00:00:00 --- log: started forth/06.04.15 00:07:02 --- quit: crc (Read error: 104 (Connection reset by peer)) 00:41:33 --- join: Cheery (i=Henri@a81-197-60-217.elisa-laajakaista.fi) joined #forth 01:05:57 --- quit: Ray_work (Read error: 104 (Connection reset by peer)) 03:25:31 --- join: Alterego (n=Alterego@88.202.136.178) joined #forth 03:37:01 --- join: crc (i=crc@pool-70-110-220-204.phil.east.verizon.net) joined #forth 03:37:52 --- join: neceve (n=Clau@unaffiliated/neceve) joined #forth 04:32:35 --- join: tathi (n=josh@pdpc/supporter/bronze/tathi) joined #forth 05:11:37 --- quit: tathi ("leaving") 05:56:00 --- join: PoppaVic (n=pete@0-1pool73-232.nas24.chicago4.il.us.da.qwest.net) joined #forth 06:16:26 --- quit: neceve (Read error: 113 (No route to host)) 06:26:04 --- join: tathi (n=josh@pdpc/supporter/bronze/tathi) joined #forth 06:38:55 --- join: oldman (n=nobody@lsi-01.bit.uni-bonn.de) joined #forth 06:55:59 --- part: oldman left #forth 07:13:40 --- join: JasonWoof (n=jason@pdpc/supporter/student/Herkamire) joined #forth 07:13:40 --- mode: ChanServ set +o JasonWoof 07:38:57 Every once in awhile I peer at some of the ficl source - and I think "Clever Girl".. It's like watching Jurassic Park and waiting for the leap from the side. 07:42:02 FICL seems to be pretty well written 07:42:08 Just wordy :) 07:42:54 Yeah... I was looking over a few files, and I was in awe of the simple, yet brilliant use of a macro-pair and a "datafile" using them - in at least 3 files and 4 places. 07:43:36 ..when I think of all the work that saves, and how that clarifies - it reminded me of a lot of insanity I was trying in CommonForth code a few years ago, to try the same sorta' thing. 07:44:26 I'm still trying to cogitate the vm, and their body-structures... 07:48:10 tathi: looking at "struct ficlWord" - how can 'code' and 'semiParen' both be native-code, yet different types? particularly before param (pfa?) 07:49:06 damn forth! all these freekin words everywhere ;) 07:49:26 that's cool :) it's not every day you find code that is written well 07:50:03 I'm trying to rethink the "words everywhere" issue.. I've been stymied the last few days. 07:50:41 vor runs great on my laptop :) 07:50:50 whats vor again? 07:51:15 variations on rockdodger 07:51:19 the best computer game ever 07:51:36 there's a beta for windoze 07:53:20 josh and I have been hacking on it for a while (mostly josh) 07:53:22 tathi: are they trying to mix together bytecode-style and func-ptr styles of pcode? 07:54:07 this laptop appears to have 4 mouse buttons, but none of them seems to be middle-click 07:54:37 this is a pIII 260Mhz 07:54:52 protege 3440CT 07:55:15 wasn't sure it could handle the game at a nice framerate, but I think it's smoother than on my desktop 08:00:17 --- join: mikett (n=nobody@80-195-208-58.cable.ubr14.live.blueyonder.co.uk) joined #forth 08:01:48 oh, lord.. it is. *sigh* 08:02:09 That is the biggest, ugliest goddamned switch/case I have seen in my life. 08:05:08 which game? 08:05:22 ficl4 08:05:40 vor 08:05:47 vor? 08:05:54 jasonwoof.org/vor 08:06:13 oh, that one 08:06:16 quick action game that tathi and I have been hacking on for quite a while 08:06:54 I need to continue xell, I'd like to do with it games. 08:07:26 xell is a chip made by ibm right? 08:07:42 can you buy a device with xell in it? 08:08:03 eh, cell is the ibm chip. xell is my forth project. 08:08:10 oh 08:08:21 xell is software? 