00:00:00 --- log: started forth/08.12.14
00:34:45 <grub_booter> what's the correct behaviour for parse-word? if the input buffer is empty, should an empty string be placed there?
00:37:47 <grub_booter> and say i have a word like : take2 parse-word parse-word ... ; and i want to define a word which uses take2, but defaults the first with a constant string - how would i do it?
01:12:12 --- join: GeDaMo (n=gedamo@dyn-62-56-77-42.dslaccess.co.uk) joined #forth
01:54:42 <ASau> It doesn't place anything.
01:54:49 <ASau> It only parses.
01:55:22 <ASau> Yes, if it reaches the end of input, it returns empty string.
02:01:11 <gogonkt_> ASau: did you using gforth for shell script?
02:01:47 <ASau> Yes.
02:02:42 <gogonkt_> can you post one?I want learn by example :)
02:10:54 <ASau> This is how simple AWK one-liner looks in Forth:
02:10:55 <ASau> http://forth.pastebin.com/m1c3bf96b
02:11:47 <gogonkt_> pastebin forth O_O
02:15:32 <gogonkt_> How can I call other script?
02:16:32 <ASau> Quick and dirty way is SYSTEM.
02:16:47 <ASau> Or you can use INCLUDE/REQUIRE
02:17:32 <ASau> Correct way is writing wrapper around vfork(2)+execvp(3) or other similar API.
02:17:36 <gogonkt_> SYSTEM. ./script.sh?
02:17:43 <ASau> Or use mine I published on clf.
02:18:02 <ASau> s" /path/to/script" system
02:18:37 --- join: qFox (i=C00K13S@132pc222.sshunet.nl) joined #forth
02:19:00 <gogonkt_> I want use git for wikis backend,so
02:19:26 <gogonkt_> I want use git for wikis backend
02:19:27 <ASau> There's a wrapper around popen(3), you can use it and build pipes.
02:20:00 <ASau> Search for OPEN-PIPE and CLOSE-PIPE in the documentation.
02:20:09 <gogonkt_> :)
02:20:46 <gogonkt_> note this in my notes,try it later
02:22:48 <gogonkt_> Thx ASau 
03:52:32 <grub_booter> hi ASau - 'place' meaning 'empty string placed on stack' - good - saw that gforth was doing that, just wanted to make sure it was standard (my own language interprets it the same way)
03:53:02 <ASau> parse-word isn't standardized yet.
03:53:12 <grub_booter> k
03:55:01 --- quit: gogonkt_ (Read error: 104 (Connection reset by peer))
03:56:38 <grub_booter> i only implemented it to see if i could define a pseudo mvcp protocol directly on the stack (i used parse instead of parse-word for brevity as well) - basically, in mvcp, you have n units - each unit can play media and you can control each (like load <unit> <video>, speed <unit> <speed> etc) - for the pause, i just have : pause speed parse 0 ; 
03:57:17 <grub_booter> that works for me, but i was curious if forth could do that, or if i'd have to have a lower level non-parsing word which they both use
03:57:55 <ASau> Well.
03:58:22 <ASau> Forth is postfix, and if you're going to use prefix notation, you have some trouble.
03:59:05 <grub_booter> yeah - mine is also postfix - it works mainly because it's all interpreted - would be more problematic if it was actually compiling stuff
03:59:06 <ASau> It is easier to write: s" video" unit load
03:59:35 <grub_booter> yup
03:59:43 <ASau> 9600 unit speed!
03:59:45 <ASau> :D
04:00:20 <grub_booter> except that doesn't implement the protocol :-) - hence the interest in hacking something up with parse-word type semantics
04:00:21 --- join: gogonkt (n=info@125.34.35.194) joined #forth
04:00:32 <grub_booter> wife is calling - back later
04:00:42 <ASau> I don't see why it doesn't.
04:01:52 --- join: pierre- (n=pierre@95-28-64-72.broadband.corbina.ru) joined #forth
04:21:05 --- quit: GeDaMo ("Leaving.")
04:30:00 <grub_booter> ASau: well, in my case, i expose the language (or, more specifically, a selectable subset of the language) over a socket - so in theory, if i can model the entire mvcp post fix protocol using parse-word type semantics and restrict the public grammar to those, that could become the server directly and thus work with existing clients...
04:30:43 <ASau> Well... You can do that too, but it is more work.
04:31:02 <grub_booter> and sure, i could have a small post fix wrapper which invokes an internal prefix grammar (: load parse parse load! ;, etc)
04:32:13 <grub_booter> implementing it made me curious about forth's approach to it 
04:33:33 <grub_booter> that would work fine for the speed/pause thing of course - : speed parse parse speed! ; and : pause parse 0 speed! ; 
04:34:48 --- join: Al2O3 (n=Al2O3@c-75-71-187-102.hsd1.co.comcast.net) joined #forth
04:35:17 <ASau> Put first words into separate wordlist, "BASE OFF" to prevent interpreting numbers,
04:35:54 <ASau> in those head words you restore BASE, do PARSE-WORD and interprete commands.
04:39:19 <grub_booter> well, in my case, the receiving server will always interpret as strings - each unit is a separate server process - so when i get a command, i identify the unit, map to the server and can then send the args/words in prefix form - iow, it's strings all the way until i hit the actual unit server
04:40:09 <ASau> How do you identify unit?
