00:06:47 -!- nooga has quit (Ping timeout: 272 seconds). 00:06:49 -!- aloril has quit (Ping timeout: 244 seconds). 00:20:04 -!- aloril has joined. 00:26:12 -!- nortti_ has quit (Quit: AndroIRC - Android IRC Client ( http://www.androirc.com )). 00:29:02 -!- aloril has quit (Ping timeout: 244 seconds). 00:36:37 -!- Tod-Autojoined has changed nick to TodPunk. 00:39:24 -!- Dovregubben has joined. 00:45:45 -!- aloril has joined. 00:47:39 -!- nys has quit (Quit: quit). 00:50:02 kmc: for (uint32_t i = 0; i <= UINT32_MAX; i++) { 00:51:04 uh oh 00:51:44 hah 01:04:48 -!- nys has joined. 01:15:59 shachaf: Psh 01:16:03 uint64_t 01:16:45 shachaf: Also wouldn't <= UINT32_MAX cause it not to terminate 01:17:22 FreeFull: Also that. 01:17:33 In addition to the other thing I meant when I pasted that snippet in here. 01:17:39 * oerjan thought that was the point 01:18:02 * shachaf did too. 01:18:38 * oerjan wonders if for (uint32_t i = 0; ; i == UINT32_MAX ? break ? i++) { will work properly 01:18:41 er 01:18:58 *for (uint32_t i = 0; ; i == UINT32_MAX ? break : i++) { will work properly 01:19:10 Huh? 01:19:15 break isn't an expression, is it? 01:19:33 i dunno, i guess if so it won't... 01:27:44 -!- DHeadshot has joined. 01:31:29 oerjan: uint32_t i = 0; do { ... } while (++i != 0) 01:32:06 -!- derdon has quit (Remote host closed the connection). 01:32:13 OKAY 01:32:38 Welp, you've stumbled upon the worst imaginable spelling of “OK”… 01:34:45 no i think elliott has clearly indicated he likes "O KAY" even less 01:36:21 i am thinking of this as a sequence of spellings and capitalizations ordered by loudness and sometimes sarcasm 01:36:46 ok okay OK OKAY O KAY 01:37:51 * oerjan shall have to start using OH KEY 01:38:26 *sobs* 01:38:37 too off key? 01:39:33 and specially for Gregor: o kaÿ 01:40:02 (for some reason my keyboard doesn't support capital ÿ) 01:40:30 O KAŸ 01:40:42 -!- mig22 has joined. 01:42:26 -!- aloril has quit (Ping timeout: 244 seconds). 01:59:34 -!- aloril has joined. 01:59:39 -!- ais523 has quit. 02:00:06 -!- mig22 has quit (Read error: Connection reset by peer). 02:00:23 -!- mig22 has joined. 02:00:25 -!- mig22 has quit (Client Quit). 02:47:58 O KAŸ 02:48:02 That looks ugly. 02:48:16 Besides, Ÿ is precomposed. I think. 02:48:23 Still looks ugly. 02:48:59 Gregor: if I ask you for suggestions about TV shows to watch, will your response consist entirely of the word "pony" repeated a few times? 02:49:15 I seem to like serious fantastical character-driven serials. 02:49:24 Well, I don't. 02:49:27 I like sitcoms and cartoons. 02:49:35 Ah. 02:49:45 So you like non-serious non-fantastical non-character-driven non-serials. 02:49:46 Gregor: Why sitcoms? 02:49:56 I grok cartoons just fine, but why sitcoms? 02:50:14 Wait, sitcoms are usually character-driven. 02:50:17 Good sitcoms are great, though bad sitcoms are abysmal. 02:50:17 At least, ELR is. 02:50:20 tswett: By the way, pony pony pony 02:50:24 Gregor: Okay, fair enough. 02:50:28 Corner Gas, Seinfeld, now I've been watching the Drew Carey show for some reason, … 02:50:42 Gregor: Sorry, usually when I hear "sitcoms" I think to bad ones, because there's so many. 02:51:21 And yet I don't for cartoons, even though there's undeniably *more* bad cartoons out there. 02:51:27 Weird. 02:51:41 Adventure Mouse! 02:51:50 (is a good one) 02:51:57 And the Adventures of Rocky and Bullwinkle! 02:52:08 Rocky and Bullwinkle is awesome. 02:52:10 And… AND… MY LITTLE PONY: FRIENDSHIP IS MAGIC! 02:52:19 Yes, that's pretty good too. 02:52:22 Rocky and Bullwinkle is perhaps the greatest cartoon show ever made. 02:52:24 pikhq_: do you have any suggestions? 02:53:30 tswett: Hmm. Serials. 'Fraid I don't watch enough *serious* stuff. I'd suggest Doctor Who, but it's anything but serious (usually). 