From: christianlott1 on
On Mar 2, 10:32 pm, BruceMcF <agil...(a)netscape.net> wrote:
> On Mar 2, 9:09 pm, christianlott1 <christianlo...(a)yahoo.com> wrote:
>
> > On Mar 2, 4:03 pm, BruceMcF <agil...(a)netscape.net> wrote:
>
> > > SO the  fastest C64 Forth was the subroutine
> > > threaded Blazin' Forth.
>
> > ok. Thanks for this :)
>
> I got several of these from the Taygeta Forth archives, but they seem
> to be down, and Volksforth is harder to find since that is the German
> forth interest group that did that one.
>
> A 64forth.zip and a Volks*.zip can be found at:
>
> http://groups.google.com/group/niclos-for-forth-94/files

hforthx.cpm would not download as it has 0 bytes.

Thanks :) Much appreciated.
From: christianlott1 on
On Mar 2, 10:32 pm, BruceMcF <agil...(a)netscape.net> wrote:

> I got several of these from the Taygeta Forth archives,

ftp://ftp.taygeta.com/pub/Forth/compilers/native/misc/commodore64/

From: BruceMcF on
On Mar 3, 9:20 am, christianlott1 <christianlo...(a)yahoo.com> wrote:
> On Mar 2, 10:32 pm, BruceMcF <agil...(a)netscape.net> wrote:

> > I got several of these from the Taygeta Forth archives,

> ftp://ftp.taygeta.com/pub/Forth/compilers/native/misc/commodore64/

Yes, it was not coming through in html yesterday, which were the links
I was using. ("used to work five or more years ago" applies to a
number of things from the 80's).

From: BruceMcF on
On Mar 3, 9:00 am, christianlott1 <christianlo...(a)yahoo.com> wrote:
> hforthx.cpm would not download as it has 0 bytes.

Good thing, too, it was buggy.

Thx, have to remember to erase that file (extending hForth for CP/M
went into the backburner ... when I get back into it, I'll probably
use Camel for CP/M instead, and give it BLOCK files ... I need more
experience with programming CP/M files in their own style before
implementing the Forth-94 FILE wordset).
From: BruceMcF on
On Mar 2, 5:03 pm, BruceMcF <agil...(a)netscape.net> wrote:
> SO the  fastest C64 Forth was the subroutine
> threaded Blazin' Forth.

Blazin' Forth is also at taygeta, thanks for the ftp address.

Only problem with a subroutine threaded forth is it compiled bigger
programs. If compact execution is wanted, a bit threaded interpreter
may be a good combination of speed and code size.

It would be built with primitives that end with "JMP NEXT", and "IP"
and "W" as two zero-page vectors (if Basic vectors are reset to run
Basic in a Box at $8000 and the Forth in $0800 to $7FFF, that would be
W=$FB and IP=$FD).

NEXT:
CLC
LDA IP
ADC #2
STA IP
BCC NEXT1
INC IP+1
NEXT1:
LDY #1
LDA (IP),Y
BEQ EXIT
BMI ENTER
STA W+1
DEY
LDA (IP),Y
STA W
JMP (W)
ENTER:
AND #$7F
STA W+1
DEY
LDA (IP),Y