00:00:00 --- log: started forth/06.07.10 00:13:53 I'll write an IRC client next, I guess. 00:34:00 Indications that your thinking might be a bit off: 00:34:07 5.0e 2.5e f- f. 2.5 ok 00:50:31 --- quit: slava () 01:08:58 --- quit: LOOP-HOG ("Leaving") 03:13:46 --- quit: ASau (Read error: 104 (Connection reset by peer)) 04:17:37 --- quit: virl (Remote closed the connection) 04:36:07 --- join: virl (n=virl@chello062178085149.1.12.vie.surfer.at) joined #forth 04:39:24 --- join: PoppaVic (n=pete@0-1pool66-67.nas22.chicago4.il.us.da.qwest.net) joined #forth 05:52:31 --- join: timlarson_ (n=timlarso@65.116.199.19) joined #forth 05:56:16 --- join: Ray_work (n=Raystm2@199.227.227.26) joined #forth 05:59:47 --- join: nighty (n=nighty@66-163-28-100.ip.tor.radiant.net) joined #forth 06:19:47 humph 06:30:33 I think parsing is part of the issue.. Damn. 06:31:34 Didn't *forth - at some point - recognize "word" as WELL as "line"? 06:32:48 Or, perhaps te question becomes: Are not control-chars as valid a "name" as anything else? 06:32:54 te/the 06:35:50 normally forth only parses words 06:36:01 I know 06:36:55 primarily, forth see's a "word" as (delimiter....)WORD(delimiter) 06:37:28 I am just glaring at the assumptions. 06:38:32 so...? 06:38:39 #1) 'delimiter' is both left and right; #2) 'delimiter' is a char - a unit. Not a set. 06:39:20 left and right? what the!? 06:39:27 #3) is likely that this is insufficient overall 06:39:46 virl: check the sources - it's a low-level assumption 06:40:22 as in "ten spaces before WORD are the same as a space and the WORD is terminated by a space" 06:40:26 and what's bad about that? 06:40:54 I'm not 100% sure. It certainly forces an RPN form 06:41:11 nothing is bad about that 06:41:35 otoh, I never said "bad" - I was cogitating/organizing. 06:42:38 parse/word/etc are just... limiting. Not sure it's "bad", am sure it's limiting. 06:44:26 virl: as an example: argc/argv or "program ...." is postfix. None of the args will usually make sense w/o the program to define and manage 'context'. 06:45:08 prefix (RPN) such as "-o test gcc" would really be weird. 06:46:05 virl: trying to see where/how to reconcile and how. In some cases, we WANT words to "look ahead" 06:46:41 the easiest to consider is "constant" and "variable" 06:47:28 You have to keep reminding yourself: "Forths are not purely RPN" 06:50:55 virl: I _am_ seeing that context-variables (and I don't care where they are) seem to work for most "Context " situations. 06:51:14 maybe even a few 3 "word" situations. 06:52:04 but, in the above: you may have "Context" as a voc, in the order, and text MIGHT have a wordlist of it's own, etc. 06:52:35 then do those words which need look-ahead do look-ahead 06:52:51 heh.. Yeah.. That's sorta' where I am 06:52:59 "immediates" 06:54:50 EXCEPT: if you have a voc of "keywords".. Now, how can you get generic help, help on the voc, and help on a word in the voc? Show me a set of decent comlines? 06:55:34 so, why don't you do it simple? 06:55:39 or explain it simple 06:55:44 I was thinking of a more efficient solution that merely searching the order for the first matching word 06:56:23 perhaps 'knowing the word'? 06:56:37 sure, but that's a lot like 'man' 06:56:53 there is also 'apropos' and 'whatis' 06:57:52 I don't know why you search for something 'efficient', meta-thing(whatever you call it) should be something for compiling, and this is something which doesn't need much runtime speed. 06:58:30 when you have a simple-to-parse description language then it's fast 06:59:24 and forth is simple to parse. 06:59:39 yep. And we get soooo many converts 06:59:57 c isn't and so they need performant compilers 07:00:07 converts? 07:00:14 * PoppaVic sighs 07:01:07 So'k, virl - you don't see a "solution" - I sorta' see a few, and I was looking for other inputs that would start a consensus. Don't sweat it. 07:01:46 Thanks for the time, though - it always helps to have a sounding-board. 