Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.forth > #26452 > unrolled thread

8-bit Forth

Back to article view | Back to comp.lang.forth

Contents

  8-bit Forth Paul Rubin <no.email@nospam.invalid> - 2013-10-12 23:41 -0700
    Re: 8-bit Forth "Elizabeth D. Rather" <erather@forth.com> - 2013-10-12 21:02 -1000
      Re: 8-bit Forth albert@spenarnc.xs4all.nl (Albert van der Horst) - 2013-10-13 12:46 +0000
        Re: 8-bit Forth Paul Rubin <no.email@nospam.invalid> - 2013-10-13 06:22 -0700
          Re: 8-bit Forth albert@spenarnc.xs4all.nl (Albert van der Horst) - 2013-10-15 02:45 +0000
    Re: 8-bit Forth albert@spenarnc.xs4all.nl (Albert van der Horst) - 2013-10-13 12:36 +0000
      Re: 8-bit Forth Paul Rubin <no.email@nospam.invalid> - 2013-10-13 06:12 -0700

#26452 — 8-bit Forth

I spent a few minutes looking at ByteForth, which appears to be a pretty
complete 8-bit Forth:

http://www.forthfreak.net/index.cgi?ByteForth

I can't really tell too much specific about it since the documentation
is written in Dutch.  But, I get the impression that it really is 8-bit
in the sense of having 8 bit data cells.  And I'd think that means
addresses don't fit in the cells.

I'm basically wondering: 1) is that really waht it's doing?  and
2) Did this method ever catch on in the 6502/8051/Z80 era?  Or did
the Forths on those systems all use 16-bit cells?

I continue to have a (masochistic?) interest in the AVR/Atmega because
there's so much cute hardware for it (femtoduino.com for example, or
http://digistump.com/products/1 ).  They seem like very plausible Forth
targets.

[toc] | [next] | [standalone]

#26454

On 10/12/13 8:41 PM, Paul Rubin wrote:
> I spent a few minutes looking at ByteForth, which appears to be a pretty
> complete 8-bit Forth:
>
> http://www.forthfreak.net/index.cgi?ByteForth
>
> I can't really tell too much specific about it since the documentation
> is written in Dutch.  But, I get the impression that it really is 8-bit
> in the sense of having 8 bit data cells.  And I'd think that means
> addresses don't fit in the cells.
>
> I'm basically wondering: 1) is that really waht it's doing?  and
> 2) Did this method ever catch on in the 6502/8051/Z80 era?  Or did
> the Forths on those systems all use 16-bit cells?
>
> I continue to have a (masochistic?) interest in the AVR/Atmega because
> there's so much cute hardware for it (femtoduino.com for example, or
> http://digistump.com/products/1 ).  They seem like very plausible Forth
> targets.

FORTH, Inc. experimented briefly with <16-bit Forths in the 70's. We had 
a 12-bit PDP-8 system that was used in a couple of applications, and 
briefly flirted with an 8-bit version when the first microprocessors 
came out. But it was really not usable, and I think everyone else pretty 
much came to the same conclusion.

The AVR/ATmega product line is just fine, and SwiftX supports several 
boards off-the-shelf, including the Arduino Uno kit, Kanda STK200, STK 
300, and STK500. It's a 16-bit implementation.

Cheers,
Elizabeth

-- 
==================================================
Elizabeth D. Rather   (US & Canada)   800-55-FORTH
FORTH Inc.                         +1 310.999.6784
5959 West Century Blvd. Suite 700
Los Angeles, CA 90045
http://www.forth.com

"Forth-based products and Services for real-time
applications since 1973."
==================================================

[toc] | [prev] | [next] | [standalone]

#26470

In article <44GdnfK__tTs2MfPnZ2dnUVZ_hadnZ2d@supernews.com>,
Elizabeth D. Rather <erather@forth.com> wrote:
>On 10/12/13 8:41 PM, Paul Rubin wrote:
>> I spent a few minutes looking at ByteForth, which appears to be a pretty
>> complete 8-bit Forth:
>>
>> http://www.forthfreak.net/index.cgi?ByteForth
>>
>> I can't really tell too much specific about it since the documentation
>> is written in Dutch.  But, I get the impression that it really is 8-bit
>> in the sense of having 8 bit data cells.  And I'd think that means
>> addresses don't fit in the cells.
>>
>> I'm basically wondering: 1) is that really waht it's doing?  and
>> 2) Did this method ever catch on in the 6502/8051/Z80 era?  Or did
>> the Forths on those systems all use 16-bit cells?
>>
>> I continue to have a (masochistic?) interest in the AVR/Atmega because
>> there's so much cute hardware for it (femtoduino.com for example, or
>> http://digistump.com/products/1 ).  They seem like very plausible Forth
>> targets.
>
>FORTH, Inc. experimented briefly with <16-bit Forths in the 70's. We had
>a 12-bit PDP-8 system that was used in a couple of applications, and
>briefly flirted with an 8-bit version when the first microprocessors
>came out. But it was really not usable, and I think everyone else pretty
>much came to the same conclusion.

That conclusion is apparently wrong. Byteforth proves that a cross
compilation Forth system with 8 bit ints is viable and can hold its
own easily against c-development systems or assembler development
systems in use. Of course it is not a true ISO Forth system and the
CPU targets may be outside of Forth inc.'s domain of interest.

>The AVR/ATmega product line is just fine, and SwiftX supports several
>boards off-the-shelf, including the Arduino Uno kit, Kanda STK200, STK
>300, and STK500. It's a 16-bit implementation.

Maybe the decision that to use 8-bit ints is what makes BYTEforth viable.

There is a 150 page application handbook (in Dutch) ranging from
Led's through LCD to motor's. Willem Ouwerkerk (and crue) are
making a version based on the MPS430 board.

