00:00:00 --- log: started forth/03.11.17
00:37:20 --- join: I440r (~mark4@12-58.lctv-a5.cablelynx.com) joined #forth
00:58:17 --- join: schihei (~schihei@ss05.co.us.ibm.com) joined #forth
03:05:02 --- quit: haroldo (Read error: 104 (Connection reset by peer))
03:06:42 --- join: haroldo (~haroldo@r200-40-212-129.adsl.anteldata.net.uy) joined #forth
04:27:45 --- quit: schihei (Client Quit)
04:28:30 --- join: schihei (~schihei@ss01.co.us.ibm.com) joined #forth
05:13:29 --- quit: schihei (Client Quit)
05:14:13 --- join: schihei (~schihei@ss02.co.us.ibm.com) joined #forth
05:35:05 --- join: rO| (rO|@p62.246.5.85.tisdip.tiscali.de) joined #forth
05:43:08 --- quit: schihei (Client Quit)
05:43:37 --- quit: I440r ("bbl")
05:43:51 --- join: schihei (~schihei@ss05.co.us.ibm.com) joined #forth
06:05:44 --- quit: schihei (Client Quit)
06:09:36 --- join: schihei (~schihei@ss02.co.us.ibm.com) joined #forth
06:16:03 --- quit: warp0x00 (calvino.freenode.net irc.freenode.net)
06:16:03 --- quit: ayrnieu (calvino.freenode.net irc.freenode.net)
06:16:04 --- quit: haroldo (calvino.freenode.net irc.freenode.net)
06:16:04 --- quit: melinda (calvino.freenode.net irc.freenode.net)
06:16:04 --- quit: Robert (calvino.freenode.net irc.freenode.net)
06:16:04 --- quit: schihei (calvino.freenode.net irc.freenode.net)
06:16:05 --- quit: arke (calvino.freenode.net irc.freenode.net)
06:16:05 --- quit: oooo (calvino.freenode.net irc.freenode.net)
06:16:05 --- quit: ianP (calvino.freenode.net irc.freenode.net)
06:16:05 --- quit: mur (calvino.freenode.net irc.freenode.net)
06:16:05 --- quit: skylan (calvino.freenode.net irc.freenode.net)
06:16:05 --- quit: MysticOne (calvino.freenode.net irc.freenode.net)
06:16:05 --- quit: chandler (calvino.freenode.net irc.freenode.net)
06:16:20 --- join: schihei (~schihei@ss02.co.us.ibm.com) joined #forth
06:16:20 --- join: haroldo (~haroldo@r200-40-212-129.adsl.anteldata.net.uy) joined #forth
06:16:20 --- join: warp0x00 (~warpzero@dsl.142.mt.onewest.net) joined #forth
06:16:20 --- join: arke (~chris@ca-cmrilo-cuda1-c3b-66.vnnyca.adelphia.net) joined #forth
06:16:20 --- join: ayrnieu (julian@206.61.132.173) joined #forth
06:16:20 --- join: Robert (~snofs@c-255a71d5.17-1-64736c10.cust.bredbandsbolaget.se) joined #forth
06:16:20 --- join: melinda (melinda@melinda.usercloak.freenode) joined #forth
06:16:20 --- join: skylan (sjh@vickesh01-4649.tbaytel.net) joined #forth
06:16:20 --- join: oooo (o@virgo.bombsquad.org) joined #forth
06:16:20 --- join: MysticOne (mysticone@mysticone.usercloak.freenode) joined #forth
06:16:20 --- join: mur (~mur@smtp.uiah.fi) joined #forth
06:16:20 --- join: ianP (ian@inpuj.net) joined #forth
06:16:20 --- join: chandler (~darmok@64-145-60-36.client.dsl.net) joined #forth
06:24:16 --- join: Serg_Penguin (~z@212.34.52.140) joined #forth
06:31:08 --- quit: schihei (Client Quit)
06:31:52 --- join: schihei (~schihei@ss07.co.us.ibm.com) joined #forth
06:35:19 --- quit: Serg_Penguin ()
06:47:43 --- quit: schihei (Client Quit)
06:48:33 --- join: schihei (~schihei@ss02.co.us.ibm.com) joined #forth
06:52:32 --- nick: rO| -> rO|bbl
06:53:07 --- join: I440r (~mark4@saturn.vcsd.com) joined #forth
06:54:41 --- quit: rO|bbl ("bbl")
06:55:58 --- quit: I440r (Client Quit)
06:58:08 --- quit: schihei (Client Quit)
07:02:00 --- join: schihei (~schihei@ss07.co.us.ibm.com) joined #forth
07:17:43 --- quit: schihei (Client Quit)
07:18:27 --- join: schihei (~schihei@ss05.co.us.ibm.com) joined #forth
07:33:45 --- quit: schihei (Client Quit)
07:34:28 --- join: schihei (~schihei@ss2.bluebird.ibm.com) joined #forth
07:40:00 --- quit: melinda (Read error: 60 (Operation timed out))
07:45:29 --- quit: schihei (Client Quit)
07:46:13 --- join: schihei (~schihei@ss01.co.us.ibm.com) joined #forth
07:59:21 --- quit: schihei (Client Quit)
08:28:43 --- join: Herkamire (~jason@h000094d30ba2.ne.client2.attbi.com) joined #forth
08:51:47 --- join: melinda (melinda@melinda.usercloak.freenode) joined #forth
08:58:42 <Herkamire> 'morning melinda
08:58:52 <melinda> morning Herkamire
09:06:56 * MysticOne pounces melinda!
09:08:58 * melinda squees!
09:10:03 <MysticOne> hehe
09:50:06 --- join: madgarden (~bughead@216.94.153.178) joined #forth
10:32:20 --- join: tathi (~josh@pcp02123722pcs.milfrd01.pa.comcast.net) joined #forth
10:42:15 --- join: I440r (~mark4@saturn.vcsd.com) joined #forth
10:43:27 --- quit: I440r (Client Quit)
10:54:09 <madgarden> Anyone here written a Forth in C?
11:05:09 <tathi> sure, why?
11:05:48 <madgarden> Just curious, since I've been working on a C Forth-like.
11:06:24 --- join: rO| (~rO|@dialin-145-254-243-038.arcor-ip.net) joined #forth
11:06:57 <tathi> probably several other people here have too...
11:06:59 <madgarden> And I figure, maybe people talk about Forth stuff in here. ;)
11:07:11 <tathi> well...
11:07:28 <rO|> have too...
11:07:28 <tathi> every once in a while
11:07:33 <rO|> what?
11:07:42 <tathi> written a forth in C
11:08:33 <rO|> why not..
11:09:01 <tathi> <it [Client Quit]
11:09:09 <tathi> oops
11:09:16 <rO|> there's a 1 fdd gui Os written in fasm, somebody interested in rewriting it in forth?
11:09:32 <rO|> (re) ..
11:09:56 <tathi> how about you? ;)
11:10:00 <madgarden> So then, any details (your C Forth)?
11:10:41 <tathi> um...don't know if I have source code around any more.
11:10:49 <tathi> at least not for a working version.
11:11:08 <rO|> tathi: ok
11:11:14 <tathi> the C type stuff is really annoying when you're trying to write something like Forth that doesn't have types.
11:11:27 <tathi> indirect threaded...
11:11:28 <madgarden> Yea, that's why I put types into mine. :)
11:11:39 <tathi> :)
11:11:39 <madgarden> I know, blasphemy.
11:11:54 <tathi> nah, people keep doing that.
11:11:55 <madgarden> But, it's more a Forth-inspired system than an actual Forth.
11:12:31 <madgarden> Don't have a contiguous memory space either. I malloc code space for words as I make 'em.
11:12:59 <tathi> gotcha
11:13:19 <madgarden> I hide away all the control flow stuff on the return stack too.
11:13:49 <tathi> well, ANS (if you're into that) does say that there may be a separate control
11:13:53 <tathi> flow stack
11:14:12 <tathi> but that you should assume that it may be implemented on the return stack for "portable" code...
11:14:14 <madgarden> Yep, I did read that when researching. Seemed like a good idea.
