[Info-vax] What are the earliest DEC operating systems you worked with ?
Phillip Helbig undress to reply
helbig at asclothestro.multivax.de
Thu Mar 4 12:19:04 EST 2021
Simon Clubley laid this down on his screen :
> What are the earliest DEC operating systems you worked with ?
Late-comer. I started with VMS 5.5mumble at the Hamburg Observatory
when working on my master's thesis in (astro)physics there. There was a
cluster of two VAXstation 3100/76 and a VAXstation 2000. There was also
another VAX which was not part of the cluster and was dedicated to the
astrometry group; they had a 1-GB disk which was huge at the time (early
1990s). The two 76 had 32 and 24, or maybe 24 and 16 MB of memory.
Several people used to do image processing on them. I used them mainly
for Fortran and LaTeX, but also for email and running a webserver (which
is why my main webserver still runs on port 8000 today!). There was one
colour and one b/w graphics terminal attached, but most people connected
from terminals via a terminal server. Not DEC terminals, but similar,
GraphOn, Tektronix, Falco, etc. Some could do sixel graphics and so on.
While this was long before there was a PC, much less internet, in most
homes, I was still somewhat unusual as a physics student with neither a
PC nor much interest in one. (This was when standard PCs had 1--4 MB of
RAM; there was one at the Observatory with 8 which was huge at the
time.) Apart from some Basic on an HP calculator in the late 1970s or
early 1980s, I had no computing experience (but a general interest in
them, along with other scientific stuff). Even when choosing my thesis
topic I chose one where it was at least not clear whether I would have
to do any programming. As it turned out, I did, and was pleasantly
surprised that I liked it. My de-factor advisor urged me to learn
Fortran and LaTeX before staring so that I wouldn't waste time should I
need them later (I was already pretty sure that I would write the thesis
in LaTeX). I was still living with a former girlfriend who was studying
surveying and hence knew Fortran; she arranged for me to be able to use
some computers at her school, which had both Fortran and LaTeX. Some
sort of HP Apollo running some sort of unix. At the Observatory, they
had some new RS/6000 (most bang for the buck at the time for
workstation-type machines), but most people, especially those I was
working with, were on the VAXes. I immediately felt a closer kinship
than to unix, and had a complete Gray Wall just ouside my office.
When I started working at Jodrell Bank in 1977, they were just moving
away from VMS for personal use (though it was still doing pulsar-data
processing and telescope control, and still was years later, perhaps
even today), so I took out a loan and bought a new 255/233 with 64 MB
RAM, a 1-GB and a 2-GB disk, and a nice monitor. I'm still using the
monitor today. 64 MB wouldn't be enough RAM today, the 21064A (EV45) a
bit slow, but in any case it died due probably to a failed capacitor on
the motherboard. Around this time, one could pick up VAXstations and
similar for free or for a token payments, so I soon built a three-node
cluster by adding two VAXen.
Someone lent me an InfoServer in return for letting him have an account.
It had a 4-GB disk in it (huge at the time), partitioned into 4 1-GB
disks. He was using only part of at most two of those. When I needed
more disk space, I---get a load of this---set up a volume set consisting
of two of the 1-GB partitions! I gradually accumulated more VAXen and
disks and tape drives and storage boxes. I had no HBVS and made backups
of everything at least daily. A couple of years later I moved to the
Kapteyn Institute in Groning and took most of the stuff with me, but
kept a VAXstation at Jodrell Bank for when I came back to visit.
At the end of 2000, I took up a VMS job at the stock exchange in
Frankfurt. They were Alpha then, the VAXen having gone at the end of
the 1990s (though there were a few still around), and moved to Itanium a
few years later. Lots of Rdb. I set up my own stuff at home, in the
early days with an internet connection via ISDN. I soon set up HBVS for
all disks, and use double-height SEAGATE 9-GB disks for a long time. I
switched off the VAXen in 2011, having accumulated enough ALPHAs
(including spares), and not long after moved completely to SBB disks in
Top-Gun Blue BA356 boxes. Never had Itanium at home. A big project for
this year (I am now in voluntary early retirement from the stock
exchange, which still uses VMS but is moving more and more to linux,
cloud, ansible, docker, openshift, and all kinds of newfangled stuff) is
to get my hard- and software up to speed to make the transition to x86
as soon as possible, which I plan to do via a mixed-architecture
cluster.
I'm now running 2 PWS and an XP1000 in the main cluster and a 5305/1200
some DS10s as satellites (only one satellite booted at a time). I
probably have enough spares to last me until I die should I stick with
Alpha, but software and perhaps licenses will soon become too old so I'm
looking forward to VMS on x86. I'll probably wait until I can get it to
run on bare metal, because life is short. According to the Python
script posted here recently, it should run on a wide range of hardware,
even though perhaps not supported. It might be slow, but still faster
than the Alphas I have.
More information about the Info-vax
mailing list