02:53:45 And you've probably gotten the recommendation before. :P 02:53:52 Anyway, the Drew Carey Show is amusing because it's like they avoided jumping the shark by trying to jump the shark in EVERY EPISODE. 02:54:07 -!- MDude has joined. 02:54:09 Doctor Who is pretty episodic, isn't it? 02:54:56 tswett: Classic Doctor Who was divided into "serials", each consisting of a few episodes, each serial being a whole story. 02:55:00 It kinda fits, at least. 02:56:48 I like non-serious fantasical non-character-driven serials 02:56:48 New Doctor Who is rather highly episodic. 02:57:32 tswett, DS9 is very character-driven. 02:57:51 I still have yet to finish season 7 of DS9 02:57:54 It's been a while, hmm 02:58:09 Yeah, DS9 is a rather series fantastic character-driven serial. 02:58:12 I like plot-driven more, I think. 02:58:15 s/series/serious/ 02:58:24 What is Babylon 5 like? 02:58:58 Sgeo: It's a 5 season story arc that executive meddling fucked up. 02:59:28 Sgeo, I understand it's structurally very similar to DS9. 03:00:40 Sgeo, season 7's the last one, right? 03:00:51 In which case: did you stop before the 10-episode finale? 03:01:11 Mmm, I'd say that the similarities between Babylon 5 and DS9 start and end with “space station” 03:01:14 Season 7's the last one. I don't think I'm in the finale yet, but not sure. 03:01:52 Well this is neat, in that I somehow made a thing that sounds like some kind of chirping insect: http://entropedia.co.uk/generative_music/#b6409DQ0CjRrCmxszM00lQ10owz0dXQ8E0sydArLiwqQZIy0dTUKlEAAA%3D%3D 03:03:02 Hmm, now this is odd… the tag on my new pants claims they're dry-clean only, but the site says “Machine washable/dry clean recommended” 03:03:05 Gregor, um, isn't it character-driven and arc-heavy? 03:03:36 Phantom_Hoover: DS9 only became arc-heavy as a response to B5 ;) 03:04:09 ...so you're saying they have no similarities whilst acknowledging that one heavily influenced the other> 03:05:01 I'm basically made of lies. 03:05:19 Yeah but come on man, at least maintain internal consistency. 03:06:32 Gregor isn't externally consistent, how can he be internally 03:08:09 -!- mig22 has joined. 03:08:30 Phantom_Hoover: See, if you can be proven consistent you are therefore not. Gregor avoids this problem by not even trying. 03:36:05 -!- aloril has quit (Ping timeout: 250 seconds). 03:36:21 -!- Dovregubben has quit (Ping timeout: 248 seconds). 03:39:13 -!- Phantom_Hoover has quit (Read error: Connection reset by peer). 03:40:49 -!- MSleep has joined. 03:42:36 -!- MDude has quit (Ping timeout: 252 seconds). 03:49:20 -!- aloril has joined. 03:55:20 -!- Dovregubben has joined. 03:57:11 -!- segorev has quit (Quit: Leaving). 04:05:07 -!- MSleep has quit (Ping timeout: 240 seconds). 04:27:22 -!- oerjan has quit (Quit: leaving). 04:39:47 -!- TeruFSX has joined. 04:51:40 -!- mig22 has quit (Quit: mig22). 04:53:02 -!- mig22 has joined. 05:10:37 Gregor: Curse thee and thy ponies! 05:34:06 tell ais523 thanks for the info! http://en.wikipedia.org/wiki/Wikipedia:Help_desk/Archives/2007_October_10#Colored_text 05:34:11 @tell ais523 thanks for the info! http://en.wikipedia.org/wiki/Wikipedia:Help_desk/Archives/2007_October_10#Colored_text 05:34:12 Consider it noted. 05:34:45 -!- coppro has quit (Ping timeout: 248 seconds). 06:14:46 -!- kallisti has quit (Ping timeout: 252 seconds). 06:14:54 -!- asiekierka has joined. 06:27:50 -!- kallisti has joined. 06:27:50 -!- kallisti has quit (Changing host). 06:27:50 -!- kallisti has joined. 06:32:17 -!- TeruFSX has quit (Ping timeout: 244 seconds). 07:02:55 -!- nooga has joined. 07:03:34 -!- epicmonkey has joined. 07:36:27 -!- epicmonkey has quit (Ping timeout: 240 seconds). 07:39:22 -!- nys has quit (Quit: quit). 