00:00:00 --- log: started forth/05.06.15 00:02:52 --- quit: nothingmuch () 00:37:04 --- join: nothingmuch (~nothingmu@212.143.92.226) joined #forth 01:07:04 --- quit: alexander_ (Remote closed the connection) 03:24:16 --- quit: saon|smgl ("Lost terminal") 03:25:27 --- quit: saon ("Lost terminal") 04:22:24 --- join: Raystm2 (~vircuser@adsl-69-149-42-251.dsl.rcsntx.swbell.net) joined #forth 04:34:36 --- join: saon (1000@c-24-129-91-106.hsd1.fl.comcast.net) joined #forth 04:38:28 --- join: saon|smgl (~saon@c-24-129-91-106.hsd1.fl.comcast.net) joined #forth 04:54:30 --- join: saon_ (1000@c-24-129-91-106.hsd1.fl.comcast.net) joined #forth 04:57:31 --- join: saon__ (~saon@c-24-129-91-106.hsd1.fl.comcast.net) joined #forth 05:03:03 --- quit: saon|smgl (Read error: 60 (Operation timed out)) 05:11:09 --- quit: saon (Read error: 110 (Connection timed out)) 05:11:59 --- nick: saon_ -> saon 05:12:07 --- nick: saon__ -> saon|smgl 05:31:10 --- join: PoppaVic (~pete@0-2pool236-226.nas22.chicago4.il.us.da.qwest.net) joined #forth 05:31:26 Mornin' 05:50:42 --- quit: KB1FYR (Read error: 110 (Connection timed out)) 05:56:49 --- quit: nothingmuch () 05:57:12 --- quit: Frek ("Leaving") 05:57:31 --- join: anvil_ (~anvil@h208n2fls31o815.telia.com) joined #forth 06:03:43 --- nick: anvil_ -> Frek 06:55:55 --- join: true-grue (true-grue@ppp143.medlux.ru) joined #forth 06:56:05 hi 06:56:16 g;day 06:56:51 Greetings. 07:00:08 Can anybody tell, which Forth is more popular today? in C and used as built-in scripting language 07:00:20 I doubt it 07:00:36 Forth has a lot more variants and a somewhat diff mindset than C 07:00:45 You could take a look at FICL. 07:00:59 It's used in FreeBSD. 07:01:00 ficl and ATLAST come to mind, yeah - so does gforth 07:01:11 madwork: really? where? 07:01:39 At boot, I think. 07:01:44 * madwork checks. 07:01:56 interesting.. 07:02:00 Yeah,its used at boot. Completely stupid imho, but whatever 07:02:29 I guess I never knew because macosx is BSD based, rather than freebsd 07:02:34 There's forth code in /boot :) 07:02:49 not remotely stupid, it's compact, complete access 07:02:51 arke, why is it stupid? 07:02:58 ok. So ficl & Gflorth, right? ...I wrote stack language similar to Forth and Postscript in C.. Very simply, fast, and with small src. 07:03:01 It seems redundant 07:03:03 :D 07:04:06 gforth - GNU forth is written in C 07:04:09 i been seen FICL.. think is TOO big and complex.. both sources and language itself 07:04:20 true-grue, I've also got a Forth-like in C. 07:04:27 ficl is a tool to get back and forth from a forth-world and a C world 07:05:01 ficl is prolly overkill, but so is a full forth-system 07:05:48 true-grue, why not just whip off your own little F83 in C? :) 07:05:57 madwork, is it standatd Forth or have unique things in it? 07:06:32 As a Rule - and this is 20 years-back Experiences - Forth is better than learning asm. And C is the current 'portable' for modern systems. Now, reconciling them? that's a whole world different. 07:06:41 My forth? It's rather unstandard. In fact, there's no actual language to it. You define that however you like. it's more like a Forthy VM. 07:06:56 Ahh, Tile Forth is a Forth-83 in C: http://www.ptf.com/ptf/products/UNIX/current/0307.0.html 07:07:27 madwork, Well, i need a language for music description.. I don't need standard Forth for it.. So, i wrote my own, simple stack language. 07:07:50 many folks do - so what? 07:07:52 Here's another location for Tile Forth: ftp://ftp.lysator.liu.se/pub/languages/forth/ 07:08:34 true-grue, ahh. So what is it you want a popular C Forth for? Or are you just curious? 