00:00:00 --- log: started forth/04.07.08
00:48:48 --- join: arke (arke@melrose-251-251.flexabit.net) joined #forth
01:14:12 --- quit: Herkamire ("piff")
01:18:26 <arke> hi from germany g00dz
01:18:55 <arke> kc5tja already in bed, i assume
01:21:53 --- join: qFox (C00K13S@cp12172-a.roose1.nb.home.nl) joined #forth
01:22:10 <arke> hi qfox
01:22:26 <arke> qFox: saw .nl from the plane
01:22:30 <arke> qFox: :)
01:22:42 <qFox> i hope you took a photograph
01:22:45 <qFox> :)
01:22:57 <arke> nope sorry
01:22:59 <arke> :)
01:23:03 <qFox> sheesh
01:23:05 <arke> i will opn teh waz back
01:23:07 <arke> way*
01:23:14 <qFox> hehe :)
01:23:18 <arke> not used to german keyboard, lol
01:23:27 <qFox> ah you in germany?
01:23:33 <qFox> their keyboard suxxxxx
01:23:41 <qFox> its even worse then a swedish keyboard :p
01:23:42 <arke> doesnt suck
01:23:45 <arke> its just different
01:23:46 <arke> .)
01:23:48 <arke> :)
01:24:49 <qFox> you get a free es-ztet and a couple of umlouded letters
01:25:04 <arke> ßäöü
01:25:09 <arke> :)
01:25:13 <qFox> that'd be the ones
01:25:14 <qFox> :)
01:25:17 <arke> :)
01:25:39 <arke> ´´´´´´
01:32:33 --- quit: arke ("bye")
01:42:23 --- nick: SolarFire-| -> Solarfire
01:50:45 --- join: foobar (fooboo@dialup-4.152.180.85.Dial1.Atlanta1.Level3.net) joined #forth
02:59:40 --- quit: aum ()
03:06:16 --- join: O3BEPH (~z@212.34.52.140) joined #forth
03:09:27 --- nick: O3BEPH -> Serg
03:23:37 --- quit: Research (Remote closed the connection)
03:24:47 --- join: Research (XINU@12-222-128-22.client.insightBB.com) joined #forth
03:41:47 --- part: foobar left #forth
03:46:05 --- join: crc (crc@0-1pool176-3.nas6.philadelphia1.pa.us.da.qwest.net) joined #forth
04:01:57 --- quit: Serg ()
04:44:22 <Robert>  /win 21
04:44:23 <Robert> oops
04:45:44 <qFox> booting to windows with your irc client? ;)
04:46:16 <Robert> irssi just doesn't like window numbers > 19
04:55:50 --- join: foobar (fooboo@dialup-4.152.180.85.Dial1.Atlanta1.Level3.net) joined #forth
04:57:33 --- part: foobar left #forth
05:05:32 --- quit: crc (Client Quit)
05:19:51 --- quit: I440r_ ("Leaving")
05:50:17 --- join: I440r_ (~mark4@168-215-246-243.gen.twtelecom.net) joined #forth
05:55:32 <mur> Robert, what, i haven't had problems past 40 windows ;)
05:56:20 <Robert> ;-D
05:57:20 <I440r_> 40 windows in what ?
05:57:50 <mur> in irssi
05:58:37 <I440r_> oh lol
06:09:21 --- join: O3BEPH (~z@212.34.52.140) joined #forth
06:09:27 --- nick: O3BEPH -> Serg
06:28:57 --- quit: mur (Read error: 104 (Connection reset by peer))
06:29:21 --- join: mur (~mur@mgw2.uiah.fi) joined #forth
06:47:29 --- join: wossname (wossname@HSE-MTL-ppp73323.qc.sympatico.ca) joined #forth
06:59:11 --- quit: Serg ()
07:17:05 --- quit: mur (Read error: 104 (Connection reset by peer))
07:39:56 --- join: warpzero (~warpzero@mi171.dn181.umontana.edu) joined #forth
07:53:28 --- join: tucknip (~Huck@170.141.109.33) joined #forth
08:10:22 <qFox> robert is spamming me pm
08:10:27 <qFox> oh wait
08:10:31 <qFox> its not spam
08:10:33 <qFox> (haha)
08:32:19 --- join: jdrake (irc_user@CPE00045afdd0e8-CM014410113717.cpe.net.cable.rogers.com) joined #forth
08:34:35 --- quit: I440r_ (Read error: 104 (Connection reset by peer))
08:34:49 --- quit: SDO (Connection timed out)
08:36:52 <tucknip> are you 7color9-blind, onetom_ ?