08:16:15 -!- jinmuxiao has joined. 08:25:57 -!- aloril has quit (Ping timeout: 244 seconds). 08:27:39 -!- jinmuxiao has left. 08:37:56 -!- Taneb has joined. 08:39:01 -!- aloril has joined. 08:42:05 Hello! 08:42:09 Oh no! 08:42:36 -!- epicmonkey has joined. 09:26:53 -!- Taneb has quit (Quit: Leaving). 09:28:59 -!- Taneb has joined. 09:29:18 -!- Taneb has set topic: Behold! The enchanting pants of Narcissus! | Just a remote control and some old gum | Help Taneb give up saying "Hello" | With creativity, anything can be a topic | except this | http://codu.org/logs/_esoteric/. 09:31:45 Great, another thing's became popular just after I get bored of it 09:31:48 :( 09:32:17 Why do have to endure the curse of the accidental hipster? 09:33:10 That makes... 09:33:12 Two things 09:33:16 Oh, it's not so bad 09:33:48 Taneb: You're giving up "Hello!"? 09:33:55 Trying to 09:34:01 * shachaf approves. 09:34:01 It's been in the topic since last night 09:34:08 monqy gave up hi and that was sad. 09:34:11 But now it's done. 09:34:17 monqy: Sorry. :-( 09:34:23 hi 09:34:23 monqy: You have 6 new messages. '/msg lambdabot @messages' to read them. 09:34:39 @ask monqy what are all those messages! 09:34:39 Consider it noted. 09:35:24 We may never know 09:35:51 they were good "don't worry" 09:36:30 monqy: "okay thanks" 09:50:18 -!- azaq23 has joined. 09:50:33 -!- azaq23 has quit (Max SendQ exceeded). 09:51:01 -!- azaq23 has joined. 10:48:30 -!- derdon has joined. 11:00:15 -!- Taneb has quit (Quit: TANEB AWAY). 11:07:23 -!- MoALTz has joined. 11:19:24 -!- mig22 has quit (Quit: mig22). 11:20:04 who here was thinking about money transfer protocol over ssh? 11:21:21 zzo38 11:43:09 -!- Phantom_Hoover has joined. 12:15:15 -!- MoALTz has quit (Quit: Leaving). 12:25:08 -!- aloril has quit (Ping timeout: 244 seconds). 12:28:15 -!- asiekierka has left ("Wychodzi"). 12:41:52 -!- aloril has joined. 13:16:03 monqy transfer protocol 13:16:09 hi 13:16:26 -!- boily has joined. 13:17:41 -!- azaq23 has quit (Quit: Leaving.). 13:19:46 shachaf: do you hilight on "hichaf" yet? 13:27:24 -!- monqy has quit (Quit: hello). 13:36:38 -!- MDude has joined. 14:14:09 -!- oonbotti has quit (Quit: oonbotti). 14:20:00 -!- DHeadshot has quit (Ping timeout: 276 seconds). 14:20:31 -!- DHeadshot has joined. 14:30:23 -!- oonbotti has joined. 14:42:00 what did pietbot do? 14:56:33 -!- aloril has quit (Ping timeout: 244 seconds). 15:02:15 -!- azaq23 has joined. 15:02:25 -!- azaq23 has quit (Max SendQ exceeded). 15:02:32 -!- Taneb has joined. 15:03:51 -!- azaq23 has joined. 15:05:17 -!- oonbotti has quit (Quit: oonbotti). 15:05:51 -!- oonbotti has joined. 15:08:07 -!- oonbotti has quit (Client Quit). 15:08:40 -!- oonbotti has joined. 15:09:35 -!- aloril has joined. 15:11:43 #/etc/passwd 15:11:43 root:x:0:0:Root Administrator:/root:/bin/ksh\nnobody:x:99:99:Unprivileged User:/dev/null:/bin/false\nwww:x:80:80:Web Server User:/var/www:/bin/false\nmessagebus:x:25:25:DBUS Daemon User:/dev/null:/bin/false\nhaldaemon:x:26:26:HAL Daemon User:/dev/null:/bin/false\njuhani:x:1000:1000:Linux User,,,:/home/juhani:/bin/ksh\n 15:20:18 -!- azaq23 has quit (Remote host closed the connection). 15:20:49 -!- azaq23 has joined. 15:21:11 -!- jlaire has joined. 15:33:12 -!- MDude has quit (Ping timeout: 244 seconds). 15:37:57 -!- Taneb has quit (Ping timeout: 248 seconds). 15:51:08 -!- soundnfury has joined. 16:09:46 nortti: probably "be written in piet" 16:10:41 -!- epicmonkey has quit (Ping timeout: 240 seconds). 16:30:48 -!- TeruFSX has joined. 16:35:59 -!- DHeadshot has quit (Read error: Connection reset by peer). 16:36:26 -!- DHeadshot has joined. 16:39:43 -!- ert3go has joined. 16:40:35 Hello , can anyone help me with brainfuck ? 