07:11:06 madmwork, Just curious :) My musical stack language have some interesting concepts, it's very small, powerfull. I publish it on sourceforge.net, if anyone come to love it 07:12:32 Do that! :) I'd like to take a peek at it. 07:12:49 I have also been thinking about musical applications of Forth, for retro video games and such. 07:13:51 madwork, thanks! can you give me your email? i will send you sources in days(after some polishing) 07:13:58 all you need is the ABI and a new voc 07:14:27 PoppaVic, ABI? What is it? 07:14:52 application binary interface - the driver 07:15:30 then, words in the new voc will either generate data for other words, or speak via the abi/driver to the 07:15:54 API is how, ABI is what 07:16:49 In my case, I'd use a new vocabulary and words to compile musical data structures which would be played by the game engine. 07:17:13 yep 07:17:14 --- quit: Frek ("Leaving") 07:17:40 the new voc is the API, the underlying out/in r/w is the ABI 07:18:16 --- join: anvil_ (~anvil@h208n2fls31o815.telia.com) joined #forth 07:18:33 * madwork is not a fan of buzzwords and acronyms. ;P 07:19:04 great, meanwhile - I live in C, dude ;-) 07:19:16 So do I. 07:19:34 C as in C language, and Canada. ;) 07:19:35 in C, the idea of API and ABI should become 2nd-hand 07:20:17 I've honestly never heard "ABI" before... "driver" works fine for me. :) 07:20:33 --- nick: anvil_ -> Frek 07:20:47 Yeah, I had to dig to learn ABI... ABI relates directly to our friggin' sizeof() and MAX* whatever 07:21:15 madwork: hence, my minish and Metabuilder is gonna' be one busy SOB 07:21:55 Minish.. Zelda: The Minish Cap, on Gameboy Advance? 07:21:58 ;P 07:22:38 no, mini-shell and metabuilder - my "I am sooo PISSED" project to get past autoshit. 07:23:51 any of you do win32 programming ? 07:24:07 the minish will NOT be a GP language and will NOT provide looping. The idea of "pick a shell" is bypassed, and the idea of "use the environment" (which can be limited) is also bypassed 07:24:13 hell no 07:24:25 I wouldn't even port my shit to doze - let a dozer port it. 07:24:34 Frek, I do some. 07:25:05 madwork: is it necessary to flush caches or anything before executing code you build on the fly ? 07:25:26 Frek, dunno, I've never done that. 07:25:36 ok ok 07:25:38 thanks anyway 07:25:45 * madwork has mostly worked with Win32/MFC etc. :-/ 07:25:52 gotcha 07:26:03 Frek, you might want to ask crc in #retro about that. 07:26:21 dynamic code?? man, yer asking for an assbite 07:26:23 madwork: ok, I'll keep that in mind. 07:27:01 --- join: nothingmuch (~nothingmu@yovalk.bb.netvision.net.il) joined #forth 07:27:11 * madwork has been wanting to look into runtime code assembly as well. 07:27:24 PoppaVic: just decided to put some work into an old project I had going on linux some years ago, but lost sources for :( ; so Instead I'll try target OS X and Windows32 this time 07:28:10 I use macosx, and the idea of KNOWING I just installed something that flauts data vs code and generates dynamic code is just scarey 07:28:33 why is that ? 07:28:54 because I don't trust folks that much 07:28:55 java does it all the time, most emulators does it too etc 07:29:05 great, and I never use java 07:29:09 --- quit: true-grue () 07:30:13 Holon forth is pretty funky. 07:31:00 --- join: sproingie (~chuck@64-121-15-14.c3-0.sfrn-ubr8.sfrn.ca.cable.rcn.com) joined #forth 07:31:31 madwork: how so? 07:31:36 * arke has never used it.. 07:33:10 http://www.holonforth.com/index.htm 07:33:41 Well, mostly it's the editor debugger. It's a hyperlinked IDE. 07:33:50 "hyperlinked"? 07:34:18 Oooh cool! 07:34:19 Yea... which is perfectly suited for forth code. The code is navigatable, sort of like a wiki. 07:35:09 I've been dreaming up something similar, and it's nice to see the concept in play. 07:35:24 more "for-pay" forth? 