08:45:16 <jdrake> when you do an ALLOC do you later have to deallocate that memory?
08:49:06 --- join: I440r_ (~mark4@168-215-246-243.gen.twtelecom.net) joined #forth
09:11:42 <jdrake> i have an operation that leaves the stack like this for a cmove: ( addr addr2 len )  but I need to make it ( addr2 addr addr2 len )  because I need to give the result of addr2. Is there any way I can do this?
09:27:32 <I440r_> >r tuck r>
09:28:21 <jdrake> i didn't know of that oen :-)
09:28:42 <I440r_> tuck pushes a copy of the top item under the second item
09:28:51 <I440r_>  ( n1 n2 --- n2 n1 n2 )
09:30:11 <jdrake> no - the >r
09:30:28 <jdrake> i just don't know all of the stack words yet
09:30:44 <Robert> jdrake: That means you're Pure(TM)
09:30:49 <jdrake> i am trying to concentrate on solving my ficl problem of giving it a string
09:31:05 <Robert> jdrake: It's not politically correct to use more than DUP, DROP and maybe OVER.
09:31:46 <jdrake> this is annoying: Error: FICL_VM_STATE_COMPILE only!
09:31:59 <jdrake> sprintf(string, "S\" %s\" %d ALLOCATE DROP SWAP >r tuck r> CMOVE", disp, strlen(disp));
09:32:57 <I440r_> oh. i dont know of ficl has either >r or tuck
09:33:19 <jdrake> it has all of tose
09:34:25 <jdrake> is there anything there that looks obvious to anyone that I am missing?
09:35:25 <jdrake> it seems to be >r that doesn't like that
09:39:57 <jdrake> it appears cmove is what ficl doesn't have
09:44:58 <onetom_> tucknip: :) no, why would i b?
09:46:08 <onetom_> anyway, gotta go 2 have a supper
09:59:50 <tucknip> bye, onetom_ 8:)
10:01:15 --- nick: TomasuDlrrp -> Tomasu
10:05:52 --- join: mur (~mur@smtp.uiah.fi) joined #forth
10:17:12 <I440r_> hrm im thinking of creating a new programming language thats a mix of brainfuck and whitespace
10:17:19 <I440r_> i can either call it whitefuck or brainspace
10:17:20 <I440r_> hrm
10:17:50 <Robert> Heh
10:18:04 <Robert> Sounds like a very good idea. Will you program a space probe in it? ;)
10:18:28 <Robert> Know what? I just discovered a neat little Forth I wrote some time ago.
10:18:40 <Robert> Basically a DOS port of IsForth.
10:18:52 <Robert> (you really burnt IsForth into my ROM ;)
10:27:51 --- quit: Research (Remote closed the connection)
10:38:44 --- join: Herkamire (~jason@h000094d30ba2.ne.client2.attbi.com) joined #forth
10:40:25 --- join: Serg (~z@193.201.231.126) joined #forth
10:45:28 * Serg looking disk from RU "Hacker" magazine, what must be named "scriptkiddie" or "comp hooligan" but no hell way "hacker" ;((
10:45:59 --- join: chandler (~chandler@chandler.registered) joined #forth
10:46:19 <chandler> I440r_: mix of brainfuck and whitespace, eh? I've heard of that. It's called "python" I hear.
10:55:35 <Serg> i just try PHPX web cmx - fail on my win32 apache
11:00:32 <Serg> today i failed to make CSS site look OK and reverted to tables ;((
11:00:45 <Herkamire> :(
11:02:44 <Serg> but anyway, it became seriosly  lighter than old one, and manageable
11:03:18 --- quit: tucknip (Read error: 110 (Connection timed out))
11:03:34 <Serg> face page 5k only
11:04:01 <chandler> I figured out css floats the other day
11:04:03 <chandler> they're not too bad
11:04:23 <Herkamire> float: left
11:04:26 <Serg> i seem they unpredictable
11:05:12 <Serg> they fall therever they want ;((
11:05:48 <Serg> if of uneven size - they pile up like wrecky cruft
11:06:35 <Serg> and CSS2 is buggy ;((
11:07:26 <Herkamire> heh
11:07:39 <Serg> inline blocks, like <b> or <span>, fail to {position: relative; right:0px;} - instead the just go after previous block
11:07:42 <Herkamire> sure it's not the implementation of CSS2?