07:10:00 for me it's the perfect solution to the problem 07:26:31 --- quit: PoppaVic ("Pulls the pin...") 07:28:01 --- join: PoppaVic (n=pete@0-1pool64-80.nas22.chicago4.il.us.da.qwest.net) joined #forth 08:18:34 --- quit: PoppaVic ("Pulls the pin...") 08:35:03 --- join: neceve (n=claudiu@unaffiliated/neceve) joined #forth 10:15:41 --- join: docl (n=docl@70-101-145-1.br1.mcl.id.frontiernet.net) joined #forth 10:17:03 hi docl 10:17:16 hi virl 11:38:10 --- join: snowrichard (n=richard@adsl-69-155-177-154.dsl.lgvwtx.swbell.net) joined #forth 11:38:36 hi 11:41:04 --- quit: snowrichard (Client Quit) 11:51:41 --- join: tlockney (n=tlockney@c-24-20-172-87.hsd1.mn.comcast.net) joined #forth 12:48:48 --- quit: timlarson_ (Read error: 104 (Connection reset by peer)) 13:13:55 --- quit: nighty ("Disappears in a puff of smoke") 14:47:46 onyx looks a lot like Postscript 14:58:32 Bok Choi looks a lot like cabbage. 15:03:27 --- quit: Ray_work (Read error: 104 (Connection reset by peer)) 15:09:45 onyx? 15:11:44 http://factor-language.blogspot.com/2006/07/onyx-programming-language.html 15:13:37 The wc example is really long. 15:17:28 The main page suggests that stacks are not first-class objects in Forth, but any new construct is a first-class object in Forth, so additional stacks certainly can be. 15:44:31 well, what are first-class objects? that's imho too much objects 15:44:51 I don't mean 'object' in any kind of OO sense. 15:46:11 so what? 15:48:02 When you define a new word or construct in Forth, there's no indirection in the calling sequence, you can just use it alongside other words that are already defined. That makes it first-class. By contrast, for instance, a function in C is not a first-class object. 15:50:49 Here's a more detailed description of what a first-class object is: http://en.wikipedia.org/wiki/First-class_object 15:52:31 aha, ok, so words can be created at runtime. c funcs not. 15:53:21 That's part of it. 15:53:47 is there such a thing as a second-class object in forth? 15:54:03 by the way, did any forth get as fast as C? 15:54:05 I suspect the Onyx documentor is saying that new stack creators/operators are part of the language definition, and that somehow that makes them 'first class objects'. 15:55:19 virl: reva is close to that of unoptimized C for many of Ron's benchmarks 15:55:27 virl, 'as fast as C' is a vague criterion. However, Pelc's Forth has an impressive optimizer; I wouldn't be surprised if it's at least as fast as C on the same platforms. 15:55:35 --- join: nighty (n=nighty@CPE00508be190fc-CM0012253ec1bc.cpe.net.cable.rogers.com) joined #forth 15:55:44 ok, then gcc C on my machine ;-) 15:55:50 Same answer. 15:56:10 Pelc's Forth? 15:56:13 link? 15:56:52 His company is MPE. 15:57:10 docl, I can't come up with one off the top of my head. 15:57:17 --- join: I440r (n=mark4@24-177-235-246.dhcp.gnvl.sc.charter.com) joined #forth 15:58:20 many Forth systems allow creating new data and return stacks at run-time 15:58:29 Right. 15:58:37 that's an ANS forth right? 15:58:48 so the stacks _are_ first-class then 15:58:51 vfx forth? 15:58:52 virl, I believe MPE makes Standard-compliant systems, yes. 15:58:56 nope, ANS doesn't define anything for it 15:58:56 segher_, right. That's what I was saying. 15:59:10 a big minus, that it's ANS 15:59:40 ANS defines a minimum with which it is possible to write useful, portable programs 15:59:49 Good grief. It's just Standard-compliant. It has plenty of implementation-specific stuff. You really need to get your head out of your ass about a system being somehow deficient just because it's able to run Standard programs. 15:59:54 well, I'd really to get some of this forth stamps, but sadly I can't buy it. 16:00:03 and which is _also_ easy to implement on basically any Forth system 16:00:28 quartus: yeah exactly 16:02:10 Quartus, I don't like the standard, because I don't like to write forth programs in this way. for me, it's horrible obsolete. and yeah, I don't like it's syntax and that's why I love to write retroforth programs. 16:02:57 You're badly confused about what the standard is, and what it requires. For you to discard a system merely because it allows the option of running Standard programs is pig-headed, and flat out wrong. 16:03:02 syntax? the only syntax in forth is whitespace. 