00:00:00 --- log: started forth/02.05.19 00:38:17 --- quit: sbk_ (Read error: 104 (Connection reset by peer)) 01:22:37 --- join: davidw (~davidw@adsl-ull-45-103.42-151.net24.it) joined #forth 01:31:18 --- join: Soap` (flop@210-54-74-28.dialup.xtra.co.nz) joined #forth 01:59:03 --- join: rob_ert (~robert@h237n2fls31o965.telia.com) joined #forth 02:30:53 --- quit: Soap` (Read error: 113 (No route to host)) 02:45:02 --- join: gforth0 (guest@adsl52049.vnet.hu) joined #forth 02:49:21 --- quit: gforth (Read error: 110 (Connection timed out)) 02:49:22 --- nick: gforth0 -> gforth 02:51:57 --- join: miket2 (Mike@62.60.64.4) joined #forth 02:57:41 hello 02:58:35 --- part: miket2 left #forth 03:00:13 --- join: divgrad (~John@194.158.217.87) joined #forth 03:11:16 --- join: Soap` (flop@210-86-38-98.dialup.xtra.co.nz) joined #forth 03:14:58 Is there any variant of colorforth for color-blind people? 03:15:21 no 03:15:29 :-/ 03:34:50 --- quit: Soap` (Read error: 110 (Connection timed out)) 03:38:17 * rob_ert wonders why I440r spells "license" as "lisence". 03:42:44 coz its much nicer 03:42:59 unfortunately, it's not english 03:43:18 :) 03:43:29 Looks ugly with a file called "Lisence" 03:43:47 Even a stupid swede who can't speak english notices that's wrong ;-D 03:44:38 as for me, who speaks semi-english only, there is no any difference b/w the two:( 03:44:59 so im a stoopid russian then 03:47:05 --- part: divgrad left #forth 04:02:10 --- join: onetom_ (tom@adsl52049.vnet.hu) joined #forth 04:25:20 heya 04:27:06 hi 04:29:49 oink: do u know about a beautified log of yesterday? 04:31:20 uhm 04:31:37 with Mr. Chuck ? 04:34:58 --- nick: onetom_ -> onetom 05:03:41 oink: sure, but jeff also count that much :) 05:03:44 count 05:03:45 counts 05:03:47 ehh 05:04:34 --- join: XeF4 (~XeF4@dsl-XIV-139.kotikaista.weppi.fi) joined #forth 05:05:05 are there logs of the 18.5 Moore interview somewhere? 05:08:38 --- join: Speuler (~l@a161161.upc-a.chello.nl) joined #forth 05:09:18 'morning 05:10:20 morning. Do you have any inkling where last night's interview logs are? 05:10:37 no idea. clog ? i wasn't there 05:11:00 neither was I. 05:12:33 XeF4: clog was certainly (how nice guy is (s)he :) 05:12:51 but im also interested in some colorful logs 05:15:25 how does one talk to clog? 05:17:43 found. 05:17:56 http://tunes.org/~nef/logs/forth/02.05.18 05:18:01 onetom: have u bn around yesterday ? 05:21:11 Speuler: partly :( my isp was evil and there was no net most of the time during the "interview" 05:29:59 bluddy telecommunicationtechnological dilletants 05:30:09 sue them 05:36:27 --- join: mur (ammu@baana-62-165-189-112.phnet.fi) joined #forth 05:41:12 hyvaa huomenta mur 05:46:19 huomenta :) 05:50:27 tak 06:00:04 kiitos in finnish :) 06:07:42 Speuler: i couldnt even call their tech support coz they r all the time busy 06:08:25 Speuler: what makes me think theres no fuckin techsupport ppl hangin around at all... 06:14:54 --- nick: mur -> murBBL 06:17:17 onetom: if you experienced technical problems, and called tech support, maybe others did so too ? 