00:00:00 --- log: started forth/06.04.18 00:06:40 --- quit: crc (Read error: 104 (Connection reset by peer)) 00:23:41 slava, why do you hate me? 01:08:48 --- join: Ray-work (n=Raystm2@adsl-68-90-192-85.dsl.rcsntx.swbell.net) joined #forth 01:08:48 --- quit: Ray_work (Read error: 104 (Connection reset by peer)) 01:20:17 --- quit: virl (Remote closed the connection) 01:24:13 --- join: virl (n=virl@chello062178085149.1.12.vie.surfer.at) joined #forth 04:04:14 --- quit: Alterego (Remote closed the connection) 04:24:28 --- join: domino-24 (n=domino@host86-132-215-57.range86-132.btcentralplus.com) joined #forth 04:26:52 --- join: Cheery (i=Henri@a81-197-60-217.elisa-laajakaista.fi) joined #forth 04:29:44 --- quit: virl (Read error: 110 (Connection timed out)) 04:29:47 --- join: virl (n=virl@chello062178085149.1.12.vie.surfer.at) joined #forth 04:48:16 --- part: xX[ReP]Xx left #forth 05:13:26 ping 05:46:43 --- join: timlarson_ (n=timlarso@65.116.199.19) joined #forth 05:51:03 hi 05:51:12 ping 05:55:11 --- join: PoppaVic (n=pete@0-1pool46-66.nas30.chicago4.il.us.da.qwest.net) joined #forth 06:28:18 pong 06:28:53 (response delayed by lisp gc) 06:30:29 well, then it's slow. 06:33:20 but why lisp? 06:34:00 'twas a joke 06:34:30 do you code lisp? 06:35:55 a reference to a classic robot ping-pong game where one of the robots was programmed in lisp, was winning, but then just stopped for several minutes seemingly for no reason, and then started playing again...turned out to be the young implementation of garbage collection that caused the delay. 06:36:56 I have only done toy projects in lisp so far. 06:39:08 ehm, how is done gc in this implementation? that's really horrible 06:52:36 data types. 06:59:58 it was a long time ago...when gc was just starting to be experimented with. 07:03:02 how did they do it? I mean that's long, I can't simply imagine how they did that. 07:05:43 Is gc a complex process? On one side there are the pro side who say that it isn't and on the other side there are the con side like me who say no. 07:05:56 so how complex is it really? 07:07:38 there are many different gc strategies...with different tradeoffs regarding how much memory overhead each takes, how much locking (if any) takes place, and how long it pauses the program (if at all). 07:08:07 mark-and-sweep, generational, etc. 07:09:13 where do you think does gc make in forth sense? 07:09:30 if the language does not make pointers (or their equivalent) an obvious type, then the gc has to be more conservative, possibly missing some potential frees. 07:10:19 not much in tiny embedded systems, but probably usable in applications for pc's, etc. 07:10:56 sideways sentence ;-) 07:12:06 as opposed to vertical? 07:14:29 I think that GC isn't a part of a programming language like today it's used. 07:14:46 hmm? 07:14:49 please expand on that 07:16:00 java & co, for example where GC is embedded into the language, it's a concept of the programming language. 07:16:29 GC, bottom-up. 07:16:39 hmm? 07:16:43 Sloppy and time-chewing 07:16:44 how would you use gc, ideally? 07:17:26 gc, makes everywhere sense where you allocate space dynamically, so at runtime. 07:17:54 I tend to believe GC is prolly more useful to _applications_, but shouldn't be "voluntary" past the point where you say "yes, I want GC" 07:17:56 the problem with this languages I think is, that they use gc everywhere where it's not really needed. 07:18:08 PoppaVic, I agree. 07:18:35 or wait, what do you mean exactly? 07:18:39 well, that says vaguely where you would use it, but not _how_. 07:18:58 I mean.. Let's take C for example.. 07:19:11 I am writing an app - not a lib, ok? 07:19:29 ok, you are writing an app and? 07:19:29 I begin with the statement "I want GC used". 07:20:02 From that point forward, you should be unable to reach malloc & friends, and EVERYTHING is done via the failsafe GC interface. 07:20:35 PoppaVic, yes, that was I tried to point out. 07:20:41 Either you need and want GC, or you do not. The moment to vary from that paradigm, things get busted 07:20:52 to vary/you vary/ 07:21:05 virl: so I thought, yes 07:23:12 when you decide between using or not using gc, you are in effect choosing between two languages; the consequences are felt all through the code, similar to picking a different language. 