11:14:40 <rO|> btw. called menuetOS, somebody ELSE interested?
11:14:57 <tathi> rO|: Oh.  I've seen that.  Looked interesting.
11:15:12 <tathi> but I don't have an x86 box, so...
11:15:47 <rO|> pity ;-)
11:15:55 <madgarden> tathi: Do you recall how you handled EVALUATE? 
11:16:17 <tathi> rO|: no it isn't, PPC is much better :P
11:17:02 <tathi> madgarden: not sure I got that far with my C forth -- I decided to switch over to asm before I got too far.
11:17:39 <rO|> tathi: isn't it more expensive? i would estimate around 100 bucks for a dev box..
11:18:13 <madgarden> It's something I've held off on implementing. I want to keep it stackless so I can't have any C recursion when evaluating strings. Almost seems like I need a seperate stack for evaluation.
11:18:14 <tathi> but I think my input loop was set up to read from a buffer, and call a function to refill it when empty.
11:19:12 <tathi> I usually just somehow save the old input source on the regular stack.
11:19:38 <tathi> and assume that the stack is going to be at the same place when I switch back.
11:20:07 <tathi> hmm, maybe that's not right.  It's been a while since I've incorporated EVALUATE into a forth.
11:20:44 <tathi> rO|: definitely more expensive.  But I have a nice dual-processor PPC box, and I don't like the x86 architecture at all.
11:20:58 <tathi> s/have/already have/
11:21:31 <madgarden> Ahh. yea, I'd thought of that, but I figured losing the string address on the data stack might not be so cool. I'm trying to automate/hide some of these functions so the user doesn't have to manage much.
11:22:03 <rO|> tathi: still the dual-forth-processor box is missing ;-)
11:22:05 <tathi> :) yeah, makes sense.
11:22:24 <tathi> rO|: or the 2500-forth-processor box :)
11:23:07 <tathi> well, a separate stack for input sources seems like the simplest way to go if you don't want the user to have to worry about losing things.
11:23:55 <rO|> tathi: just a dream, just a dream.. ;-)
11:24:07 <madgarden> I do have stacks as a basic type. The vocabulary uses a stack as well.
11:25:16 <tathi> that's cool
11:27:29 <madgarden> Yea, this doesn't really specify how/where the input source is stored: http://forth.sourceforge.net/standard/dpans/dpans6.htm#6.1.1360
11:32:26 <tathi> or whether you can do it recursively
11:32:40 <tathi> though I can't imagine that _I_ would ever do that
11:33:29 <madgarden> I've tested recursion with EVALUATE on other systems, and it generally seemed to work. Not sure how yet, though. ;)
11:34:06 <madgarden> But I can see where it might be used, when using an "include" or "uses" type syntax, and the files included include other files, etc.
11:35:37 <tathi> ok, right, I wasn't thinking.
11:43:51 <Herkamire> woh, peaple talking about forth ;)
11:44:07 <Herkamire> I definately recomend having a seperate input stack
11:44:42 <Herkamire> and using the return stack for control stuff
11:45:05 <madgarden> Great, that seems to fit. :)
11:45:09 <Herkamire> I wish there was something like menuetOS for ppc
11:45:30 <Herkamire> I can't find an example of a simple, opensource OS for ppc
11:47:02 <madgarden> That menuetOS looks quite spiffy
11:50:58 <Herkamire> spiffy is a good word for it
11:52:00 <Herkamire> unfortunately the gui is completely un-original
11:52:27 <Herkamire> seems to me if people are going to go to all the trouble of writing their own OS they shouldn't make it exactly the same as existing systems
11:52:43 <madgarden> Well, on the flip side, if they want people to use it, they might need to play the familiarity card.
11:53:14 <Herkamire> I don't think so
11:54:20 <Herkamire> if it's more fun and/or easier to use, and you can show someone how to use it it 5 secconds, then I  don't think it would be a problem for it to be quite different
11:54:54 <madgarden> Like Forth? ;)
11:55:22 <Herkamire> what?
11:55:37 <Herkamire> forth is not a GUI
11:55:45 <madgarden> Yes, I know.
11:57:03 <madgarden> I just get the same "familiarity" argument from people when I mention Forth. 
11:58:41 <Herkamire> farmiliarity and progress usually don't go so well together
11:59:13 <madgarden> Agreed.
11:59:14 <Herkamire> GUIs are worse than they were 10 years ago
11:59:26 <Herkamire> user interfaces are worse than they were 20 years ago
11:59:38 <madgarden> Incidentally, this also looks spiffy: http://brix-os.sourceforge.net
12:10:07 <Herkamire> brixos does look cool
12:48:21 --- join: wossname (wossname@HSE-QuebecCity-ppp80438.qc.sympatico.ca) joined #forth
13:44:21 --- part: madgarden left #forth
14:32:21 --- join: gilbertdeb (~gilbert@fl-nked-ubr2-c3a-37.miamfl.adelphia.net) joined #forth
14:32:53 --- part: gilbertdeb left #forth
14:49:35 --- quit: tathi ("leaving")
14:52:05 --- quit: wossname ("warez are hot like sex")
14:58:54 --- join: jamc (dne@as3-6-8.asp.s.bonet.se) joined #forth
15:12:16 --- quit: haroldo (Read error: 60 (Operation timed out))
15:13:29 --- join: haroldo (~haroldo@r200-40-167-2.adsl.anteldata.net.uy) joined #forth
15:36:06 <rO|> re herkamire
15:37:43 <rO|> ack brix. not usable yet though
15:38:36 <rO|> gotta change for better, not all change is better though
15:41:54 <rO|> apropos "spiffy", first comes usability, of course... your suggestions for something "better"?
15:50:29 --- quit: ayrnieu (Remote closed the connection)
15:51:10 --- join: Nurf (~ray@soggy199.drizzle.com) joined #forth
15:57:25 --- join: I440r (~mark4@12-58.lctv-a5.cablelynx.com) joined #forth
16:01:01 --- quit: jamc ("As President I have to go vacuum my coin collection!")
16:13:31 <arke> hello everybody :)
16:14:02 <Nurf> Isss very quiet in here
16:14:07 <Nurf> 'lo :-)
16:14:42 <Nurf> Would you happen to know anything about forth?
16:14:57 * Nurf is a bit of a noob and is a bit stuck
16:15:11 <arke> yes, i do, although I'm not that good :)
16:15:22 <Nurf> Well, chances are you are better than me :-)
16:15:24 <arke> shoot, i might be able to help (but I don't guarantee it..)
16:15:32 <Nurf> Thanks :-)
16:15:50 <Nurf> I want to make a word that has similar semantics for compile time and for interpretation time
16:15:57 <Nurf> The word is 0x
16:16:14 <Nurf> It parses the number directly after the word as being a hex value, regardless of BASE
16:16:30 <Nurf> but I want it to work all the time, like this:
16:16:40 <Nurf> 1 2 3 0x 100 4 <enter>
16:16:47 <Nurf> puts 1 2 3 256 4 on the stack
16:16:52 <Nurf> And
16:17:02 <Nurf> : test 1 2 3 0x 100 4 ;
16:17:15 <Nurf> compiles as ":test 1 2 3 256 4;"
16:17:42 <Nurf> I can make a word that does one or the other, but I'm buggered if I can figure out how to make one word work for both situations :-)
16:17:51 <I440r> according to dvd+rw-tools dvd+rw's have a 1000 write limit - is that right and does it also apply to cdrw's ?
16:18:04 <Nurf> l44: I dunno. Sorry
16:18:25 <I440r> nurf you know forth ?
16:18:56 <Nurf> Um. Sorta. :-) I am still learning the more complicated stuff
16:19:14 <I440r> nurf look at isforths number input - if the first char of the number is a $ then the number is taken to be hex
16:19:21 <I440r> you could do the same with 0x easilly
16:19:37 <I440r> what forth are you using in linux ?
16:19:46 <Nurf> Hm. I'm using pForth
16:20:05 <I440r> im not familiar with it hhe
16:20:07 <Nurf> I've actually ported to a FireWire embedded link layer. I use it to drive UIs and stuff
16:20:17 <I440r> is it coded in c or in assembler ?
