpected end of file 07:23:38 `` cc $'#include \n#include \nmain() { srand(time(0)); for(;;) putchar(rand()); } | dd of=/dev/null bs=1M count=1K 07:23:40 bash: -c: line 0: unexpected EOF while looking for matching `'' \ bash: -c: line 1: syntax error: unexpected end of file 07:23:41 Oops 07:25:33 well thanks 07:25:35 `` cc $'#include \n#include \nmain() { srand(time(0)); for(;;) putchar(rand()); }' | dd of=/dev/null bs=1K count=100K 2>&1 07:25:43 102400+0 records in \ 102400+0 records out \ 104857600 bytes (105 MB) copied, 6.6342 s, 15.8 MB/s 07:25:47 That's unusually slow 07:26:10 i wouldn't say "unusually"... 07:27:03 you're throwing away most of the data you get from rand and you're printing one character at a time 07:32:05 `` cc $'#include \n#include \nmain() { char buf[4096]; for(;;) fwrite(buf, sizeof buf, 1, stdout); }' | dd of=/dev/null bs=4K count=100K 2>&1 07:32:11 102400+0 records in \ 102400+0 records out \ 419430400 bytes (419 MB) copied, 5.78028 s, 72.6 MB/s 07:33:38 Oh, syscall overhead 07:33:50 -!- stalem has joined. 07:33:50 `` cc $'#include \n#include \nmain() { char buf[1<<20]; for(;;) fwrite(buf, sizeof buf, 1, stdout); }' | dd of=/dev/null bs=1M count=1K 2>&1 07:33:52 dd: warning: partial read (4096 bytes); suggest iflag=fullblock \ 469+555 records in \ 469+555 records out \ 782241792 bytes (782 MB) copied, 0.911976 s, 858 MB/s 07:34:02 `` cc $'#include \n#include \nmain() { char buf[1<<20]; for(;;) fwrite(buf, sizeof buf, 1, stdout); }' | dd iflag=fullblock of=/dev/null bs=1M count=1K 2>&1 07:34:04 1024+0 records in \ 1024+0 records out \ 1073741824 bytes (1.1 GB) copied, 1.17474 s, 914 MB/s 07:50:45 putchar() doesn't really print "one character at a time" unless you specifically get an unbuffered stream. 07:52:11 well there's the function call overhead 07:52:58 And HackEgo's performance characteristics aren't probably very good if you're looking for something "normal", since UML. 07:53:21 what's that cc ? 07:55:03 -!- AnotherTest has quit (Ping timeout: 252 seconds). 07:55:42 `cat bin/cc 07:55:44 #!/bin/sh \ echo "$@" | sed 's/\\/\n/g' >/tmp/a.c && gcc -w -Wfatal-errors -std=c11 -O2 /tmp/a.c -o /tmp/a.out && /tmp/a.out 07:56:18 -!- PinealGlandOptic has joined. 07:59:20 that's pretty convenient 08:00:21 I don't like it using a useless temporary file for the source. 08:00:43 !show tell 08:00:43 sh xargs printf "%s: "; sed 's/.*# *//g' interps/$1 08:01:01 fizzie: what would you have it use instead? 08:01:38 * stalem suddenly remembers show n tell 08:02:10 i'm not sure we have/had that in norway 08:02:27 be glad you didn't/don't 08:02:59 `` sed -I 's{>/tmp/a.c &&}{|};s|/tmp/a.c|-x c -|' bin/cc 08:03:00 sed: invalid option -- 'I' \ Usage: sed [OPTION]... {script-only-if-no-other-script} [input-file]... \ \ -n, --quiet, --silent \ suppress automatic printing of pattern space \ -e script, --expression=script \ add the script to the commands to be executed \ -f script-file, --file=script-file \ 08:03:11 `` sed -e 's{>/tmp/a.c &&}{|};s|/tmp/a.c|-x c -|' bin/cc 08:03:12 sed: -e expression #1, char 37: unterminated `s' command 08:03:13 i don't even get the whole point of it. 'look at my stuff, this is what i own'? 08:03:22 ugh 08:03:44 `` sed -e 's,>/tmp/a.c &&,|,;s|/tmp/a.c|-x c -|' bin/cc 08:03:44 #!/bin/sh \ echo "$@" | sed 's/\\/\n/g' | gcc -w -Wfatal-errors -std=c11 -O2 -x c - -o /tmp/a.out && /tmp/a.out 08:03:57 `` sed -i -e 's,>/tmp/a.c &&,|,;s|/tmp/a.c|-x c -|' bin/cc 08:04:00 No output. 08:05:56 stalem: wikipedia suggests it's to "teach public speaking", which is admittedly something no one ever taught me afair. 08:06:04 `cc #include \ main(){puts("you ok?");} 08:06:06 you ok? 08:06:08 so maybe i should regret we didn't have it 08:06:17 I'd use that instead. 08:06:43 fizzie: haha good point! 08:07:15 oerjan: that's a pretty valid reason, but then again i must be the exception to the rule or that system doesn't work 08:07:34 imo there are far better ways to teach public speaking than showing off material things your parents got you 08:09:17 -!- AnotherTest has joined. 08:12:49 Our `cc is much less convenient than ##c's candide's ,cc, which does all kinds of intelligent stuff, like having an automated list of includes, handling #include without manual newlines, wrapping a main automatically if not present (even extracting other functions out of the wrapper if you have mixed functions and code), supporting input, automatically printing local variables and the last ... 08:12:55 ... statement's value if you don't output anything, and so on and so forth. 08:12:58 I mean, compare `cc #include \ int main(void) { printf("%zu", sizeof (int)); } vs ,cc size_t s = sizeof (int); which adds includes, wraps in main and auto-prints s. 08:13:24 how is candide implemented? 08:13:24 fizzie: you should improve it twh 08:14:00 @tell hppavilion[1] I should stop before I make someone very powerful* angry <-- maybe too late, according to /whois, walrus has a redhat cloak. 08:14:00 Consider it noted. 08:14:02 I don't think we need all of that, but maybe I could add the "automatically include standard headers" + main-wrapping parts. 08:14:15 `cat bin/cc 08:14:16 #!/bin/sh \ echo "$@" | sed 's/\\/\n/g' | gcc -w -Wfatal-errors -std=c11 -O2 -x c - -o /tmp/a.out && /tmp/a.out 08:14:34 ashl: Huge gobs of Perl parsing code with regexps + lots of plumbing to run the stuff in a separate qemu VM. 08:14:42 :| 08:14:52 Oh, and gdb for the "print locals and last statement" parts. 08:15:16 (You can also call into gdb from the code, to e.g. print types and such.) 08:15:34 Can't we just have candide join this channel? 08:15:42 That's what we do with all the other bots. 08:16:21 fizzie: `interp's cc already does a lot of that stuff, doesn't it? 08:16:34 `! cc printf("hi") 08:16:34 /hackenv/bin/!: 4: exec: ibin/cc: not found 08:16:40 `! gcc printf("hi") 08:16:40 /hackenv/bin/!: 4: exec: ibin/gcc: not found 08:16:47 `! c printf("hi") 08:16:50 hi 08:17:13 oerjan: Badly, but it does make an attempt, yes. 08:18:31 It broke for some things, although I don't exactly remember what. 08:20:04