07:35:31 yeah 07:35:40 Some of it is free. 07:35:41 yup. Uninterested 07:35:56 Forth "for pay" might as well be Doze 07:36:01 * madwork is interested in concepts, not purchasing. 07:37:48 I really find the "for pay" idea of Forth stupid. 07:38:12 --- quit: saon (Read error: 110 (Connection timed out)) 07:38:15 --- quit: saon|smgl (Read error: 110 (Connection timed out)) 07:38:18 factor has a webby browser front end that links every word 07:38:18 The only time I wouldn't mind seeing something in Forth for pay is if it really is a very large and complicated program 07:38:21 :D 07:38:24 Why? People have to make a living somehow. 07:38:34 i believe it supports editing, though i've never managed to activate it 07:38:36 ns 07:39:09 madwork: it's true in C too, and forth was grassroots. So, paying for forth lately is just about stupid.\ 07:39:26 A very large and complicated Forth program is kind of an oxymoron, isn't it? At least in spirit... 07:39:27 An end-user APP, a turnkey, written in Forth - THAT is another issue. 07:39:40 any app can be large and complicated 07:39:42 PoppaVic, sure. 07:39:45 madwork: rr, sgiagoborgbogbu 07:39:48 what PoppaVic said 07:39:49 :D 07:40:08 Plus, I live on freenode for a reason. 07:40:28 How about donations then? Or is money just plain BAD? 07:41:06 * madwork goes to join the Forth mountain monks, casting off all material wealth... 07:41:07 ..I don't do homework, or save someones Job. And I am all for helping anyone writing GNU LP or P licenseing,let alone freebsd, etc 07:41:09 * arke remembers the following happening all the time: bar bar? sure? foo err, yeah, I meant what PoppaVic said 07:41:30 :D 07:42:37 I've ben coding for decades, I never get paid, and I enjoy it. I'd LOVE to be paid, even more - wealthy. But I have no inclination to save dufii's jobs or grades. 07:43:13 :) 07:43:27 off to grad practice 07:43:59 and I love helping folks in ##C, but - again - I refuse to write for them to pass a class, save a job, or do their research. 07:44:10 Wurd. 07:45:00 my "save my job" questions are usually like "why is XYZ mysteriously breaking, and where can i get a chainsaw at this hour and what's the home address of the person who wrote XYZ?" 07:45:03 Of course, this makes me unpopular - almost as much as suggesting forth alternatives or that C needs a less-than-ivory-tower release. 07:45:25 * sproingie has comments in his latest python project like # bring me the freshly polished skull of Guido Van Rossum 07:45:53 damn thing uses threads, and is now losing state on global objects. in one thread, a property is populated, in another it's blank 07:45:59 yep, I looked at other lang. C suits me pretty well, Forth for interaction, and I'd LOVE to reconcile the two 07:46:47 isn't Python the one uses tabs for blocks? 07:46:56 yep. that's the least of its damage 07:47:14 nai: the tabs were simple, like a Makefile 07:47:28 Folks bitch because it isn't like {...} 07:47:32 it's probably my fault. somehow this language has me making mistakes other languages don't 07:47:53 the indent thing i can get used to. the mysterious and idiotic scoping rules i cannot 07:48:05 I was interested in python, but I can work confidently in C and gforth - so it offers me nothing of interest. 07:48:16 get this: if you reference a variable in an outer scope, it's fine as long as you read it 07:48:26 ok. 07:48:27 the moment you assign it, it creates a NEW scope for it 07:48:33 ahh, that's dumb 07:48:52 oh my yes. but that's only for globals. in closures, you can read them as normal 07:48:59 and only read. it throws an error if you write 07:49:00 OK, lemme throw my favorite pony in here, since ##C already knows it... 07:49:30 * sproingie should never have converted the code from perl, but perl's threading model is *really* weird 07:49:43 Why the hell are we NOT evolving, advancing in such a way that A xlates to B xlates to C and hence to ASM local to the machines? 