On 25 Dec 97 07:47am, Mike Bilow wrote to George Fliger:
 MB> George Fliger wrote in a message to Mike Bilow:
 MB> My PC networking days go back to truly ancient kludges such as
 MB> the Corvus Omninet -- remember those?
 GF> My God!  Somebody actually remembers those???  I thought I
 GF> was the only one who dealt with that misfortune. 
 MB> Yes, and not fondly.  I first had to deal with the Corvus boxes
 MB> in 1980 or so, when they were introduced as shared hard drives
 MB> for TRS-80 networks.  Corvus supplied two big white boxes, one
 MB> containing a 10 MB hard drive and power supply, and the other
 MB> containing a 16-port multiplexer for the TRS-80 expansion
 MB> interface bus.  It was a horrible mess, but it worked for
 MB> years. The Corvus Ominet was actually a later variation for the
 MB> IBM PC.
I just converted one of my clients from a Corvus Omninet Netware setup
to Netware 3.12 a few years back.  That was the first I'd seen of
Corvus-anything in almost 10 years.  Did I mention I was surprised to
see it?
 MB> The fundamental idea behind the Corvus architecture, that of
 MB> electrically extending machine buses with differential drivers
 MB> and ribbon cables, has mercifully been totally forgotten.  As
 MB> soon as it became economical to use fairly intelligent
 MB> microcontrollers on cards, the inherent advantages of real
 MB> networking systems such as token ring and Ethernet became
 MB> obvious.
Ah yes, the days of spooling excess cable around a wooden dowel and
tying it all down as neatly as possible within the box with rubberbands
or nylon ties.  It was interesting!
 MB> I was a registered developer (or whatever it was called) for
 MB> the TRS-80, which Tandy was then actively promoting as an ideal
 MB> business solution.  That was not a very good idea, since it
 MB> took them a while to make such conveniences as printer ports
 MB> and serial ports standard equipment, and they never did
 MB> implement a competent serial port.  In fact, we finally
 MB> designed our own serial port boards for the TRS-80, and I wrote
 MB> what would pass for drivers. We used the 8250, then a $50 chip
 MB> that was hard to find in anything other than the Heath/Zenith
 MB> line of Z-80 machines, but our experience with it came in very
 MB> handy when IBM later chose it for their PC, which also drove
 MB> the price down to about $15.
I never got that far into it.  I was mainly installation and support.  I
could never get my mindset fixed on learning to program.  Batchfile
manipulation and enhancement is about as far as I went.  As for the
8250's, I remember how popular they got when the end-user, particularly
corporate types, started using modems -- back then, 75, 150 and if you
were really lucky and had the bucks to spend, 300 baud acoustic couplers
or REAL modems.
Heath/Zenith Z80's.  I remember them.  Most of what I saw were running
CPM for an OS.
 GF> You probably pre-date me networking wise (no shot at your age,
 GF> really!).
 MB> I started dealing with microcomputers around 1977.  In fact, I
 MB> was at the show where the TRS-80 was introduced, and it was the
 MB> talk of the place because it cost only $600 and you did not
 MB> have to solder it yourself.  It was also the first real coup by
 MB> Microsoft, since their BASIC was in the system ROM.  Until the
 MB> TRS-80, Microsoft was just one among several players in the
 MB> BASIC market, and they were probably not the best technically.
We started about the same time although I didn't get into networking
until the mid 80's.  I was in the military at the time and the only
computer systems we had at the local Battalion/Company level were mainly
word processing machines (IBM DisplayWriters and CompuCorp).  You could
find a few IBM Selectrics with MAG-Card readers but nothing more fancy
than that.  You had to be assigned to the post AG office in order to get
your hands on anything big.  We were still dealing with punch cards and
mark-sense readers back then.  I was able to get my hands wet on some of
the equipment since I had friends that worked in the computer
department.  They were not allowed to do any real programming, just
input via punch-cards provided by the programmers (sent to them via
dispatch from Ft. Ben Harrison, Indiana).  The rest was pretty much
daily logistical card input.  If the equipment broke, IBM was on-call 24
hours a day to be onsite to get things back up and running.
My first TRS-80 wound up being a CoCo with 16KB of RAM.  Big machine for
color in those days.  With the first 160k floppy drive and external
plug-in controller, the entire package ran about $1200.  I later bought
a 64KB upgrade and did the trace and wiring changes myself to install
the upgrade.  What an awesome machine.  It was unbelievable what you
could do with 64KB back then. This was one of the first machines that
would allow you to change the memory mapping and copy its ROM code to
RAM, turn off the ROMs and run a straight 64KB memory bus.  The 6809E
was a nice processor, although slow.  .9Mhz in normal mode but, if you had
a replacement 68B09E you could poke a few memory addresses and double
your clock speed to 1.8Mhz.  At that rate, you were still only pushing
the chip at about a 75% rating.  Although an 8-bit processor, it had the
capability of allowing you to combine it's A & B accumulators
(registers) into a D register, thus giving the ability to process 16-bit
instructions.  I seriously considered learning how to program after
being impressed with this little processor but at the time, assembler
was all that was available and I, quite frankly, couldn't get into it.