16:42:38 quintopia: I'm under the impression it also did "connect to IRC", but I'm not entirely sure. 16:44:54 hi ert3go 16:45:01 hi quintopia 16:45:10 can I get help with brainfuck ? 16:45:27 everyone here is familiar with brainfuck 16:45:34 "most likely" 16:45:44 okay , I've tried to write a simple interpreter in C 16:45:49 mind taking a look ? 16:47:39 ok 16:48:26 quintopia, http://pastie.org/4450659 16:49:24 does it work? 16:49:28 what am i looking for? 16:49:57 I'm looking at magical constants 16:50:02 ert3go: I'd suggest omitting the printf("enter your input"); thing, as well as the default: case in the switch. 16:50:09 am i handling the case of nested loop properly ? 16:50:25 then how can an input be given pikhq_ ? 16:50:34 In Brainfuck characters other than [].,><+- are *comments*, not invalid. 16:50:46 As someone who's more familiar with C than brainfuck, I'd like to make a stylistic note or two 16:50:54 Your getchar (well, scanf("%c")) thing does that just fine. 16:50:56 i was gonna say what pikhq_ said 16:50:59 it's conventional to put a space between #include and the < 16:51:19 Regarding the brainfuck side, [ looks wrong. 16:51:20 and also either a space or a newline after the closing } of typedef struct {blah;} Stack; 16:51:43 [ is supposed to jump to matching ] when current cell is 0, not just be a pair for ]. 16:51:44 Hmm, yeah, it's not conditional. 16:52:01 in fact generally your coding style seems inconsistent 16:52:12 (it doesn't matter what it is, but it should be consistent) 16:52:30 ert3go: A [ should have basically the same condition as ] there. 16:52:52 there's no bounds-checking in the functions push() and pop() 16:52:53 i am confused pikhq_ 16:53:07 ert3go: if(*p) go to next instruction else go to matching ] 16:53:13 That's basically the logic [ has. 16:53:36 Or, equally, [ ... ] is while(*p){ ... } 16:53:40 doesn't ] just jump back to previous [? 16:53:48 lines 60-61 don't have the (presumably) desired effect in the case of two successive newlines 16:54:09 (it maybe ought to be while() rather than if()) 16:54:32 i am afraid I am confused . Can anyone modify the code ? 16:54:33 although, that line would be removed if you treated "invalid" characters as comments 16:54:34 soundnfury: Yeah, but he should just skip unknown characters *anyways*. 16:55:04 nortti: You can implement it that way if you want. But since the "previous [" (i.e. matching [) will just jump to one past the ] if !*p, quite commonly ] gets done as "jump to right after matching [ if *p, else nothing". 16:55:07 ert3go: No, we're kinda just telling you what you need to fix. We're finding bugs for you, not writing your Brainfuck interpreter. :P 16:55:55 ok then let me get it correct 16:56:12 if i encounter a [ , and if(*p) , then i should execute next instruction ? 16:56:21 more C issues: you're not bounds-checking 'p' (eg. take the program "<+", this will write to tape[-1]) 16:56:34 yeah true soundnfury 16:56:50 is that c++ comment on line 92? 16:56:50 soundnfury: I generally don't care about handling that in a Brainfuck interpreter though. 16:56:53 really you shouldn't use a char*p=tape; but instead a size_t p and use tape[p] 16:57:05 pikhq_: yeah but segfaults are bad 16:57:09 It's undefined behavior in Brainfuck. So, eh. 16:57:14 if i encounter a [ , and if(*p) , then i should execute next instruction ? 16:57:15 nortti: Also C99. 16:57:49 ert3go: yes and if you encounter a [, and if(!*p), then you should execute the instruction after matching ] 16:58:08 -!- MoALTz has joined. 16:58:33 so that's a while loop as someone correctly mentioned 16:58:45 line 87 may as well be putchar(*p) rather than using printf 16:59:01 and the puts() at the end should really put a '\n' 16:59:15 because of buffered stdouts and stuff 16:59:21 hmm 16:59:25 soundnfury: puts adds a newline. 