16:20:22 <I440r> and is it for linux ?
16:20:27 <Nurf> Its ansi C
16:20:37 <Nurf> Compiles comfortably on Linux
16:20:42 <I440r> woo that sounds cool (the firewire stuff)
16:20:43 <Nurf> I've also compiled it on IRIX
16:20:52 <I440r> i dont like c forths personally, i dont consider them to be real forths
16:21:02 <I440r> <-- a purist :)
16:21:03 <Nurf> Yeah, well.. It was what I could find :-)
16:21:12 <I440r> isforth.clss.net :)
16:21:17 <Nurf> It just implements a virtual machine
16:21:23 <I440r> if your fire wiare suff is open source port it to isforth :)
16:21:27 <Nurf> Its interpretive and everything...
16:22:05 <Nurf> I'll be implementing my own forth soonish
16:22:11 <Nurf> pForth is too big
16:22:40 <I440r> the isforth kernel is coded in x86 assembler
16:22:41 <Nurf> Hm. isforth.ciss.net doesnt resolve for me
16:22:48 <I440r> clss.net
16:22:49 <I440r> sorry
16:22:51 <I440r> typo
16:23:22 <Nurf> Hm. Thats actually a disadvantage for me
16:23:29 --- join: Sonarman (~matt@adsl-64-160-165-193.dsl.snfc21.pacbell.net) joined #forth
16:23:31 <Nurf> The link layer I am running on is an ARM processor
16:23:41 <I440r> ouch heh
16:23:53 <Nurf> I plan to write a tiny VM in C and write my own forth in the VM assembler in that
16:24:00 <Nurf> In the meantime pForth is ok, but big
16:26:24 <Nurf> ciss.net doesnt resolve for me either - its a parked domain
16:26:49 --- quit: rO| ("bye")
16:29:34 <I440r> nurf its CLSS.net
16:29:35 <I440r> heh
16:29:42 <arke> pforth is quite small compared to gfoth
16:29:49 <I440r> gforth isnt small hhe
16:29:52 <arke> I440r is too pure, imho :)
16:29:53 <I440r> isforth is only 50k
16:30:02 <I440r> so far heh
16:30:14 <Nurf> Well, size is really important to me
16:30:46 <Nurf> It has to fit in around 60k including all dictionaries and application code
16:30:48 * arke is not familiar with ARM...
16:31:05 <arke> what size processor is it? (as in, bits)
16:31:15 <Nurf> Would anyone be interested in seeing my current implementation of 0x ?
16:31:18 <Nurf> 32 bit
16:31:42 <arke> sure!
16:31:50 <Nurf> Its a very pleasant processor to program. Has a decent instruction set unlike x86 :-P !
16:32:11 <Nurf> : (0x) ( -- <hexval> )
16:32:12 <Nurf> \ 0x ABC123 puts ABC123 on the stack regardless of current base
16:32:12 <Nurf>   0 0            \ Double precision value to start with
16:32:12 <Nurf>   32 word count  \ read in number as a string
16:32:12 <Nurf>   base @ >R      \ save the old base temporarily
16:32:12 <Nurf>   16 base !      \ set base to hex
16:32:13 <arke> STC is smaller on the interpreter side, but (for most CPUs) larger on the app code side
16:32:14 <Nurf>   >number        \ convert value
16:32:16 <Nurf>   R> base !      \ restore base
16:32:18 <Nurf>   drop drop drop \ get rid of cruft
16:32:22 <Nurf> ; IMMEDIATE
16:33:02 <Nurf> Hm... well pForth is 32 bit, which is wasting me space - it uses a full cell for tokens when a byte would be plenty
16:33:06 <arke> seems ok...
16:33:11 <I440r> so you have to do 0x 123 ?
16:33:14 <I440r> not 0x123 ?
16:33:16 <Nurf> So my dictionary is 4x larger than I like
16:33:18 <Nurf> yes
16:33:20 <Nurf> 0x 123
16:33:35 <Nurf> Now my problem is that that above wont work in compilation mode
16:33:37 <arke> but you also have to do " hello" instead of "hello", so no reason to complain :)
16:33:39 <I440r> i dont like 0x anywahy - i use $ (old 6502 convention)
16:33:42 <I440r> $123 is hex 123
16:33:46 <I440r> no space required
16:33:54 <arke> neat
16:33:59 <Nurf> Yes, but that would mean I have to modify the interpreter
16:34:03 <I440r> and it works in interpret and compile
16:34:04 <Nurf> Anyway
16:34:18 <I440r> nurf you can probably port that to pforth from isforth with NO problems
16:34:20 <arke> $ has to be hardcoded tho
16:34:29 <I440r> so does 0x
16:34:32 <arke> (i think)
16:34:40 <Nurf> l44or, nope
16:34:41 <arke> no, i menat into the number detector
16:34:45 <Nurf> 0x is a std word
16:34:55 <Nurf> $123 would have to be a modification of the forth engine
16:35:01 <arke> Nurf: whats your IRC client?
16:35:04 <I440r> isforth accepts $123 = hex /123 = octal %1010 = binary
16:35:11 <Nurf> Anyway, the above was a definition of (0x)
16:35:16 <Nurf> : 0x 
16:35:16 <Nurf>   postpone (0x) postpone literal
16:35:16 <Nurf> ; IMMEDIATE
16:35:21 <Nurf> That is the definition of 0x
16:35:33 <I440r> ick
16:35:36 <Nurf> arke: X-Chat-2
16:35:38 <I440r> i dont like postpone either
16:35:44 <Nurf> Join the club
16:35:47 <arke> Nurf: use tab-completion then :)
16:35:50 <Nurf> I  had never used it until today
16:36:05 <I440r> i especially do not like postpone x postpone y postpone z postpone x postpone foo ; immediate
16:36:10 <arke> : compile postpone ; immediate   <--- would that even work? :)
16:36:10 <I440r> its realy realy relay relay bad code
16:36:18 <I440r> way too much visual clutter
16:36:34 <Nurf> l44: Well, that is why I am on this channel
16:36:36 <arke> I440r: ] x y z foo [
16:36:39 <Nurf> I want an 0x that works :-P
16:36:54 <I440r> arke no the point is that if you define : foo postpone x postpone y ; immediate
16:37:04 <I440r> where ever you reference foo it will comile x and y inline
16:37:07 <arke> Nurf: its an 'i' that starts I440r, just type the I and the 4 and a tab
16:37:11 <I440r> that sort of makes foo a macro
16:37:14 <Nurf> arke: For some reason, tab completion doesnt work on l44's nick
16:37:15 <I440r> a very poor one
16:37:36 <Nurf> arke: Ah. In this font I cant tell. I thought it was L44
16:38:07 <Nurf> arke: I have been using IRC since 1993, btw :-)
16:38:20 <arke> Nurf: use tabs then :)
16:38:46 <Nurf> arke: Well, I would if I could read the nicks I was trying to complete :-P
16:39:02 <I440r> nurf inside isforth theres a file that does macro colon definitions
16:39:07 <I440r> m: foo x y ;m
16:39:19 <I440r> foo when referenced injects x and y inline
16:39:30 <I440r> but the code will NOT work on a native compiler
16:39:34 <I440r> only on a TRUE forth
16:39:41 <Nurf> I440r: Can they be implemented in an ANS forth?
