23/05/2016
Machine firmware
Whilst there is a continued
need to maintain legacy video machines, there is an ever larger
problem looming.
Although the documentation
for example, from Sony is considered very good, there is a missing
element to the whole system – namely the source code that was used
to create the software that runs on the processor chips that make
these machines viable. Why the importance ? Because eventually the
program code will cease to work due to component ageing.
I assume the software has
been kept closely guarded to prevent copy-cats from redeveloping
their systems as the manuals go into great depth in how these
machines are assembled and designed.
Short of reverse engineering
the op-codes in the Eeproms or even burnt in micro-controllers
(thankfully not utilised until in much later tech), it is next to
impossible to replicate the precise actions of these machines; and
the overall point of restoring these machines is that it is
considerably less effort than starting from scratch. Time to bring
out that old disassembler that apparently is the only version
available that can do the job, and that will only run on some old PC
hardware that is 30 years old?
Obviously much of this
software will still be in some copyrighted state, and will probably
never see the light-of-day even if was available, possibly lost,
with no magnetic media records, or perhaps the odd surviving paper
record. How interesting would it be if an “open-source” architecture
could be imagined, that could be revised and improved upon, e.g.
adding new interfaces, taking advantages of new processors and the
like. An original design that was built using the Z80 uP and
supporting discrete logic gates could now-a-days be replaced with
one simple FPGA running a “soft” uP for good measure. It may be
beneficial to be able to look at some of the original algorithms
employed, even though much of the processing was done in the
“analogue” domain, a sort of real-time computing done via discrete
electronics with op-amps that were essential for creating reliable
filters and mathematical functions.
So, if Sony and similar
manufacturers documented their products so well, where are the
software program listings, and why were they never published?