08:08:49 ok, I feel much better about that. I thought you were obsessing about some hardware that wasn't even available yet 08:10:15 PoppaVic: the so-called "switch threading", eh? :) 08:10:37 the cell chip would be also a thing that would be very cool. when it's sold, I'd like really a workstation with one in it. 08:11:33 arke: Apparently they 'code' a shitload of 'primitives' - using an enumeration; then they want to run a master "inner interpreter" - which can fault-out to just executing a func-ptr. 08:12:01 the 'master' looks like it has multiple Kilo-lines of cases to handle. 08:12:29 stupid design decision 08:12:38 it would be faster with a single function pointer to behin with 08:12:45 one thing I did correctly, using func ptrs instead of switches. 08:12:54 Yeah... I'm thinking a func-ptr is prolly a lot more useful. 08:13:41 I've pretty-well concluded yer further ahead with the cfa being a f-ptr and let THAT code know how to "interpret" it's own "body" 08:17:56 the only thing which I dislike about my current implementation is that it only saves func-ptrs to primitives instead of the machine code. iirc that's some kind of indirect threading with func-ptrs. 08:19:05 but that should be portable, when visual C shit can compile gcc -ansi proved C code. 08:19:13 If I want REAL mc, I let C generate asm to generate the mc - otherwise, I'd have to learn all about the guts of libs and shit. 08:20:00 whats real microcode 08:20:14 machine-code; not micro 08:24:49 how do you do calls to user-defined functions? 08:26:11 well, for distribution for pix(=portable implementation of xell) I use C which produces a xell vm for the installation and then I will feed it with xell code which describes machine code with this it will create the real target xell vm. 08:27:36 That confused me. Only C and Asm and such _really_ understand the files/libs/PIC-nonPIC/.o stuff. I don't want to create a new version of .o to create .so or whatever 08:28:35 I sometimes think that what I want is 2 interpreters: one is interactive and the other just spews .c and .h for building objects 08:29:14 ..and that latter sure sounds like the most generic part of cpp 08:29:18 I don't want to create a special object file format for xell, java does this, xell not. 08:30:38 And, mind you: I have nothing against a vastly improved cpp - I think that's sorta' a side-effect of my current plans. 08:31:43 PROBLEM is in getting cpp source for a compare - there are several 3rd-party tarballs, but gcc (those fucks) integrated it. 08:34:07 Hmm... maybe I can see why I'm boggled, the last few days.. I think I've done all the "bottom-up" I can and it might be time for "top-down". 08:34:19 brb 08:34:24 --- quit: PoppaVic ("recycles ISP") 08:36:03 --- join: PoppaVic (n=pete@0-1pool65-101.nas22.chicago4.il.us.da.qwest.net) joined #forth 08:36:49 Hmm... Doesn't STATE end up being more like trinary - rather than boolean? 08:37:24 interp vs compile; but, the word flags end up changing this? 08:39:24 --- part: mikett left #forth 08:54:24 * tathi thinks boolean can be enough 08:54:45 I do too. if you fork enough 08:54:57 I'm just trying to see where it all leads. 08:55:20 but word flags aren't part of the state 08:55:44 nope, they operate on it. And add at least 2 more substates (each) 08:55:58 but it's not state 08:56:00 ..if yer thinking of bits 08:56:12 it's part of the word definition really 08:56:18 tathi: no, it's behaviors. 