16:59:43 but err , do i need a stack ? 16:59:57 fizzie: oh right 16:59:57 puts(somestring); is equivalent to printf("%s\n", somestring); 17:00:11 ert3go: yes unless you write recursive interpreter 17:00:11 * soundnfury doesn't know /why/ it does that, but you're right that it does 17:00:14 (fputs doesn't!) 17:00:52 ert3go: 'i' shouldn't be an int, but a size_t 17:00:59 okay 17:01:04 but i am confused 17:01:11 *when* to use stack ? 17:01:14 sizeOfFile really ought to be an off_t or perhaps a ssize_t 17:01:27 ert3go: use it as a call stack 17:01:35 line 52: DON'T cast the return value of malloc. Ever. 17:01:44 malloc returns a void *, which casts implicitly 17:01:47 but it returns a void * 17:01:56 Yes... so you don't need to cast it 17:01:59 push the location of [ to a stack when encountered during execution and while searching for the matching ] 17:02:04 soundnfury: except when you want C++ compatibility 17:02:05 hang on, there's a C-FAQ about this 17:02:09 jlaire: fuck C++ 17:02:11 srsly 17:02:14 :D 17:02:17 lol 17:02:28 yeah. fuck c++ 17:02:39 sounds like an orgy 17:02:46 http://c-faq.com/malloc/mallocnocast.html 17:03:14 that doesn't convince me 17:03:18 " The compiler is likely to assume that malloc is a function returning int, which is of course incorrect, and will lead to trouble. " 17:03:20 that's what's broken 17:03:27 location of i should be pushed only if (*p) ? 17:03:55 at least the compilers I've been using lately warn about that 17:04:04 implicit declarations of functions 17:04:18 nortti: I've seen quite a few non-recursive no-stack brainfuck interpreters. The "common" inefficient way to implement [/] is just "seek to matching ] or [ when necessary", and you don't need a stack for that, just a counter of current depth. 17:04:26 or can someone atleast tell the pseudocode for case '[' : ? 17:04:34 ert3go: why don't you just look at other interpreters if you can't see how to handle []? 17:04:43 it's no less "cheating" than asking us 17:05:03 quintopia, the other interpreters are little confusing , i'd like to try without seeing them 17:05:07 jlaire: well, if you want to take up the argument, ask about it in comp.lang.c. While wearing flame-proof clothing ;) 17:05:18 ert3go: then try without asking too :P 17:05:25 soundnfury: I've seen this particular point debated a number of times there o_o 17:05:28 please quintopia , i really need help :( 17:05:40 most folks on comp.lang.c seem to hate c++ with a passion 17:05:47 with good reason 17:05:50 jlaire: what point? 17:05:57 nortti: casting malloc's return value 17:06:02 oh 17:06:03 to (char *) or (int *) or whatever 17:06:21 ert3go: If you don't want to do any work in advance, a common [ is "if *p, go to next instruction; else { depth = 1; while (depth > 0) { go to next instruction; if [, depth++; if ], depth-- } }". 17:06:29 I think you should probably not do it unless you find very good reason to do so 17:06:59 (Though it's also quite common to do a first pass to find all matching [] pairs and put them somewhere, so that it doesn't need to scan forward for the matching ].) 17:07:15 sometimes compatibility with C++ is a good reason, sometimes not; it just irks me to hear C-programmers saying you should *never* cast void* to another pointer type 17:07:36 (You can't use a stack maintained during execution to find the matching ], since it could be anywhere in the future program where you haven't even been in yet.) 17:07:43 * FreeFull casts (void *) to (void **) 17:07:44 What now 17:08:50 jlaire: I say again, fuck C++. 17:09:19 jlaire: It's bad style in C. And that's it. 17:09:19 soundnfury: you can say that again 17:09:28 pikhq_: why is it bad style 17:09:34 Can we pretend C++ doesn't exist 17:09:36 soundnfury: (I like C++) 17:10:08 jlaire: For the reasons already mentioned in the comp.lang.c FAQ? The C++ compatibility aspect doesn't magically make it good style. 