home .. forth .. colorforth mail list archive ..

[colorforth] ColorForth in colorForth...

Hi Albert + friends,

I would be delighted if you would disassemble and re-assemble the binary
image cfdos4.blk !  ( blocks 0 to 17 )
Its available from my website www.inventio.co.uk/cfdos.htm .
I think I have got the assembler files for it, but its development has been
long and complicated...

I agree with all of the recent batch of posts here, from Nick, Albert, Mark,
Chris, Ray and Philip.

Creating a group of collaberating programmers is never easy. We definitely
need the right tools for this.

My motivation to create cfdos.com was to drag colorForth source and binaries
into the WinDOS world, at least until we have full Internet connectivity and
the ability to run on "nearly all" PCs. At the very least I can name and
archive different versions...

Around October 2004 I published cfdos3.1 - an attempt to make the colorForth
names closer to ANS.
Whatever your views on ANS Forth per se, I think it is foolish to use ANS
names in colorForth with different functionality.
So I renamed them by hand-editing the binary :

Was     is now called. function
.
?dup    /dup    compiler - decides whether a dup is needed
-       invert  -1 XOR
stop    /flop   turns off the floppy disk drive motor
or      xor     bitwise XOR ( New in V3 )
--      or      bitwise OR    ( New in V3 )

This was ( as I suspected it would be, ) a big mistake.

So in cfdos4.0 I put the original binary back and re-defined the offending
words. This is less efficient at run-time, but meant that the high level
source could be ported to other systems. This allowed me to support other
binaries, such as the 800x600 pixel video versions.

What became clear was that the next step is to port colorForth to
colorForth. I have done most of the Editor - but I must add, with the
essential assistance of many others. ( Please follow the link above for
credits ). To create a coherent system, I have gathered other people's
applications and made the necessary adjustments for them to run with my
published system.

There is an inherent conflict : how to create compatibility without limiting
freedom. cfdos3.1 would have been fine if my name was Bill Gates and I had
the resources to persuade everyone to adopt my binaries, source and
standards. But I can't ( and shouldn't ) force others to follow me - this is
after all a collaborative effort, so I back-tracked.

I have no inclination to change assembler files and re-assemble.
( I have an irrational aversion to assembler ).
Maybe its because I do that sort of thing for a living ( in C mostly ),
or maybe its because I want to try to find an alternative.
We already have file based OS's...

ColorForth is magic. I will definitely keep investigating it until I find
out why my code for the Editor written in colorForth is so much better than
the same code written in conventional Forth. Part of it is that a second
attempt will always be better, but part of it is the lean environment which
keeps the code lean.

So a Mocha Latte to the first person to publish a colorForth compiler in
colorForth. This should be the next major part of the project.

BTW please keep all variables and compiled code out of blocks 0 to 17 - one
of
the problems with the current colorForths is that the binary image changes
as soon as you run it. ( I developed chk.com which calculates a checksum but
avoids the variables - I think ).
I have already got "key" coded in colorForth. The PC keyboard hardware is
very fiddly, and Chuck's assembler for this correspondingly clever, but
undocumented ( cfdos4.blk block 196 ). I haven't documented it either...
yet.

Progress should be slow and organic,

Regards

Howerd  8^)


-----Original Message-----
From: Albert van der Horst [mailto:albert@xxxxxxxxxxxxxxxxxx]
Sent: 07 April 2005 23:13
To: colorforth@xxxxxxxxxxxxxxxxxx
Subject: Re: [colorforth] New Linux 4word


On Thu, Apr 07, 2005 at 06:17:04PM +0200, Fr?d?ric DUBOIS wrote:
[snip]

Is there is a binary dump from the latest officialy colorforth?
I would love to regenerate an assembler file, like I did for
retroforth. We could proceed from there.

It is a good workout for the
computer_intelligence_assembler_disassembler.

>  Amicalement,
>   Astrobe

Groetjes Albert.

--
Albert van der Horst,Oranjestr 8,3511 RA UTRECHT,THE NETHERLANDS
Economic growth -- like all pyramid schemes -- ultimately falters.
albert@xxxxxxxxxxxxxxxxxx http://home.hccnet.nl/a.w.m.van.der.horst



---------------------------------------------------------------------
To unsubscribe, e-mail: colorforth-unsubscribe@xxxxxxxxxxxxxxxxxx
For additional commands, e-mail: colorforth-help@xxxxxxxxxxxxxxxxxx
Main web page - http://www.colorforth.com