16:39:52 <Nurf> All I want is a variation of 0x that works in both modes
16:39:56 <Nurf> I've shown you the code :-)
16:39:56 <I440r> nurf probably no. i am diametrically opposed to the ans standard
16:40:03 <Nurf> *shrug*
16:40:08 <I440r> in fact if you read the comment on that file it flams ans BIG TIME heh
16:40:15 <Nurf> Too bad. I have an ANS forth to work with
16:40:19 <Nurf> End of story. :-)
16:40:36 <I440r> the ans forth standard does not describe the forth language but a language of the same name
16:40:38 <arke> I440r: which file is it? :)
16:40:40 <I440r>   -- chuck moore
16:40:43 <I440r> inline.f
16:40:59 <Nurf> It doesnt matter how good iisforth is - it doesnt run on the hardware I have
16:41:09 <I440r> true :)
16:41:15 <I440r> but some of the extensions might work for you
16:41:19 <I440r> you would have to port them tho
16:42:39 <Nurf> *nod*
16:42:40 <Nurf> Ta
16:43:01 <Nurf> I'll look, but I would be more interested in finding out how to do what I need using standard Forth semantics
16:43:20 <Nurf> basically I am asking how the heck to specify two behaviours for one word
16:43:26 <I440r> it doesnt use ANS semantics, it uses forth semantics
16:43:33 <I440r> well
16:43:37 <I440r> you make it state smart
16:43:38 <Nurf> one for interpretation and the other for compilation
16:43:45 <I440r> something YOU as the USER are not allowed to do 
16:43:56 <I440r> ans seperates users from system
16:43:58 <I440r> your the users
16:44:01 <I440r> its the system
16:44:06 <I440r> never the twain shall meet
16:44:15 <Nurf> I dont believe you
16:44:21 <Nurf> What I read tells me I can do it
16:44:24 <I440r> you CAN do it
16:44:29 <Nurf> What I know is that I dont know exactly how to do it
16:44:32 <I440r> hyour just not allowed to and be ans compliant
16:44:41 <Nurf> I dotn care about ans compliance
16:44:45 <I440r> state if do this alse do that then
16:44:52 <Nurf> I am not doing something that requires ANS compliance
16:44:59 <Nurf> I am doing stuff that uses std forth semantics
16:45:12 <Nurf> the thing I am using HAPPENS to be an ANS compliant forth
16:45:33 <Nurf> I dont care about religion, I just want to know how to make a word have two behaviours :-P
16:46:05 <Nurf> All I have atm, is the "Forth Programmers Handbook" and this stuff is clear as mud :-)
16:46:48 <I440r> thats ans for ya
16:46:49 <I440r> lol
16:47:01 <arke> ANS is teh suck
16:47:04 <I440r> lets look at an example
16:47:17 <Nurf> I pasted one above
16:47:28 * arke doesnt like "immediate"
16:47:30 <I440r> : (foo) postpone x postpone y ;
16:47:40 * arke would prefer maybe i: and ;i
16:47:47 <I440r> : foo state if (foo) else x y then ; immediate
16:48:16 <Nurf> arke: Yeah, i: would be cool, I think
16:48:37 <Nurf> Ok. That checks between compile and runtime, right?
16:48:47 <I440r> i considered it for isforth but rejected - its just more cruft :)
16:48:48 <Nurf> What about interpretation and compile states?
16:48:54 <I440r> yes
16:49:02 <I440r> nooo
16:49:15 <I440r> if you do foo on the command line you are in interpret mode
16:49:21 <Nurf> Yup
16:49:23 <I440r> not compile so foo runs x and y
16:49:55 <I440r> if your inside a : definition and refernce the immediate word foo - it sees you are in compile mode and calls (foo) which compiles x and y inline
16:50:10 <I440r> state is a variable. if its false your in interpret
16:50:15 <I440r> if its true your in compile
16:50:23 <Nurf> and runtime?
16:50:39 <Nurf> Ah
16:50:45 <I440r> : blah 1 2 3 foo ;
16:50:47 <Nurf> (foo) never runs at runtime, its immediate
16:50:55 <I440r> will be compiled as : blah 1 2 3 x y ;
16:51:23 <I440r> its BLAH's run tiome your interested in
16:51:37 <Nurf> *nod*
16:51:38 <Nurf> Ta
16:51:39 <Nurf> I will play
16:51:47 <I440r> :)
16:52:22 * arke hates chemistry
16:52:23 * arke hates chemistry
16:52:23 * arke hates chemistry
16:52:23 * arke hates chemistry
16:52:36 <arke> chemistry hell !
16:52:38 <arke> :)
16:53:18 <arke> : compound-hell-store 0 do hell ! next ;
16:53:36 <arke> chemistry ansi-forth microsoft 3 compound-hell-store
16:53:39 <arke> :)
16:54:31 <Nurf> Been a while since I had to worry about chem :-)
16:55:03 <arke> ANS defines some pretty superfluous words
16:55:06 <arke> like */mod
16:55:17 <arke> : */mod * /mod ;
16:55:22 <arke> grr'
16:55:42 <Nurf> *nod*
16:55:58 <Nurf> Yes... Its one of the reasons I want to redo the engine
16:56:08 <Nurf> I hav almost 1000 words when I start up
16:56:08 <arke> actually
16:56:15 <Nurf> I only need about 30 :-)
16:56:21 <arke> : */mod rot * /mod ;
16:57:08 <arke> compile, emit, dup, drop, swap, over, key, key?, open, load, pick, +, -, *, /mod
16:57:37 <Nurf> ! @
16:57:43 <arke> oh yeah, those too
16:58:17 <arke> check out "levels of forth" (link following)
16:58:21 <Nurf> and some looping constructs and conditionals
17:00:20 <arke> http://www.taygeta.com/forthlev.html
17:22:39 <Herkamire> Nurf: did you get 0x to work?
17:24:09 <Herkamire> something along the lines of:   : 0x state @ if 0x-compile exit then 0x-runtime ;
17:24:39 <Nurf> Oh, I have actually been dragging into other stuff :-)
17:25:23 <Nurf> I have to do dinner and stuff
17:25:24 <Herkamire> oh, and that has to be immediate
17:25:30 <Nurf> *nod*
17:25:33 <Nurf> Thanks
17:25:46 <Nurf> It's logged. I'll have to try it later
17:26:04 <Nurf> *wave*
17:26:10 --- quit: Herkamire ("Changing server")
17:26:22 --- part: Nurf left #forth
17:27:57 --- join: Herkamire (~jason@h000094d30ba2.ne.client2.attbi.com) joined #forth
17:29:15 --- quit: Herkamire (Client Quit)
17:29:27 --- join: Herkamire (~jason@h000094d30ba2.ne.client2.attbi.com) joined #forth
17:29:49 <Herkamire> &*^*&#$ irc clients
17:29:52 <arke> :)
17:29:57 <arke> which one do you usE?
17:30:01 <Herkamire> irssi
17:30:05 <Herkamire> and yes I've tried bitch-X
17:33:38 <arke> irssi is so much better than BX
17:33:41 <arke> i love irssi
17:33:51 <arke> also, i screen it
17:33:56 <arke> so its always there
17:34:01 <arke> screen irssi -n arke -c freenode
17:34:03 <arke> :)
17:34:06 <Herkamire> me too, only problem is I can't get it to keep the server messages for the opn server out of this window
17:34:36 <arke> I use separate irssi sessions for separate servers
17:35:04 <Herkamire> I was thinking about doing that, but then I'd have to check both, and I'd end up being even more spacy :)
17:35:11 <arke> o.O
17:35:21 <arke> what window manager and terminal do you use?
17:37:07 <Herkamire> ion, rxvt
17:37:18 <arke> ion?
17:37:24 <arke> i keep hearing about ion ... link?
17:38:27 <Herkamire> kc5tja and tathi use it too
17:38:51 <arke> link?
17:38:52 <arke> :0
17:38:56 <arke> :)
17:40:17 <Herkamire> ahh there it is:  http://modeemi.fi/~tuomov/ion/
17:40:40 <Herkamire> gentoo's link to the homepage is old. freshmeat.net to the rescue
17:40:53 <Herkamire> I am _very_ happy with ion
17:42:57 <Sonarman> why do you like it?