06:26:29 sure sure, but in the whole day? 06:26:56 anyway, ive just managed 2 reach them 06:27:07 and they say: we know about the problem 06:27:35 and it will go away after we have finished 06:28:24 expanding our system, but the completion date is absolutely unknow 06:28:36 great... :/ 06:34:34 --- part: XeF4 left #forth 06:46:44 --- nick: murBBL -> mur 06:50:02 --- join: gforth0 (guest@adsl52006.vnet.hu) joined #forth 06:55:15 --- quit: onetom (Read error: 60 (Operation timed out)) 07:03:31 --- join: onetom (tom@adsl52006.vnet.hu) joined #forth 07:05:04 --- quit: gforth (Read error: 110 (Connection timed out)) 07:05:04 --- nick: gforth0 -> gforth 07:35:43 --- quit: davidw (Remote closed the connection) 07:35:49 --- join: davidw (~davidw@adsl-ull-45-103.42-151.net24.it) joined #forth 07:54:40 as im browsing machine forth tutorial ive discovered a common misunderstanding - ithink @ least 07:59:23 what would that b ? 07:59:49 when chuck talks about stack depth constraints 08:00:15 he probably assumes a machine forth enviroment 08:00:34 where there is an address register available 08:01:22 tho its a global reg, u cant use those "tricky" looping technics 08:01:27 aiming 4 stacks w/o random xs capabilities ? 08:01:38 random xs capabilities? 08:01:58 random xs on stack elements, like with pick 08:01:58 4 stacks? or forth level stacks? 08:02:06 hmm 08:02:09 oh, I think Speuler means random access 08:02:30 xs - k got it 08:02:31 i assumed you'd understand your favorite abbrevs well 08:02:40 im slow 2day a bit 08:03:00 y r u abbreviating so much anyway ? 08:03:05 yes, thats what im refering 2 08:03:16 4 speed? 08:03:21 10-4 08:03:21 its a habit 08:03:39 feeling el8ed 08:03:57 deltab: you're right 08:04:12 x-s onetom talk :) 08:04:43 merely adapting to :) 08:04:56 :p 08:05:05 u r over adapting 08:07:25 what about the stacks ... ? 08:10:49 sorry - /me multitasks 08:11:18 trad 4th: 08:11:19 0 a2 a1 DO DUP I C! 1+ LOOP DROP 08:11:19 a1 a2 OVER DO I OVER - I C! LOOP DROP 08:11:19 a1 a2-a1 0 DO I OVER + I SWAP C! LOOP DROP 08:11:19 0 a1 08:11:19 BEGIN 08:11:21 OVER OVER C! 08:11:23 CHAR+ SWAP 1+ SWAP 08:11:25 DUP a2 = 08:11:27 UNTIL DROP DROP 08:11:32 these r various looping constructs 08:12:05 adapted 2 constant memory range fillings 08:12:37 the respective machine 4th code is: 08:12:48 a1 a! 0 DUP 08:12:48 BEGIN 08:12:48 DROP DUP !+ 1+ 08:12:48 A a2 XOR 08:12:48 WHILE REPEAT DROP 08:13:45 it is much more efficient 08:17:08 BUT 08:17:44 what do u do if a1 comes from the stack as a parameter? 08:18:26 what do u do if u wanna use the above code 08:18:46 as the kernel of an outter loop 08:18:57 what also scans an address range? 08:19:35 hmmm. 08:19:40 rewrite the word ? 08:19:41 u can rely anymore on the address register, since its global 08:20:01 what do u mean by rewrite? 08:20:11 modify the definition of the word? 08:20:24 yep 08:20:31 the compiled thread? 08:20:54 show me 08:21:32 anyway, code rewrite also avoids reentrancy :/ 08:22:09 so its not useful in multitask environments 08:22:37 all in all i stil cant understand how can b machine forth so 08:22:55 ... omnipotent 08:23:10 so optimal 08:23:24 how can u avoid C-like techniques 08:23:42 a 10 0 do X loop 08:23:51 certainly can b replaced by 08:23:55 X X X X X X X X X X 08:24:17 but what if that 10 is a parameter? 08:25:13 there's a for...next in machine forth ? 08:25:30 so the overhead u pay in the case of do-loop-like constructs 08:25:40 returns in flexibility 08:25:53 and in being more general purpose 08:26:13 sure, for next is better 08:26:50 does it utilizes the return stack for cycle vars? 08:27:30 on the novix, it used an own register 08:27:46 j was on r 08:28:23 u mean it was nestable 08:28:37 via pushing the cycle var onto the return stack 08:28:44 but was efficient 08:28:45 yes 08:29:02 by holding the top of cycle vars in a reg 08:29:08 right? 