fryers at gmail.com
Wed Jul 13 07:50:53 CDT 2005
On 7/13/05, Jules Richardson <julesrichardsonuk at yahoo.co.uk> wrote:
> On Wed, 2005-07-13 at 12:37 +0100, Simon Fryer wrote:
> > Most of the Apollo range happily talk to a serial console so a
> > screen/keyboard isn't really necessary. However in my not so humble
> > opinion, I am yet to see anything that is anywhere near as nice as the
> > GUI. .
> Agreed! I haven't tested all our earlier Apollos yet anyway, so it's
> possible that one might be beyond repair and so become a spare machine -
> freeing up a keyboard (I've not tried an early-style Domain keyboard on
> a later 4xx machine yet though; electrically they're compatible but I
> don't know if the protocol's the same)
I seem to get the impression that the protocol is the same but I don't
have any reference or experience to confirm this.
> > From memory, they also boot happily across Ethernet. Although, 12Mb/s
> > Token ring is far nicer then 10Mb/s CSMA/CD.
> I've got a feeling that the earlier 3000-era machines don't have
> Ethernet as standard anyway (although a card could be added, but I have
> no idea what boards the OS would recognise).
The 3K, 4K and 5K didn't have Ethernet as standard. Ethernet seemed to
become a more common option in the later machines.
There was only one board that the Apollos recognised - an early 16 bit
3 Com card. Finding the card is only part of the problem. Apollo had a
custom boot ROM on the card to enable network booting.
> Using the ATR would seem
> like the way to go to make a little domain, with one of the later 4xx
> machines that also has an Ethernet interface being used to get to the
> outside world.
This was quite a common setup.
> Memory's perhaps going to be a problem though - I think a couple of our
> earlier machines are running rather short on RAM (and of course they're
> proprietary modules) - 10.4's the only OS release we have, which might
> not run on a memory-starved machine...
I think there was a minor release after 10.4. There is also a set of
patches for the November 1997 feature.
For the reference of the November 97 feature. DomainOS calculates the
time from a 1/4 second count, with 0 at 1 Jan 1980. The count is cast
as an unsigned integer. The Pascal implementation was incorrect and
treated unsigned integers as signed integers. Novermber 1997 (from
memory) is the time the MSB is set. There are a couple of small parts
of the OS that rely on libraries written in Pascal - which I am lead
to believe have some interesting results.
"Well, an engineer is not concerned with the truth; that is left to
philosophers and theologians: the prime concern of an engineer is
the utility of the final product."
Lectures on the Electrical Properties of Materials, L.Solymar, D.Walsh
More information about the cctech