17:43:44 <Herkamire> gimp is the only program where I _ever_ have to move windows. and I _never_ resize windows. they always take up all the available space
17:43:49 <Herkamire> (generally the whole screen)
17:44:30 --- quit: melinda (Read error: 54 (Connection reset by peer))
17:44:59 <Herkamire> I often have around 10 programs open, and with ion I can get to the one I want in two or three keystrokes (counting alt)
17:45:04 <arke> naaah, im not liking this
17:45:30 <arke> i have every app in a different space automatically, but i group certain things
17:45:44 <Herkamire> arke: I group things with ion
17:45:48 <arke> like, my rxvt's for one thing go to one place, and to another place for another
17:45:51 <Herkamire> I have 6 desktops
17:45:56 <arke> ion would drive me crazy :)
17:46:05 * arke has 12, but usually doesnt use them all
17:46:06 <Herkamire> on is split into three sections for the toolbars in gimp and sodipodi
17:46:52 <Herkamire> s/on/one/
17:46:56 <arke> gaim would drive me crazsy, too
17:47:08 <Herkamire> I use centericq 
17:48:15 <Herkamire> ion is much better with applications that use just one window (like any text-mode program) or programs that you only use one window at a time (like web browsers)
17:48:45 --- join: ASau (~asau@158.250.48.197) joined #forth
17:49:10 * arke nods
17:49:16 <arke> terve, ASau 
17:49:24 <ASau> Dobroe utro!
17:50:27 <Herkamire> most people waste soooo much time moving windows around and resizing them and hiding them and finding them
17:51:22 <Herkamire> my OS will not have windows
17:52:37 <Herkamire> you will probably be able to divide up the screen to work with multiple things at once, but I see no point in making it so you can have views that overlap eachother
17:53:56 <Herkamire> arke: oh right, I think you asked (a few hours ago before I disapeared to aikido) how I would improve GUI design
17:53:59 <Herkamire> this is one
17:54:12 <Herkamire> I think I would do a zooming interface instead of file system.
17:54:35 <Herkamire> undoo across whole system
17:55:07 <Herkamire> computer comes back to exactly where it was when you shut down and start back up (like palm pilot)
17:55:34 <Herkamire> I can't believe I typed "undoo"
17:55:55 <Herkamire> automatic saving (but you can still undo, even after rebooting)
17:56:30 <arke> Herkamire: you stole my friend's idea! lol
17:56:31 <arke> :)_
17:56:44 <Herkamire> none of these are original
17:56:44 <arke> yes, a 3d, circular, filesystem.
17:56:54 <Herkamire> not 3d
17:56:54 <ASau> arke: Persistancy is not a new idea.
17:56:57 <arke> well, not circular :)
17:57:13 <Herkamire> no, I wasn't thinking circular
17:57:36 * arke imagines AppDirs and a 3D filesystem browser as the shell
17:58:13 <Herkamire> I think 3d is pointless because the display is 2D. you can't see through stuff (at least not well)
17:58:31 <ASau> Herkamire, have you tried to live under Emacs with "read-desktop" in ".emacs"?
17:58:56 <ASau> And "emacs" in your ".profile"?
17:59:19 <Herkamire> ASau: no, what does it do?
17:59:30 <arke> like, if you wanted to edit a textfile called "foo", you move to the /Documents room, then either look around for "foo", or move into the Text room, which contains a path to the Foo room. the when you enter the Foo room, You've got several choices, like "View", "Edit", "Change Doors"
17:59:37 <arke> the latter being a way to "move"
17:59:38 <arke> :)
17:59:46 <ASau> You just get a persistency this way.
18:00:33 <Herkamire> arke: sounds like it could be fun, but would be time consuming, and you would still have to make rooms/directories 
18:00:48 <Sonarman> ASau: is the "read-desktop
18:00:57 <Sonarman> ASau: is the "read-desktop" function built into emacs?
18:01:10 <Herkamire> I was thinking (as described in The Humane Interface) of just having one gargantuous desktop that you can view at any magnification. 
18:01:16 <ASau> Herkamire: You do you work in Emacs. C-x C-c out of it at end. Login next day and there you're.
18:01:23 <Herkamire> the desktop has the documents right on it (not an icon and a caption, but the actual document)
18:01:58 <Herkamire> ASau: cool. I have heard of that. (then forgot about it)
18:02:18 <ASau> Sonarman: Never bothered. I've installed the full packaging in my Slack 9.
18:02:42 <Sonarman> sorry, by "built into emacs
18:02:43 <ASau> Anyway I have to use MULE.
18:02:52 <Sonarman> dammit stupid quotation next to the enter key
18:03:05 <Herkamire> Sonarman: use dvorak :)
18:03:12 <Herkamire> then you can have it near tab
18:03:52 <arke> :)
18:04:10 <Herkamire> the zooming desktop saves you the trouble of making directories (you can just cluster documents near eachother)
18:04:10 <Sonarman> sorry, by "built into emacs" i didn't mean built into the emacs binary, i meant included in the official package.  anyway, i can't find read-desktop in my emacs; i'll look for it on emacswiki
18:04:26 --- join: madgarden (~madgarden@Kitchener-HSE-ppp3576567.sympatico.ca) joined #forth
18:04:32 <Sonarman> Herkamire: the other day i was trying to use dvorak :)
18:04:49 <ASau> Sonarman: You should setq first, IIRC.
18:04:56 <Herkamire> and it saves you the trouble of naming your files (because the file is displayed)
18:05:31 <Herkamire> Sonarman: it's worth re-learning to type. it's so much more relaxing on the hands. and faster
18:05:39 <ASau> Sonarman: Read info. Search for desktop in "Concept Index"
18:05:59 <arke> Herkamire: i did it for a few days, then switched back..
18:06:30 <Herkamire> :) yeah, I think it takes a week or two to be comfortable
18:07:27 <arke> B BAck
18:07:29 <arke> Ack
18:07:36 <arke> I don't understand ANSI
18:07:48 <Herkamire> :)
18:07:53 <ASau> Herkamire: Try not to invent a new but worse Emacs.
18:08:13 <arke> It wants to standardize everything, yet copies of the standard cost several hundreds of dollars
18:08:28 <Herkamire> ASau: don't you worry about that. maybe a new but worse vim... ;)
18:08:31 <arke> and the standard is sompletely superfluous, anyway
18:08:38 <arke> */MOD is _not_ needed
18:08:43 <arke> POSTPONE is ugly
18:08:51 <arke> it needs macro defs
18:09:04 <madgarden> What's wrong with POSTPONE?
18:09:22 <arke> madgarden: COMPILE is nicer, and also more flexible, and also more user-aware
18:09:26 <ASau> Herkamire: "What user should know about <vi>?"  "Esc :q! Enter"
18:09:41 <arke> Emacs sucks.
18:09:43 <madgarden> Hmm. But don't you then need both COMPILE and [COMPILE] ?
18:09:46 <arke> Vi is the best :)
18:09:52 <arke> madgarden: exactly.
18:09:59 <arke> madgarden: prettier, more to the point.
18:10:01 <madgarden> So, more explicit is good.
18:10:08 <arke> yes.
18:10:24 <arke> */mod is a good naming for it, but the word itself is completely uneccesary
18:10:29 <madgarden> I take it you're more a fan of Forth 83 perhaps?
18:10:34 <arke> :)
18:11:01 <ASau> Wah! I haven't seen non-ANTI fans here!
18:11:12 <arke> er?
18:11:26 <ASau> Most folk is accustomed to POSTPONE, but not COMPILE. 
18:11:46 <Sonarman> ASau: hehe it's desktop-read :)
18:11:57 <madgarden> I used POSTPONE in mine, but I might consider splitting it into the 2 COMPILEs.
18:12:22 * Herkamire wanders off to teach his forth editor to act like a command line
18:12:26 <ASau> Sonarman: I don't rewrite my ".emacs" in any two days.
18:12:55 <arke> also, this is a personal thing ... i don't like the placement of immediate
18:12:59 <arke> its just not forth-like
18:13:00 <arke> :)
18:13:14 <madgarden> Where should it go?
18:13:25 <arke> well, theres several options
18:13:35 <arke> a good ones (i guess) to use would be
18:13:39 <arke> i: and ;i
18:13:47 <arke> and them m: and ;m for macros
18:13:55 <Herkamire> my forth doesn't have immediates or state :)
18:13:59 <madgarden> What about compile-only?
18:14:06 <arke> Herkamire!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!11
18:14:10 <arke> madgarden: just : ;
18:14:12 <ASau> arke: No need. IMMEDIATE is sufficient. It's done well.