07:23:41 not really "languages", more like a voc, wordlist or api 07:25:25 true, but only from the standpoint of a forth programmer...and since we were talking about multiple languages... 07:26:11 right. I tend to confuse C'ers with such terms, but I've tinkered in C for decades, and predicated that with FIG-forth and friends.. They can catch up ;-) 07:27:40 I've noticed for years that C'ers want to cry about "namespaces", while breaking rules left, right and indifferent. 07:28:08 what do you mean? 07:28:22 about the "breaking rules" part? 07:28:52 oh, all the sorts of things about underslashes, restricted-names, compound-names, contexts, globals. 07:29:54 I spent about 2 hours yesterday, trying to explain how C can really deal with Public, Protected and Private - like C, but with more work. 07:30:31 creating own datastructs.. 07:30:34 I think I finally got the guy to see what I was saying, but I was amazed none of the "gurus" jumped all over me with hobnails. 07:30:52 virl: and headers and source and build versus install 07:31:46 ok 07:32:10 and do they gurus jump regularly over you? 07:32:45 Oh, sure - same with the ANS folks here. We get nit-picky about terms, "because the standard says..." 07:33:35 otoh, I respect even those gurus - and the rest at least tolerate me with a smile and a greeting, so - it all balances out. 07:34:19 the ans forth standard is ugly, so no reason to follow it. 07:34:59 with the C ans standard I can live but not with the forth ans standard. 07:36:30 well, C uses ANSI and C-99 - so, it's just as messy (if not moreso) 07:53:38 interrupts are fine, aren't they? 07:57:57 I never touch them 07:58:12 well, ok: I kill SIGPIPE often 07:59:22 I think they are a nice way to request services in assembler. 08:02:45 I avoid them, and don't write "asm", so I can't tell you more than that. 08:04:06 I know pthreads used to take-over the USR1 signal/interrupt, and because of such issues: tis better I avoid them entirely. 08:10:14 brb - lemme' recycle early 08:10:17 --- quit: PoppaVic ("Pulls the pin...") 08:11:54 --- join: PoppaVic (n=pete@0-2pool236-140.nas22.chicago4.il.us.da.qwest.net) joined #forth 08:24:37 Hmm.... I think I'm getting snarfled. 08:28:19 --- quit: madwork ("?OUT OF DATA ERROR") 08:28:26 --- join: madwork (n=foo@derby.metrics.com) joined #forth 08:30:05 OK, someone give me their views on 'words' and 'state' and 'state-sensitive'. 08:32:18 Seemed to me that, w/i any wordlist - given a word - the word looks at "state" and it's own flags for GROSS behaviors. Question is, are they supposed to use the same "payload" of code or data, and use differing "interpreters" - or what? 08:33:30 NOW, it seems like the 'name' needs to track the 2 main-states and 1 or two body-pointers and a name-flags. 08:52:10 when can a language describe itself? 09:00:08 metacompilers. 09:00:43 1:1 versus expansion into 1:N 09:01:37 or, the latter into X:Y ratios 09:02:03 everything is filters and transliterators/conversion 09:40:15 yes, metacompilers. but I meant it at a more abstract level 09:40:41 virl: well, GCC is a "metacompiler" 09:41:26 I know 09:42:02 well, there is no "metainterpreter" - unless you consider sh and make under it. 09:42:30 I only wanted to start a discussion about the facilities a language needs to describe itself. 09:43:30 yeah, I know - and in #forth, "starting" a discussion is usually heretical and often ignored. 09:45:17 * PoppaVic fights the heretics in ##C 09:54:57 --- join: OrngeTid1 (i=orange@orangetide.com) joined #forth 09:56:24 --- part: OrngeTid1 left #forth 10:00:55 the c channel doesn't like you, ehm why? 10:01:36 some days - like #forth - I challenge their "standards". Otherwise, I'm fine there. 10:03:01 virl: prolly for the same reason "#asm" got tired of me: I challenge their basic assumptions. 10:03:22 that assembler is the only one truly language? 10:03:46 right 10:04:05 ..ignoring that "asm" is variated on any platform. 