> On Aug 5, 2023, at 2:25 AM, Nigel Williams via cctalk <[email protected]>
> wrote:
>
> On Sat, Aug 5, 2023 at 10:16 AM Bill Gunshannon via cctalk <
> [email protected]> wrote:
>
>> Which version of RSTS is MicroRSTS V2.1 equivalent to?
>>
>
> From here:
> https://groups.google.com/g/net.micro/c/_HXPyIyrSwo/m/MuWDLNE8P48J
>
> "Micro-RSTS comes with RSTS/E V8.0". This means that while buying a
> Micro/PDP-11 will get you a general OS license, you are expected to
> get the actual media through other means, like having a bigger RSTS
> system. The idea here is to allow OEM's to market the Micro/PDP-11 -
> and give them REAL control. The end user can't buy Micro-RSTS on RX50's.
> (Although one can take his RSTS/E tapes to a Dec office and change
> media for a slight fee.)
>
> Previously I was told Micro-RSTS comes pre-genned. I am now told that
> the OEM has to make a sil on HIS machine and transfer it to the
> Micro-PDP/11.
> At present the only compatible media is RL02, with RC25's coming soon.
> (RX02's are available too, but make life difficult as they're NFS.)
>
> Micro-RSTS is a genuine RSTS V8.0, minus a few cusps Dec decided
> weren't so common. There is a execute-only version of Basic+, and
> yyy]}2o}ur swapfiles are most likely very small compared to those on a
> full 63 job system. There is nothing stopping you from implementing
> full RSTS with Basic+, BP2, etc. I am told the spooling package
> IS now only one job... I'll wait to see that.
I wonder who came up with that story about Micro-RSTS, because it is in no way
correct.
RSTS never was sold on a license-only basis that I know of, it always came as a
binary kit, with several choices for media. What those choices were shifted
over time, but generally they were whatever would allow the number of media to
be just a handful. For example, the usual RSTS-11 V4A kit was three DECtapes.
(There was also an RK05 plus a large stack of paper tapes kit, what a pain).
Later there were RL02 or RK06 kits, and of course mag tape (800 or 1600 BPI).
When the MicroPDP-11 appeared with its RX50 drives that was problematic because
those devices are so tiny that normal RSTS kit procedures of the time would not
work. So Micro-RSTS was a packaging (kitting) exercise, with some small bits
of new machinery added to the installation process to make it possible to split
a RSTS kit into 400 kB pieces. One example is the INIT "COPY" operation which
transfers the initial RSTS bits from the kit to the output device. In
Micro-RSTS those don't all fit on one disk. So, if the copy operation
encounters a xyz.EOV file, it prints the contents of that file and waits for
another disk, with label xyz, to be mounted, then continues. A disk with no EOV
file on it is the last disk for the COPY operation (which means that, after
it's copied, the output disk is booted).
Micro-RSTS was a DEC supplied kit. I have never heard of OEM-supplied kits;
RSTS wasn't really an OEM product. And the kitting procedures were DEC
internal tools, hard to use and not at all documented. The notion that OEMs
would be expected to make their own kits makes no sense at all.
I've spent some time with the RSTS build scripts trying to run an actual
standard kit-building procedure (for V10.1). I haven't succeeded yet; it is a
very unfriendly process that depends on a whole lot of magic. And
unfortunately my wife, who actually did some of that stuff in the 1980s,
doesn't remember any of it.
I'm pretty sure there were several Micro-RSTS releases, the version number 2.1
that was quoted suggests as much. I once took a later one (9.3 or 9.6 based, I
don't remember) and constructed three replacement disks for it to allow it to
be installed on a Pro. That was just an experiment for my early Pro work.
Actually, part of my reason for tinkering with the kit procedures is that I
want to create a 10.1 based Micro-RSTS so I can do a Pro version for that one,
but even getting the basics is very hard.
paul