18:14:24 <arke> ASau: but its _after_ the word is defined
18:14:28 <ASau> madgarden: ?EXEC
18:14:31 <arke> which makes things not so nice.
18:14:46 <ASau> arke: Think postfix.
18:15:19 <madgarden> ASau: True. Mine's in C though, so I put the state in flags to speed things up a touch.
18:16:21 <arke> .
18:17:38 <ASau> Herkamire: That you have no state, it's a problem at your site. You can't do some non-trivial compiling words.
18:19:03 <ASau> arke: Once I used to be a vi-fan. Emacs is better. Modes how they're done in "vi" are evil.
18:19:25 <arke> ASau: emacs gives me cramps
18:19:37 <ASau> arke: Also vim is not so programmed as Emacs.
18:20:12 <ASau> is not so _easily_ programmed
18:20:31 <Herkamire> let's not argue emacs vs vim
18:20:51 <chandler> emacs!
18:21:23 <ASau> Well, anyway the only of my friends stayed with vi are sysadmins.
18:21:40 <ASau> They have vi installed everywhere by default.
18:21:57 <ASau> Delelopers just install Emacs.
18:23:41 <Herkamire> ASau: I have a seperate definition for immediates. the editor does all the work. instead of:    : dup state @ if $95edfffc , else [ $95edfffc , ] then ;
18:23:58 <Herkamire> I have:   : ,if $95edfffc , ; : if if ;
18:25:20 <ASau> When your editor does just what a Forth has to compile, it's good.
18:25:26 <Herkamire> the editor does the magic. words starting with , are sortof "macros" in this way: when the editor sees the word in green (compile) and has a corresponding word (same name, but with comma prefix) it replaces it, but in the execute color
18:25:49 <ASau> But how am I to handle user extensions?
18:25:50 <Herkamire> so the interpreter sees this:   : if [ ,if ,;
18:26:20 <Herkamire> ASau: huh?
18:28:10 <ASau> I define words that simulate a primitive syntax with validity test done at compile time by immediate words.
18:29:44 <ASau> That is: user enters text, Forth compiles, runs occasionally.
18:30:41 <ASau> User is supposed to have a simple syntax, for not to make a system-wide crash.
18:31:09 <Herkamire> mmmmmm. all those example definitions of "if" above are supposed to read "dup"
18:31:14 <ASau> As Forth compiles it is also time to check.
18:33:48 <Herkamire> Unix was  
18:33:49 <Herkamire> +not designed to stop you from doing stupid things, because that would also stop you from doing clever things.  &mdash; Doug G 
18:33:51 <Herkamire> wyn
18:34:06 <Herkamire> I believe the same about forth
18:34:28 <Herkamire> If you want to stop the machine from crashing, use memory protection
18:35:46 <ASau> Herkamire: memory protection is less efficient than safe language.
18:36:35 <ASau> Just look at ML family, you can't crash system because of a wrong subscription index.
18:37:15 <ASau> All "possible buffer overrun"-type errors are caught at compile.
18:37:17 <Herkamire> ahh, now editor works as a command line too :)
18:37:33 <Herkamire> ASau: forth is not a safe language.
18:37:38 <ASau> Without any efficiency lossage.
18:38:17 <ASau> Herkamire: That doesn't mean I can't make it restricted for not to harm system in case it should not do so.
18:40:05 <ASau> If you have tools for secure programming, why not do so?
18:40:47 <madgarden> WORD ;)
18:41:24 <ASau> ?
18:42:07 <madgarden> I agree.
18:47:20 <Herkamire> ASau: you can stop forth from harming the system through protected memory
18:47:51 <Herkamire> safe languages do not have direct access to memory. I can't amagine/accept a forth without direct access to memory
18:48:17 <ASau> TO-words.
18:48:22 <madgarden> My C Forth-like works that way.
18:48:59 <ASau> Protected memory is invented for those who can't imagine safe-language-based OS.
18:49:50 <ASau> It's a solution of the kind "let's build a wall".
18:50:42 <ASau> There is more efficient "this does what it's meant -- proven" approach.
18:50:59 <ASau> "Proven" is not "tested". It's proven.
18:51:07 <madgarden> Got some examples?
18:58:01 <Herkamire> I agree that protected memory seems sorta like cheating, but it saves me a hell of a lot of time.
19:00:55 <Herkamire> has anyone actually implemented a "proven", compiled language?
19:02:28 <Sonarman> isn't ML a "proven", compiled language?
19:02:35 <madgarden> O'Caml?
19:03:24 <ASau> The simplest example is records in Forth and bit-fields' names.
19:03:42 <ASau> Herkamire: /join #sml  /join #ocaml
19:04:03 <Herkamire> no thanks
19:09:17 --- join: tathi (~josh@pcp02123722pcs.milfrd01.pa.comcast.net) joined #forth
19:09:37 <ASau> Dobroe utro, tathi!
19:16:36 <tathi> Hi
19:18:29 <madgarden> Hello again.
19:23:13 --- join: melinda (melinda@melinda.usercloak.freenode) joined #forth
19:23:29 <ASau> Dobroe utro!
19:24:38 <tathi> ASau: are you taking over from Robert as official greeter? :-P
19:25:48 <ASau> No! It's just a matter of style.
19:26:35 <ASau> It used to be a good practice.
19:46:07 --- join: kc5tja (~kc5tja@66-91-231-74.san.rr.com) joined #forth
19:46:08 --- mode: ChanServ set +o kc5tja
19:46:33 <ASau> Dobroe utro!
19:47:09 <kc5tja> I don't know what that means.
19:47:38 <ASau> Good morning!
19:48:42 <ASau> Dobroe -- "good", "kind", "big". Utro -- "morning".
19:49:21 <kc5tja> Ahh
19:49:23 <kc5tja> What language?
19:49:48 <ASau> Russian.
19:50:11 <madgarden> ANS Russian?
19:50:25 <ASau> GOST Russian.
19:54:06 <ASau> Russkijj GOST. ;)
19:54:45 <madgarden> Heh.
19:54:54 <ASau> GOST XXXX-YY
19:55:37 <ASau> I can't recall what year last time the writting rules has been revised.
19:56:09 <ASau> It was more than 20 years ago.
19:58:47 <ASau> How "swap drop" is now called?
19:59:02 <ASau> "nip"?
19:59:33 <madgarden> Yea.
19:59:53 <ASau> I.e. : nip  swap drop ;
20:00:06 <ASau> Thanks.
20:02:09 <madgarden> swap drop would potentially be faster anyway
20:03:03 <tathi> ?
20:03:05 <ASau> It depends.
20:03:12 --- join: topaz8 (~topaz@node-d-9180.a2000.nl) joined #forth
20:03:27 <ASau> Dobroe utro!
20:03:41 <topaz8> nastrovje
20:03:44 <madgarden> Well, it's faster in my ITC system anyway. ;)
20:04:00 --- join: warpzero (~warpzero@dsl.142.mt.onewest.net) joined #forth
20:04:11 <ASau> Dobroe utro!
20:04:14 <tathi> ah.  I do native code, and cache the top of stack in a register, so nip is just incrementing the stack pointer.
20:05:09 <ASau> In my "minimal" system, "swap" is not a primitive. :)
20:05:18 <ASau> >r drop r>  is faster.
20:06:06 --- quit: warp0x00 (Read error: 110 (Connection timed out))
20:09:07 <madgarden> Mine's a C Forth-like, I hardly have any primitives, no language. The user just registers the words they want to use and call them whatever they want.
20:09:20 <ASau> How Gforth or PFE would like a sequence of this kind:
20:09:30 <ASau> 0= if nip ;s then
20:09:35 <ASau> ?
20:09:42 --- quit: skylan (Read error: 113 (No route to host))
20:10:10 <madgarden> What's ;s ?
20:10:46 <ASau> The word, that is compiled by ";" to return from a call.
20:11:17 <ASau> It returns.  ip = *(sp++)
20:11:42 <ASau> Sorry,  ip = *(rp++)
20:11:47 <ASau> If you prefer C.
20:12:11 <ASau> BTW, speaking about C-Forth.