I'm not that big of a number cruncher.  Later, of course, higher
programming languages and utilities came to pass but the CoCo and its
amazing little processor faded off into the sunrise like so many other
good ideas.  I think Tandy should have continued the line but upgraded
the process all the way to the Motorola 68000.  That would have made for
an awesome machine -- way before its time.
 MB> One serious contender for technical supremacy was Benton Harbor
 MB> BASIC, written by Gordon Letwin to run under his HDOS operating
 MB> system on the Heath/Zenith machines, but Heath would not
 MB> license it to competitors.  BH BASIC even had the beginning of
 MB> dynamic memory management, a real innovation in a 64 KB address
 MB> space.  Gordon Letwin later became frustrated with Heath and
 MB> joined Microsoft as employee number 10 or so, eventually
 MB> inventing OS/2.
I was never exposed to BH BASIC and this is the first time I'd heard of
it.  Most of the systems I was exposed to back then were Tandy and IBM.
Since computers were not my field of expertise back then I was rather
limited in the resouces I could access.
 GF> I can remember the first Netware system I
 GF> installed back in 1986.  IBM PC XT's and AT's (35
 GF> workstations total) with 2.5 Megs expanded memory, Token
 GF> Ring NICs and QuarterDeck's DesqView for a front-end.  The
 GF> server was an AT (6Mhz) with two Seagate 4096 80Meg MFM
 GF> drives.  The whole bundle (including the programmer time)
 GF> cost the company around $250,000 to imple- ment.  And to
 GF> think, all they had for the entire company was a bit less
 GF> than 160Megs total server space.  They never thought they'd
 GF> run out. :)
 MB> NetWare evolved very strangely.  I think someone here mentioned
 MB> a couple of days ago the original system which depended upon a
 MB> dedicated 68000 CPU box. By 1986, though, things had greatly
I never heard or saw this configuration.  I'm sure I came onboard with
Netware long after that scenario.
 MB> matured and NetWare was widely used.  The Seagate ST-4096 was
 MB> among the larger drives you could get, but had a tendency to
 MB> develop mispositioning error, so we stayed away from it.  I do
 MB> still have three or four of them sitting around, though.  We
Yes, we did have a bit of a time getting the server initially online
because of drive problems.  We had to have one or two drives overnighted
to us because the ones we had developed some problems.  We had a
redundant server (they didn't want to pay the extra cash to attempt SFT
III) and both setups required one drive to be replaced.
 MB> moved to Maxtor, which in those days made some of the finest
 MB> and most reliable stuff around, although I certainly would not
 MB> say the same about them today.  For many years, my personal
 MB> desktop ran a Maxtor 180 MB drive that we pulled from a file
 MB> server because it "sounded funny," but we couldn't send it back
 MB> for that -- and it did run flawlessly for about five years,
 MB> 24-hours a day, in my desktop machine until it was retired in
 MB> favor of a larger drive.
I still have my original Maxtor ESDI drives from my home server packed
away.  They never died and still have my original server configuration on
them.  I never could get myself to part with them. :)
 MB> The largest network we did in that era had three NetWare
 MB> servers and about 300 workstations, and it was also token ring
 MB> since Ethernet was not really a proven technology for more than
 MB> a few workstations, nor was coax practical for many physical
 MB> office layouts.  We did use Ethernet for the interserver
 MB> backbone, which even then we had the good sense to keep
 MB> separate.  Each server had about 400 MB using Seagate ST-4383E
 MB> ESDI drives, usually running on Western Digital WD1007 ESDI
 MB> controllers.  It took about 4 MB RAM to cache these drives, and
 MB> both the drive capacity and the physical memory were considered
 MB> enormous for the time. -- Mike
Oh yes, the WD 1007 controller.  Probably the best controller Western
Digital ever made.  I still get an occasional call to have one repaired
for one of my County government clients who still have older 2.15 and
3.11 servers still running in obscure departments.  As they die, they
get replaced but they'll keep them running till then.
"Four Hundred MB and 4MB cache per drive!  Why, that should last you till
the turn of the century."  Wasn't that the claim back then? :)
It's been a pleasure.
George
... A pedestrian hit me and went under my car.
--- Via Silver Xpress V4.4P [Reg]
---------------