00:00:00 --- log: started retro/06.07.29
00:21:13 --- join: Cheery (n=Cheery@a81-197-20-242.elisa-laajakaista.fi) joined #retro
00:23:28 <Cheery> How do I install that help file in linux?
00:24:04 <Cheery> cp /rf/help/ /????
01:47:27 <Cheery> I think those shallow stacks would be really useful.
01:48:03 <Cheery> I mean, I've always smashed up if I've not done something data-driven rather from code-driven.
01:48:25 <Cheery> and if you have lots of stuff in your stack, you are usually doing code-driven code.
01:48:48 <Cheery> because in another case you let your data drive your machine.
01:50:55 <Cheery> and oh boy, the data can drive the machine better than you.
03:40:29 <virl> code-driven, datas-driven? what do you mean exactly?
03:42:18 <Cheery> I mean that instead I've encoded the functionality in code, I've distributed it into stuff which detects data.
04:00:30 --- quit: virl (Read error: 104 (Connection reset by peer))
04:06:18 --- join: virl (n=virl@chello062178085149.1.12.vie.surfer.at) joined #retro
04:40:27 --- join: neceve (n=claudiu@unaffiliated/neceve) joined #retro
06:05:51 <Cheery> virl
06:05:59 <Cheery> bother to see where am I now going to?
06:06:00 <Cheery> http://retroforth.net/paste/?id=86
06:06:27 <Cheery> I'm about to write the rest logic, then make the [eax] etc. smarter.
06:06:44 <Cheery> all code you see in that page is composition from code made yesterday and today.
06:07:20 <Cheery> I also put the 32 maximum width, it's enough, I think.
06:08:37 <Cheery> I'm going to do it smarter in a way that, for example, if you get   eax -register, it automatically determines you are going to use 32 bit addressing.
06:08:58 <Cheery> then if you try write something like:   eax bx it gets angry.
06:11:41 <Cheery> because you wanted something it cannot give.
06:32:16 <docl> lookin' good
06:55:54 <docl> some experimental code I've been working on: http://retroforth.net/paste/?id=87
06:59:25 <docl> 10 following: .s  ( executes .s immediately after the first word of each following line for 10 lines )
06:59:45 <docl> maybe too complicated
08:08:53 <crc> good morning
08:10:06 <crc> Cheery: the easy way is to put a copy of help.data in a known location (e.g., your home directory) and load it when retro starts (retro -f ~/help.data)
08:24:54 <crc> =f  is the same as 'off'
08:25:01 <crc> =t  is the same as 'on'
08:30:20 <Cheery> crc: yeh, I think I should yet once rethink this damn thing.
08:30:52 <Cheery> btw. How could I get stuff from registers to TOS?
08:31:43 <Cheery> The current design almost was good already, it requires only tuning.
08:33:01 * crc will do some code for that in a few minutes
08:47:02 <docl> crc: were you going to include is: or some variant?
08:47:44 <docl> makes grepping the source easier, since you can use ": name"
08:47:48 <crc> http://retroforth.net/paste/?id=88
08:48:12 <docl> nice!
08:48:14 <crc> docl: I can
08:48:37 <crc> docl: I just pushed a patch to the repo shortening the lines in rx/bootstrap to 32 chars max
08:51:09 <docl> cool, looks nice
08:52:35 <Cheery> crc... is eax the TOS?
08:53:06 <crc> Cheery: yes
08:53:19 <crc> esi is the data stack pointer; [esi] is the NOS
08:53:20 <crc> docl: http://retroforth.net/paste/?id=89
08:54:15 <Cheery> crc, what was the prefix for octal and binary?
08:54:44 <crc> & for octal, % for binary
08:55:06 <Cheery> ok, I'll use them to prevent confusion with myself.
08:55:21 <crc> # for decimal, ' for ascii, and $ for hex
08:55:27 <crc> that covers the most common bases :)
09:14:21 <Cheery> I think I'm really missing a unificator.
09:15:49 <crc> ?
09:17:08 <Cheery> it is a cool component which allows you to use pattern matcher on lists and other such structures.
09:17:39 <Cheery> oh well, maybe I'm doing without.
09:17:57 <Cheery> I just must find out how could I possibly make them order neatly.
09:23:48 <Cheery> Or then I could use lookup tables.
09:24:16 <Cheery> they are quite simple to create, and may be even an good idea.
