[SGVLUG] April Presentation Online

bobjaffray at juno.com bobjaffray at juno.com
Thu Apr 15 13:57:29 PDT 2010


BJ: Thanks for the update on the language
used for the project (C).

BJ: The reason why FORTH is ideal for
tight places in 16bit memory is that 
indirect threaded interpreted code
doesn't need opcodes for all 
intermediate to high level routines
(words) beyond a minimal set as low
as 18 rather small machine code
primatives. So full applications
can go into the 2k memory space
(wasn't that it?).

BJ: But I am not sure how to make
it work for a user of the device
with only the 4 buttons or even
an external keyboard. Another 2K
(total of 4K) is needed for 
a minimal interactive FORTH
system. Basic would require
more than that.

BJ: But I remember using a 
. prompt "monitor" in 256
bytes on a chip of nonvolitle
memory with a TTY type video 
to enter octal 8080 processor
machine code into memory. So
that is doable. The native 
memory for that later verion 
of the MITS Altair was 2K.
The next step was to get a
board for the S-100 bus to
talk to and save code on 
cassette tape. With this
system and the interface with
USB exteral memory, that is
passe. The problem is the
limitations of internal memory
for exploration out in the
boondocks.

Bob 
 



____________________________________________________________
Free Credit Score
A bad credit score is below 598. Click here to see yours for $0. Checking won't affect your score. By Experian&#174
http://thirdpartyoffers.juno.com/TGL3141/4bc77e03156c42dff0bst05vuc


More information about the SGVLUG mailing list