11:08:15 <Serg> sure - both IE5 and Mozilla 1.7 have same bug ;)))))
11:08:28 <Herkamire> lol
11:08:34 <Herkamire> that proves it
11:08:35 <Serg> i mean inline blocks
11:08:38 <Herkamire> browsers suck
11:09:04 <Serg> all ex Lynx, ha-ha
11:09:22 <Herkamire> there's lots of pages about how the lates versions of mozilla and IE don't support CSS1 properly
11:09:38 <Serg> "this site is optimized for Lynx at resolution 40x25"
11:09:39 <Herkamire> yeah, I like lynx, w3m, dillo
11:10:51 <Herkamire> firebird has some nice features, but also has some (#*&$#@*&$^@#*&$@#$ annoying ones
11:11:02 <Serg> my "hamster" (home page) is _really_ enhanced for Lynx, despite it has `www.anybrowser.org` tag
11:11:16 <Serg> ;))
11:11:27 <Herkamire> well, if it works in links, it should work in firefox
11:11:36 <Herkamire> oops I said bird
11:11:47 <Herkamire> maybe they'll stop changing the name sometime
11:11:59 <Serg> firerat ;))
11:12:13 <Herkamire> fireturtle
11:12:22 <Serg> ha-ha-ha
11:12:39 <Serg> once great fire was near my home - damn competitor wars
11:12:41 <Herkamire> fireannoythepissoutofyouwithdialogboxes
11:12:58 <Serg> we were afraid of firebirds and firerats ;((
11:14:33 <Serg> 4-storey house burnt down to ground, 10+ firetrucks and foam attack 
11:14:47 <Herkamire> wow
11:15:20 <Serg> and phones off in a district ;( so fatal latency in firemen call
11:15:38 <Herkamire> yikes
11:16:27 <Serg> it was at wild 90-s, when only mafia could afford cellphones
11:17:15 <chandler> Herkamire: have you seen the "firesomething" extension?
11:17:21 <chandler> it randomly names your firefox windows
11:17:43 <chandler> I have two open right now, one is Mozilla Lightningcow, the other is Mozilla Spacesheep
11:17:56 <chandler> http://www.hellinahandbasket.org/blog/ is my recent adventure in css float
11:18:10 <chandler> paste.lisp.org is still tables for the layout
11:20:22 * Serg recalls "afterburners" joke ;))))
11:21:15 --- quit: Serg ("GOTO sleep....")
11:23:21 <Herkamire> chandler: lol
11:30:43 <Herkamire> firesomething is funny!
11:31:02 <Herkamire> it can even change the user-agent
11:32:08 <chandler> yup :-)
11:33:46 <Herkamire> download is .xpi... isn't that an image format?
11:34:09 <chandler> no, it's Cross Platform Installer. which is "xpi" because the Mozilla team can't spell.
11:34:32 <Herkamire> I see
11:34:37 <Herkamire> I'm scared
11:34:44 <chandler> dont' be
11:36:50 <Herkamire> what do I do with it?
11:37:06 --- join: tucknip (~Huck@170.141.109.33) joined #forth
11:38:03 <Herkamire> well, that went well. firebird crashed instantly
11:39:02 <qFox> lol
11:39:33 <Herkamire> oh whatever
11:39:37 <Herkamire> I got a good laugh
11:39:47 <Herkamire> my WM doesn't show window titles anyway
11:40:11 --- quit: Tomasu (Connection timed out)
11:40:39 --- join: Tomasu (~moose@d66-222-216-222.abhsia.telus.net) joined #forth
11:41:12 <Herkamire> I'm not feeling motivation to work
11:48:04 <slava> hey
11:54:31 <chandler> slava!
12:01:08 --- quit: slava ("Leaving")
12:11:57 --- join: slava (~slava@CPE00096ba44261-CM000e5cdfda14.cpe.net.cable.rogers.com) joined #forth
12:12:00 <slava> hey
12:41:45 --- quit: Tomasu (Connection timed out)
12:42:17 --- join: Tomasu (~moose@d66-222-216-222.abhsia.telus.net) joined #forth
13:13:01 --- join: tathi (~josh@pcp02123722pcs.milfrd01.pa.comcast.net) joined #forth
13:29:48 --- quit: tathi ("leaving")
13:57:11 <qFox> anyone know a bootable, (semi) indirect threaded, small, 32bit forth? :)
13:57:33 <qFox> oh yeh, add traditional to that list
13:57:36 <qFox> :)
13:57:58 <I440r_> define traditional... "not ans" ?