07:50:12 PoppaVic: eh? plenty of languages have several levels of translation 07:50:12 We have too many haqueers all going their "own way" 07:50:34 ocaml compiles to C 07:50:38 sproingie: I know of C->asm->mc - now, what runs that path? 07:50:53 ok, that's ABOVE C, I presume? 07:51:19 of course. haskell also compiles to c, though it's also added a c-- backend 07:51:26 ahh 07:51:42 c-- is more well suited for backends than c 07:51:48 so, haskell can generate C; and so can ocaml - which 'C' flavor? Any more? 07:51:56 presumably ansi c 07:52:07 ahh, ansi isdeliberately lamed. 07:52:18 it's C. of course it's lame 07:52:23 nope 07:52:26 ok question 2 then, any of you know of an extremely rare x86 instruction that's virtually never used ? 07:52:30 I said "deliberately lamed" 07:52:37 Frek: HCF - Halt and Catch Fire 07:52:42 heh 07:52:43 Frek: wtf for? 07:52:56 Frek: the BCD instructions tend to be pretty rare 07:53:02 kill the away or get ignored 07:53:13 sorry forgot about it 07:53:16 anyway 07:53:26 I need to tag the code I generate 07:53:44 bad idea to consider an opcode 07:53:53 i suspect the pascalish call/ret equivalents, forgot what they're called, are pretty rare 07:53:59 i wouldn't use an opcode tho 07:54:06 I can't use anything else 07:54:07 use a struct 07:54:11 bullshit 07:54:13 can't do that 07:54:23 still bullshit 07:54:26 what's the purpose of this "tagging"? 07:54:43 make a call to a 'tagger' - like forth nest/denest 07:54:54 i'm even more confused 07:54:58 I use "sc" on the powerpc; it's a supervisor only instruction that would never occur in userlevel mode (and if it's encountered it will cause an exception) 07:55:12 oh goody 07:55:13 you are insane 07:55:30 why not just use some made up number then, it'll generate an illegal instruction trap 07:55:33 sproingie: basically I'm using the compiler to generate the platform dependent code; by inserting small volatile assembly macros I can seperate the segments 07:55:47 ok, that's it... 07:55:49 Frek: and you're using instructions for this. astonishingly lame idea 07:56:00 so much for that noise 07:56:01 sproingie: it's the only way 07:56:17 i doubt this 07:56:23 It's GOT to be nationalistic-stupidity 07:56:35 Frek: no other compiler needs this 07:56:54 perhaps this is a usage of "segment" i'm unfamiliar with 07:56:59 I had it going before; although then I used pointer to labels to "cut up the code" but this compiler don't support that gcc extension 07:57:16 sucks to be you 07:57:18 this is either object file "sections" or memory segments 07:57:40 sproingie: ignore, it's really relaxing 07:58:52 Frek: just insert 0xDEADBEEF and a jump over it right before. or something like that. of course if that happens to actually occur in the code, you're screwed 07:59:21 sproingie: think of it like a clip book; say I have "void cutbook(){ asm("noncommon instruction"); vREG1++; asm("noncommon instruction"); vREG2++; asm("noncommon instruction"); ...}" 07:59:49 the clipbook metaphor doesn't seem to be taking hold with me 08:00:39 using these noncommon instruction (supervisor instructions for example, I can enumerate the actual routines, and using memcpy I can copy it into a newbuffer and that way generate an entire new program at runtime, do you follow ? 