00:00:00 --- log: started forth/03.07.21 00:13:47 * Serg_Penguin is shutting all apps to fry a backup 00:13:50 --- quit: Serg_Penguin () 00:22:28 --- quit: onetom (leguin.freenode.net irc.freenode.net) 00:22:32 --- quit: Robert (leguin.freenode.net irc.freenode.net) 00:22:34 --- quit: a7r_ (leguin.freenode.net irc.freenode.net) 00:22:35 --- quit: ianni (leguin.freenode.net irc.freenode.net) 00:22:54 --- join: a7r_ (~a7r@206.72.82.135) joined #forth 00:22:54 --- join: onetom (~root@cab.bio.u-szeged.hu) joined #forth 00:22:54 --- join: Robert (~snofs@h126n2fls31o965.telia.com) joined #forth 00:22:54 --- join: ianni (ian@inpuj.net) joined #forth 00:24:35 --- join: CrowKilr (CrowKilr@Ottawa-HSE-ppp3652913.sympatico.ca) joined #forth 00:25:46 * CrowKilr advice of the week: listen to Shpongle - The Sixth Revelation while on Salvia Divinorum 00:26:49 my bincopy tokens simply becomes a different color of comment, with a slight twist: the count is stored in memory shifted 00:26:54 aka decoded 00:27:08 to be able to use it as a sign-change counted string 00:28:36 * kc5tja votes that we rename freenode.net to netsplit.net 00:28:48 [03:27] -lilo- [Global Notice] Hi all. Some quick rehubbing, which we hope will resolve our current problems. It's worth reminding you that freenode needs more hubbing, http://freenode.net/sponsoring_servers.shtml .... thanks 00:28:57 lilo is a jackass 00:29:23 --- quit: onetom (leguin.freenode.net irc.freenode.net) 00:29:25 --- quit: Robert (leguin.freenode.net irc.freenode.net) 00:29:30 * kc5tja doesn't have any beef with lilo. 00:29:39 --- join: Robert (~snofs@h126n2fls31o965.telia.com) joined #forth 00:29:39 --- join: onetom (~root@cab.bio.u-szeged.hu) joined #forth 00:30:02 --- quit: a7r_ (leguin.freenode.net irc.freenode.net) 00:30:03 --- quit: ianni (leguin.freenode.net irc.freenode.net) 00:30:14 --- join: a7r_ (~a7r@206.72.82.135) joined #forth 00:30:14 --- join: ianni (ian@inpuj.net) joined #forth 00:30:40 i got now the almighty coma word that solves all my troubles! 00:31:30 by using movs to AH and keeping the bytecount in al, i get tight code, cleaner design, no unnecessary move in memory and self-confidence that might future code will be almost optimal 00:31:41 i knew there was something I could do! 00:31:51 :) 00:32:20 yay! viva Sally! SHe appears to me most of the time as a dancing female being 00:32:21 --- quit: a7r_ (leguin.freenode.net irc.freenode.net) 00:32:21 --- quit: ianni (leguin.freenode.net irc.freenode.net) 00:32:21 --- quit: onetom (leguin.freenode.net irc.freenode.net) 00:32:21 --- quit: Robert (leguin.freenode.net irc.freenode.net) 00:32:27 mostly distotionned because of teh pystrance 00:32:35 distortionned 00:33:09 --- join: onetom (~root@cab.bio.u-szeged.hu) joined #forth 00:33:09 --- join: Robert (~snofs@h126n2fls31o965.telia.com) joined #forth 00:33:09 --- quit: onetom (leguin.freenode.net irc.freenode.net) 00:33:09 --- quit: Robert (leguin.freenode.net irc.freenode.net) 00:35:38 --- join: Robert (~snofs@h126n2fls31o965.telia.com) joined #forth 00:35:38 --- join: onetom (~root@cab.bio.u-szeged.hu) joined #forth 00:35:48 --- join: a7r_ (~a7r@206.72.82.135) joined #forth 00:35:48 --- join: ianni (ian@inpuj.net) joined #forth 00:35:59 --- quit: onetom (leguin.freenode.net irc.freenode.net) 00:35:59 --- quit: Robert (leguin.freenode.net irc.freenode.net) 00:36:34 --- join: onetom (~root@cab.bio.u-szeged.hu) joined #forth 00:36:34 --- join: Robert (~snofs@h126n2fls31o965.telia.com) joined #forth 00:36:34 --- quit: onetom (leguin.freenode.net irc.freenode.net) 00:36:34 --- quit: Robert (leguin.freenode.net irc.freenode.net) 00:36:48 --- join: onetom (~root@cab.bio.u-szeged.hu) joined #forth 00:36:48 --- join: Robert (~snofs@h126n2fls31o965.telia.com) joined #forth 00:36:49 --- quit: onetom (leguin.freenode.net irc.freenode.net) 00:36:49 --- quit: Robert (leguin.freenode.net irc.freenode.net) 00:37:11 * CrowKilr give my vote to the netsplit.net renaming endeavour 00:37:23 lets hack some DNSes ;o) 00:38:07 --- quit: a7r_ (leguin.freenode.net irc.freenode.net) 00:38:07 --- quit: ianni (leguin.freenode.net irc.freenode.net) 00:38:22 --- join: onetom (~root@cab.bio.u-szeged.hu) joined #forth 00:38:22 --- join: Robert (~snofs@h126n2fls31o965.telia.com) joined #forth 00:38:22 Heh 00:38:31 my countedstring word is now 00:38:32 Now, see, if we had a server network written in Forth, we wouldn't have this problem. 