13:58:08 <qFox> well, "forth"
13:58:13 <qFox> and works
13:58:14 <qFox> ans
13:58:30 <qFox> but i would not call ans small
13:59:09 <I440r_> i wouldnt call it forth
14:00:40 <Robert> I440r_: qFox needs to know about VGA hardware...
14:00:53 <I440r_> oh i forgot to send you that shit didnt i lol
14:00:53 <Robert> You know that stuff, right?
14:01:00 <qFox> oh my god man, what have you done!
14:01:01 <qFox> :p
14:01:05 <I440r_> ive done some vga programming yes
14:01:08 <Robert> Yeah, but it's no hurry
14:01:32 <Robert> Not like I need it today :)
14:23:55 --- quit: Tomasu (Connection timed out)
14:24:26 --- join: Tomasu (~moose@d66-222-216-222.abhsia.telus.net) joined #forth
14:33:27 --- quit: Tomasu ("SIGWTF")
14:33:42 --- join: Tomasu (~moose@d66-222-216-222.abhsia.telus.net) joined #forth
15:25:11 --- join: aum (~aum@port-204-54-210.fastadsl.net.nz) joined #forth
15:35:49 --- part: tucknip left #forth
16:31:35 --- join: Sonarman (~matt@adsl-64-169-93-27.dsl.snfc21.pacbell.net) joined #forth
16:32:38 <Sonarman> qFox: http://membres.lycos.fr/astrobe
16:32:49 <Robert> Hi
16:32:50 --- quit: Sonarman (Client Quit)
16:32:53 <Robert> Bye
16:32:57 --- join: Sonarman (~matt@adsl-64-169-93-27.dsl.snfc21.pacbell.net) joined #forth
16:33:01 <Robert> Hi
16:33:04 <Sonarman> Hi Robert.
16:33:07 --- quit: Sonarman (Client Quit)
16:33:10 <Robert> Bye
16:37:12 <slava> hi
16:37:39 <Robert> Hi slava 
16:38:15 <slava> i'm debugging my garbage collector function by function.
16:38:23 <slava> its 140 lines of code but its tricky :)
16:39:37 <Robert> Fun. I'm debugging the "parse" word of an old Forth I found.
16:42:25 --- quit: qFox ("meh ircii's netsplit detection appearantly sux ^^")
16:43:27 --- quit: I440r (Nick collision from services.)
16:43:33 --- nick: I440r_ -> I440r
16:43:38 <Robert> Hey :)
16:43:39 --- mode: ChanServ set +o I440r
16:43:40 --- join: I440r_ (~mark4@216-110-82-1.gen.twtelecom.net) joined #forth
16:43:41 <Robert> Guess what?
16:43:44 <aum> hi I440r
16:43:57 <I440r> hi
16:44:04 <Robert> qFox just sent me a LONG list of (sometimes prettu advanced words, requiring a lot of support code)
16:44:09 <Robert> And told me to implement them for him.
16:44:20 <Robert> In x86 asm.
16:44:59 <Robert> _After_ he rejected my Forths because they use VARs instead of VARIABLEs.
16:45:03 <Robert> Now THAT is picky.
16:45:26 <aum> anyone know if there's any primitives for operating on 'addr u' strings? eg, concat, toupper, split, strip, replace etc?
16:45:37 <fridge_> HAR HAR
16:45:53 <aum> or, should i ask, can anyone point me to some forth repository sites?
16:48:38 <I440r> ftp://ftp.taygeta.org
16:50:00 <aum> .com actually
16:54:50 <I440r> oh yea :)
16:54:59 <I440r> .org used to work too
16:56:47 <aum> I440r: was it you writing to ask me about netForth?
16:57:46 <I440r> aum lol thats you ?
16:57:50 <aum> yep
16:57:51 <I440r> i didnt know your nick lol
16:58:23 <aum> to answer your question - i haven't yet got a site for Forpy
16:59:10 <aum> but, the python integration is pretty much complete and working well
16:59:25 <aum> (even if my code does lots of nested 'evaluate's)
17:00:00 <I440r> :)
17:00:17 <aum> has a 'python stack', separate from float/data stacks
17:01:05 <aum> i'm very tempted to remove the float/data stacks entirely, and just have the python object stack
17:04:38 <I440r> :)
17:04:47 <I440r> that would be a simplification
17:04:51 <I440r> and im all for those
17:04:59 <aum> makes sense in my mind - everything on the stack is an object
17:05:27 <aum> where the '->' operator fetches an attribute
17:12:15 <I440r> stack based OOP
17:12:21 <I440r> :)
17:12:25 <I440r> OOPswap
17:12:26 <I440r> OOPdup
17:12:27 <I440r> hehe
17:12:29 <aum> no
17:12:31 <aum> pswap
17:12:33 <aum> pdup
17:12:34 <aum> pdrop
17:12:37 <aum> p-rot etc
17:13:03 <aum> python int literals have a 'p' in front, eg 'p1', 'p655', 'p2.3e+45'
17:13:21 <aum> python string literals start with 'p"', no space required
17:13:45 <aum> eg 'p"Hello, world, my name is \"Fred\"\n"
17:14:32 <aum> felt very lame to me that 'S"' doesn't handle backslash escapes
17:17:08 <aum> i'm now using the Forpy shell as my nix shell (instead of bash)
17:17:33 <slava> heh
17:17:51 <slava> in my language " parses escapes, #" doesn't.