09:40:02 * crc pushed a few more patches, and 9.2.4 is done
10:03:16 <virl> Cheery, how far are you know with your assembler?
10:06:49 <Cheery> virl: I've encountered problems with R/M and SIB byte.
10:06:58 <Cheery> It's hard to make a working combination.
10:30:31 <virl> that's a challenge..
11:23:13 <Cheery> virl: this is actually really annoying, a moment I feel I got it, then it suddenly disintegrates... again.
11:25:21 <Cheery> crc: let me correct, literal, is actually just writing a duplication command and five byte long instruction containing the address given to literal, .
11:25:30 <Cheery> crc: erp.
11:25:41 <Cheery> I meant, please correct If I am wrong.
11:27:11 <crc> essentially correct as to the end effect, though the technical details are different
11:27:18 <crc> a literal looks like:
11:27:26 <crc> call dolit
11:27:29 <crc> dd value
11:27:38 <crc> where dolit is a function something like:
11:28:02 <crc> : dolit r> dup cell+ >r @ ;
11:29:26 <crc> though dolit is actually defined in the assembly part of rx, and is not a named word
11:29:42 <Cheery> I'm just wondering how could I possibly do a haskell-style function I could continue.
11:30:20 * crc has no experience with haskell
11:30:27 <Cheery> just that mechanism how could I change the address to call.
11:30:40 <Cheery> I mean, I first have function, like:
11:31:10 <Cheery> : foobar baaz ebbxblaa 0 0; execute ;
11:31:31 <Cheery> I'd like later to make sth like this:
11:32:23 <Cheery> : baaz same-trash-etc... 0 0; execute;
11:32:23 <Cheery> and alter function foobar so that end result would be somewhat like this:
11:32:48 <Cheery> : foobar baaz ebbxblaa [ ' baaz literal, ] 0; execute ;
11:32:51 <Cheery> you get the idea?
11:33:10 <Cheery> (btw, that 0; is a brilliant control flow tool. ;))
11:33:34 <crc> JasonWoof (#forth, developer of HerkForth introduced me to 0;)
11:34:26 <crc> are baaz and ebbxblaa are normal forth words, not macros?
11:34:30 <Cheery> how could I possibly 'tag' the address, store it somewhere and later change it?
11:34:33 <Cheery> yes, normal words.
11:34:51 <Cheery> but I wouldn't mind if it'd work with macros as well.
11:35:08 <Cheery> or wait.
11:35:11 <Cheery> would it be this simple?
11:35:15 <crc> macros are trickier since they often inline machine code... changes the counts 
11:36:02 <Cheery> : foobar baaz ebbx [ 0 literal, here cell- variable: foobar-next ] 0; execute ;
11:36:18 <Cheery> lets try that.
11:36:29 <Cheery> isn't it looking cool? :P
11:36:34 <Cheery> at least.
11:36:47 <crc> won't quite work... I don't think
11:38:02 <Cheery> arg! it says illegal instruction. :s
11:38:08 <Cheery> ok.
11:38:11 <Cheery> next try.
11:38:19 <Cheery> : foobar baaz ebbx [ 0 literal, here cell- constant foobar-next ] 0; execute ;
11:39:54 <Cheery> hehe. but that works. :)
11:40:22 <Cheery> ok, lets see if I could make it cleaner.
11:40:29 <Cheery> lets try macros.
11:41:14 <Cheery> macro : reloc ( 'name' -- ) 0 literal, here cell- constant ;
11:41:29 <Cheery> : foobar baaz ebbx reloc foobar-next 0; execute ;
11:41:44 <Cheery> oh, not forgetting to change back...
11:42:03 <Cheery> or were there a function for changing the class of the last defined function?
11:42:08 <crc> reclass:
11:42:22 <crc> or reclass...
11:42:29 * crc doesn't use them that much
11:42:46 <crc> reclass
11:42:50 * crc tries:
11:42:51 <crc> loc:
11:42:51 <crc>  : dolit r> dup cell+ >r @ ;
11:42:51 <crc>  self: var c: dolit 
11:42:51 <crc>    create , ;
11:42:51 <crc>  ~var
11:42:53 <crc> ;loc is var
11:44:02 <crc> http://retroforth.net/paste/?id=91
11:44:19 <crc> is a better approach I think
11:48:47 <Cheery> hmm?