00:38:44 mov al, [ESI-1] 00:38:49 jmp SkipComment 00:38:54 as easy as that 00:39:01 eh 00:39:30 servers are trivials to program in forth 00:39:45 since there's no filesystem they're faster 00:39:56 Servers aren't bound by filesystem speed. 00:40:14 first arg that came to mind 00:40:22 --- join: a7r_ (~a7r@206.72.82.135) joined #forth 00:40:22 --- join: ianni (ian@inpuj.net) joined #forth 00:40:40 They are bound primarily by upstream network connectivity, largely due to the end-to-end nature of TCP. :( 00:40:49 i cant believe how that small change change everything lol 00:40:52 This means, the more hops in between two nodes, the longer it takes to exchange data. 00:41:00 Yet another argument for using ATM instead of IP-based technologies. 00:41:35 asynchronous trame m? ?? 00:41:44 transfer mode? 00:41:57 Note: I'm not talking about using telco-provided ATM technology; I'm talking about using the ATM cell relay structure as defined by open, industry standards. (In case anyone was questioning my rationale) 00:42:05 Yes -- asynchronous transfer mode. 00:43:01 Using a dedicated peer-to-peer reliable transport protocol over ATM would improve data throughputs immensely. 00:43:16 my system is complete!!! cant believe it!!! finaly the ugly bincopy is gone lol 00:43:29 it was so simple! 00:43:43 yeah adsl get a facelift when prioritizing acks 00:44:11 * kc5tja nods 00:44:33 You know, I'm seriously thinking about not using a target compiler for the Linux version of FS/Forth, and just writing it raw in assembly from the get-go. 00:44:45 =) 00:44:55 now i can provide a nice encoding scheme 00:45:17 little limited but very efficient, extend to 16 bits to get all the range you need 00:49:57 * CrowKilr still thinking if he couldnt think something better without the use of an hallucinogen ;p 00:50:35 Hmm 00:50:49 Maintaining and building the vocabulary arrays by hand is not my idea of a good time. 00:51:07 * kc5tja will continue with the target compiler. 00:51:22 ha i cant do that, it breaks everything 00:51:57 definition then countedstring token 00:52:11 if i do a jump token on that def ill land straight in the count.... 00:52:26 idem for call 00:53:18 im talking nonsense 00:53:24 wait a second ;p 00:53:47 OK, I am going to bed. It is 1:00AM here. 00:53:55 here its 4 lol 00:54:07 --- quit: kc5tja ("THX QSO ES 73 DE KC5TJA/6 CL ES QRT AR SK") 00:54:30 --- quit: CrowKilr ("gonna think how this thing's gonna work, cause now it doesnt") 01:25:36 hi 01:44:22 --- join: Serg_Penguin (Serg_Pengu@212.34.52.140) joined #forth 02:02:24 --- quit: Serg_Penguin () 05:00:37 --- join: mur (jukka@baana-62-165-186-178.phnet.fi) joined #forth 05:17:57 --- quit: a7r_ (Read error: 110 (Connection timed out)) 08:03:58 --- join: flyfly (~marekb@ip164.ktvprerov.cz) joined #forth 08:08:47 --- quit: flyfly (Client Quit) 08:08:53 --- quit: onetom (leguin.freenode.net irc.freenode.net) 08:08:54 --- quit: Robert (leguin.freenode.net irc.freenode.net) 08:09:33 --- join: Robert (~snofs@h126n2fls31o965.telia.com) joined #forth 08:09:33 --- join: onetom (~root@cab.bio.u-szeged.hu) joined #forth 08:49:19 --- join: TreyB (~trey@cpe-66-87-192-27.tx.sprintbbd.net) joined #forth 08:57:49 --- quit: mur (leguin.freenode.net irc.freenode.net) 08:58:10 --- join: mur (jukka@baana-62-165-186-178.phnet.fi) joined #forth 08:58:55 --- quit: Fractal (leguin.freenode.net irc.freenode.net) 08:59:03 --- quit: XeF4 (leguin.freenode.net irc.freenode.net) 08:59:19 --- join: XeF4 (xef4@lowfidelity.org) joined #forth 08:59:19 --- join: Fractal (bron@i.either.got.mad.cow.from.alberta.beef.or.strongLSD.com) joined #forth 08:59:54 --- quit: TreyB (leguin.freenode.net irc.freenode.net) 08:59:56 --- quit: skylan (leguin.freenode.net irc.freenode.net) 09:00:21 --- join: TreyB (~trey@cpe-66-87-192-27.tx.sprintbbd.net) joined #forth 09:00:21 --- join: skylan (sjh@vickesh01-4504.tbaytel.net) joined #forth 09:01:50 --- quit: onetom (leguin.freenode.net irc.freenode.net) 09:01:53 --- quit: Robert (leguin.freenode.net irc.freenode.net) 09:01:56 --- quit: ianni (leguin.freenode.net irc.freenode.net) 09:01:59 --- quit: TreyB (leguin.freenode.net irc.freenode.net) 09:02:04 --- quit: skylan (leguin.freenode.net irc.freenode.net) 09:02:05 --- quit: Fractal (leguin.freenode.net irc.freenode.net) 09:02:09 --- quit: XeF4 (leguin.freenode.net irc.freenode.net) 09:02:12 --- quit: mur (leguin.freenode.net irc.freenode.net) 09:02:14 --- quit: ChanServ (leguin.freenode.net irc.freenode.net) 09:02:38 --- join: ChanServ (ChanServ@services.) joined #forth 09:02:38 --- join: skylan (sjh@vickesh01-4504.tbaytel.net) joined #forth 09:02:38 --- join: TreyB (~trey@cpe-66-87-192-27.tx.sprintbbd.net) joined #forth 09:02:38 --- join: Fractal (bron@i.either.got.mad.cow.from.alberta.beef.or.strongLSD.com) joined #forth 09:02:38 --- join: XeF4 (xef4@lowfidelity.org) joined #forth 09:02:38 --- join: mur (jukka@baana-62-165-186-178.phnet.fi) joined #forth 09:02:38 --- join: onetom (~root@cab.bio.u-szeged.hu) joined #forth 09:02:38 --- join: Robert (~snofs@h126n2fls31o965.telia.com) joined #forth 09:02:38 --- join: ianni (ian@inpuj.net) joined #forth 09:02:38 --- mode: leguin.freenode.net set +o ChanServ 10:22:29 --- nick: mur -> mur_masterchef 11:01:02 --- join: a7r_ (~a7r@206.72.82.135) joined #forth 11:03:50 hi 11:04:07 szerusz onetum :) 11:04:30 * onetom is almost done w his color editor 11:04:33 hi mur 11:04:38 :) 11:04:41 so 2 say, terve 11:06:35 my father is experimenting w SOL 4 x86 :) 11:06:55 http://t3x.dyndns.org/odds_ends/SOL/index.html 11:07:54 he was searching 4 a standalone 4th running on a 386 laptop 11:27:22 --- nick: mur_masterchef -> mur 12:41:47 --- join: CrowKilr (CrowKilr@HSE-Windsor-ppp251557.sympatico.ca) joined #forth 12:42:16 hi everyone 12:43:07 hi 12:43:26 hey 13:04:40 * CrowKilr tries to change his binary string (formely bincopy) handler 13:06:14 the age old literal problem.... 13:10:02 --- join: ASau (~asau@158.250.48.196) joined #forth 13:10:18 Dobryj noczer! 13:10:52 privet aleksej! 13:11:31 Any news? 13:11:52 tram accident in helsinki 13:11:56 at least one dead 13:12:16 :))) 13:13:44 You remind me of our beloved movie "Moscow doesn't trust tears". 13:15:04 ahaa 13:15:05 :) 13:15:09 is it good or bad? :) 13:15:47 You should answer: "There is no stability in the world... It was raining in the morning. Terrorists have taken airplane..." etc 13:15:59 ;) 13:16:03 That's not bad. 13:16:16 at the time there were about 1000 people at the area 13:16:19 it was 5 pm 13:16:40 the market it collapsed has a bit status 13:16:54 it coudl be called a bit like "harrods" 13:17:09 you can buy anything, but the price is at the same quality too :) 13:18:31 I'm not interested in such accidents. 13:18:40 :) 13:18:54 It's like "stone has fallen here not there". 13:19:06 okay 13:19:07 :) 13:19:24 While stones fall everywhere. 13:19:30 WIth constant density. 13:19:42 ;) 13:19:56 when i was in army one dude shot himself 13:20:06 and downstairs (i wasn't in th ebuilding then :) 13:20:10 at 13:20:12 not and 13:21:28 Since january 7 guys have gone into windows of our hostile. 7 deathes, of course, 6th floor and upper. 