On 2013-02-07 16:30, Paul_Koning at
Dell.com wrote:
On Feb 7, 2013, at 10:10 AM, Bob Armstrong wrote:
That would mean at least RSX-11M V3 (or M+ V2),
I found an SPD (no software kits, sadly!) for DECnet-11M Phase III and
RSX-11M v3.2, but it's dated 1984. That's later then I would have expected
- I could have sworn that Phase IV was out by then. Actually by the fall of
84 I was working at DEC and I'm sure we had Phase IV on VMS by then. Do you
know when Phase III first shipped?
That 11M V3.2 would have any SPD from 1984 seems more and more strange. 3.2 was introduced
in 1979, I think. Unless I remember wrong RSX-11M V4.0 was introduced in 1982.
The different operating systems got their DECnet upgrades at different times. For
example, RSTS got Phase III
around 1981 (that was my first job in RSTS development). Phase IV took a very long time
because of some benighted
notion around DEC management that Phase IV was a "big system" thing and
therefore could not be considered for
lowly PDP11s.
Well, RSX had it. I'm pretty sure about that one.
As I recall, RT lagged even more, it stuck with Phase II longer than most. Or maybe
I'm confused with another
unloved OS, like one of the 36 bit ones. Then again, I remember DECnet/10 doing Phase IV
support rather early.
You should ask Peter if you want more on DECnet/10, but as far as I remember, TOPS-10 and
-20 DECnet support was rather later coming. They had this horrible solution where the
machine actually thought it was still talking Phase III even when it was Phase IV. Lots of
trickery in the FE.
Phase IV development started not long after Ethernet appeared, though early on there was
something (on paper only)
that looked much more like what ended up being Phase V. The "long header
format" in Ethernet packets for Phase IV
is a vestige of that. The actual Phase IV originally was called IIIe (extended) and was
concocted by Paul Beck
and myself in an attempt to make something that wasn't so hard. Looking up node
addresses to get MAC addresses was
considered too hard by DECnet/VMS developers at that time.
That is definitely more than I know, except that yes, I understood that Ethernet was a
major reason for Phase IV.
Johnny