08:30:00 cycle var in reg is quicker to test and decrement 08:30:13 but tor was a reg too 08:30:19 sure sure i understood 08:30:21 just as tos and nos 08:30:30 nos? 08:30:34 next of stack 08:30:37 second 08:30:40 ah 08:31:20 do loop took about 10 times more cycles 08:32:00 for general iteration, for next is perfect 08:32:32 comparable to decrement-branch-if-not-zero, implemented by many cpus 08:32:38 tho, in the case of 0 a2 a1 DO DUP I C! 1+ LOOP DROP 08:32:55 there is a very short cycle kernel 08:33:06 and its general 08:33:26 coz it can receive the bounds via stack 08:34:39 and the so called overhead is not absolutely a waste of time 08:34:50 coz it also does useful things 08:35:09 what should b done manually in the machine 4th version 08:35:16 am i right? 08:35:59 depends 08:37:02 if you have use for the results of extra operations 08:37:23 resembles a bit the risc/cisc discussion 08:37:53 ur right too 08:37:59 i prefer risc 08:38:43 i like things b reduced 08:38:46 forth tends to offer riscy words, to put together to ciscy ones yourself 08:38:56 like the length of words ;) 08:39:00 easier to pile up than dicard from existing words 08:39:15 sure 08:39:38 just these specific comparision 08:39:39 s 08:41:32 08:41:42 can b generalized 08:42:06 like: u can create apps 10x smaller & 100x faster 08:42:18 in machine forth than in trad forth 08:42:50 coz its just true in the case of tiny problems 08:43:03 however 1 can learn a lot from these examples... 08:43:25 look: 08:43:34 : number ( table -- ) M M * 0 do i over c! char+ loop drop ; 08:43:48 its an MxM matrix filler 08:44:10 how can we get rid of do-loop? 08:44:19 for-next doesnt play this time 08:44:45 just the non-return stack utilizer 08:44:49 forgot a dup somewhere ? 08:44:56 begin-while-repeat & begin until 08:45:08 ah 08:45:10 Speuler: no i didnt. i worx 08:45:17 me blind 08:46:00 it would look like: 08:46:01 : number ( tbl -- ) 0 swap begin 2dup c! char+ swap 1+ swap dup dup M M * + xor until 2drop ; 08:46:07 in my variation 08:46:32 : 2dup over over ; of course 08:47:00 M M * tuck + swap for 1- i over c! next drop 08:47:06 probably its faster than using 2dup, in some cases 08:47:13 [05:44] for-next doesnt play this time 08:47:41 thought you meant "can't be used because won't work" 08:48:08 "He had gone from DO LOOPs to FOR NEXT in cmForth for 08:48:08 the Novix because it was simpler and closer match to 08:48:08 how the software worked." 08:48:28 "Chuck explained that DO LOOP 08:48:28 was often used to increment a loop or manipulate 08:48:28 pointers and that it was a clumsy, resource wasteful, 08:48:28 and slow way to do it compared to BEGIN and that 08:48:28 using a third as many different words would make the 08:48:29 source code cleaner and clearer." 08:52:47 chuck "variable names only serve to confuse" moore will henceforth be known as : chuck_moore adam eve 5000000000 0 do dup loop ; 08:54:22 IF stack and data mem are shared, it shouldn't make much difference whether data is put on stack, or into mem 08:54:39 in terms of mem consumption 08:55:32 may end up with need to compact multi-char into cell on stacl 08:55:34 stack 08:56:22 seperate stacks is a diff matter though 08:56:34 and for performance, you want those 08:59:59 hmm 09:01:58 dont u think that begin-until solution is much more unreadable than do-loop? 09:05:27 "much more" ... no . i like to use begin ... while/until for an undetermined number of iterations, where the exit condition develops