13:21:47 I knew one of them. 13:21:54 :P 13:21:55 that's bad 13:22:45 LAN's phorum had long discussions on such incidents. 13:23:08 well every dead people saves earth little 13:23:15 (if no new ones born) 13:23:38 there are too many people on earth anyways 13:24:37 After 4th or 5th one such opinion has been told, of course. 13:25:05 It had become usual by that time. 13:25:51 News were told as: "O, yet another one..." 13:26:20 I'm mistaken, since Jan. 9 guys. 13:26:45 And 5th-6th. 13:27:09 Well, let's close topic. It is not interesting. 13:28:20 :) 13:28:25 what other? 13:29:19 Just to close to Forth. 13:29:38 What are you doing in it now? 13:30:07 ASau: my editor is almost done 13:30:07 Or, for it. 13:30:33 Dobryj veczer, onetom! 13:31:28 I've not yet looked at it, but I'll definatly look. 13:31:29 terve 13:31:49 jo ejt (not sleeping yet thogh:) onetom :) 13:31:55 its already much more sophisticated 13:32:04 mur: oh :) 13:32:40 i ddin't remember evening :) 13:32:47 word for "evening" :) 13:33:01 este 13:33:08 jó estét 13:33:44 May I ask to write umlauts as "ae", "oe"? 13:34:05 I can't understand as I have another encoding. 13:34:06 it's 'o and 'e :) 13:34:16 not ae and oe 13:34:20 Aha. 13:34:27 yup 13:34:31 and if you can't show ulauts in finnish, they are ignored and you use a and o etc 13:34:32 Accented one also. 13:34:55 but i dont use accented chars usually 13:35:23 onetom, you write jo8 instead? ;) 13:35:44 sure :) 13:36:09 but only 4 ppl who r good @ code switching 13:36:21 eg experienced chatters 13:36:40 ? 13:36:47 I don't like to see j-"e-backwards" staying together. 13:36:48 code switching? 13:37:08 code switch = switching between languages 13:37:20 :) 13:37:22 its a linguistic term 13:37:28 ahaa :) 13:37:46 ASau, perhaps because j and e-backwards make normal e in russian :) 13:38:05 I don't like switching at all. And here I don't know codepage you use. 13:38:16 No, that's 13:38:47 "e-oborotnoje" (Ukr/Bel) not "e" (Rus/...) 13:39:34 And neither Ukr. nor Bel. has letter "j" 13:40:39 --- join: gilbertdeb (tvmn@fl-nken-u2-c3b-152.miamfl.adelphia.net) joined #forth 13:40:46 Good evening! 13:40:47 terve gilbertdeb :) 13:40:49 hello 4th smiths. 13:40:53 terve mur :) 13:41:11 whats new in 4th? 13:41:26 We're looking for it too. 13:42:00 hahaha. 13:42:21 Maybe you have news? ;) 13:42:31 not really. 13:42:40 I just started using plan9, if thats news enough. 13:43:13 And what's you impression? 13:44:02 I have to finally learn C to read through its sources etc to understand it. 13:44:28 it is a minimalistic environment compared to linux or freebsd. 13:44:47 I sitll haven't done anything interesting enough in it, but it has my attention. 13:47:24 Hm. I thought of minimal Forth/Linux instead of GNU/Linux. 13:47:41 Linux is really big. 13:48:21 what would your Forth/linux do though? 13:48:22 Here I call "Linux" the kernel only. As it is in real. 13:48:29 oh okay. 13:49:26 have you ever looked at plan9? 13:50:21 I think that these tools can be replaced by Forth: bash, awk, sed, perl, dc/bc, cat, cut, cc and others. 13:50:35 I have no machine to run it. 13:50:57 you can run the demo. just burn the iso, it'll give you the option of either installing or running its demo. 13:51:00 I only read papers (pages). 13:51:15 ah good. the papers have some 'stuff' in there as well :) 13:52:36 Their idea of FS view is good one. 13:52:50 --- join: kc5tja (~kc5tja@ip68-8-206-137.sd.sd.cox.net) joined #forth 13:52:50 --- mode: ChanServ set +o kc5tja 13:52:58 Linux and BSD are inconsistent in it. 13:53:06 Good evening! 13:53:12 hi kc 13:53:24 re 13:53:40 how's your project coming along? 13:53:52 Which? 13:53:59 dolphin 13:54:17 ASau, I think plan9 has a potentially sane environment to play in. 13:54:21 i *think*. 