16:03:33 ok, then the words to clarify it 16:04:06 Nothing about a Standard-compliant system forces you to write programs in any particular way. You can write non-standard stuff on a standard system. Being a standard-compliant system does not prevent any system from having implementation-specific extensions; every one I've seen does. 16:04:08 standards are good, but this one is imho bad. I like more the C standard. 16:04:30 sp! rp! 16:04:32 :) 16:05:00 it forces me to use a word convention which I think sucks 16:05:07 i have a depth! and an rdepth! instead 16:05:41 but i'll change it to sp0 sp! rp0 rp! soon i guess, to support multiple stacks easier 16:06:19 virl, I have no idea what 'forces me to use a word convention' means. 16:08:03 well, fine. 16:20:02 the think i dislike about the ans standard is actlally the word names the chose. 16:20:11 why did they rename !> to TO for instance 16:20:19 When was it !> ? 16:20:25 ooh dear, a non forth coder wont know what !> means !!! 16:20:35 ! store 16:20:36 > to 16:20:39 fpc used !> 16:20:45 I thought FPC was F83. 16:20:53 no it was based on it 16:21:02 F83 had TO. 16:21:03 f-83 was by laxen and perry 16:21:23 Laxen and Perry did a popular implementation based on the 83 standard. 16:21:40 I'm referring to the standard, not the implementation. 16:21:43 it didnt have eithr 16:22:01 I thought it did. Let me check my docs. 16:22:18 store to was invented in f-pc 16:22:39 as was "value" 16:23:11 I think TO is a greatly improved name over !>. 16:23:34 no way 16:23:40 its incomplete 16:23:43 STORE to 16:23:46 says it all 16:23:54 On the other hand I rarely use either values or locals. 16:24:18 i dont call them values either. and locals are non forthlike so i dont even support them 16:24:46 I like renaming words for my own use, at least til I learn them 16:24:56 i think value is a bad name it doesnt suggest a variable 16:24:57 e.g. rot and -rot to dig and bury 16:25:01 so i called them var 16:25:43 i have variable and constant with traditional definitions but i also have var and const 16:25:58 which are my prefered means of doing variables and constants 16:26:07 but some ppl have issues with my const word 16:26:23 it creates a state smart immediate word that either returns the value or compiles a literal 16:26:36 i think thats the way variable and constant should have worked from day 1 16:27:04 I do that with normal CONSTANTs, but without a state-smart word. 16:27:43 so a constant is compile only ? 16:27:45 In Quartus Forth, 5 constant foo is equivalent to : foo 5 ; inline 16:27:54 No, works fine interpreting too. 16:27:59 aha ok 16:28:04 No state issues. 16:28:25 likewise VARIABLEs are inlined. 16:28:39 So they compile as an address literal. 16:28:54 as a constant address 16:28:58 Yes. 16:29:24 So they still tick as you'd expect, behave as they should, don't check state, but still optimize. 16:29:27 you cant do n !> foo above 16:29:33 tho thats probably a good thing :) 16:29:36 No, you can't. It's a constant. :) 16:29:58 in isforth you can still do a !> a const but thers little point in that kind of trickery 16:30:03 If you want a VALUE, it's provided too. I use it so seldom I'd have to go check to see if I bothered to do any optimization. 16:30:10 i dont restrict you in any way usually 16:30:25 tho i enforce proper factoring of case statements 16:30:27 you cant do 16:30:28 case 16:30:38 x of 500 lines of code endof 16:30:44 y if 1000 lines of code endof 16:30:46 etc 16:31:02 No, my VALUE is just : VALUE ( x "name" --) 16:31:02 create , does> @ ; 16:31:04 you can only make a reference to a word 16:31:08 --- part: tlockney left #forth 16:32:00 Quartus Forth is standard-compliant, so of course I provide CASE and friends as they're expected, but I also have a SELECT/XT/END-SELECT that is a just an indexed list of words, a jump-table. 16:32:06 i think my case: is the ONLY time i actually enforce anything 16:32:16 and thats only as a result of how my case construct works 16:32:27 it creates tables 16:32:40 not a bunch of nested if/else statements 16:33:13 Right, that's why I provided the select/xt/end-select -- though in my case it's also because a simple comma'd list of xts doesn't turnkey. 16:33:43 err how come ? 16:33:48 Single-pass turnkeying. 