20:12:25 <ASau> I haven't succeded to compile one.
20:12:31 --- part: topaz8 left #forth
20:12:35 --- join: skylan (sjh@vickesh01-4820.tbaytel.net) joined #forth
20:13:25 <madgarden> Have they all been broken?
20:13:36 <ASau> It seems so.
20:13:55 <madgarden> Mine's pretty simple to build.
20:13:59 <madgarden> Not released yet though.
20:14:03 <madgarden> And not the fastest, either.
20:14:05 <ASau> Maybe they require a particular lex/yacc to compile, not a GNU ones.
20:14:13 <madgarden> But, the speed seems comparable to Lua.
20:14:59 <ASau> I've looked once at Lua.
20:15:07 <ASau> I don't take it. 
20:15:28 <ASau> It's the same as everything else of this kind.
20:15:49 <madgarden> Yea, that's why I made my system.
20:15:59 <madgarden> Don't need to be stuck with some new syntax.
20:16:19 <madgarden> So, I can roll my own depending on the problem.
20:16:27 <kc5tja> Forth cannot be described in lex/yacc.
20:16:27 <madgarden> Mostly, I'm using it for game projects.
20:17:19 <ASau> It's a reimplementation of the same old thing with new fancy names and "curly braces".
20:17:37 <madgarden> Yep, I agree.
20:17:52 <madgarden> But, Lua is one of the more popular ones.
20:17:53 <ASau> I don't like C Forthes.
20:18:04 <madgarden> Ahh well. ;)
20:18:46 <ASau> I tried to get acquainted with that one, but it couldn't be even built.
20:19:25 <madgarden> WHich one?
20:19:38 <ASau> Gforth and PFE are just plain MONSTERS.
20:19:55 <ASau> Alan Pratt's derived.
20:19:57 <madgarden> I've mostly been playing with FICL as a reference for my work.
20:22:32 <ASau> Hm. Maybe I'll consider FICL as next one to build.
20:22:40 --- quit: tathi ("leaving")
20:23:16 <ASau> For today I've seen just one useful utility, that is portable between DOS and Linux.
20:24:45 <ASau> The most interesting thing is that it doesn't run with "make", but does pretty well with "cc -o XXX *.c".
20:25:29 <madgarden> Which utility is this?
20:25:58 <ASau> GeMa macropreprocessor.
20:26:44 <ASau> Does Gforth have ";s"?
20:27:54 <ASau> Or it was thrown off by ANTI-standard?
20:28:22 <madgarden> Looks like it does.
20:28:58 <ASau> Don't frighten me so.
20:29:13 <madgarden> gforth has ;s
20:29:24 <madgarden> At least, the version I have.
20:29:44 <ASau> I've been afraid I am to search this file for replacement.
20:31:04 --- join: arke_ (~chris@ca-cmrilo-cuda1-c3b-66.vnnyca.adelphia.net) joined #forth
20:34:38 <I440r> whats ;s
20:35:32 <ASau> It pops R-stack top to IP.
20:36:07 <I440r> what ?
20:36:12 <madgarden> EXIT?
20:36:20 <I440r> ooh
20:36:30 <ASau> : goto ( n -- ) r> drop >r ;
20:36:34 <I440r> into ip - not into where ip points heh
20:37:13 <ASau> : ; compile ;s [compile] [ ;
20:37:37 <madgarden> immediate
20:37:47 <ASau> yes, immediate
20:38:12 <I440r> err ;s is just exit then
20:38:29 <madgarden> That's what I gather...
20:38:36 <I440r> where does ;s come from
20:38:40 <ASau> Hm. Nevertheless they've renamed it.
20:38:48 <I440r> ;s is totaallt non descriptive
20:38:51 <madgarden> I guess ;s is just more obvious?
20:38:52 <madgarden> :)
20:39:01 <I440r> who renamed it
20:39:21 <I440r> ans ?
20:39:22 <ASau> ANTI-standartization team, it seems
20:39:29 <I440r> ugh
20:39:34 <kc5tja> I440r: No, heavens no.
20:39:35 <madgarden> It's rather hard to do a google search for ;s
20:39:42 <kc5tja> ANSI insists on the word "EXIT" instead.
20:39:55 <I440r> in this case ans is correct
20:40:19 <kc5tja> I440r: I direct you to the case where the most often used words are shorter.  Consider @ instead of FETCH, or ! instead of STORE.
20:40:27 <arke_> kc5tja: !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
20:40:30 <arke_> :)
20:40:32 <kc5tja> It's entirely possible that EXIT is used so often that they created a short-hand name for it.
20:40:39 <I440r> i refer you to the historical names
20:40:50 <kc5tja> Fuck the historical names.
20:40:52 <I440r> ;s is a rename for a word that already exists - why 
20:41:05 <arke_> because ans can suck it
20:41:05 <kc5tja> I440r: Because it's shorter, easier to type, and hence, less error-prone.
20:41:09 <I440r> i always prefered unnest to exit
20:41:19 <kc5tja> unnest is non-obvious.
20:41:32 <kc5tja> Unnest from what?  A loop?  The current definition?  Does it depend on context?
20:41:37 <I440r> nope - you nest into : definitions so you unnest from them :)
20:41:39 --- quit: arke ("Lost terminal")
20:41:53 <kc5tja> I440r: That is a completely arbitrary definition of the word nest and unnest.
20:42:02 <ASau> Hm. If I have not nested, what should I unnest?
20:42:41 <madgarden> This is why Forth is fun.
20:42:43 <kc5tja> Anyway, I won't bother getting into any religious wars here.
20:43:14 <kc5tja> Not today at least.
20:43:34 <madgarden> Why not ;x
20:43:41 <I440r> i dont object to ;s - i just dont object to exit either
20:44:04 <ASau> Because it is ";s" already.
20:44:16 <I440r> s is usually indicative of a string
20:44:26 <arke_> alright
20:44:30 <I440r> so ;s looks like a string termination 
20:44:31 <ASau> I think ";"-semantics is the explanation.
20:44:34 <arke_> whats the signal thats gonna kill a process no matter what?
20:44:46 <madgarden> -9
20:44:47 <I440r> kill -9 
20:44:52 <ASau> ";" is an end of a word definition.
20:44:55 <I440r> but that still doesnt always work
20:45:04 <arke_> kc5tja: what do you like better ---     : foo ... ; immediate    or    i: foo ... ;i
20:45:14 <I440r> it wont kill sox if there is a chile play running
20:45:27 <kc5tja> arke_: None of the above.  I prefer two separate vocabularies, one for compiler words, and one for normal words.
20:45:54 <I440r> kc5 im still undecided on that one :)
20:45:56 <arke_> well, i meant for _defining_ them
20:45:56 <kc5tja> That being said, of the two choices given to me, I prefer : ... ; immediate
20:45:58 <I440r> but it sounds neat
20:46:07 <arke_> I440r: i prefer the second one. probably easier to implement too.
20:46:10 <madgarden> immediate is the most obvious to read.
20:46:12 <kc5tja> arke_: So do I.
20:46:23 <arke_> kc5tja: you just said you didnt...
20:46:28 <arke_> i like i: ... ;i better
20:46:34 <kc5tja> arke_: Re-read, very, very carefully, what I wrote above.  :)
20:47:02 <arke_> er....
20:48:19 <arke_> ...
20:49:01 --- nick: arke_ -> arkje
20:49:04 --- nick: arkje -> arke
20:51:05 <ASau> Well, maybe I'll return soon.
20:51:14 <ASau> Do svidanija!
20:51:16 --- quit: ASau ("Toffee IRC client for DOS v1.0/b535")
20:51:36 * kc5tja busts out his O'reilly & Associates book on using Make (though, not GNU Make, unfortunately).
20:51:42 * kc5tja will be needing that soon enough.
20:52:02 <arke> make is easy.
20:52:04 <arke> but
20:52:07 <arke> a book is good.
20:52:19 <arke> kc5tja: can i borrow it, when you're not using it?
20:52:35 <kc5tja> Make is easy only for easy projects.
20:52:47 <kc5tja> Ever look at even a moderately complex project's make file?  :)
20:52:50 <kc5tja> It can get pretty hairy.