13:54:38 I haven't worked on it in years. Forth has superceded it. 13:55:34 kc whats new in 4th then? 13:55:46 as in, what have you been up to regarding 4th. 13:55:59 Nothing I've made public recently, because no code has been written yet. 13:56:15 hi 13:56:16 But I've completed the initial designs of FS/Forth's new vocabulary structure. 13:56:25 kc,gil 13:56:46 cool. 13:56:49 hi onetom. 13:57:26 It should offer roughly the same compile-speeds as ColorForth, but without pre-encoding or pre-tokenizing source text. 13:58:39 Will you share ideas or code? 13:58:40 what is the "FS" part? 13:58:58 ASau: I've been sharing them since day one. The question is, are people willing to accept them? 13:59:17 :) 13:59:22 Where can I read? 13:59:26 gilbertdeb: Initials for an older company I created, called Falvo Software Solutions, which later became just Falvo software. 13:59:38 ok. 13:59:58 ive written some code in my own 5 colors bytesized colorforth, but im now rethinking the whole thing, since ambiguities arose 14:00:03 we really gotta do something about them 4th titles. 4thsmith or 4thrite ;) 14:00:17 ASau: Well, I haven't written anything up on a website yet, but I described the system several times in this channel already 14:00:21 CrowKilr: wow. what were the 5 colors? 14:00:31 Basically, each vocabulary consists of three arrays. 14:00:48 The arrays are the length array, the name array, and the CFA array. 14:01:20 red to define, green to compile call, yellow to execute, magenta for "binary citation" and white for comments 14:01:41 Each element of the length array stores the length of the word's name, plus the first three characters of the name. As you might expect, each element is 32-bits wide. 14:02:10 The length array is used to scan for potential matches. But, three significant characters in a word name isn't enough. So the name array elements contain 16 additional characters. 14:02:45 Thus, there are 19 significant characters to a word's name. 14:03:03 onetom: its quite easy to code in it, but it might be made better, since the source is loaded in ram and the bin citation duplicate information already present, thats my main grief 14:03:13 Each element of the CFA array is just a pointer to the word's definition. 14:04:31 aha 14:04:49 Hm. 14:05:15 It can be one array of records LEN-NAME-CFA 14:05:30 It can, but that is inefficient. 14:05:48 It's slower, because less significant data is packed into a cache line taht way. 14:06:27 Ah, I thought you'd rely upon this. 14:06:43 Scanning the vocabulary database would cause a lot of cache line refills. 14:06:50 * kc5tja nods -- why not? ;) 14:07:13 So, SCASD? 14:07:16 I even considered inserting PREFETCH instructions in the scan code, to let the CPU prefetch the next cache line while it was busy searching the current one. 14:07:35 Nope -- too slow (CISC instructions are very slow on x86 architectures now) 14:08:07 prefetch makes a huge difference 14:08:22 look at amd optimization guide, even intel have some data 14:08:22 kc now? 14:08:24 how is that? 14:08:27 CrowKilr: Yeah, but not all CPUs support those instructions. :( 14:08:36 I mean, in fact it's old SCASD, but coded maybe differently. 14:09:54 gilbertdeb: SCASD requires read and write access to a whole multitude of registers. But the register file is optimized for two reads, one write, per cycle. Hence, in order for SCASD to complete, it must consume multiple cycles to affect all the register reads and writes. PLUS it's dependent on memory access to boot. 14:10:46 You can achieve the same functionality as SCASD by hand-coding it in regular, simpler x86 instructions. 14:11:06 ack. 14:11:09 These instructions have the benefit of conforming to traditional RISC behaviors, which makes them much, much more adaptable for optimizing. 14:12:02 Many of the instructions used to implement SCASD are, in fact, pairable with each other, thus making more effective use of the CPU's time by completing the scan twice or three times as fast as a normal SCASD would, even assuming a SCASD could compare one memory per clock cycle. 