Origin of 'Straight 8' name

Jim Carpenter jim at deitygraveyard.com
Fri Dec 21 14:51:45 CST 2018


On 12/21/18 1:10 PM, Al Kossow via cctalk wrote:
>
> On 12/21/18 10:03 AM, Al Kossow via cctalk wrote:
>
>> "Straight-8" seems to be a fairly modern name coming from collectors
>>
>> I never heard it called that before then.
> Anyone feel like doing a alt.sys.pdp-8 search for it by date?

The PDP8-LOVERS mailing list predates alt.sys.pdp8 by a couple years. I 
just checked the archives and the earliest usage of 'straight-8' is from 
Charles Lasner in an e-mail introducing himself to the still new mailing 
list on August 10th, 1990. I've pasted his complete message at the 
bottom. The 'straight -8' is mentioned in the second sentence of his 
second paragraph.

A quick check shows that it was common for cjl to use the term 'straight-8'.

> I don't feel like going down the rathole of trying to find a way to
> search Usenet by date right now.
>
>

I miss DejaNews. Damn I hate Google.

Jim


Received: from ELI.CS.YALE.EDU by BUGS.SYSTEMSY.CS.YALE.EDU; Fri, 10 Aug 
90 05:37:28 EDT
Received: from life.ai.mit.edu ([128.52.32.80]) by ELI.CS.YALE.EDU; Fri, 
10 Aug 90 05:36:07 EDT
Received: from watsun.cc.columbia.edu by life.ai.mit.edu (4.1/AI-4.10) 
id AA00254; Fri, 10 Aug 90 05:12:04 EDT
Received: by watsun.cc.columbia.edu (5.59/FCB)
         id AA20618; Fri, 10 Aug 90 05:12:20 EDT
Date: Fri, 10 Aug 90 5:12:19 EDT
From: Charles Lasner <lasner at watsun.cc.columbia.edu>
To: PDP8-LOVERS at ai.mit.edu
Subject: Belated Mail Reply
Message-Id: <CMM.0.88.650279539.lasner at watsun.cc.columbia.edu>

Date: Fri, 10 Aug 90 5:12:19 EDT
From: Charles Lasner <lasner at watsun.cc.columbia.edu>
To: PDP8-LOVERS at ai.mit.edu
Subject: Belated Mail Reply

 >Received: from AI.AI.MIT.EDU (CHAOS 3130) by MC.LCS.MIT.EDU 18 Feb 89 
10:32:35 EST
 >Date: Sat, 18 Feb 89 10:32:31 EST
 >From: "Robert E. Seastrom" <RS at AI.AI.MIT.EDU>
 >To: pdp8-lovers at MC.LCS.MIT.EDU
 >Message-ID: <540513.890218.RS at AI.AI.MIT.EDU>
 >
 >
 >Well, folks, it's finally here.  The PDP8-LOVERS mailing list is now
 >reality!  Messages for the list go to PDP8-LOVERS at MC.LCS.MIT.EDU;
 >requests to be added to or deleted from the list go to
 >PDP8-LOVERS-REQUEST at MC.LCS.MIT.EDU (case is _not_ critical here).
 >
 >Perhaps we ought to all introduce ourselves to each other...
 >
 >          This one's for you <7001>
 >
 >                      -Rob

From:   cjl

     Let me introduce myself.  I am Charles J. Lasner.  I only use the
J to get my initials, which I usually go by, thus cjl.

     I am a PDP-8 programmer.  I don't know how many of us there are
left, but I started in 1968 with a straight -8 table-top machine at
Brooklyn PolyTechnic Institute.  The school probably has another name
by now, due to academia's answer to the business world's phenomena of
acquisitions and mergers.  The machine in question is quite
legendary.  The work done on it is responsible in LARGE part for why
all of us are here reading this, since this is the "original" PDP-8
used by the legendary Richard Lary and company.  If it can be said
that the PDP-8 created the phrase "mini-computer", then it is THIS
PDP-8 that made the "mini-computer" into something other than a
paper-tape machine!

     This PDP-8 was originally configured by the academics in charge
as a programmer's disaster: 4K, EAE, a model 33 teletype, AF01A A-D
converter with 16-channel multiplexor,  AA01A D-A converter with
(wow!) THREE channels.  A Bud blue rack cabinet housed the A-D and
D-A with lots of empty space.  All empty slots had those wonderful
super-thick zinctone panels, and the fronts were all in place; they
were held in place with those pressed-in heavy shiny metal threaded
bosses that most of you can't figure out the purpose of on your
wire-wrap racks.  As far as I know, these were the only style of
cabinet that the bosses were supplied for.  The plates used heavy
nickel-plated knurled knobs with a screw-driver slot you could turn
with a dime.

     Soon after all of this arrived, and got nowhere due to the
enormous waste of time paper-tape can be (especially at 110 baud and
unreliable at that!), a DEC salesman suggested a high-speed
reader/punch be added.  Fortunately for all of us, THIS NEVER
HAPPENED, for if it had, no further work of external significence
would have been done.  (The EE department would have been very happy
to just develop their diddly A-D and D-A experiment programs.)