17:18:32 <aum> inserted a word into the parse chain, so if i type an unknown word like 'grep', the parser does 'which grep' and, if anything comes back, invokes the cmd with args
17:18:41 <slava> ficl?
17:18:47 <aum> a ficl fork
17:19:27 <aum> so 'grep something somefile' runs the grep via 'system()', dumps result to stdout
17:19:51 <aum> whereas '<grep something somefile' runs grep via 'popen()', puts result onto stack
17:20:26 <aum> and '>grep something' pops string from the stack, feeds it to grep, dumps result to stdout
17:20:40 <aum> and '<>grep something' pops string from stack, feeds to grep, puts result back on stack
17:22:17 <aum> and, stack doesn't get cleared on errors (which annoys me about most forths - a dumb typo can waste a lot of effort)
17:25:57 <slava> are python objects garbage collected?
17:26:17 <aum> slava: yes, through reference counting
17:26:26 <slava> aum, what if there is a reference on the forth stack?
17:26:45 --- quit: warpzero ("Tried to warn you about Chino and Daddy Gee, but I can't seem to get to you through the U.S. Mail.")
17:26:53 <aum> i increase the reference count when pushing to stack, and decrease it after i'm finished with an item popped off
17:27:02 <slava> ah
17:27:03 <aum> for instance, 'p+'
17:27:09 <slava> i don't like reference counting for this reason
17:27:15 <slava> its too much bookkeeping
17:27:22 <aum> not too bad actually
17:27:24 <slava> miss an unref or a ref somewhere, and it breaks.
17:27:36 <aum> i've audited the refcounts, and it's holding up
17:27:38 <slava> also there is overhead for each heap object
17:28:17 <aum> objects have an overhead, yes, but so do any data structures 
17:29:51 <slava> also python heap objects have next/prev links
17:29:55 <slava> that is a lot of overhead
17:30:24 * aum with 1GB of RAM is surprisingly unfrightened by overheads
17:37:29 <I440r_> back :)
18:12:13 --- part: jdrake left #forth
19:10:19 --- join: Sonarman (~matt@adsl-64-169-93-27.dsl.snfc21.pacbell.net) joined #forth
19:10:57 <Sonarman> Robert: what do you use for editing and viewing RTF files?
19:11:31 <aum> OpenOffice.org is probably best for rtf
19:11:40 <Sonarman> NO
19:11:42 <Sonarman> :-)
19:11:48 <aum> ??
19:11:58 <Sonarman> i don't have the hard drive space for that behemoth
19:12:10 <Sonarman> literally
19:12:11 <aum> abiword then perhaps?
19:12:18 <Sonarman> good idea
19:12:24 <aum> or, get the 'unrtf' util
19:12:42 <aum> though unrtf severely uglifies the doc
19:12:46 <Sonarman> though i was just wondering if there was something nice and small dedicated to rtf files
19:13:03 <aum> 'wine wordpad.exe' ?
19:13:41 <Sonarman> :P
19:14:35 <Sonarman> ooh, freshmeat has some nice stuff
19:15:21 <Sonarman> rtf2ps!
19:20:13 <I440r_> http://www.agoracom.com/nonmemforum/main.asp?pgnum=1&topic=Patriot+Scientific&tr
19:20:22 <I440r_> i didnt even know about this
19:20:29 <I440r_> anyone in here know aobut it ?
19:21:13 <aum> which particular thread are you referring to?
19:21:47 <I440r_> the patriot suit
19:22:08 <I440r_> patriot is claiming 100% ownership on a patent that was issued to chuck and his partner
19:22:25 <I440r_> his partner sold his interest which patriot now owns and they say the own 100% of it
19:22:31 <aum> and suing *users* of fast chips?