08:01:02 this is a C program, no? 08:01:05 yes 08:01:24 why not just walk the symbol table and do it that way? swipe the relocation code from the linker if you have to 08:01:37 but it's purpose is to demostrate how to build a portable recompiler. 08:01:56 by using a grossly incompatible hack? 08:02:05 incompatible how ? 08:02:07 to say nothing of unreliable 08:02:24 as I said; I've done it before with success 08:04:08 but then I used pointers to labels to "cut up" the generated code into smaller segments, I'm testing a new way now. I could of course write a single function for all operations but that would end up being more of a function pointer list than an actual "compilation" 08:04:11 portable way to do it is to go into the symbol table. should be portable to any ELF system 08:04:14 same principles apply to PE 08:04:55 yes but it requires knowledge of the file format; I want to make it as automatic as possible; ie. at most I should flush the instruction cache 08:05:45 back 08:06:25 you're meddling with blacker magic than I. all i can tell you is that inserting short "magic" delimiters will break the instant they actually get used in real code 08:06:39 as a test, try using this trick on the code that implements this trick 08:06:54 'markers' need a checker - complicating shit w/o a reason. 08:06:55 sproingie: it's not supposed to be reused 08:07:09 i could tell you to keep a table of functions getting this treatment and their offsets 08:07:15 but the linker already did it for you 08:08:02 why the FUCK is he fighting a call??? 08:08:27 best advice i can give then is to just use some sequence that's an illegal instruction 08:08:33 it's rather unlikely to be generated 08:08:44 it'll screw the compile 08:08:59 sproingie: well a supervisor instruction would be the best. 08:09:19 yer still insane, but I expect that from fornats and frogs 08:13:37 http://www.intel.com/design/pentium4/manuals/245471.htm 08:13:42 there ya go, pick one 08:13:45 this is sort of what I mean... http://paste.lisp.org/display/9139 08:14:40 anyway I didn't ask for opinions about this project; I'm very well aware it works already as I've done it before. however I asked if anyone knew of an uncommon instruction (possibly supervisor only instruction) that's all. 08:14:45 Like I said waay back: use a call, check the stack - tweak as required, return. 08:15:18 No sane person would use your "approach" 08:16:45 once again I didn't ask for your opinions, I just ask if you happen to know an instruction, but Instead of simply answering yes or no, you start an entire investigation regarding it being sane or not... 08:16:59 keep jawing 08:17:35 use a BCD instruction then, possibly an illogical combination of BCD instructions 08:17:41 those are rare 08:17:46 sproingie: ok, thanks 08:17:52 unless of course you're writing financial apps 08:18:11 you cannot insure SOMEONE isn't using them 08:18:18 most of those still don't use BCD. you never know of course 08:19:09 a call & return is dead easy. an interrupt isn't, and he'sstill talking trash 08:19:15 no, if it's not obvious already (besides I said it twice I think) it's a concept test. then again I don't give a fuck if other apps outside mine uses this instruction; it's only going to affect a single routine in this application 08:19:26 PoppaVic: are you just plain right dumb or did you even look at the example I made + 08:19:26 ? 08:20:04 looking to jump into the middle of a sub? use duff's device 08:20:27 nope. I ignored you about the 3rd or 4th time you talked crap. setjmp/longjmp are documented as well 08:20:34 tho i suspect optimizers make a dogs breakfast out of duffs device these days 08:20:46 PoppaVic: well by that you just admit you didn't understand a shit of what I said so far 08:21:10 sproingie: duffs device has no use here though; it's function local, I need to access the entires from outside the function 08:22:42 sproingie: have you ever written an emulator ? 