17:10:24 fizzie: the FAQ itself says that compilers warn about it 17:10:39 if that's the only argument against it, it's a very weak one 17:11:17 jlaire: It's god damned clutter and noise is all! 17:11:43 pikhq_: now that is something I'll accept 17:12:10 jlaire: Maybe in C++ the more characters you type the better, but down here we believe in characters *spent*. :) 17:12:39 pikhq_: that's not the C++ I know 17:13:27 Pssst, Boost. 17:13:50 ok. 17:14:07 * jlaire says no more 17:14:27 Just because (some) compilers are "increasingly likely" to warn about implicit declarations doesn't make the "cast hides an important warning" argument meritless either; especially since (considering only C here) it doesn't really have drawbacks either. 17:15:04 fizzie: the few compilers I've used have all warned about it 17:21:38 Perhaps you've been using too good/new compilers, then. Anyway, GCC only warns in the special case of function it knows about, by default. If you're in the habit of casting void *'s from malloc, you'll probably cast them from zqalloc too, and that'll cause problems when you forget to #include "zqlib.h". 17:21:43 (Hypothetical library.) 17:22:03 (You need -Wall or -Wimplicit-function-declaration to get warnings for all implicit declarations.) 17:22:14 -!- nys has joined. 17:22:15 I use -Wall 17:22:51 Not everyone does. 17:23:21 I think I need to do some eating. 17:26:58 http://patriciopalladino.com/blog/2012/08/09/non-alphanumeric-javascript.html 17:27:35 -!- epicmonkey has joined. 17:32:44 I use -Wall -Werror -ansi -pedantic 17:32:48 'cuz hell yeah. 17:33:04 I replace -ansi with -std=whatever 17:33:11 but otherwise, highfive 17:33:26 I like to go with c89 because I'm hardcore like that. 17:33:31 Features are for losers. 17:34:21 I don't usually write C 17:34:26 but when I do, I write C89 17:34:38 -!- azaq23 has quit (Ping timeout: 248 seconds). 17:44:32 -!- Taneb has joined. 17:45:14 fungot 17:45:15 Taneb: no, the boy, do i? we're nazis, but we are not, ja, of the house! the entire earth has surrendered to you, our helpful. ' is phone is still workin' class, guv. 17:45:19 -!- coppro has joined. 17:45:25 iwcs 17:45:28 ^style 17:45:28 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 17:45:33 Oh, I'm good 17:45:39 I probably set it as that 17:45:49 ^style alice 17:45:50 Selected style: alice (Books by Lewis Carroll) 17:45:53 fungot 17:45:54 Taneb: let " persons" be universe; fnord fnord y="that can sit still. the professor took it as a general law, who do not drink coffee."' 17:46:05 fungot 17:46:06 FreeFull: " all fnord are uncanny ( fnord all fnord are equal', i suppose. if not, why do things have names at all?' alice ventured to remark. 17:50:29 -!- Eladith has quit (Quit: leaving). 17:56:05 Now why did I write yet another Brainfuck implementation? 17:56:22 I guess talking with someone about it compelled me. Bad damned habit there. 17:56:31 It's not even an interesting one. 17:56:58 Practise? 17:57:22 I'd hesitate to call such a thing practice. If it were I would've at least, y'know, parsed. 17:57:25 :P 17:58:10 Make your own language 17:58:19 I have! 17:58:29 Make your own speaking language 17:58:31 This was literally a waste of 20 minutes. 17:58:51 Gregor: Where's your -Wextra, huh? 17:59:07 Or "-W" for the oldskool way of spelling it. 17:59:27 "-Wall" -- the worstly named option? 17:59:35 -Ceiling 17:59:39 (Since it's so far from 'all'.) 18:00:00 If I ever write a compiler, it will have -Ceiling and -Floor options 18:00:09 About the only thing it has going for it is simplicity. 