during execution of looping code 09:06:07 and if the number of iterations is known beforehand, do loop 09:06:48 but i'm not extremely consequent in doing so 09:07:13 c 09:07:14 thx 09:07:48 no cause .. thank you too 09:23:29 --- quit: onetom (Read error: 60 (Operation timed out)) 09:27:18 --- quit: gforth (Read error: 104 (Connection reset by peer)) 09:38:32 --- join: onetom (tom@adsl52006.vnet.hu) joined #forth 09:40:11 --- join: jim2 (~jim@12-233-225-152.client.attbi.com) joined #forth 09:41:59 --- part: jim2 left #forth 09:58:57 --- join: kc5tja (~kc5tja@stampede.org) joined #forth 10:08:59 --- join: tathi (~josh@ip68-9-58-81.ri.ri.cox.net) joined #forth 10:09:05 --- join: _svara (~svara@pD950B1E4.dip0.t-ipconnect.de) joined #forth 10:09:29 --- quit: svara (Read error: 110 (Connection timed out)) 10:09:41 --- nick: _svara -> svara 10:10:17 --- quit: tathi (Client Quit) 10:10:57 --- join: tathi (~josh@ip68-9-58-81.ri.ri.cox.net) joined #forth 10:32:38 --- quit: jim (carter.openprojects.net irc.openprojects.net) 10:32:38 --- join: jim (~jcassady@adsl-64-171-52-226.dsl.snfc21.pacbell.net) joined #forth 11:02:09 agh 11:02:13 i have so many windows right now 11:02:48 That's OK -- nothing much at all is going on here. 11:02:54 * kc5tja is just pondering a number of things. 11:07:58 hehe 11:08:03 but i have like 25 of them: ) 11:26:46 --- join: Fare (fare@samaris.tunes.org) joined #forth 11:58:55 --- join: GilbertBSD (~gilbert@max1-34.dacor.net) joined #forth 12:14:12 Hi 12:14:26 hello 12:14:33 hey! :) 12:14:40 hej rob_ert 12:15:06 Hej Gilbert. 12:21:53 --- quit: Speuler (Read error: 113 (No route to host)) 12:31:14 --- join: miket2 (Mike@62.60.91.127) joined #forth 12:32:58 Hi. 12:33:22 re 12:34:23 Hi kc5tja :-) 12:34:51 * kc5tja is documenting his networking protocol design. 12:35:04 * kc5tja decided to take a first principles approach to designing a wireless networking protocol. 12:36:13 Although it is very minimal (the code for it is rather quite trivial!), it fulfills the data link, network, and transport layers of the ISO model, in one implementation. 12:36:28 In particular, it's equivalent to UDP/IP/Ethernet. 12:36:44 Only, it's a heck of a lot easier to implement, both in software, and in hardware. 12:39:17 :)) 12:39:27 I'm trying to implement sockets in IsForth ;) 12:42:10 * rob_ert pets Mark with http://linuxassembly.org/linasm.html#Sockets 12:44:59 --- quit: onetom (Read error: 110 (Connection timed out)) 12:45:22 --- join: onetom (tom@adsl52005.vnet.hu) joined #forth 12:46:07 wb onetom :) 12:46:52 --- part: miket2 left #forth 12:51:11 --- join: I440r (~mark4@1Cust174.tnt2.bloomington.in.da.uu.net) joined #forth 12:51:24 Hey! 12:51:43 * rob_ert has started with his own sockets lib for IsForth now... 12:51:51 eep! 12:51:58 heh 12:51:59 btw, www.google.com saves you a lot of reverse-engineering. 12:52:06 i MIGHT be releasing mine some time 12:52:09 Oh, btw, license is spelled "license", not "lisence" :) 12:52:12 no it doesnt 12:52:14 Yeah, _might_ 12:52:18 google is full of FLUFF now 12:52:28 http://linuxassembly.org/linasm.html#Sockets 12:52:31 My second hit. 12:52:34 ;-) 12:52:36 you search for something and you get 3 useful pages in the results and 28946598273469823475698237465982747958724 pages of FLUFF 12:52:47 i been there 12:53:00 so look at the useful pages 12:53:21 im not JUST reverse engineering but that site does NOT tell you how the parameter to the socketcall syscall should be formatted 