20:53:32 <madgarden> Are there any Forth-like shell replacements?
20:55:01 <kc5tja> None that I'm aware of, but then again, there wouldn't be much point.
20:55:02 <kc5tja> :)
20:55:22 <madgarden> Heh, why not?
20:55:24 <kc5tja> A Forth shell would itself be Forth, at least in my mind, so may as well use a regular Forth interpreter.  :)
20:56:22 <madgarden> Well that's the Forth ideal. But, when you're already got an OS, and don't feel like scripting in bash, wouldn't Forth be better?
20:56:38 <arke> kc5tja: I hate automake/conf
20:56:46 <arke> kc5tja: it produces major bloatware
20:56:47 <kc5tja> arke: I don't use automake/autoconf.
20:56:54 <arke> :
20:56:55 <arke> :)
20:57:10 <kc5tja> madgarden: That depends.  I've written some scripts in Forth, some in Python, and still more in Bash.  It just depends on what I'm trying to do.
20:58:18 * arke likes csh
20:58:58 <Herkamire> how would/could a forth shell work?
20:59:41 <Herkamire> would you hack the interpret loop so if it doesn't find a word in the dictionary it searches the $PATH for an executable?
20:59:56 <arke> prolly
21:00:11 <kc5tja> arke: http://www.faqs.org/faqs/unix-faq/shell/csh-whynot/
21:00:22 <kc5tja> May as well use REXX for that.
21:00:38 <kc5tja> I find REXX is overwhelmingly the best scripting language, which uses *precisely* that same principle.
21:01:10 <Herkamire> what about piping and stuff?
21:01:14 <Herkamire> globbing
21:01:41 <kc5tja> The shell interpreter handles that stuff, so it's still not an issue.
21:02:19 <kc5tja> Commands that REXX doesn't grok are converted to a string, and sent to the currently addressed "environment."  (Read: program.)
21:02:39 <kc5tja> By default, REXX addresses the system's command interpreter (conveniently named SYSTEM in REXX).
21:02:41 <Herkamire> so a forth shell, would just pass the whole rest of the line to the interpreter if it found that the word you typed was not a word but a program in $PATH
21:03:30 <madgarden> Just an extra parse step.
21:03:42 <arke> kc5tja: can you give me a few pointers to REXX, and maybe why its so good?
21:03:57 <arke> btw, i think tcsh supports most, if not all, the things that are listed there
21:03:59 <madgarden> Could possibly compile a path into words, too.
21:06:34 <Herkamire> I really can't imagine forth making a good shell, unless the system is forth
21:06:59 <arke> Herkamire: i agree.
21:07:36 <madgarden> Well, I'm stuck with command.exe over here. ;)
21:11:17 <kc5tja> arke: http://www.rexxla.org .  The tutorial they have is crappy, but check out some of the pages under the Links section.  I'm sure someone has a better tutorial.
21:12:13 <arke> ok
21:26:16 --- quit: madgarden ("*frotz*")
21:28:20 <arke> hrm
21:28:36 <arke> kc5tja: how do you find my idea of a loop stack? :)
21:28:58 <kc5tja> Personally, I don't have any need for one.  I just use the return stack for a loop stack.
21:29:49 <arke> i thought they werent supposed to use the return stack?
21:31:06 <arke> you're gonna kill me for saying this, but...
21:31:14 <arke> rexx looks like basic
21:32:41 <kc5tja> No, I'm not going to kill you for it.
21:32:50 <kc5tja> REXX has some similarities with BASIC.
21:33:00 <kc5tja> Though they evolved completely separately.
21:33:25 <arke> :)
21:33:35 <arke> kc5tja: can i still borrow that make book someday? 
21:33:52 <arke> :)
21:33:56 <kc5tja> Why not just get one yourself?  :)
21:34:03 <kc5tja> It's not that expensive.
21:37:39 <arke> because i wanna save my money for drving
21:37:40 <arke> :)
21:38:09 <arke> i wonder if theres ever gonna be a forth convention close to here
21:38:20 <arke> before i move back to germany, in < 2 years
21:38:51 <kc5tja> Dunno.
21:39:17 <arke> would be neat
21:39:21 <arke> might meet you :P
21:40:43 <kc5tja> I won't make it.
21:40:48 <kc5tja> You're WAY too far north for me.
21:41:38 <arke> http://bash.org/?56558
21:41:44 <arke> kc5tja: im not even that far lol
21:41:54 <arke> and it would prolly be in LA or such
21:41:57 <arke> a good compromise :)
21:42:24 <kc5tja> SVFIG meetings are the closest Forth conventions to me, and Silicon Valley is just too far north for me.
21:44:49 <arke> kc5tja: well, i'll just pick you up or something :)
21:45:08 <kc5tja> If you're willing to drive down practically to the Mexican border, sure.
21:45:09 <arke> and when and how often are they?
21:45:09 <kc5tja> :)
21:45:16 <kc5tja> arke: I have no idea.
21:45:23 <kc5tja> I don't even know if they're still meeting.
21:45:27 <arke> lol
21:45:32 <arke> well, find out for me
21:45:41 <kc5tja> Sorry, I have more important things to do at the moment.
21:46:56 <arke> bash.org/?5300
21:49:10 <arke> :)
21:54:11 <kc5tja> Well, I think I'm going to go to bed.
21:54:17 <kc5tja> I have school in the morning.
21:55:11 <arke> ok :)
21:55:15 <arke> good night
21:56:49 --- quit: I440r ("Leaving")
21:59:14 --- quit: Herkamire ("bedtime for bonzo")
22:11:01 --- join: segher (~segher@blueice1a.de.ibm.com) joined #forth
22:13:39 --- join: Serg_Penguin (~z@212.34.52.140) joined #forth
22:14:08 <Serg_Penguin> hi !
22:14:19 <segher> howdy
22:14:20 <Serg_Penguin> did one implement XML parser in Forth ?
22:14:33 <segher> why would you want to do that?
22:15:02 <Serg_Penguin> website generation
22:15:26 <Serg_Penguin> what i think :
22:15:33 <segher> why choose XML if you're using Forth, though?
22:15:40 <Serg_Penguin> generally, echo STDIN to STDOUT
22:16:07 <Serg_Penguin> if <, stop echo and start collecting chars into buf
22:16:20 <Serg_Penguin> if >, execute word in buf and go on
22:16:46 <segher> that's way too simplistic, but okay
22:17:35 <Serg_Penguin> coz XML is more standard than Forth, and XML site src can be processed by other tools, even written in Forth but more complex way
22:17:59 <segher> so why rewrite those tools in Forth, then?
22:18:27 <Serg_Penguin> coz 2-mode echoer won't grok parameters, but what if i'll want em ?
22:18:53 <Serg_Penguin> <book pages=217 weight=0.3>
22:18:57 <segher> i mean, there are plenty of existing XML tools already
22:20:13 --- quit: Sonarman ("leaving")
22:21:00 <Serg_Penguin> by now, i use M$ via VBscript, and it sucks like turbojet
22:22:09 <Serg_Penguin> i try to write pages in XML and convert to HTML _before_ upload
22:22:21 <segher> sure -- so use some better tool
22:23:10 <Serg_Penguin> i find XSLT obscure and counter-intuitive
22:23:57 <Serg_Penguin> what i want is plain file #INCLUDE (or inline output, or null-conversion to self)
22:25:40 <Serg_Penguin> : #incl ... create ... does> ... ; ( gets it's name, and echoes to stdout file w/ that name )
22:25:55 <Serg_Penguin> #incl <sidenote>
22:26:39 <Serg_Penguin> or just : <sidenote> ." <div class=sidenote>" ;
22:27:04 <Serg_Penguin> or maybe, just find a good XSLT tutor ?
22:37:08 --- quit: kc5tja ("THX QSO ES 73 DE KC5TJA/6 CL ES QRT AR SK")
23:15:24 --- quit: melinda (Read error: 110 (Connection timed out))
23:44:37 --- quit: Serg_Penguin ()
23:59:59 --- log: ended forth/03.11.17