1) I am writing in response to your plea for comments on Mr.
re:standardised extensions to Nas-Sys.
I agree entirely with hie motives in providing standard calls for
add-ons but I would like to suggest a few alterations to his
a) Graphic boards
Rather than separate SCAL calls for line, circle, etc.,
would it not be better to use one SCAL GRAF and load up the
arguments and/or registers?
b) Floating Point Handling
Whilst floating point routines have been published in many
places, they are still not widely available to everyone.
I believe extended calls to Nas-Sys should include floating
point arithmetic and functions as well as perhaps extended
precision integers. In doing this it will turn Sys into s
‘software bus’ onto which can be tacked all manner of language
and arithmetic applications.
c) Nas-Sys (COM) Naughties
Many moons ago, in the days when Z**l’s weren’t around,
there was an article in INMC (issue 5 or 6 I think) entitled
The purpose of it was to ward off people
making direct calls to Nas-Sys as they had done with the
Nas-Bugs. May I now suggest another naughty – Direct Keyboard
Many games programs and certain languages, ie BASIC and BLS
Pascal, scan the keyboard directly. (BLS Pascal even makes a
direct call to Nas-Sys). This practice is all very well and good
on a Nascom but it is perfectly possible to run Sys on a Gemini
by fudging the video RAM (see Richard Beal’s program in 80-Bus
So, in conclusion, I would repeat my wish that Sys should act as
a ‘software bus’ running the ‘hardware bus’ we all know.
2) The proposal by Mr. O’Farrell in the Nascom Newsletter Vol.3,
No.1 was a very good idea. A lot of people have already added
new routines to their Nas-Sys. I know that Polydos uses its own
set of new routine numbers and I think that Nasdos does the
same. It would be nice if we could get some standard out of it.
But I think that there is another problem. Where in memory
should these extra routines be placed? Somewhere in the top of
memory most people would say, but I think that this is only half
a solution. Why not place the whole monitor on another page?
This is what I have done. Nas-Sys is placed on
board (but I think that it should be a simple matter to modify
the CPU-board such that it could be paged in and out). The
monitor pages itself in and out when Nas-Sys routines are