14:12:02 :) 14:12:22 kc, How do you want to manage memory for voc.? 14:12:39 ASau: I just pre-allocate a chunk of RAM to hold the arrays. 14:13:04 I allocated the Forth dictionary space to hold 1024 entries (comes to 24KB of memory), and the Compiler vocabulary to hold 256 entries (6KB of memory). 14:14:39 Is there possibility to extend, to truncate, to set pre-allocation size manually? 14:15:09 Once the image is compiled, it is fixed. 14:17:43 You should diminish pre-alloc. size for Forth voc., 1024 is too many. There should not be so many of words. Let programmers break in word sets. 14:18:14 VLists of 1000 words! 14:18:32 I don't understand it. 14:18:53 Word sets are broken by design -- they encourage large, monolithic, everything-including-the-kitchen-sink style of programming. 14:19:16 No, you're wrong. 14:19:33 Don't believe me? Look at gforth. 14:19:40 Look at most any ANSI-compliant Forth. 14:19:55 If you switch vocabularies often, you can write everything like in natural. 14:19:58 kc isn't that the sort of thing onyx (?) tries to solve? 14:20:07 I know Gforth. They are wrong. 14:20:39 ASau: The fact is, no matter how you skin the cat, you still need the cat to skin in the first place. 14:20:49 I use tree-style vocabularies. 14:20:55 And I don't. 14:20:57 Deal with it. :) 14:21:23 When I need a tool, it is compiled right when I need it. 14:21:47 When I don't need a tool anymore, or if I switch to another tool, it is removed from the dictionary. 14:21:59 I have zero need for hierarchial, infinite vocabularies. 14:23:57 If I have one word, consider "open", for different operations, like file or window opening, I want to call it "open" and compiler should guess what I want. 14:24:04 Having implemented several Forths with hierarchial vocabularies, I can say with utter impunity that they suck. They're hard to implement, they make recycling the dictionary significantly harder to support than is necessary, and they don't offer any help in organizing code. At least, none that I've found. 14:24:20 I don' want to invent new words like "file-open" and "window-open". 14:24:22 Hence, I'm abandoning them in favor of the fixed, two-vocabulary structure (FORTH and COMPILER). 14:24:28 ASau: Then don't. 14:25:16 In fact, what you should be doing is taking extreme pride in the fact that you can redefine words at will, to mean precisely what you want them to mean, at any time. 14:25:32 kc, That's why I have nested in FORTH vocs FILE and WINDOW ;) 14:25:34 There's no reason why open can mean "open file" at one point in the program, and "open window" in another. 14:25:55 ASau: But to use those words, you still have to switch vocabularies. 14:25:58 There is CONTEXT :) 14:26:01 ASau: But to use those words, you still have to switch vocabularies. 14:26:13 At some point, some how, you're telling Forth which open to use. 14:26:21 So what difference is there? 14:26:45 Vocabularies are a red herring. 14:26:46 Yes, but I set context once and stay in it till I need another context. 14:27:01 And I don't need "context" (the variable) at all. 14:27:07 kc5tja: what is that COMPILER voc? immediate words/macros? 14:27:27 At one point in the program, open will open files. At a later point, open will open windows. At a still later point, open will open treasure chests in a game. 14:27:32 Context is implicit. 14:28:10 onetom: Yes. FORTH words are interpreted/called, and COMPILER words are executed immediately by the compiler. COMPILER words are NOT visible to the interpreter. 14:28:29 Do you provide opening files, windows and chests in mix? 14:28:40 ASau: i dont know wheather is kc right, but come on, give us a real life example where vocs r useful & unavoidable 14:28:42 ASau: No, but neither do you. 14:28:57 I do really. 