11:48:53 <Cheery> let me show what I did. ;)
11:49:09 <Cheery> http://retroforth.net/paste/?id=92
11:49:56 <crc> ahh, clever
11:49:59 * crc likes it :)
11:50:00 <docl> this channel is so much fun to watch :)
11:50:09 <crc> that is a good word
11:51:11 <Cheery> it creates relocatable pointers.
11:51:44 <Cheery> sick, but I can suddenly invent much more use for that thing...
11:51:50 <docl> nice
11:52:12 <Cheery> it also works as a kind of special variables inside word.
11:52:23 <Cheery> you can set from outside.
11:52:57 <docl> sounds like an oop-ish construct
11:53:14 <crc> Cheery: you can use macro: rather than changing the active class to macro, then back to forth
11:53:41 <Cheery> If you think a function as object you can think it as an oop-ish construct. :)
11:53:52 <docl> how does guux-> get defined?
11:54:02 <Cheery> it is defined with constant.
11:54:10 <docl> ahh, I see
11:54:21 * docl forgot that was a parse-word
11:54:35 <Cheery> crc: do I must put it in front or back?
11:55:20 <docl> just take out the space in macro :
11:55:39 <crc> macro: <name> <code> ;
11:55:45 <crc> replacing : in this case
11:56:22 <Cheery> is it new word? I can't find it and ? says 9.1
11:56:33 <crc> get 9.2
11:56:40 <crc> 9.2.4 is the current version
11:56:49 <docl> it's very new :)
11:57:02 <Cheery> what's new this time?
11:57:09 <Cheery> except the fabulous macro:
11:57:24 <docl> prefix: class: and similar
11:57:28 <crc> lots of cleanups, anonymous, nestable definitions, normal conditionals, prefixes
11:57:33 <crc> among other things
11:57:39 * crc lost count a while back
11:58:20 <Cheery> do you keep a changelog?
11:58:32 <crc> only of the checkins
11:58:42 <crc> http://retroforth.net/changes.php
11:59:30 * crc will have to go through it and make a list of new stuff
12:03:31 <Cheery> thank you, this will help me a lot.
12:03:35 <Cheery> and many others.
12:05:04 <Cheery> btw. if you added new words, I should patch my vim plugin.
12:05:13 <Cheery> *syntax script.
12:05:15 <crc> I did
12:05:27 * crc is trying to put together a more detailed changelog
12:05:37 <Cheery> kewl, I think I'll download it then.
12:08:07 <Cheery> uh, where do I find the upgrader script?
12:08:13 <Cheery> *upgraded.
12:08:19 <crc> http://retroforth.org/getretro
12:08:34 <crc> that is the download/installer script for BSD and Linux
12:09:03 <Cheery> found already that, but I mean the upgraded vim syntax script.
12:10:12 <crc> http://retroforth.org/contrib
12:22:08 <crc> http://retroforth.org/changes.txt
12:22:13 <crc> should cover most of the new stuff
12:22:18 <Cheery> ok.
12:22:41 <Cheery> btw, do you know how could I change \r\n to \n from text files?
12:22:43 <Cheery> in unix
12:24:15 <crc> sed :)
12:24:32 <crc> cat filename | sed 's/.$//'
12:24:39 <crc> should work
12:24:51 <Cheery> thanks, but where does it stuff the result?
12:25:01 <crc> stdout
12:25:06 <crc> redirect to a new file
12:25:30 * crc has experienced data loss redirecting to the same file that is being dumped by 'cat'
12:26:21 <Cheery> I'm happy I had quick backups of that file. :)
12:26:47 <crc> backups are good
12:27:26 <Cheery> ok, now it works, but macro: doesn't highlight.
12:27:33 <Cheery> oh well, nvm.
12:27:49 <Cheery> yeh, especially when it is so easy.
12:28:07 <Cheery> cp xx<tab> xx<tab>.bak
12:28:42 <crc> darcs revert
12:28:44 <crc> :)
12:28:55 <Cheery> :D
12:29:04 <Cheery> I think I should learn use darcs.
12:29:19 <crc> I don't use most of the features it provides
12:29:37 <crc> but it's trivial to create a repo, record, send, and apply patches 
12:29:56 <Cheery> The project will be big what I'm starting, so it'd be nice to see what it looked like three years ago.
12:30:23 <crc> yeah
12:30:37 <crc> I love darcs 
12:30:53 --- join: snoopy_1711 (i=snoopy_1@dslb-084-058-186-078.pools.arcor-ip.net) joined #retro
12:31:03 <Cheery> I've heard it is the most trivial to use from repo software available.