16:34:27 --- quit: segher_ (Connection reset by peer) 16:34:31 erm i dont understand what that means 16:34:41 turnkey in isforth is just a SAVE method 16:34:48 i either save with headers or without headers 16:34:57 More sophisticated in Quartus Forth. It recursively extracts only dependent code, without headers. 16:35:06 oh yea thats right 16:35:46 So xts change, and I'd either have to go multipass on the turnkey, or provide an A! and A, if I wanted to be able to identify which data items needed relocating, both of which seemed ugly to me. So I added select/end-select. 16:35:50 when you buy quartus forth do you get all the sources to everything ? 16:36:11 It's closed-source, most of which you can browse with the disassembler, but the sources are not provided. 16:36:43 k 16:37:25 The various additional libraries are all provided in source form. 16:37:34 --- join: segher_ (n=segher@dslb-084-056-136-095.pools.arcor-ip.net) joined #forth 16:50:22 I440r. are you considering a recursively-extracting turnkey? 17:20:16 --- quit: neceve (Remote closed the connection) 17:20:35 --- join: nighty_ (n=nighty@CPE00508be190fc-CM0012253ec1bc.cpe.net.cable.rogers.com) joined #forth 17:21:00 --- quit: nighty (Read error: 104 (Connection reset by peer)) 17:25:32 --- quit: madwork (Read error: 110 (Connection timed out)) 17:26:04 no not yet 17:26:09 maybe in the future 17:26:22 i think my highest prioritis are an assembler and documentation 17:26:34 i want online help 17:26:38 ill use block files for that 17:27:13 i dont know if i can code the assembler i want tho - not been able to code it since starting isforth 17:29:06 How is it elisive? 17:29:08 oops 17:29:09 elusive? 17:29:34 i have certain requirements for it 17:29:50 i.e. i want to be able to take the existing kernel sources AS IS 17:30:00 with maybe some spaces injected here and there 17:30:21 but i do not want to have to totally rewrite my kernel just to support some assembler im not going to like 17:31:18 the macroified : defs will be converted to real colon defs 17:31:29 but the coded defs MUST be able to be taken AS IS 17:31:43 that totally counts out a 5 # ax mov bullshit assembler 17:31:50 i do not like traditional forth assemblers 17:32:19 also. i want the assembler to support all opcodes from 386 to p6 plus mmx mmx2 sse sse2 fpu etc etc etc 17:32:26 3dnow blah blah 17:32:34 huge task writing it 17:32:44 and ive not even been able to get off the ground with it 17:33:10 Ah. 17:33:41 Maybe you could retool your existing assembler to output 5 # ax mov style code. 17:33:44 so far the only assembler ive been able to write myself is my 8051 assembler 17:33:57 i dont have an existing assembler 17:33:57 brb 17:34:04 I thought you had existing kernel sources. 17:34:42 i do 17:35:04 but im not going to butcher them to support an assembler i dont like 17:35:55 and i do not like backwards asm 17:40:03 I was suggesting you alter your existing assembler to generate suitable intermediate source, not to butcher your existing sources. 17:40:29 Certainly the postfix style of assembler has ease-of-implementation as a benefit. 17:40:47 injection of spaces is the only modification i would accept 17:40:56 Right. Again, not suggesting you modify your sources. 17:41:04 ok 17:41:14 Suggesting a possible alternate course of action. 17:41:26 well i dont have an existing assembler 17:41:35 i am currently assembling my kernel with nasm 17:41:36 Then what do you assemble your sources with? 17:42:01 Ok. What I'm suggesting is adding a back-end to or modifying nasm. 17:42:50 the point of that would be ? 17:44:21 The ability to assemble your sources inside your own Forth. 17:45:27 ya and that would be a non event if it was not the forth itself doung the assembling 17:45:47 and it would be diffdicult to do 17:45:55 bow would u assemble 17:46:15 mov eax, [ ' myconstant >body ] 17:46:44 [ and ] signifying "contents of" not compile/interpret 17:46:48 --- nick: segher_ -> segher 17:46:59 I don't follow. But not to worry, if that's not your goal then my suggestion isn't useful for you. 17:48:08 im not trying to be anti, i just have an image of what i want 17:48:22 Sure, that's ok. I don't know what you're after. 