|
|
|
Date: Mon, 9 Apr 2001 14:02:28 -0500 From: Bob Brown Subject: Re: HP 2000 TSB status update Awsome! Here's hoping you can get 'er all up and running! I'm quite envious...The first computer that I ever used was an hp2000 system (owned by the local high school district...I was 10 years old...our local public library had a public access terminal that could be used to dial into the hp). A while back I was a computer operator here at the college, and we also had an HP2000 system (that I got to backup etc).....loved keying in the paper tape loader (in octal) from the panel and reading the paper (plastic) tape to get the system bootstrapped. Someday, if I ever have the chance and the room, I'd love to play with some old hp equipment....sigh........ If you get your system online, and decide to make it accessable via the internet, a few old hp2000 friends of mine and I would love to be able to telnet in and try running some old programs that I saved from the good old days! Good Luck and Have Fun!!!!! - -Bob
>At long last, the 7900A drive is working reliably - and passes all I thought my response might be of interest to others, so here's the scoop... The oldest systems in the "1000" series are the M series. They consisted of the 21mx 2112a, 21mx 2108a, and the 21mx 2105a. Later HP introduced the E series with the new 21mxe 2113a and 21mxe 2109a boxes. The E series did not replace the M series though. Then there was a major revision to the power supplies used in the cpu which was indicated by the B version (21mx 2112a became 21mxe 2112b, 21mxa 2108a became 21mx 2108b, 21mxe 2113a went to 21mxe 2113b, and 21mxe 2109a went to 21mxe 2109b). The 21mx 2105a stayed 21mx 2105a though. Finally there was the introduction of the F series with the 1000F 2117f and the 1000F 2111f. When the F series was introduced, all the systems became known as 1000's model such and such, with different designations. So finally there was the HP1000 F series 2117F, the HP1000 F series 2111F, the 21mxe 2113b became the 1000E 2113e, the 21mxe 2109b became the 1000E 2109e, the 21mxe 2112b became 1000M 2112m, and the 21mx 2105a finally changed to 1000M 2105a. More to the point, most of the upgrades along the way in a given branch were related to the power supply, the addition of more instructions to the instruction set (most related to better memory management), and faster memory boards. The HP 2000 systems did not designate a particular cpu. HP2000 or access, 2000a, 2000b, 2000c, 2000c', 2000e, 2000F (option 200/205 or 210/215), and the third party (NON-HP) 2000G designated a particular combination of peripherals. The hp2000 family generally used the pre-M series 2114 or 2116 cpu's. Later the 2100A and 2100S cpu's were used. Finally, even though HP still used the 2100 cpu's, many customers switched over to the 21mx 2108 and 21mx 2112 from the M series because they were slightly faster and used the new semiconductor memory. Few did setup HP2000 on the E series because they were faster enough to introduce time dependent problems in the TSB code. Generally the HP2000 designation meant the system was running TSB (Timeshare BASIC) which is my particular target of collecting since it was the first computer system I ever learned. If the same hardware was running DOS/RTE, etc. is was called an HP 1000. Hope this helps! Jay West ==============
Date sent: Wed, 31 Oct 2001 11:11:28 -0500
The HP 2000 is a very old machine, the last of which rolled off the
assembly line in 1978. George Mason University (the State University in
Northern Virginia) had (it is long gone) the next-to-last HP 2000
produced. It ran a version of the operating system with date code
1812. The operating system on this system was "HP 2000 ACCESS BASIC". The
system GMU had (which was fairly typical of the HP 2000's of the day)
consisted of the following hardware:
- -- one 21MX-E processor with 64K bytes RAM, used as the "System Processor"
The operating system supported 32 users. As the "HP 2000 ACCESS BASIC"
name implies, the system supported a single language, BASIC, but it was a
good one, and the old BASIC/V on the HP 3000 appears to be an expansion of
the HP 2000 ACCESS BASIC. There were interfaces RJE, and many HP 2000's
supported users who created/edited batch jobs on the HP 2000 which were
then submitted through the RJE interface to an IBM, CDC, Univac, or other
mainframe (as GMU did).
There was no print spooler or spooler for the card reader, so people wrote
their own, in BASIC, some supporting GE Terminets or DECwriters as remote
spooled printers.
The accounting structure was based upon account names consisting of four
characters - a letter followed by a three digit number. The "A000" account
was the equivalent of the HP 3000's "PUB.SYS". It was the system library
account, and had special privileges. It was here that you placed the
"HELLO" program that every user ran when they logged in, whether they knew
it or not.
Group library accounts on the HP 2000 were those accounts where the three
digits in the account name were "000". For example, all users in the
accounts D301 through D399 would be allowed special access to
programs/files in the D300 account.
The Z999 account was used by HP for special purposes. It has been so long
now that I don't remember what the unique capabilities of Z999 were.
The HP 2000 also had the capability of running other operating systems in
stand-alone, single-user mode. One of these was Fortran, but I never used
it so I can't comment on it.