14:28:58 let this example b simple 14:29:27 Show me where vocabularies are absolutely necessary, and I'll show you a way in which to do the exact same thing without them. 14:29:55 ASau: yeah! show us. im also very interested 14:30:08 kc, you can write everything in asm or hex, it is sufficient, hence forth is not needed too. 14:30:23 ASau: Stop changing the argument. 14:30:55 Last I checked, writing things in asm or hex had nothing to do with what we're currently discussing. 14:31:01 That's why I use vocabularies, I don't want complex words. 14:31:08 * kc5tja sighs 14:31:11 You're impossible to deal with. 14:31:52 I just say "FILE" and since then everything I say is said about files. 14:32:08 If I "open", "close", "do-this 14:32:13 Yeah . . . so? 14:32:16 Your point? 14:32:21 "do-that", all deals with files. 14:32:21 : open ...something about files... ; 14:32:26 : foo ... open .... close ; 14:32:32 : open ....something about windows .... ; 14:32:36 : close ....ditto.... ; 14:32:41 : bar ....open.....close.... ; 14:32:46 I mean, c'mon!!! 14:32:57 What is so frigging hard, OR COMPLEX, about that? 14:33:26 its not the problem 14:33:38 But that's what ASau is claiming the problem to be. 14:33:39 ASau would ask: 14:33:58 how would u use the 1st open & the 2nd open in 1 definition 14:34:04 U would ask: 14:34:09 He didn't ask that. 14:34:18 why should i use in 1 definition 14:34:56 asau should show a nice real life example where u should do that 14:35:02 "Do you provide opening files, windows and chests in mix?" 14:35:14 He didn't ask that, because he knows that the result would be far worse than not having file-open and window-open. 14:35:25 Here's why: 14:35:33 using wordlist is a very object oriented ... whatever.. view? 14:35:46 : foo [ FILES ] ...open... [ WINDOW ] ...open... [ CHESTS ] ...open... ; 14:35:53 So where is the advantage to vocabularies here? 14:35:55 kc: : ed ... ; ? 14:36:56 ASau: Please, if you're going to ask a question, ask it. That means nothing to me. 14:37:12 kc, You have no need to write: [ FILE ] open 14:37:22 each time you want to open file. 14:37:30 I didn't say you did. 14:40:07 That's one definition where you have to open (given) file and window. (Assuming window environment, either graphics or text). 14:40:17 No it's not. 14:40:33 Any sane programmer will factor his code into many different modules. 14:40:40 One module will be dedicated to the user interface. 14:40:43 One, to file. 14:40:55 The top-level of an editor will invoke the services of subsidiary modules. 14:41:45 So, since we know full-well that the UI-related module will know that 'open' will open a window, WHY have any need for vocabularies at all? *ALL* the UI-code will be grouped into one place in the system, and *ALL* will depend on one definition of open. 14:42:10 When it is time to start the files module, we redefine open to work with files instead. 14:42:26 Now all the file-related code is fully aware that open opens a file, not a window. 14:43:08 You get the same benefits of vocabularies without the implementation overhead of them. 14:43:25 And until you write your own Forth, you just won't appreciate the sheer volume of code that goes into making vocabularies work. 14:43:56 You just substitute: : A V1 B V2 B ; with : B B1 ; : B B2 ; : A B1 B2 ; 14:44:06 Nothing else. 14:44:36 I think your example is a little skewed. 14:44:43 I replace: : A v1 B v2 B ; 14:44:44 with: 14:45:17 Yes I did mistake. 14:45:19 : B ... ; : w1 .. B .. ; : B ... ; : w2 .. B .. ; : A w1 w2 ; 14:45:25 But you caught idea. 14:45:42 Basically correct. 14:46:19 kc5tja write an article!!! 14:46:31 Now, if you want to use vocabularies in your Forth environment, by all means, please do. I'm not arguing against using the features of your host environment. 14:46:43 ASau: roughly, but the view of the problem is totally different this way 14:46:43 If you have same level of abstraction, then you have to mix words dealing with objects different classes. 