18:00:43 What language 18:00:52 C 18:01:19 Taneb: my compiler has -AnotherWall 18:01:32 pikhq_, that's... the third letter of the alphabet 18:01:35 IT MUST BE A CLUE 18:01:40 Hmm... 18:01:42 "Clue" 18:01:48 That's the name of 2 esolangs... 18:02:10 "2", that's precisely the number of dimensions Befunge runs in! 18:04:16 fizzie: Oh that's it. 18:04:29 fizzie: I'm goin' back and making Fythe compile with -Wall -Wextra -ansi -pedantic now. 18:04:51 Gregor: why not -Werror? 18:05:02 Whoops, missed -Werror 18:05:03 That too. 18:05:22 If the policy is "no warnings are acceptable", -Werror is a bit superfluous. 18:05:29 Oh yeah, because it complains about unused parameters. That's stupid. 18:06:00 -!- asiekierka has joined. 18:06:05 You can IIRC quiet those with a "(void)a;" statement. 18:06:29 At least that's a common thing, I don't recall if it helped on GCC. 18:07:14 you can use unnamed parameters, int f(int x, int b, int) 18:07:15 * jlaire ducks 18:07:39 There was a commonly-ishly done warning-silencing trick that on GCC had the opposite effect, it just caused new warnings. (With sufficiently high warning levels.) 18:08:22 I think (void)a works to remove unused params warnings, but (void)(a,b) will cause a no-side-effects warning 18:09:28 -!- Taneb has quit (Quit: AWAY). 18:10:49 OK, I'll go with -Wall -Wextra -Werror -Wno-unused-parameter, because the last one, although understandable, is just silly. 18:10:56 -!- asiekierka has quit (Remote host closed the connection). 18:11:24 yeah, the unused parameter warning is not very useful 18:11:26 nah, -Wunused-parameter is fine 18:11:34 -!- pikhq has joined. 18:11:49 -!- pikhq_ has quit (Ping timeout: 244 seconds). 18:11:52 it seems reasonable to me to decorate your function with __attribute__((unused)) if you really do want an unused parameter 18:12:28 if you only want to compile on gcc, sure 18:12:34 "Reasonable" to decorate your otherwise pure-ANSI... right. 18:12:49 Anyway, you could have an UNUSED macro to do that. 18:12:49 of course things like that are hidden behind macros 18:12:59 UNUSED_PARAM or some such 18:13:04 right 18:13:07 And autoconf it up? 18:13:08 well, for other compilers you can pass -D__attribute__\(\(unused\)\)= 18:13:18 pikhq: no, never autoconf 18:13:32 soundnfury: You can't have a macro with (( in the name. 18:13:34 because autoconf is horrendous 18:13:38 fizzie: oh right 18:13:43 other way round then, with an UNUSED macro 18:13:45 just #ifdefs 18:13:48 like you said 18:13:57 soundnfury: It does suck, but it sucks less than most alternatives when you need/want to check for certain features. 18:14:22 * soundnfury prefers custom Makefile addenda 18:14:43 typically by having something like "-include config.mak" strategically placed in the Makefile 18:14:56 and cunning use of deferred assignment 18:14:56 Also, for the love of all that's sane don't cargo cult autoconf... 18:15:12 why? 18:15:31 i.e. don't use autoscan, it just makes Bad autoconf files. 18:15:47 soundnfury: why? 18:15:51 anyone compiling from source is intelligent enough to know what to put in the config.mak 18:16:15 well yeah 18:16:15 as for distro packagers, it's their whole /raison d'etre/ to deal with that kind of thing 18:16:27 autoscan basically generates an autoconf file that checks for every god-damned ISO C feature. 18:16:46 AC_PROG_CC_C99 and we're done. 18:16:51 ordinary package users shouldn't have to run a script that tests that the cc creates runnable executables, ffs 18:16:53 -!- ert3go has left ("Leaving"). 