     Due to the efforts of Richard Lary, Jack Burness, Hank Maurer,
Lenny Elekman, and Joseph R. Fischetti (to name a few legends I
knew), the EE department was convinced to spend MORE money on some
new-fangled beast the salesman had vaguely heard of; he was fairly
certain it was called a MicroTape.  This was, of course, an early
name for DECtape.  So the EE department shelled out another $8k and
got itself another Bud blue rack cabinet complete with sides, another
power controller, 11 buss cables, and a TC01 and one (yes one!) TU55
DECtape drive.  The academics thought that the drive was custom made
for the PDP-8, because the numeral "8" always appeared on the drive
select.  (No need to change drives when you have only one of them!) I
later found out why the "8" was there (not "0"):

     This is a throwback to the earlier Microtape drives used mostly
on the PDP-6 which are really DECtape drives, but with an entirly
different interface to the PDP-8, which really belongs to the PDP-5
era.  I believe that the PDP-5 doesn't support 3-cycle data-break,
because this is a single-cycle interface.  This limits programming
possibilities. (Another subject entirely!)

     I believe the only programming for this earlier hardware was an
original version of the DECtape Library System which I had contact
with on the TC01/TU55 (an incompatible later version, but obviously
less obscure).  The interesting point is that while the TC01 has a
three-bit drive select register, where all cases are viable, this
earlier hardware had a FOUR-bit register, so drives had to be
selected as 1-8.  Clearing the register deselects all drives, whereas
when talking to the "Solid-State DECtape" drive, as the TU55 was
originally known, you are selecting drive zero.  Apparently there
were hardware adaptations to theoretically connect the TU55 as a
microtape drive (using RELAY LOGIC level converters instead of the
jumper cards usually in the drives), so they thought in terms of the
original microtape nomenclature of 1-8.  Years later (I think first
done for the benefit of the PDP-12 LINCtape) they had an ECO to glue
a 0 over the 8.

     Soon after the TC01/TU56 cabinet arrived, the 33 teletype started
failing (which is normal!) from "normal" use (reading paper-tapes!).
They had their first "run-in" with DEC Field Service, whose members
sort of knew what to do with the -8, but they KILLED teletypes!
Eventually, they purchased a PT08, the only hardware on this machine
to use ICs (RTL chips only, TTL came later!).  With it came an
additional teletype, a model 35 (without paper-tape).  The 35 became
the 03/04-connected one, and the 33 was "banished" to the PT08. (And
over the years it "walked" out of the PDP-8 room to be "borrowed" by
various groups, only to be "retrieved" periodically.)

     Fortunately, a private teletype maintenance company contacted the
school and eventually won a teletype-only maintenance contract, over
DEC's objections.  Since these teletypes were so over-used, it became
necessary to rotate them between the PT08 and the console to keep
them working.  We even created software to SLOW DOWN the teletype,
since it was noticed that an out-of-adjustment 35 will work FINE if
run at 89% of maximum speed (that's 89% of 110 baud!) when it starts
jumbling characters at 100% speed.

     This machine was setup on a large desk, with the teletypes
nearby, and a Bud cabinet at either end.  We didn't have the nice
Formica table option.  So much software was toggled into this machine
that I personally replaced the slide switches which wore out!  I also
repaired the CPU several years later, after DEC Field Service swore
they couldn't get it to work after 9 months of trying.

     The problem was that there are isolation diodes in the wires that
run from various registers to the light panel, which has lamp drivers
and soldered in lamps on it.  Periodically we would replace the
occasional blown transistor driver and numerous blown bulbs.
(Touching the panel was a calculated risk since you killed even more
lamps by doing it!)

     The diodes are located on the card edge where all of the wires
are soldered close together, and plugged into the backplane; there
are several such cards.  The wires break off easily, and hasty
repairers would sloppily solder them back to the board lugs.
Eventually some diodes shorted, so the capacitance of the wires now
loaded the affected driver.  Also there now was the possibility of
cross-talk between the wires collector-triggering the gates!  This is
why the machine was so prone to random failure.  (It worked in
single-step mode.)

     Replacing the diodes didn't completely work, because there was
yet another problem: the diodes were photo-sensitive!  DEC attempted
to light-shield the diodes by painting them.  This paint got scraped
off of them due to the soldering, etc., so the machine was sensitive
to its own lamps being lit!  It worked better in room darkness, so we
got the idea about inadvertent photo-diodes!  Painting the diodes
solved the last remaining problem.  I almost single-handedly got the
machine back up (there were other random problems as well, but they
were easier to solve, like the tendancy for a PT08 to burn up a land
UNDER a chip where you couldn't see it!), and got the thanks of the
faculty members of the EE department, who would have scrapped the
machine if it was unrepairable!

     Although this wasn't my first computer, it was my first DEC
computer.  I have stayed with the PDP-8's to this day.  I was then
and now a programmer, but we all participated in keeping the machine
up.  I can trace my extensive DEC hardware background to this
beginning (as well as DEC software!).

cjl (Charles Lasner)
     (lasner at watsun.cc.columbia.edu or lasner at cunixf.cc.columbia.edu)



More information about the cctalk mailing list