19:22:39 <aum> they're like the SCO of hardware! :(
19:22:55 <I440r_> i hope chucm makes $028546294569824569234876254982736495263478596 
19:22:59 <I440r_> chuck even
19:24:27 <aum> ahh, chuck moore has a stake?
19:24:31 <I440r_> yes
19:24:40 <I440r_> his name is on that patent TOO!!!!
19:24:45 <aum> hopefully the $$$ will help him to develop and market an ultra-fast Forth chip!!!
19:25:06 --- quit: wossname (Client Quit)
19:25:13 <I440r_> i consider chuck to be the equal of tesla
19:25:16 <aum> imagine a 4.8GHz Forth chip with all primitives implemented in on-chip hardware
19:25:25 <aum> 64-bit cells
19:25:26 <I440r_> and i hope he dies a multi billionaire
19:25:32 <I440r_> but not too soon
19:28:13 <htp123> what frequency do his other chips run at?
19:28:25 <I440r_> dunno
19:28:57 * aum hates those websites that plant cpu-sucking javascript and animated gifs
19:32:51 <slava> cpu sucking animated gifs?
19:32:54 <slava> what kind of cpu do you have :)
19:32:56 <slava> 386?
19:33:05 <aum> athlon 1700xp+
19:33:53 <slava> kc5 hasn't been around lately
19:34:28 <aum> it's just that the javascript on some sites is very aggressive - if the browser refuses to open a popup, the js goes into this tight loop
19:37:18 <aum> ahh, just built a set of words for implementing user-defined stacks
19:39:25 <slava> aum, hehe
19:39:35 <slava> aum, i have this too, there is 1 library-defined stack, the namestack
19:39:58 <aum> ?
19:40:03 <aum> you mean, the 'stack' of defined words?
19:40:04 <slava> in my interpreter
19:40:07 <slava> no
19:40:10 <slava> a stack of bindings
19:40:15 <slava> with get/set operations
19:40:28 <slava> you can implement almost any kind of inheritance of scoping discipline using this
19:41:31 <aum> kinda reminds me of some of those 8-line forth OO engines
19:41:45 <slava> yeah this is a pretty minimal OO sort of thingy :)
19:43:13 <aum> well, writing these stack-manipulation words was an interesting exercise
19:43:29 <aum> shows me there's definitely a trade-off between efficient factoring and efficient code
19:44:23 <slava> how so?
19:44:44 <aum> heavily-factored words incur the overhead of word call/return
19:45:04 <aum> can make all the difference inside tight loops
19:46:52 <I440r_> robert if my old vga code isnt on my laptop ive lost it. i had a drive crash not long ago :/
19:47:07 <I440r_> it MIGHT be on there but i wong guarantee it
19:47:24 <slava> aum, not really
19:47:34 <slava> aum, you could probably make your forth inline some words etc
19:47:57 <aum> how do you do that?
19:48:22 <slava> well i dunno about *your* language :-P
19:48:41 <slava> in factor you can do something like : 2* 2 * ; inline
19:48:44 <slava> but its kidn of pointless
19:48:47 <slava> :)
20:07:24 <aum> quick quiz
20:07:46 <aum> for a word which looks up a symbol in a dynamic lib, and calls it with n arguments, what's the most intuitive stack signature:
20:08:07 <aum> 1) ( arg1 ... argn n addr u dynlib-handle -- retval )
20:08:25 <aum> 2) ( dynlib-handle addr u arg1 ... argn n -- retval )
20:08:43 <slava> when i did this in java, i had ( arg1 ... argn instance method-proto method-class method-name -- result )
20:08:59 <slava> but in C replace 'instance method-proto method-class method-name' with whatever
20:09:18 <aum> 3) ( dynlib-handle arg1 ... argn n addr u -- retval )
20:09:52 <slava> i'd say the dynlib handle goes at the top
20:09:58 <slava> since its almost never computed
20:10:10 <aum> top, as in last-pushed?
20:10:22 <slava> yes.
20:10:26 <slava> right most in ( ) diagram
20:10:49 <aum> so that would make it ( arg1 .. argn n funcname-addr funcname-len handle -- retval )
20:11:00 <aum> or alternatively,
20:11:17 <aum> ( funcname-addr funcname-len arg1 ... argn handle -- retval )
20:11:47 <aum> hm, the former feels more intuitive to me
20:12:24 <Sonarman> i would find ( arg1 ... argn n addr u dynlib-handle -- retval ) most intuitive
20:12:37 <Sonarman> all the items related to located the function are on top.