12:31:05 * crc should figure out how to use a cron job to automatically record changes in his project directories...
12:31:13 <crc> that's true in my experience
12:31:26 <crc> cd project
12:31:30 <crc> darcs initialize
12:31:33 <crc> darcs add -r *
12:31:44 <crc> darcs record --look-for-adds
12:32:06 <crc> darcs send
12:32:10 <crc> darcs apply <filename>
12:32:20 <crc> that's my working command set 
12:33:07 <Cheery> thanks from quick lesson! :)
12:33:25 <crc> np
12:35:09 <Cheery> oh, forgot, does that include work now in 9.2?
12:35:33 <crc> yes
12:35:46 <crc> at least, it was when I last tried it...
12:35:57 <Cheery> ok.
12:36:10 <crc> it works
12:36:51 <Cheery> then I think I replace my run.sh with retroforth script which includes the stuff.
12:37:09 <Cheery> then add it, rarely needing to look in. ;)
12:38:39 <Cheery> beautiful job.
12:39:01 --- quit: Snoopy42 (Read error: 145 (Connection timed out))
12:39:11 --- nick: snoopy_1711 -> Snoopy42
12:43:07 <Cheery> crc: forths normally doesn't have error handling I know, but I accidently may invent one when doing this assembler then.
12:43:27 <crc> beyond the syntax handling?
12:43:35 <Cheery> yep.
12:43:38 <crc> ok
12:43:48 <crc> I did some code for handling signals under linux
12:44:14 <Cheery> actually to make sure user doesn't give foobar stuff for assembler, if it is necessary to do such.
12:45:53 <Cheery> btw. good job!
12:46:11 <Cheery> if I didn't mention.
12:46:14 <crc> thanks
12:47:38 <Cheery> especially like the prefixes.
12:47:47 <Cheery> more logical than postfixes in memory accessing.
12:48:16 <crc> I think so too
12:58:00 <Cheery> is c: similar to my reloc function in way it peeks a word from forfard and makes a compile for it?
12:58:49 <Cheery> so. c: blah would compile a call to 'blah'
13:01:53 <crc> no
13:01:57 <crc> c: blah 
13:02:00 <crc> is the same as:
13:02:05 <crc> ['] blah compile
13:02:32 <crc> x: blah
13:02:36 <crc> compiles a call to blah
13:04:09 <Cheery> ok, so I must use x:
13:08:59 <Cheery> except... c: was the correct. :P
13:20:20 * crc is going out for a bit; will return in an hour or two
13:21:00 --- join: Raystm2- (n=NanRay@adsl-69-149-36-229.dsl.rcsntx.swbell.net) joined #retro
13:23:58 * Cheery is going to sleep. soon...
13:26:18 --- quit: Raystm2 (Read error: 60 (Operation timed out))
14:25:38 --- quit: neceve (Read error: 113 (No route to host))
14:33:33 --- quit: Cheery (Read error: 110 (Connection timed out))
14:36:40 <crc> back
14:52:03 --- join: Raystm2 (n=NanRay@adsl-69-149-36-229.dsl.rcsntx.swbell.net) joined #retro
14:52:26 --- quit: Raystm2- (Read error: 104 (Connection reset by peer))
16:38:26 --- join: jas2o (n=jas2o@WNPP-p-203-54-32-197.prem.tmns.net.au) joined #retro
18:19:02 <Raystm2> http://retroforth.net/paste/?id=93 
18:19:11 <Raystm2> now with interactive help system.
18:20:45 <crc> very nice :)
18:23:43 <Raystm2> A little something everyday goes a long way.
18:24:44 <Raystm2> I should remember to format the backslashes before accepting the paste.
18:24:59 * Raystm2 does this
18:28:57 <Raystm2> http://retroforth.net/paste/?id=94 got rid of unnecessary backslashes and doubled the pertinant one in the modules\help.data
19:05:10 --- join: nighty_ (n=nighty@CPE00119576a9c5-CM0012c90d36fc.cpe.net.cable.rogers.com) joined #retro
22:21:00 --- quit: jas2o ("Chatzilla 0.9.74 [XULRunner 1.8.0.1/2006012608]")
23:30:08 --- join: Cheery (n=Cheery@a81-197-20-242.elisa-laajakaista.fi) joined #retro
23:59:59 --- log: ended retro/06.07.29