08:22:51 naw, friend of mine did 08:23:05 gameboy emulator 08:23:07 ok, but you're aware how you can use function pointers for the actual emulation 08:23:09 ? 08:23:26 usually some sort of opcode table, yes 08:23:33 like opcodetable[instruction&kOpcodeMask](instruction); ? 08:24:05 sure 08:24:33 * sproingie usually uses a Big Honking Switch statement, the compiler does a better job at it 08:24:56 step two to make it slightly more efficient would be to simply memcpy all the instructions into a long codebuffer, thus reducing some jumps, right ? 08:25:10 QEMU does this sort of 08:25:26 memcpy and executing code in data is not my idea of efficient 08:25:35 no ? 08:26:27 if one is looking to preserve locality... well, there's always goto 08:26:54 i suppose that doesn't work from the outside. of course by definition that's not local 08:27:21 gcc does a bangup job with computed goto tho 08:27:36 sure, it's a local jmp 08:27:55 yes; but goto also adds one branch extra all the time; by simply cutting the routine and reorder it as I need I can remove that branch entirely 08:28:59 sounds like you'd be better off with JIT code generation. hell of a lot more flexible in the end 08:29:09 just generate the code on the fly and cache it if you have to 08:29:28 and that's as portable as it gets 08:29:29 sproingie: well this is JIT code generation; except I'm reusing the code the compiler generated instead of handrolling my own 08:29:39 DLopen ;-) 08:30:05 Frek: any amount of optimization could trash your approach completely 08:30:35 sproingie: yes it could; that's why I have to compile that particular routine without any optimization what so ever 08:31:34 but the big win is that I would avoid (atleast) 1 branch and lots stack frames by gluing the code together in a new memory block then execute it. 08:32:03 that's like avoiding a detour in town by going around the state 08:32:11 not really 08:32:22 i suspect you'll get better answers in #asm 08:32:42 maybe even better in some MC/cpu site 08:32:45 executing code in the data segment is usually a BIG lose 08:32:52 once the code block been generated it can be reused as many times as one wish. it does indeed has a high setup/construction overhead; but then again all JITs has that overhead 08:32:53 I think he's digging a deeeeeep hole 08:33:22 code on the stack is usually just fine tho 08:33:25 sproingie: data segment ? I would allocate (malloc) the buffer of course 08:33:40 JIT's are unamusing. THey either generate shit to save and dlopen, or they run pcode/icode 08:34:08 eh? most use trampolines of some sort 08:34:21 yea I donno what JITs PoppaVic used 08:35:14 well, you could slice up the code that way, and since it already works, more power to ya, but i have my doubts about the net gain if any. anyway, i gotta go 08:35:31 sproingie: ok bye. 08:35:54 see what debuggers add for breakpoints. those are known to be "harmless" but recognizable instructions 08:35:56 later 08:36:04 good point. 08:49:23 --- join: Herkamire (~jason@c-24-218-95-147.hsd1.ma.comcast.net) joined #forth 08:49:23 --- mode: ChanServ set +o Herkamire 08:49:32 lo, Herk 08:49:43 hey :) 08:49:51 Whazzup? 