10:05:20 everyone wants to ignore everything they need to assume, and hell - they never want to assume a dozen of the same things. 10:06:35 they'd love to assume six things - and then assume the rest as MAYBE available. 10:06:52 so, this is why I fight my mountain. 10:09:40 I can ignore asm, and use C - and (worst case) is: I need GCC. I'd prefer plain old C topmost. 10:10:06 ..and that means either ignoring or codifying lower shit. 10:11:11 This is why I want to settle on source vs dlopen() and friends. 10:12:19 there is a thick line between "compile me" and "plugins" 10:22:21 --- quit: PoppaVic ("Pulls the pin...") 10:27:35 --- quit: domino-24 ("Leaving") 10:44:25 ping 10:45:03 pong 10:47:26 hi JasonWoof 10:49:50 hi 10:49:54 what's hangin'? 11:06:49 did you do any metacompiler? 11:11:34 --- join: PoppaVic (n=pete@0-2pool198-86.nas30.chicago4.il.us.da.qwest.net) joined #forth 11:35:05 moi? 11:35:24 hmm? 11:46:53 --- join: timlarson____ (n=timlarso@65.116.199.19) joined #forth 12:02:30 hi timlarson____ :) 12:02:50 --- quit: timlarson_ (No route to host) 12:13:50 --- quit: PoppaVic ("Pulls the pin...") 12:17:05 --- quit: Cheery ("Leaving") 12:17:50 --- join: snoopy_1711 (i=snoopy_1@84.58.190.181) joined #forth 12:25:24 --- quit: Snoopy42 (Read error: 145 (Connection timed out)) 12:25:37 --- nick: snoopy_1711 -> Snoopy42 12:43:05 --- join: Lars_G (n=lars@unaffiliated/lars-g/x-000001) joined #forth 13:34:21 hi...I have a great net connection, don't I? ;) 13:36:56 reall? 13:36:59 really? 13:41:37 --- quit: timlarson____ ("Leaving") 13:42:31 lol 13:46:39 and JasonWoof what'S hangin'? 13:47:35 just sitting around pretending I'm working 13:49:30 you are working... you are working... you are working... you are working... you are working... 13:52:18 that's what I need ;) some good affirmations :) 13:57:20 : I'm? 1000 for ." you are working... " cr next ; 14:01:48 : this looks like code ; 14:04:11 looks? 14:04:13 like? 14:04:15 code? 14:04:38 --- part: Lars_G left #forth 14:05:25 I don't know what I should put into the core api of xell, hell that's shitty. 15:04:31 --- join: crc (i=crc@pool-70-110-221-58.phil.east.verizon.net) joined #forth 15:04:52 hi crc 15:04:58 how are you? 15:05:13 tired and sick 15:10:01 oh? why? 15:10:15 work, and I'm still trying to get rid of this flu 15:10:33 * slava is fiddling with some register allocation code 15:10:50 crc: have you ever exprimented with optimizing out stack traffic between primitives? 15:11:09 for example 2 swap - could become just a couple of instructions 15:11:30 I have done a few experiments, but nothing much 15:13:30 what are you doing slava? trying to stuff a circle through a quad? 15:13:58 so translating stack operations to register based? 15:14:23 yeah, at the lowest levels of the call tree 15:15:22 for that it needs to know how much arguments the word takes. 15:15:36 this is known for primitives 15:15:59 i cannot do anything with registers across subroutine calls since their contents will not get preserved 15:25:51 so you have a table for primitives where the argument count is listed? ehm, can't that be a very complex task? 15:26:22 in fact i have stack effect inference for arbitrary words but its used for something else 15:27:22 doesn't seems very clean 15:27:23 primitive stack effects are given and compound words, conditionals and recursion use various rules 15:27:42 its a debugging aid 18:18:01 --- join: Raystm2_ (n=Raystm2@adsl-69-149-37-5.dsl.rcsntx.swbell.net) joined #forth 18:18:47 --- quit: Raystm2 (Read error: 104 (Connection reset by peer)) 18:36:20 --- quit: uiuiuiu (Remote closed the connection) 18:36:24 --- join: uiuiuiu (i=ian@dslb-084-056-231-214.pools.arcor-ip.net) joined #forth 19:51:24 --- join: Quartus (n=trailer@CPE0001023f6e4f-CM013349902843.cpe.net.cable.rogers.com) joined #forth 23:19:50 --- join: segher_ (n=segher@dslb-084-056-140-006.pools.arcor-ip.net) joined #forth 23:27:06 --- quit: segher (Read error: 110 (Connection timed out)) 23:57:51 --- join: ayrnieu (i=julian@pdpc/supporter/sustaining/ayrnieu) joined #forth 23:59:59 --- log: ended forth/06.04.18