20:12:49 <Sonarman> /located/locating
20:12:59 <aum> that's the former one, so you agree
20:13:03 * aum implements
20:13:21 <Sonarman> better put me in the credits, bitch! :P
20:13:25 <aum> haha
20:13:43 <aum> shared libs are probably the best computing invention since the subroutine
20:15:03 <aum> now, i just gotta figure out this C varargs shit for calling the routine
20:15:20 <Sonarman> why don't you go another route, though...
20:15:28 <aum> suggestions?
20:15:33 <Sonarman> one sec
20:16:13 <Sonarman> instead of providing the name and handle every time you want to call a function, why don't you get the function's address in advance (like gforth does)
20:16:23 <slava> yeah
20:16:40 <slava> actually my java ffi primitives took the method name each time but the compiler optimized it
20:17:00 <aum> guess i could provide an additional 2 words - one to fetch func addr, and one to call the func with n args
20:17:14 <slava> make a parsing word
20:17:22 <slava> that defines a new word that does an ffi call
20:17:24 <aum> ahh, nice idea
20:17:27 <slava> you pass it the word name, function name, and # of args
20:17:32 <Sonarman> yes, make the defined word make the call automatically
20:17:43 <Sonarman> with DOES>
20:17:46 <slava> yup.
20:24:50 <aum> so then, invocation would be something like:
20:25:05 <aum> arg1 arg2 arg3 dynlib-handle dynlib-call functionname
20:25:21 <aum> oops
20:25:25 <aum> arg1 arg2 arg3 3 dynlib-handle dynlib-call functionname
20:33:22 <Sonarman> no, just arg1 arg2 arg3 functionname
20:34:47 <Sonarman> here's a complete example of what would probably be the best way (of course the minor details of the syntax could we different; i'm just trying to give you a general idea) :
20:35:01 <Sonarman> "libfoo.so" library: libfoo
20:36:02 <Sonarman> 4 "MyFunction" libfoo import-foreign: my-function
20:36:16 <Sonarman> arg1 arg2 arg3 arg4 MyFunction
20:36:41 <Sonarman> actually, import-foreign: should probably also take an arg specifying the function's return type
20:36:57 <Sonarman> this is the way it's done in gforth:
20:37:16 <Sonarman> library sdl libSDL.so
20:37:29 <Sonarman> sdl SDL_Init int (int) SDL_Init
20:37:35 <Sonarman> later...
20:37:50 <Sonarman> SDL_INIT_VIDEO SDL_Init
20:38:37 --- join: Teratogen (leontopod@67.40.69.97) joined #forth
20:38:40 <Teratogen> greetings, earthlings
20:38:43 <Sonarman> in the gforth implementation, "library handle" words do the function importing themselves
20:39:02 <Sonarman> Mk'tvlokcal Teratogen
20:39:36 <Sonarman> aum: the details could be different, but youg et the general idea, right?
20:41:04 <slava> hi Teratogen
20:43:05 <Teratogen> slavish slava
21:08:58 <Teratogen> slava <3 java
21:10:33 <slava> eh?
21:13:23 <Teratogen> you are unbanned from efnet #java, sirrah
21:13:46 <slava> the channel is +i now wtf
21:14:15 <slava> your perl irc scripts broke? :)
21:52:48 --- join: poingie (~chatzilla@64-121-15-14.c3-0.sfrn-ubr8.sfrn.ca.cable.rcn.com) joined #forth
21:53:05 --- join: LOOP-HOG (~jdamisch@sub22-119.member.dsl-only.net) joined #forth
21:53:43 <Sonarman> poingie: are you still using bigforth/
21:53:45 <Sonarman> >
21:53:46 <Sonarman> >
21:53:47 <Sonarman> ?
21:53:57 <poingie> Sonarman: yep
21:54:14 <Sonarman> have you discovered VIEW and W@/W! ?
21:54:19 <poingie> egad
21:54:22 <poingie> never heard of em
21:54:27 <poingie> url?
21:55:38 <Sonarman> well, i remember the other day you were complaining that SEE doesn't provide much information when it comes to words like !. use VIEW foobar to view the actual definitoin of a word in the editor
21:55:53 <Sonarman> in the case of !, you'll see the assembly definition
21:56:06 <slava> SEE should disassemble
21:56:20 <Sonarman> and then i remember your wondering about 16-bit stores and fetches. use W! and W@ for that purpose
21:56:24 <Sonarman> slava: but it doesn't :)
21:56:41 <poingie> oh those are words :)
21:56:48 <Sonarman> hehe, yeah
21:56:49 <slava> ok, my gc finishes one collection, but if you try collecting again it dies :(
21:57:17 <poingie> W@ does 16 bit ... i guess W stood for "word" once
21:57:27 <Sonarman> yesh
21:57:30 <Sonarman> -h
21:57:49 <poingie> lemme check it out...