18:16:57 pikhq: I once found one that checks for fortran 77 features of my system 18:17:34 soundnfury: Part of that is autoconf being overly complex in a few areas, but most of that is that people have no clue how to use it. 18:17:35 Anyway, since some other compilers will need the "(void)a;" thing, if you do want the __attribute__, you'll end up with something like void f(UNUSED_PARAM int a) { UNUSED(a); ... } where GCC gets #define UNUSED_PARAM __attribute__ ((unused)) #define UNUSED(x) /* empty */ and the others get #define UNUSED_PARAM /* nothing */ #define UNUSED(x) (void)x, or the like. 18:18:41 which is just so much clearer than void f(int) { ... } 18:19:24 jlaire: yeah but usually you shouldn't /have/ functions with unused parameters 18:19:29 unless there exists a good reason 18:19:40 in which case maintainability demands that you draw attention to it 18:19:42 soundnfury: Such as writing a callback. 18:19:48 the exception is for callback-heavy code 18:19:50 jlaire: that requires c++ 18:19:56 Which is used less than it should be in C. 18:19:56 olsner: my subtle point, yes 18:20:13 like if you're writing a GTK+ gui 18:20:22 I wish there was Go but without garbage collection 18:20:42 though in that case I /still/ choose to leave -Wunused-params active, and write the __attribute__s 18:22:15 * soundnfury quite likes that void f(int) { ... } approach 18:22:22 and is amazed to learn that C++ gets something /right/! 18:23:17 They have more of the "implement this inherited thing with a fixed prototype" kind of thing going on, perhaps there was more call for such a feature. 18:23:20 c++ also adds single-line comments and mixed code and declarations, those are nice 18:23:42 olsner: C also has those. 18:23:50 However, the IE6 of compilers doesn't implement it. 18:23:57 yeah, but it's new for C99? 18:24:07 What's the IE6 of compilers? 18:24:08 ... Dude, C99 is *13 years old*. 18:24:12 Sgeo: Microsoft Visual C++. 18:24:18 Sgeo: They don't support C99. 18:24:20 At all. 18:24:23 It's got "++" in the name. 18:24:54 They added C++ support to it, and let the C frontend go untouched for over a decade. 18:24:56 C99 is obsolete, long live C11! 18:25:12 It still functions as a C compiler, but it was last updated in the 90s. 18:25:22 The C compiler in it is just for compiling old code, after all. 18:25:27 (That's what they said.) 18:25:34 yeah, who would write c these days? 18:25:54 With Microsoft's attitude, nobody on Windows. :P 18:26:05 because c is good programming language 18:26:22 Also, Microsoft's @#%@@ API support. 18:26:57 Which means any code on the platform is basically doing #if WINDOWS ... #else ... #endif 18:27:12 s/any/any portable/ 18:27:31 when I have to port something to Windows, I just cross-compile with mingw on my (linux, naturally) box 18:27:50 and then throw DLLs into the tarball until it works :/ 18:27:51 butts 18:28:27 Also, whoever did the LPSTR nonsense deserves to be shot. 18:28:39 LPCTSTR 18:28:40 soundnfury: I used to use mingw msys. then I no longer ported anything to windows 18:29:14 Good old TCHAR stuff. 18:29:29 with all that "long pointer" support in the windows APIs, they should've done their x64 stuff using two pointer sizes 18:30:16 WPARAM wParam, LPARAM lParam. 18:30:48 Also, whoever's responsible for *not supporting UTF-8* should be beaten. 18:31:32 If you're using a char* on Windows it's a legacy charset. 18:32:20 You only get Unicode with wchar_t, which is UTF-16, which is what Microsoft calls "Unicode", because they're a bunch of mouth-breathers. 18:32:51 both microsoft and java suffer from being "early adopters" 18:33:14 oh, wparam and lparam... reminds me of working with a code base that decided to build their own message passing framework where every callback gets one INT16 and one INT32 parameter 18:33:15 that's not to say that they should've changed things in the past 15 years :| 18:33:20 shouldn't have* 18:33:25 Java at least has the benefit that their *normal APIs* support Unicode, and it "just works". 18:33:33