Bob Armstrong wrote:
Ethan Dicks (ethan.dicks at
usap.gov) wrote:
On Sat, Mar 01, 2008 at 08:35:14PM -0500, Charles
H Dickman wrote:
...
Software development for a 20 year old OS is not easy.
And in the case of the PDP-11, several 20-year-old OSes, or in the case
of the VAX, a few old OSes, of many versions.
All true, although it might not be as bad as it sounds. There are a lot
of people on this list, and some of them used old DEC computers and
operating systems. Some of them might already know how to write a driver
for one of these OSes - get a group of them together and you'll have all the
old OSes covered. And it is fairly well documented how to write device
drivers for at least RT, RSX and VMS. And the first thing you should do is
hack up simh to include emulation of your proposed new "RQATA" interface
card; then you could start writing drivers before the hardware was even
done, and simh is a great tool for debugging the driver software. If you
actually had a group of motivated (i.e. they were getting paid for it :-)
people together it would not be that hard.
Jerome Fine replies:
I have been writing new RT-11 device drivers and making bug fixes and
enhancements to
old, existing RT-11 device drivers for more than 20 years. If a new
device driver is needed
for an RT-11 (Qbus to CF) device, I would be glad to write a driver for
RT-11 without any
pay. If anyone else wants to get together for RSX, RSTS and VMS, I
would enjoy the
challenge. However, I have only Qbus hardware if any testing is needed.
As Allison has mentioned elsewhere, RT-11 has a 16 bit limitation for
the device size (as
did V3.3 of DOS), so the solution was to use partitions with 65536
blocks. Any physical
device in RT-11 that is larger than 65536 blocks is then split into
partitions of exactly
65536 blocks each. In practice, DEC wrote only the MSCP device driver
to be capable
of handling devices with more than 65536 blocks. In RT-11, it was the
DU(X).SYS
device driver which initially (starting with V05.03 of RT-11) was able
to handle disk drives
with up to 8 GB of capacity, although with only 8 active RT-11
partitions could easily
see only 256 MBytes at a time. With V05.05 of RT-11, 64 active
partitions could
easily see up to 2 GB at a time. As has been noted elsewhere, MSCP is a
complex
protocol and a Qbus to anything controller which handles the MSCP
interface is not
trivial.
My suggestion is to take a page from John Wilson's Ersatz-11 and use the
HD: interface
which is MUCH easier to implement in real hardware / firmware for a host
adapter on the
Qbus or Unibus for a PDP-11. While the flexibility in an MSCP protocol
is missing,
the speed of drives these days probably makes that unnecessary. If
anyone produces
a controller for the Qbus, I will further enhance the HD(X).SYS device
driver to use
a translation table (in a manner similar to the DU.SYS device driver in
RT-11) which
can handle drives up to 2 TB. I already have a partially complete
HD(X).SYS device
driver which is able to read and write 64 contiguous partitions (i.e. up
to 2 GB) within a
file or device that is up to 2 TB, although in practice, the largest
file under W98SE
which I am using for testing is 2 GB in any case.
The mention of SIMH leads me to ask that someone write the emulation
code to allow
SIMH to use the HD: device with the SIMH emulator. If anyone does write
the code
for SIMH for an HD: device, I will help debug the code from the RT-11
perspective.
Once SIMH has an HD: device for RT-11, it will be almost trivial to
modify one of
the other device drivers for RSX and RSTS (probably the RL02) to handle
the HD:
device.
OTOH, the SBC6120 was sort of my experiment in open
source old computer
development. I always hoped people would expand and add to it, both in
hardware and software, but for the most part that never really happened.
Most people wanted to buy a turnkey system that they could plug in, turn on,
and start using it - people seem to have jobs, families and real lives and
just don't have the time to do a lot of development.
I guess that being retired now - I am the exception. I have been
making bug fixes and extensive enhancements to RT-11 for more than
25 years including my own set of Y2K bug fixes - which eventually
I want to expand to Y9K. I think the problem with any 3rd party
changes was that most companies felt that DEC was the only supplier
which could be counted upon. By 1994, DEC had sold the PDP-11
software to Mentec - who supported RT-11 even less than DEC with
good reason, of course, since there was really very little call
for any bug fixes except for Y2K and even less call for enhancements.
At present, I can't seem to find anyone else who is making changes
and fixing bugs in RT-11. If you have made any recent changes to
RT-11, I would be VERY interested in hearing about them!!!!!!!!!!
Sincerely yours,
Jerome Fine
--
If you attempted to send a reply and the original e-mail
address has been discontinued due a high volume of junk
e-mail, then the semi-permanent e-mail address can be
obtained by replacing the four characters preceding the
'at' with the four digits of the current year.