17:48:36 a forth assembler :) 17:48:43 thats not backwards :P 17:49:03 mov eax, variable <-- point eax mov eax, [ variable ] 17:49:24 mov contents of variable into eax 17:49:43 mov eax, ' my-var >body 17:50:03 poingt eax at my-var (a value] 17:50:09 I put a prefix mode in my postfix assembler. In prefix mode each opcode first parses and evaluates the rest of the line before contiuning. It's not identical with a full prefix assembler, but it's pretty close. 17:50:16 mov eax, [ ' myvar >body ] 17:50:32 dont need to parse 17:50:44 parsing makes it more difficult 17:50:56 : mov >asm ..... ; 17:51:01 I sometimes think you jump to a contrary response without actually considering that I'm not being antagonistic. :) 17:51:09 all opcodes start with >asm 17:51:25 If you have an existing postfix assembler, of which there are lots and lots, you can add a 'simulated prefix' to it that's pretty handy and might take you some distance to your goal. 17:51:30 : >asm r> last-op @ >r last-op ! ..... ; 17:52:39 * ayrnieu grumbles. 17:54:00 ive actually contemplated paying someone to help me code this 17:55:05 pretty good traffic on comp.lang.forth , recently 17:55:22 not really :) 17:55:28 mostly just bleh postings lol 17:55:56 I've liked Elizabeth Rather's and John Passaniti's exchanges, 17:59:40 er usually only hard sells forth inc hehe 17:59:56 but sometimes she is very helpfull 18:00:04 tho i tend to not agree with her on most things :) 18:00:49 quartus - at what level to you refill ? 18:01:03 i put my refill in word itself but i think i need it in parse 18:01:15 Not sure what you mean by 'level'. 18:05:02 I don't use WORD as part of the interpreter. 18:05:26 I use PARSE-WORD. 18:06:24 Simpler, returns the c-addr u of the next word in-place, without copying it to a buffer the way WORD does. 18:06:25 ya my word calls parse-word 18:07:25 you don't have WORD ? 18:07:30 I have WORD. 18:07:37 I just don't use it in the interpreter. 18:07:44 I440r - you don't have WORD ? 18:08:28 yes i have word] 18:08:44 word does a refill if tib is empty and then calls parse-word 18:11:32 unrelatedly, Factor is quite strange. 18:12:44 that worked great when i counted the entire file memory map as tib and the file size as #tib 18:13:08 but ive modified it to make tib the start of the next line of source within the memory mapping 18:13:19 and #tib being the size of the line 18:13:50 but when parse-word runs out it barfs because it cant refill 18:13:57 so im thinking of moving refill 18:14:23 because doing it line by line gives me the ability to count lines and flag where an error is more precisely 18:15:03 My REFILL is called from inside INCLUDED. The interpreter doesn't use REFILL, it has its own input loop based on ACCEPT. 18:19:05 Well, that might be confusing. I mean to say, the QUIT loop uses ACCEPT. INCLUDED clearly interprets too, but it uses REFILL. 18:20:51 --- join: JasonWoo1 (n=jason@c-71-192-33-206.hsd1.ma.comcast.net) joined #forth 18:22:02 --- quit: nighty_ (Read error: 110 (Connection timed out)) 18:22:30 --- join: nighty (n=nighty@CPE00508be190fc-CM0012253ec1bc.cpe.net.cable.rogers.com) joined #forth 18:24:12 --- quit: JasonWoof (Nick collision from services.) 18:24:19 --- nick: JasonWoo1 -> JasonWoof 18:24:23 --- mode: ChanServ set +o JasonWoof 18:26:30 y the difference ? 18:28:07 Simple historical precedent, the interpreter was coded before REFILL was part of the kernel. :) 18:28:31 heh 18:29:01 None of my parsing words parse into the next line of source, though. 18:29:02 i often suffer from "too lazy to fix something that isnt broken" lol 18:29:08 :) 18:29:30 i do not want to count end of lines as an end of input 18:29:37 it would disallow things like 18:29:49 10 blah xxx yyy zzz aaa bbb 18:29:58 ccc ddd eee fff ggg 18:31:03 On the other hand ignoring line-ends may silently allow certain typos to get through. At any rate while it may be deemed a limitation, theoretically, in practice it isn't something I notice. 18:31:11 How would \ work if you line-endings are ignored? 18:31:53 they arent ignored they are counted as white space by words that need to parse beyond them 18:32:36 So in some instances they're end-of-input, and in some they're not. I prefer the consistency of having them always be the end-of-input. 