21:58:24 <poingie> still learning real forth, so i'm having problems getting past trivial stuff
21:58:52 <poingie> woo, there's VIEW all right
21:59:26 <Sonarman> well, if you have any bigforth question, please feel free to ask me. i don't know if i'll be able to answer everything, but i've used bigforth do i may be able to help
21:59:29 <Sonarman> s/do/so
21:59:57 <slava> is bigforth good? :)
22:00:05 <poingie> mostly looking at getting sdl bindings working
22:01:08 <Sonarman> slava: it's OK i guess
22:01:26 <poingie> people say bigforth is real real fast
22:01:28 <slava> Sonarman, how does it compare to gforth?
22:01:50 <poingie> only forth that works on win32 and linux that has a working FFI on both
22:02:17 <Sonarman> slava: well, it's faster than gforth. and it uses a GUI environment (which, while nice, has an 'unpolished' feel to it)
22:02:20 <poingie> and dragon.m is nifty, really shows off the power of the opengl turtle gfx
22:02:33 <Sonarman> slava; it has some cool things :)
22:02:41 <poingie> i'm thinking of porting Io's ion UI to bigforth
22:02:45 <poingie> it's an opengl ui
22:02:55 <slava> what is ion like?
22:03:04 <poingie> hm, hard to find screenshots
22:03:04 <slava> i want to do a zoomable opengl gui for factor
22:04:25 --- quit: LOOP-HOG (Read error: 60 (Operation timed out))
22:04:47 <poingie> i have this weird fantasy of writing nethack in forth
22:05:05 <Sonarman> get help
22:05:06 <poingie> using opengl to draw "craggy" looking walls and such
22:05:11 <Sonarman> j/k; sounds neat :)
22:05:31 <poingie> well it sounds neat ... but i probably do need help too :)
22:05:58 <poingie> seriously, something more like telengard or temple of apshai comes to mind
22:08:59 <poingie> way ambitious project but i think i can break it into manageable chunks
22:26:39 --- join: lalalim (~lalalim@p508AA795.dip.t-dialin.net) joined #forth
22:43:33 --- quit: lalalim_ (Read error: 110 (Connection timed out))
22:55:59 <aum> why doesn't 's" hello, " s" world" 2swap type type' work?
22:57:13 <Sonarman> it should, and it does in gforth
22:58:14 <Sonarman> maybe your forth stores s" strings" in a transient buffer?
22:58:54 <aum> seems to
22:59:04 <Sonarman> does it ouput world world, ?
22:59:05 * aum tries it in another forth or six
22:59:08 <Sonarman> /ouput/output
22:59:11 <aum> Sonarman: yes, it does
22:59:25 <Sonarman> yep, then a transient buffer it is
22:59:33 <Sonarman> maybe it's PAD
22:59:57 <aum> minForth gets it right
23:00:27 --- join: SolarFire-| (SolarFire@pD9E59D99.dip.t-dialin.net) joined #forth
23:00:47 <aum> gforth and kforth also get it right
23:10:36 <Herkamire> http://www.schlockmercenary.com/d/20000911.html
23:17:30 --- quit: Solarfire (Read error: 110 (Connection timed out))
23:17:49 <aum> ahh, found the problem - my forth was copying to HERE
23:18:47 <aum> general question - is frequent use of EVALUATE considered bad form
23:18:47 <aum> ?
23:19:25 <slava> yes
23:20:02 <aum> any particular reason?
23:21:09 <slava> makes code harder to understand
23:21:14 <slava> also late binding is not the forth philosophy
23:24:09 <aum> k
23:24:26 <aum> personally i avoid it unless i can't find another way
23:31:07 <Sonarman> aum: by the way, s" hello, " s" world" 2swap type type  should work in that forth when it's inside a definition
23:31:28 <aum> Sonarman: correct, i tried it before
23:32:39 <aum> my main use for 'evaluate' is for a command shell i've written in forth
23:33:35 <slava> oh
23:33:37 <slava> that's ok
23:45:19 --- quit: Herkamire ("bed")
23:53:45 --- quit: Sonarman ("leaving")
23:59:59 --- log: ended forth/04.07.08