08:50:32 trying to get a ride to my mom's house, but can't get the right people to answer their phones 08:50:47 I did actually find out what time lunch is though. that's progress I guess 08:51:04 heh 08:53:24 --- join: saon (1000@c-24-129-91-106.hsd1.fl.comcast.net) joined #forth 08:53:56 --- join: saon|smgl (~saon@c-24-129-91-106.hsd1.fl.comcast.net) joined #forth 08:56:44 --- quit: sproingie (Remote closed the connection) 09:24:49 --- quit: PoppaVic ("stay well all") 09:24:51 --- quit: nothingmuch (Read error: 60 (Operation timed out)) 09:29:00 --- join: nothingmuch (~nothingmu@yovalk.bb.netvision.net.il) joined #forth 09:29:47 --- quit: [Forth] ("abort" Reality Strikes Again"") 09:36:23 --- quit: I440r_ (Read error: 104 (Connection reset by peer)) 09:36:57 Herkamire: harroth is working, like for real 09:37:16 not branching yet 09:37:26 but boy have a learned a lot of haskell 09:37:33 anywho, dinnertime 09:42:38 harroth? isn't that an rpg game? 09:44:00 no, guess not, that must be somethingelse-oth 09:44:13 glad to hear it :) 09:44:42 my ride is showing up any minute now 09:45:20 heh :) there it is. 09:45:21 laters 09:52:52 when you return! Harrorth Revealed - what is the TRUE nature of the said project? after these messages 09:52:55 /msg 09:53:15 harrorth is my attempt at writing a forth interpreter in haskell, so that I can learn haskell and forth 09:53:59 i'm off too... ciao! 10:03:32 --- quit: nothingmuch () 10:33:27 forth in haskell? sounds tricky 10:33:37 --- join: swsch (~s@p54924610.dip.t-dialin.net) joined #forth 10:35:12 --- part: swsch left #forth 11:51:33 --- join: tathi (~josh@tathi.bronze.supporter.pdpc) joined #forth 11:59:01 --- join: KB1FYR (~Alex@196-220.suscom-maine.net) joined #forth 12:14:14 --- join: snoopy_1711 (snoopy_161@dsl-084-058-152-157.arcor-ip.net) joined #forth 12:14:14 --- quit: Snoopy42 (Read error: 131 (Connection reset by peer)) 12:14:23 --- nick: snoopy_1711 -> Snoopy42 12:21:56 --- nick: Raystm2 -> nanstm 12:53:51 Today's interesing picture [ NOT SAFE FOR WORK { Hentai/Tentacles } ] 12:53:53 http://cgi.2chan.net/o/src/1118850873505.jpg 12:54:08 good morning 12:56:21 um, eew? 13:24:29 --- join: Sonarman_ (matt@adsl-64-169-94-21.dsl.snfc21.pacbell.net) joined #forth 13:24:42 --- quit: Sonarman (Nick collision from services.) 13:24:55 --- nick: Sonarman_ -> Sonarman 13:27:40 --- join: onetom (~tom@ns.dunasoft.com) joined #forth 13:31:25 --- quit: Sonarman ("Lost terminal") 13:31:43 --- join: Sonarman (matt@adsl-64-169-94-21.dsl.snfc21.pacbell.net) joined #forth 14:22:51 --- join: nothingmuch (~nothingmu@yovalk.bb.netvision.net.il) joined #forth 14:31:11 --- quit: vitaminmoo ("Leaving") 14:44:38 hi 14:48:11 --- join: vitaminmoo (~vitaminmo@dsl-94-128.peak.org) joined #forth 14:54:24 --- part: _ziggy_ left #forth 17:37:21 --- join: sproingie (~chuck@64-121-15-14.c3-0.sfrn-ubr8.sfrn.ca.cable.rcn.com) joined #forth 17:51:14 --- join: slava (~slava@CPE0080ad77a020-CM000e5cdfda14.cpe.net.cable.rogers.com) joined #forth 17:51:27 anybody know about the calling convention that OS X uses? 17:51:34 it seems to have a few differences from the sysv powerpc abi used by linux/ppc 17:53:18 not me, sorry. 17:53:20 slava: http://www.osnews.com/story.php?news_id=10366 has some details on OSX calling conventions 17:53:36 i can make ffi calls fine for the most part 17:53:45 but if i call a function that calls printf, i get a SIGILL 17:53:54 and what's worse, gdb doesn't work properly on OS X 10.3 17:53:56 :/ 17:54:52 hmm, are you setting that condition-register flag that indicates that you're calling a varargs function? 17:55:01 * crc doesn't use gdb 17:55:03 i'm not calling a varargs function... the function that i'm calling is 17:55:06 void foobar(float x, int y, float z) 17:55:07 { 17:55:07 printf("%f\n",x); 17:55:07 printf("%d\n",y); 17:55:07 printf("%f\n",z); 17:55:07 } 17:55:21 when i call this via ffi, the y parameter is corrupt, and after printing z it crashes 17:55:23 oops, didn't read carefully. 