18:36:07 I love the ease with which you can hack the sources in gentoo, and still get your stuff installed with the package manager 18:36:18 :) 18:36:25 that was always a pain in debian 18:36:51 I've done it with rpms once 18:36:56 it was quite a nightmare 18:37:01 rpm is next to useless 18:37:10 worse than useless, you give up on useless 18:37:27 wait... I actually didn't even change the sources in that case. I just needed to rebuild the srpm so it would link properly 18:37:47 but on gentoo... 18:37:56 I440r, one solution might be to keep end-of-line as end-of-input, and write a PARSE-REFILL or something of that ilk for the times it might be useful. 18:38:03 I uncommented the following: #define NO_BELL 18:38:19 in rxvt, and no more annoying beeps every time I hit tab or hold bs too long or something 18:39:05 I just added a single line to libsdl to make it ignore the SDL_FULLSCREEN flag in SDL_SetVideoMode 18:42:38 watching the closer 18:43:02 normally i dont like this sort of crap but this is VERY funny 18:45:18 oops, that didn't work 18:48:06 this is how you clear a flag it C right? 18:48:17 flags &= ~SDL_FULLSCREEN; 18:48:40 or is it ^ 18:50:48 ~ 18:53:32 oh, right. I forgot to enable X support :) 18:53:33 hehe 18:59:18 --- quit: uiuiuiu (Remote closed the connection) 18:59:21 --- join: uiuiuiu (i=ian@dslb-084-056-229-172.pools.arcor-ip.net) joined #forth 19:25:43 --- join: snoopy_1711 (i=snoopy_1@dslb-084-058-135-254.pools.arcor-ip.net) joined #forth 19:33:10 --- quit: Snoopy42 (Read error: 145 (Connection timed out)) 19:33:20 --- nick: snoopy_1711 -> Snoopy42 19:52:51 --- quit: I440r ("Leaving") 19:57:57 --- quit: nighty ("Disappears in a puff of smoke") 20:58:29 --- quit: JasonWoof ("brb") 20:58:33 --- join: JasonWoof (n=jason@pdpc/supporter/student/Herkamire) joined #forth 20:58:33 --- mode: ChanServ set +o JasonWoof 20:59:07 --- quit: JasonWoof (Client Quit) 21:01:31 --- join: JasonWoof (n=jason@pdpc/supporter/student/Herkamire) joined #forth 21:01:31 --- mode: ChanServ set +o JasonWoof 21:02:02 --- quit: JasonWoof (Client Quit) 21:03:09 --- join: JasonWoof (n=jason@pdpc/supporter/student/Herkamire) joined #forth 21:03:09 --- mode: ChanServ set +o JasonWoof 21:18:19 --- join: ASau (n=user@home-pool-173-2.com2com.ru) joined #forth 22:10:09 cool, the gforth ebuild (gentoo linux package) now has ffcall as a prerequisite 22:10:31 dunno why it's not optional, but it's nice people will start having support for the ffi built in 22:12:21 There's FFI even without ffcall. 22:30:58 really? 22:31:03 how? 22:31:19 ouch, herkforth doesn't work right onder my new linux install 22:31:35 All I can do is move the cursor around 22:31:39 can't save or type 22:32:03 well, actually it seems to save and then flip out 22:32:14 the new saved image behaves the same as the first 22:53:30 --- join: slava (n=slava@CPE0080ad77a020-CM000e5cdfda14.cpe.net.cable.rogers.com) joined #forth 22:53:49 --- quit: Snoopy42 (clarke.freenode.net irc.freenode.net) 22:53:49 --- quit: JasonWoof (clarke.freenode.net irc.freenode.net) 22:53:51 --- quit: ohub (clarke.freenode.net irc.freenode.net) 22:55:07 --- join: JasonWoof (n=jason@pdpc/supporter/student/Herkamire) joined #forth 22:55:07 --- join: Snoopy42 (i=snoopy_1@dslb-084-058-135-254.pools.arcor-ip.net) joined #forth 22:55:07 --- join: ohub (n=oherrala@sikw1.oulu.fi) joined #forth 22:55:07 --- mode: irc.freenode.net set +o JasonWoof 22:55:56 hi 23:01:54 library libc cygwin1.dll 23:01:54 3 (int) libc getopt getopt 23:01:54 (addr) libc optarg optarg 23:01:54 (addr) libc optind optind 23:01:58 23:02:14 Privet, slava! 23:02:43 Jason, that's the code. 23:02:50 i benchmarked factor running on four machines, three different cpu architectures: http://factor-language.blogspot.com/ 23:14:34 this is bizare 23:14:40 TYPE is crashing in herkforth 23:24:35 hah! got it 23:24:37 bastards 23:25:20 the write() syscall was ocationally messing up the count register 23:28:11 --- join: johnnowak (n=johnnowa@user-0cev1ia.cable.mindspring.com) joined #forth 23:32:39 --- quit: slava () 23:59:59 --- log: ended forth/06.07.10