04:40:12 <ASau> Number?
04:40:14 <ASau> String?
04:40:36 <grub_booter> the real protocol stipulates a string U<n>
04:41:27 <ASau> Why don't you change protocol to easier one?
04:42:16 <grub_booter> standards in the industry - lots of existing client implementations from various vendors to leverage
04:42:33 <ASau> ???
04:42:47 <ASau> Is it your custom protocol or existing one?
04:42:56 <grub_booter> existing
04:43:05 <ASau> Ah.
04:43:10 <ASau> That makes sense then.
04:43:22 <ASau> Is the set of units fixed?
04:43:49 <ASau> I.e. can it be configured at build time? At startup time?
04:44:33 <grub_booter> clients have control of that to a degree - they can can ask a server of a list of types of playout devices they support - and if memory serves, then can do something like create U0 <type>
04:44:49 <grub_booter> and they can also ask for existing units
04:44:52 <ASau> Hm.
04:45:34 <ASau> I'd create two wordlists: one for verbs, one for units.
04:46:30 <ASau> At start of input search order includes only verbs and BASE is reset to prevent interpreting numbers.
04:47:17 <ASau> Each verb sets search order to units and leaves BASE reset.
04:48:04 <ASau> Thus the second word is interpreted as unit, unless verb is special (like "create").
04:48:31 <ASau> Thus unit action is passed to unit itself.
04:48:48 <ASau> The rest of string unit word may interprete as it wishes.
04:50:40 <grub_booter> well, i was actually considering having the create U<n> ... thing create a new word called U<n> and when executed, i could get (say) a socket which is open to that server... 
04:51:44 <grub_booter> then i can map the <postfix-word> <unit> <args...> quite naturally
04:53:23 <grub_booter> in essence it means that i don't need to enforce the U<n> conventions - any unique id which doesn't clash with existing words will work (but from the clients pov, they're all U<n> oriented so that flexibility would be lost on them)
04:56:38 <grub_booter> ie: : load parse execute parse ." load" 3 send ; (where send takes an a count of the number of arguments + socket)
04:57:26 <ASau> If you have predefined set of commands, just pass magic numbers.
04:57:49 <grub_booter> true
04:57:55 <ASau> Or simply enumerate.
04:58:07 <ASau> It will be easy to use simple optimizations.
04:58:15 <grub_booter> indeed
04:58:38 <ASau> Like Eaker's CASE, or even call vectors.
04:59:26 <grub_booter> well, i better go back up - thanks for the chat - will probably have another crack at this this evening :-)
05:00:51 --- join: JasonWoof (n=jason@c-66-31-44-71.hsd1.ma.comcast.net) joined #forth
05:00:51 --- mode: ChanServ set +o JasonWoof
06:40:29 --- quit: aguai (Remote closed the connection)
06:40:41 --- join: aguai (n=aguai@122-116-183-8.HINET-IP.hinet.net) joined #forth
08:37:10 --- join: GeDaMo (n=gedamo@dyn-62-56-77-42.dslaccess.co.uk) joined #forth
09:42:02 --- quit: Al2O3 ()
10:49:07 --- quit: GeDaMo ("Leaving.")
11:07:13 --- quit: xjrn (Read error: 145 (Connection timed out))
13:04:26 --- join: xjrn (n=jim@c-69-181-213-99.hsd1.ca.comcast.net) joined #forth
13:04:38 --- quit: pierre- (Read error: 145 (Connection timed out))
14:03:58 --- join: Al2O3 (n=Al2O3@c-67-166-31-15.hsd1.co.comcast.net) joined #forth
15:26:56 --- quit: Al2O3 ()
15:31:18 --- join: Al2O3 (n=Al2O3@c-67-166-31-15.hsd1.co.comcast.net) joined #forth
15:35:51 --- quit: Al2O3 (Client Quit)
15:57:24 --- quit: qFox (Read error: 104 (Connection reset by peer))
16:15:05 --- quit: xjrn (Read error: 110 (Connection timed out))
16:35:49 --- join: Al2O3 (n=Al2O3@c-71-196-180-177.hsd1.co.comcast.net) joined #forth
16:42:46 --- quit: Al2O3 ()
17:19:41 --- quit: Deformati (Remote closed the connection)
17:19:59 --- join: Deformati (n=joe@c-68-62-76-160.hsd1.mi.comcast.net) joined #forth
17:24:55 --- join: xjrn (n=jim@c-69-181-213-99.hsd1.ca.comcast.net) joined #forth
17:27:17 --- join: nighty__ (n=nighty@210.188.173.245) joined #forth
18:12:58 --- quit: xjrn (Read error: 110 (Connection timed out))
19:44:11 --- join: proteusguy (n=proteusg@61.7.144.97) joined #forth
19:45:31 --- quit: JasonWoof (Read error: 113 (No route to host))
20:07:18 --- join: xjrn (n=jim@c-69-181-213-99.hsd1.ca.comcast.net) joined #forth
23:57:03 --- join: pierre- (n=pierre@95-28-64-72.broadband.corbina.ru) joined #forth
23:59:59 --- log: ended forth/08.12.14