17:55:27 gotcha 18:06:39 --- quit: tathi ("leaving") 18:08:07 --- join: LOOP-HOG2 (~chatzilla@sub22-119.member.dsl-only.net) joined #forth 18:25:04 --- quit: nothingmuch (Read error: 110 (Connection timed out)) 18:39:36 --- nick: nanstm -> Raystm2 19:37:38 --- quit: LOOP-HOG2 ("ChatZilla 0.9.61 [Mozilla rv:1.7.1/20040707]") 20:30:37 --- quit: sproingie (Remote closed the connection) 20:43:42 --- part: slava left #forth 20:43:43 --- join: knoppix (~knoppix@c-24-2-135-15.hsd1.ct.comcast.net) joined #forth 20:44:38 no hablo en ingles, pero necesito hablar contigo... 20:48:38 porque el forth esta muy bueno y me gusto usar-lo, 20:49:05 --- quit: Sonarman (brown.freenode.net irc.freenode.net) 20:49:48 --- quit: onetom (brown.freenode.net irc.freenode.net) 20:49:48 --- quit: Snoopy42 (brown.freenode.net irc.freenode.net) 20:49:48 --- quit: saon|smgl (brown.freenode.net irc.freenode.net) 20:49:48 --- quit: Raystm2 (brown.freenode.net irc.freenode.net) 20:49:48 --- quit: warpzero (brown.freenode.net irc.freenode.net) 20:49:48 --- quit: crc (brown.freenode.net irc.freenode.net) 20:49:48 --- quit: ianp (brown.freenode.net irc.freenode.net) 20:49:48 --- quit: KB1FYR (brown.freenode.net irc.freenode.net) 20:49:48 --- quit: OrngeTide (brown.freenode.net irc.freenode.net) 20:49:48 --- quit: madwork (brown.freenode.net irc.freenode.net) 20:49:48 --- quit: saon (brown.freenode.net irc.freenode.net) 20:49:48 --- quit: arke (brown.freenode.net irc.freenode.net) 20:49:48 --- quit: ccfg (brown.freenode.net irc.freenode.net) 20:49:48 --- quit: cmeme ("Client terminated by server") 20:49:48 hola, estais aqui ahora? 20:49:48 necesito hacer sexo con tu mama 20:49:48 porque esta muy bonita 20:49:48 y estais hijos de putas 20:49:48 uauauauau 20:49:48 --- quit: madgarden (Read error: 104 (Connection reset by peer)) 20:49:48 que pasado, no se, este cosa que pasado esta loco en la cabeza de IRC 20:49:48 --- join: crc (crc@pool-151-197-232-212.phil.east.verizon.net) joined #forth 20:49:48 --- join: Sonarman (matt@adsl-64-169-94-21.dsl.snfc21.pacbell.net) joined #forth 20:49:48 --- join: onetom (~tom@ns.dunasoft.com) joined #forth 20:49:48 --- join: Snoopy42 (snoopy_161@dsl-084-058-152-157.arcor-ip.net) joined #forth 20:49:48 --- join: KB1FYR (~Alex@196-220.suscom-maine.net) joined #forth 20:49:48 --- join: saon|smgl (~saon@c-24-129-91-106.hsd1.fl.comcast.net) joined #forth 20:49:48 --- join: saon (1000@c-24-129-91-106.hsd1.fl.comcast.net) joined #forth 20:49:48 --- join: Raystm2 (~vircuser@adsl-69-149-42-251.dsl.rcsntx.swbell.net) joined #forth 20:49:48 --- join: warpzero (~warpzero@wza.us) joined #forth 20:49:48 --- join: OrngeTide (~orange@rm-f.net) joined #forth 20:49:48 --- join: madwork (~madgarden@derby.metrics.com) joined #forth 20:49:48 --- join: arke (f2@bespin.org) joined #forth 20:49:48 --- join: ianp (~ian@inpuj.com) joined #forth 20:49:48 --- join: ccfg (ccfg@dsl-roigw3de0.dial.inet.fi) joined #forth 20:49:48 --- mode: irc.freenode.net set +o crc 20:49:48 --- quit: knoppix (Remote closed the connection) 20:50:50 --- join: cmeme (~cmeme@216.184.11.2) joined #forth 21:17:30 --- join: true-grue (true-grue@ppp135.medlux.ru) joined #forth 22:19:28 --- join: Serg[ICQ] (~Miranda@212.34.52.140) joined #forth 23:06:14 --- join: nothingmuch (~nothingmu@212.143.92.226) joined #forth 23:09:56 --- quit: cmeme (Read error: 54 (Connection reset by peer)) 23:10:48 --- join: cmeme (~cmeme@216.184.11.2) joined #forth 23:11:04 --- quit: cmeme (Remote closed the connection) 23:11:51 --- join: cmeme (~cmeme@216.184.11.2) joined #forth 23:14:54 --- quit: true-grue (Read error: 145 (Connection timed out)) 23:57:01 --- quit: Herkamire ("off to bed") 23:59:59 --- log: ended forth/05.06.15