12:53:34 and there are about 9 idfferent sub functions on the socketcall syscall 12:53:51 the info in there is not enough 12:53:56 Just look in net.g 12:53:59 net.h 12:54:05 heh 12:54:30 dood c header files are no good. looking at a disassembly of the object code tells you EVERYTHING 12:54:40 header files bury everyting in a gazillion #defines 12:54:43 they tell you NOTHING 12:55:25 besides. i dont want to look at c code. c is ugly 12:55:28 :) 12:55:41 c makes me angry every time i hear its name 12:55:47 :P 12:55:54 if i could SHOOT it - i would :P 12:55:57 kerbanga :P 12:56:12 12:56:12 heh 12:58:06 #define SYS_SOCKET 1 /* sys_socket(2) */ 12:58:07 #define SYS_BIND 2 /* sys_bind(2) */ 12:58:07 #define SYS_CONNECT 3 /* sys_connect(2) */ 12:58:08 etc. 12:58:12 A #define is equivalent to an assembly language EQU. 12:58:15 Do I need to tell more? :) 12:58:23 For the purposes of defining an interface, they're everything. 12:58:28 Rather %define in NASM :-) 12:58:38 or const in isforth 12:58:44 0 const sys_socket 12:58:54 Yeah... 12:58:55 using the sys_socket const will compile a literal 12:59:10 and the actual definition for sys_socket will eventually be discarded on turnkey 12:59:18 so a const is a compile time constant 12:59:20 :} 12:59:41 but im opposed to defining 2785782458762438725 different consts and structs 12:59:42 Isforth _is going to_ be great ;) 12:59:48 it will make forth as ungly as c 13:00:01 all that inter dependancy (ugh) 13:00:03 How far have you come since 1.05b? 13:00:33 one of isforths design goals is to keep each sourcefile autonomous. keep dependancies to an absolute MINIMUM 13:00:49 i REFUSE to support a "needs" type word 13:01:49 ; i shall probably rename fload when i can think of a better name but the 13:01:49 ; word 'needs' and its variants will never be part of isforth. add them 13:01:49 ; yourself if you realy gotta (sorry mrreach :) 13:01:49 ; 13:01:49 ; rationale... 13:01:50 ; 13:01:52 ; foo.c includes foo.h 13:01:54 ; foo.h includes bar.h and futz.h 13:01:56 ; bar.h includes whussit.h and bleh.h 13:01:58 ; futs.h includes the already included bleh.h and.... 13:02:00 ; 13:02:04 ; what you end up with is an include tree. you have absolutely no idea 13:02:06 ; which files included which files and it is highly probable that you 13:02:08 ; are totally oblivious to some of the files that 'were' included. 13:02:10 ; 13:02:12 ; what you shoud have is a single load file including all the required 13:02:14 ; source files. if one of those files relies on some other file then 13:02:20 ; it should not include it itself. 13:02:22 ; 13:02:24 ; if your compile aborts with "bleh ?" you should know which file it was 13:02:26 ; that you forgot to include. if you dont then grep is your friend 13:02:43 --- quit: GilbertBSD ("xchat exiting..") 13:02:58 im not going to have a gazillion library include files defining everything except the kitchen sink 13:03:09 or sometimes even including the kitchen sink 13:03:16 thats one of the worst things about c 13:03:57 Hmm... same thing about assembly if you don't keep everything in one file ;) 13:04:12 no. keeping it all in one file isnt required 13:04:33 You will still get dependency trees. 