In my basement I still have dozens of paper tapes of HP 2000 BASIC
programs. In 1999 and 2000 I converted some of those to run on the Classic
(16 bit) HP 3000 in BASIC/V. The most difficult part of converting HP 2000
BASIC programs to run on the HP 3000 is that the disc files on the HP 2000
were based on 512 byte "blocks", which means that any HP 2000 application
"smart" enough to know the algorithm used to calculate how many free bytes
there are remaining in a block (there are overhead bytes for each string,
etc.) has to be painfully rewritten to run on the HP 3000.
That's about all I have time for. I hope that gives you some feel for the
HP 2000. There were many of us who loved that little system. As to
whether there are any still running, I don't know, but I doubt it.
John
==============
From: Jay West (jlwest@TSEINC.COM)
Jay West here....
Yes, the HP2000 systems definitely predate the HP3000's. I don't have the
luxury of access to my library here at work, but I know that HP2000 TSB was
around in the late sixties... at the very least, 1969, but from foggy memory
I recall it being a bit earlier than that even.
Terminology Misnomers... The computers you are referring to were called the
"HP 2000 Computer System", with emphasis on the SYSTEM part. That
designation did not refer to the cpu, or the operating system, it referred
to a complete setup. The HP2000 computer system used standard off the shelf
components, that also could run standalone environments (Algol, Fortran,
etc.) and other operating systems (RTE, DOS, MTOS, and another one or two
that escape me at the moment). The major commonality was the operating
system, which was called TSB or TimeShared BASIC. The systems were, as I
said, off the shelf cpus and peripherals that HP already had. However, they
did some tweaking and customizing to the hardware that was used for TSB,
most notably, the dual processor systems had special firmware in the IO
processor for handling the async terminals.
The TSB operating system came in different varieties: 2000A, 2000B, 2000C,
2000C' (aka 2000C high speed), 2000E, 2000F, 2000F' (aka. 2000F high speed),
and 2000/Access. I have heard unconfirmed rumors that when HP discontinued
the HP2000 stuff, some third party company produced a lookalike called
2000G. No idea how reliable that rumor is though. Some of the 2000 variants
listed above required only one cpu, most required two. Off the top of my
head, I know the 2000E was a single cpu version. I think the rest were all
multiple cpus, but perhaps 2000A was a single cpu I think. I can dig that up
for sure tonight if anyone is interested.
Hardware - Some variants ran on the early 2116 cpus (and possibly
2114/2115). Later the 2100A & 2100S cpus were used, and finally the 21MX
M/E/F line. The processors in dual cpu configurations were linked via 4
12566 cards (or 12655, don't have my notes here), two in each system. The
main cpu held the disk controller, the tape controller, and the system
console. The FP option was required. It also held the main portion of the
operating system and actually ran the user programs. The IOP (I/O processor)
had the async mux cards in it, and along with the custom firmware to add
some high speed mux instructions, to handle all terminal I/O. You could also
add additional devices to the IOP in a non-sharable fashion - ie. paper tape
readers, punches, card readers, etc. These devices would be available for
users to attach to and use exclusively. 2100 based systems required a paper
tape reader in order to load the mag tap bootstrap into the main processor
from the paper tape reader that was connected to the IOP (across the
processor interconnect kit) via a special crosslink loader (not the standard
crosslink loader, which worked the opposite direction). 21MX systems COULD
do it this way, but there was a special TSB loader rom for the IOP that made
the paper tape reader no longer required (the 2100's had no loader rom
capability of course).
Interesting to note - some of the TSB variants had particular hardware
requirements. Going from memory, I seem to call that the early flavors
(2000A, 2000B...) required "disk file" devices, or Fixed Head disks (low
seek times huh). I don't believe the later variants would even support that
hardware. Later ones used 13037 controllers for example, which the earlier
variants certainly didn't support. There's a whole lot of "gotcha's" with
regards to hardware requirements for the different variants.
Yes, I do have a 2000/Access system running, it is the centerpiece of my
collection. Right now I am letting anyone who wants to dial into it via
direct analog call, but in the not too distant future I hope to put a unix
front end on it so anyone can telnet to it from the internet. If anyone is
interested, drop me a line at west@tseinc.com
Lastly - the base operating system does not include the A000 library. I'm
going to start keying in all the programs from the David Ahl's Computer
Games (Startrek, Wumpus, Hanoi, etc.), so there should be some nostalgic
stuff there soon. But - if anyone comes across any SLE or HIB tapes from the
2000, or even paper tapes - that contain user library programs I would
greatly appreciate getting a copy. Once I get as many library programs as I
can from people, I will gladly cut a full tape of the total library for
anyone that wants.
Oh - and I'm always on the prowl for more hardware to keep these systems up
and alive for nostalgic and historical reasons... If anyone knows where
there's any 2100 cpus (2100, NOT 21MX, I have plenty of those) or related
cards, 7900A disc drives, etc. I'd appreciate it if you'd drop me a line.
Regards,
Jay West
|