08:56:32 I begin to think there is something we all miss. 08:57:01 ok, behaviors. but it's still not part of the interpreter's state 08:57:32 we can presume a lookup, but I begin to believe the returns should execute something - based on word-flags and body-ptr 08:58:40 I'm not sure how to phrase it, because we deal with so MANY "states".. This might be why ficl4 decided to have two forms of "native-code" references per body 08:59:47 I think most anywhere we'd have a "ascertain_state" using the body to determine what to DO - compile a ref, run code, change states, etc 09:04:49 tathi: sorta' a machine-state, plus a word-found-states, plus EITHER "smart cfa" - or something like "primitive run-on-find; primitive handler;" 09:07:52 sounds complicated 09:08:22 heh.. It all is. 09:08:28 * tathi thinks he should re-read Matthew Flatt's article _Composable and Compilable Macros_ 09:08:58 yeah.. I suggested that quite awhile ago and got flak like a mofo 09:09:21 I don't think Forth needs that much complexity 09:09:35 Though I am partial to a dictionary with action/value instead of cfa/flags 09:09:44 Forth doesn't - but forth is not what we are all working with - just the basic concepts 09:09:45 IIRC Retroforth takes that approach 09:09:57 Forth is what I'm working with. 09:10:02 I don't know about you... :) 09:10:03 action/value? 09:10:24 There is no "forth" - ANS is just a variant - they are all "variants"/ 09:11:17 OK, I'll rephrase -- I don't think any forth-like system needs as much complexity as you seem to be talking about. 09:11:53 I'd agree, except I am trying to integrate C and overcome a pile at the same time. 09:12:17 For "dirty shit", I'd just as soon write C or gforth. 09:12:39 Yeah, and I think that the amount of C integration that you can do is relatively limited and not particularly hard to deal with. 09:12:46 can do portably 09:12:51 perhaps 09:13:31 Certainly I have no interest in thinking I am generating .o that would transport clandistinely elsewhere 09:22:13 --- join: neceve (n=Clau@unaffiliated/neceve) joined #forth 09:41:09 Hmm... maybe heads should NOT contain flags... 09:45:10 --- join: mikett (n=nobody@80-195-208-58.cable.ubr14.live.blueyonder.co.uk) joined #forth 09:49:20 I like package managers 09:49:24 well, I like portage anyway 09:49:40 just told it to install vlc with support for loads of codecs 09:49:46 and now it's installing 27 packages for me 10:08:52 humph... reconciling heads/bodies to a key/value system is cute. 10:23:19 --- quit: PoppaVic ("Pulls the pin...") 10:29:05 --- join: Ray_work (n=Raystm2@adsl-68-90-192-85.dsl.rcsntx.swbell.net) joined #forth 10:41:00 what did he mean? 10:42:31 PoppaVic? who knows? 10:49:58 god? 10:54:37 --- part: mikett left #forth 11:13:34 --- mode: ChanServ set +o crc 11:32:36 --- quit: Alterego (Read error: 110 (Connection timed out)) 11:44:04 --- join: Serg_Penguin (i=Serg_Pen@ppp85-140-29-66.pppoe.mtu-net.ru) joined #forth 12:16:19 --- join: snoopy_1711 (i=snoopy_1@84.58.152.183) joined #forth 12:24:22 --- quit: Snoopy42 (Read error: 145 (Connection timed out)) 12:24:33 --- nick: snoopy_1711 -> Snoopy42 12:26:56 --- quit: tathi ("leaving") 13:25:19 --- quit: Serg_Penguin (Read error: 104 (Connection reset by peer)) 14:20:43 virl: how much have you done on xell so far? 14:39:14 lots of theoretically blurb, some practically and then there is pix where I'm working on. 14:45:51 but I'm only a human and so nothing compared to our holy god slava. 14:49:14 * virl nows sees that his current way of xell is simpler than his original way 15:00:15 --- quit: neceve (Remote closed the connection) 16:14:58 --- quit: Cheery ("Leaving") 18:30:43 --- quit: uiuiuiu (Read error: 104 (Connection reset by peer)) 18:35:52 --- join: uiuiuiu (i=ian@dslb-084-056-234-002.pools.arcor-ip.net) joined #forth 20:32:31 --- quit: nanstm (Read error: 110 (Connection timed out)) 22:06:22 --- join: neceve (n=Clau@unaffiliated/neceve) joined #forth 22:38:17 --- quit: JasonWoof ("off to bed") 23:59:59 --- log: ended forth/06.04.15