>
>Cheers,
>Elizabeth
>
>--
>==================================================
>Elizabeth D. Rather   (US & Canada)   800-55-FORTH
>FORTH Inc.                         +1 310.999.6784
>5959 West Century Blvd. Suite 700
>Los Angeles, CA 90045
>http://www.forth.com
>
>"Forth-based products and Services for real-time
>applications since 1973."
>==================================================
-- 
Albert van der Horst, UTRECHT,THE NETHERLANDS
Economic growth -- being exponential -- ultimately falters.
albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst

[toc] | [prev] | [next] | [standalone]

#26473

albert@spenarnc.xs4all.nl (Albert van der Horst) writes:
> Maybe the decision that to use 8-bit ints is what makes BYTEforth viable.

I think a 16-bit cross-compiled Forth for the ATTiny85 would be viable.
It might not be viable for the smallest of the ATTiny series but I'm not
sure BYTEforth claims to work on those either.  E.g. the ATTiny10 has
something like 25 bytes of ram (not sure if that means the cpu registers
and there's really no additional ram).  But it's available in an
ultra-small 6-pin package which is useful for some things.

> Willem Ouwerkerk (and crue) are making a version based on the MPS430 board.

Bit the MSP430 is a 16-bit processor that maps very well to conventional
Forth.  An 8-bit Forth for it seems surprising.

[toc] | [prev] | [next] | [standalone]

#26537

In article <7xr4bpb90r.fsf@ruckus.brouhaha.com>,
Paul Rubin  <no.email@nospam.invalid> wrote:
>albert@spenarnc.xs4all.nl (Albert van der Horst) writes:
>> Maybe the decision that to use 8-bit ints is what makes BYTEforth viable.
>
>I think a 16-bit cross-compiled Forth for the ATTiny85 would be viable.
>It might not be viable for the smallest of the ATTiny series but I'm not
>sure BYTEforth claims to work on those either.  E.g. the ATTiny10 has
>something like 25 bytes of ram (not sure if that means the cpu registers
>and there's really no additional ram).  But it's available in an
>ultra-small 6-pin package which is useful for some things.
>
>> Willem Ouwerkerk (and crue) are making a version based on the MPS430 board.
>
>Bit the MSP430 is a 16-bit processor that maps very well to conventional
>Forth.  An 8-bit Forth for it seems surprising.

They're making a version of the 'Egel application manual, adapted to
more modern processors, notably noforth of the same authors,
not ByteForth. Sorry if I didn't make that clear.

Groetjes Albert
-- 
Albert van der Horst, UTRECHT,THE NETHERLANDS
Economic growth -- being exponential -- ultimately falters.
albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst

[toc] | [prev] | [next] | [standalone]

#26468

In article <7xppr9d65u.fsf@ruckus.brouhaha.com>,
Paul Rubin  <no.email@nospam.invalid> wrote:
>I spent a few minutes looking at ByteForth, which appears to be a pretty
>complete 8-bit Forth:
>
>http://www.forthfreak.net/index.cgi?ByteForth
>
>I can't really tell too much specific about it since the documentation
>is written in Dutch.  But, I get the impression that it really is 8-bit
>in the sense of having 8 bit data cells.  And I'd think that means
>addresses don't fit in the cells.
>
>I'm basically wondering: 1) is that really waht it's doing?  and
>2) Did this method ever catch on in the 6502/8051/Z80 era?  Or did
>the Forths on those systems all use 16-bit cells?
>
>I continue to have a (masochistic?) interest in the AVR/Atmega because
>there's so much cute hardware for it (femtoduino.com for example, or
>http://digistump.com/products/1 ).  They seem like very plausible Forth
>targets.

Byteforth has been around for a very long time. It is intended for the
very smallest chips like 4-8 pins and 1K flash.
Its largest problem is that the system only runs on CHForth, which
was for a long time the standard Forth of the Dutch Forth Chapter.
This must be run in a Dos Box or a windows 98 (!) vintage machine.
(The second largest problem apparently is the documentation, I
didn't even realize it was only in Dutch.)

It is eminently practical, and its author has done quite some
commercial applications with it.
Willem Ouwerkerk is also a co-author of noforth for the MPS430.
At least that has all its documentation in English.
Despite being superior to the well known 4e4th, it appears on the
3th page if you google for "MPS430 Forth".

Groetjes Albert
-- 
Albert van der Horst, UTRECHT,THE NETHERLANDS
Economic growth -- being exponential -- ultimately falters.
albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst

[toc] | [prev] | [next] | [standalone]

#26471

albert@spenarnc.xs4all.nl (Albert van der Horst) writes:
> Byteforth has been around for a very long time. It is intended for the
> very smallest chips like 4-8 pins and 1K flash.

Thanks!  Hmm, interesting, does that mean it's a cross compiler?  What
is the runtime model on the target?  Does it really have 8-bit data
cells and what does that do to the concept of storing addresses in
cells?

> Its largest problem is that the system only runs on CHForth...
> This must be run in a Dos Box or a windows 98 (!) vintage machine.

That might be less of a problem today than a few years ago, since
virtualization is everywhere now, so running a dos box is easier than it
was.  But yeah, being able to run under Linux would be really nice.

> (... documentation, I  didn't even realize it was only in Dutch.)

There is a little bit in English.  I can also sort of decipher some of
the Dutch by squinting at it and pretending it's garbled-up English.

> Despite being superior to the well known 4e4th, it appears on the
> 3th page if you google for "MPS430 Forth".

Noforth looked nice but I think I want to try FlashForth for the MSP430
and the Atmega.  There doesn't seem to be much for the ATTiny series
other than ByteForth.

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.forth

csiph-web