14:47:27 I've been programming for more than 18 years. I have NEVER done anything close to what you described above. 14:47:30 ASau: it relies on flux -- the constantly changing of modules in memory 14:48:05 onetom: If you have abstractions of the same level, that becomes simple formal renaming. 14:48:09 Literally: 14:48:10 ASau: and it assumes a fast compiler, 2 make it possible 2 compile modules before using them 14:48:16 : B1 B ; 14:48:21 I HIGHLY encourage you to research software engineering practices. They will *ALL* agree on this: KEEP LIKE FUNCTIONALITY LOCAL. 14:48:50 onetom: Even without a fast compiler. PygmyForth is NOT a fast compiler. It's horrifically slow. 14:49:08 But the key is to NOT pre-compile the modules you Think You'll Need(tm) into the Forth kernel. 14:49:31 For example, consider the case of a typical Forth for Windows. 14:49:48 It will quite often have the entire damn Win32 API in place as a vocabulary inside its image. 14:49:55 kc, matrices, integers, long integers, float point numbers of various precision make a production. 14:50:12 It's the same functionality. 14:50:36 ASau: What is the same functionality of what? 14:50:58 * CrowKilr is trying to understand Asau and now feel a glimpse of all the suffering kc5tja endured on this chan; he's so sorry ;p 14:51:20 ASau: they r only said 2 b the same 14:51:22 * CrowKilr using the me command at the thired person refering to himslef 14:51:28 The only operation is "*". But it applies different objects. 14:51:56 yeah, objects 14:52:18 * CrowKilr 's typing is ugly and silly, he'll return to what he's been doing 14:52:22 ASau: Wonderful. But what happens when you want to multiply two matrices together? A matrix and a vector? A matrix and a scalar? Do you have *THREE* distinct vocabularies to handle these situations? 14:52:25 but if u need OO concepts, then use a real oo env 14:52:36 And in all the equations I've ever worked with, I'm CONSTANTLY changing data types. 14:52:53 Which means all my code would be literally peppered with words to change vocabularies. 14:53:00 not fake it w such a weak hack as wordlists 14:53:13 kc, No, I will write three different words: 14:53:24 MUTIPLY-TWO-MATRICES 14:53:35 MULTIPLY-MATRIX-BY-VECTOR 14:53:41 and so on. 14:53:47 ooff!!!! I'm a fan of long, descriptive names, but those are a bit overkill for me. :) 14:54:49 If I switch temprorarily to quaternions, I want that "*" be production in q. algebra. 14:55:02 I'd just define Q* or something. 14:55:04 Not "Q*" or like. 14:55:09 But "*". 14:55:22 But the higher the dimensionality of numbers, the less sense * makes. 14:55:34 Because, I can use practically the same code for it. 14:55:55 Joy. 14:56:03 But that doesn't affect me. 14:56:04 eg.. 14:56:33 If I switch to another model, "*" should be linear operators composition. 14:56:38 And so on. 14:57:02 OK, I'll agree with this. 14:57:03 It is the same operation, with the same axioms. 14:57:04 This is not in contention. 14:57:07 Has never been in contention. 14:58:27 That easily explains why I need "open" dealing files and windows under the same name. 14:58:46 I find abstractions of the same level. 14:58:55 ASau: if u would like 2 think in a global namespace, then, yes, u need some subsetting mechanism what allows the recycling of names 14:58:58 Jesus Christ -- I fucking give up. 14:59:05 thats what java does, eg 14:59:08 lol 14:59:21 ASau: Have you read and understood anything I've written so far? 14:59:26 And sometimes I have to mix them together. 14:59:28 Besides the swear words? 14:59:37 ASau: Give me ONE EXAMPLE of where you mix them up. 14:59:38 kc, I've read. 14:59:40 Give me ONE EXAMPLE. 14:59:57 I mean, actual source. 15:00:03 ASau: a real life example 15:00:14 Because I can guarantee you, there is a 100% better way of doing things. 15:00:27