LETTERS TO THE EDITOR.
Right of Reply.
May I claim the accused’s right of reply to
Mr Perkins review of HS-1N
(80-BUS Vol. 1, Issue 4)? But first, congratulations on a very thorough piece of
detective work on the hardware and firmware!
We designed HS-1N for our own use in mid 1980, because we had Philips
DCRs and Nascom 2s, and no prospect of ‘official’ disk drives. Following a
demonstration of the system at the Scottish Amateur Computer Society in Dec ’80,
we were approached by Microspares and we negotiated a license for them to
manufacture the product. We accept now that further work should have been done on
the design at that stage, but we simply passed the prototype designs to
Microspares and let them deal with the PCB, manufacture, manuals etc.
Design bugs:– We had already implemented NASIO on our N2s, and did not
need DBDR, and hence did not implement these on-board in our prototype. They are
implemented on current boards. The conflict between the SIO and Page Mode memory
arose because we received assurance from Nascom in mid 1980 that no existing or
planned product would use ports F8-FF. Current boards are supplied using ports
Firmware bugs:– The bugs in ‘Initialise’ are real – no excuse is
available, except that if the system is properly used, and no tape is removed
without ‘eXiting’, the bugs remain hidden! The criticism of the error handling
routines (which use the HS-1N warm start after issuing the error message) applies
equally to Nascom ROM BASIC when used by machine code calls, but the revised
operating system (HS-1N+) avoids this by using a RAM vector, which can be patched
to divert error routine calls back to a user’s program.
I accept the criticisms of the absolute calls to Nas-Sys, and the use of
address 0 to discard unwanted data. It takes several months of use to be
reasonably sure that there are no major bugs in a system, and in this case
pragmatism defeated the desire to rewrite the program and and purge its
ZEAP files were not directly supported simply because we do not use ZEAP,
but the faster but apparently no longer obtainable Z2, which we have patched to
allow direct access to HS-1N+. Relocation of files as they are loaded is
supported by HS-1N+, and is an essential feature of HS-Kit, the toolkit for Xtal
BASIC and HS-1N+ which Mr Perkins referred to in his review. This allows access
to files by name from Xtal BASIC, but the additional code for this is just too
much to be incorporated into a 2K operating system ROM. In our initial
specification we deliberately chose access by number, not by name.
Dr M.D. Hendry, HS Design Ltd.,
Linley, East Rd., Cupar, Fife. KY15 4HR.
Pascal facilities and PIO ports
Please allow me to comment on some items that cropped up in the
Nov/Dec ’82 issue.
raised a couple of queries with regard to Pascal. He claimed
that Pascal does not have the same simple string-handling facilities that BASIC