13:04:37 yes 13:04:53 but part of my design philosophy is to keep that to an absolute bare minimum 13:04:55 for instance 13:05:08 color.f in the isforth source files is dependant on term.f 13:05:12 terml.f IS optiona 13:05:14 optional 13:05:26 http://www.hackles.org/cgi-bin/archives.pl?request=65 13:05:39 the tui i want to write will also require term.f 13:05:56 but the less of these dependancies i can get away with the better 13:06:08 color.f was actually stripped out of term.f 13:06:24 what i might end up doing is put subdirs under the forthsrc directory 13:06:32 one might be isforth/forthsrc/term 13:06:41 inside the tirm directory we can have 13:06:42 color.f 13:06:48 term.f 13:06:59 and window-system-thing-of-a-good-name-here.f 13:07:09 dependancies within this directory are fine 13:07:28 but NOTHING in the sockets directory sould be dependant on anything in the term directory 13:07:31 etc etc etc 13:08:10 the sockets directory might have a unix-domain.f 13:08:16 tcp.f 13:08:18 telnet.f 13:08:22 dns.f 13:08:27 etc etc etc 13:09:06 hrm maybe not telnet.f - thats an application not a library thingie 13:09:26 a telnet application might be dependant on things in the sockets AND the term directories 13:11:32 how far did you get with your sockets code 13:12:50 Hm 13:12:51 wtf 13:12:57 When I modify syscalls.1 13:13:03 And do make clean/make linux 13:13:10 The file is restored to what it was before. 13:13:12 Why? 13:13:17 lol 13:13:29 because syscalls.1 is either linux.1 or freebsd.1 13:13:40 make linux copies linux.1 to syscalls.1 13:13:41 It could be automatically generated from other source code. 13:13:41 Hmm... darn. 13:13:46 Fuck, heh... 13:13:48 make frebsd copies freebsd.1 to syscalls.1 13:13:50 lol 13:13:53 Why didn't you tell me?! ;) 13:13:58 * rob_ert goes back to coding 13:14:01 Murr... 13:14:02 that might change :) 13:14:08 why are you modifying syscalls.1 ? 13:14:14 you can add a new syscall at compile tim 13:14:29 #parameters syscall# syscall 13:14:36 the word syscall creates a new handler 13:14:42 it takes 2 parameters 13:14:48 Hmm... OK. 13:14:50 the number of parameters the syscall needs 13:14:54 and the number of the syscall 13:14:57 * rob_ert changes it anyway :-) 13:15:01 lol 13:15:10 what change ??? :P 13:15:27 linux.1 is PERFECT!!!! 13:15:37 its the best syscall code ive seen on any application 13:15:50 no need for 2984762794 different syscall handlers for different number of parameters 13:15:56 what a fscking MESS 13:15:56 lol 13:16:06 the linux syscall macros are C R A P 13:16:17 and you can quote me directly to linus t on taht one 13:16:37 so how would you do it? 13:17:50 the way i did do it 13:17:55 deltab has isforth sources ? 13:18:06 look in isforth/asmsrc/linux.1 13:19:07 hello =) 13:20:06 the word "syscall" (the creating word) is in compile.1 - at the bottom i think 13:20:57 isforth doesnt define every single syscall there is. the kernel only defines those calls it actually uses 13:21:15 but it gives the user the ability to add a handler for any syscall at compile time 13:23:16 hi gosh! 13:23:56 hey hey, i4 13:23:59 how are you, today? 13:24:17 not too bad 13:24:24 good to hear, what's new? 13:24:56 not much been working on my n^2-1 puzzle solver 13:25:10 on theh 3x3 level it gets to within 3 moves of solved and segfaults :P 13:25:11 lol 13:25:39 ouch 13:25:55 that's cool, though...where did you encounter those kind of puzzles? 13:26:56 i used to get them in these cheezy xmas stocking things heh 13:27:00 when i was a kid :) 13:27:07 plastic 4x4 things 13:27:52 oh, yeah...I know now 13:28:19 I didn't know that was what you called them 13:28:24 neat 13:28:27 i could never solve them when i was a kid but i can solve a 6x6 one easilly heh 13:28:48