About 1401 Design Memos
Date: Wed, Aug 26, 2015 11:15 pm
To: Carl Claunch < Carl.Claunch@gartner.com >, Ken Shirriff < ken.shirriff@gmail.com >
Cc: ...
Carl, Ken,
As I was scanning Fran's 5th and 6th design memos this evening, I noticed this excerpt in the 4th memo, SPACE OPERATOR'S MANUAL,
(now how's that for projecting at the dawn of the Space age?! ;-) It looks like "d" may have been an abbreviation for "single-digit", as in < "d" indicates a single-digit character" >.
Here's a snapshot of thee explanation from the manual's page 9:
On 8/23/2015 8:43 PM, George R Ahearn wrote:
> Yes, memos were the way to communicate technical issues within the 1401 development team.
In general, memos were frequent throughout IBM. That is until the attorneys handling the several antitrust law suits were required to produce them to opposing counsel as evidence.
> I remember Tom Barr of Cravath Swain & Moore letting everyone know how he felt about that. Tom was lead outside counsel in the Dept of Justice v. IBM case.
> George
> On Sun, Aug 23, 2015 at 2:51 PM, EARL BLOOM
> Amen, he "was" one of the "very" best! >
> Best regards, Earl Bloom
> On Sunday, August 23, 2015 2:44 PM, John Pokoski
> Robert,
> Thank you for forwarding this. It is fascinating. I have three thoughts.
> I have no idea why the modifier is called "d".
> The IBM 1401 project being my first job, I was introduced to the ideas of memos immediately. I found it was a great way to keep everyone in the loop and stay informed. Many memos were "to file" and provided hard copy evidence, in one spot, of any actions or decisions.
On a subsequent IBM assignment, I didn't find such a rigorous procedure. I don't know to this day if other companies used memos as consistently. I suspect this is one reason why the 1401 group seemed to have a minimum of misunderstandings. The memos that I scanned and forwarded to Ed which he put on the 1401 site were largely those of Paul Farbanish, who saved them in a binder and gave them to me for background information and historical records when he left for grad school in the early sixties. They and the ones I accumulated are still in my attic.
Probably they are also in an IBM storage facility too. It must be remembered that producing such a memo was not as simple as it would be today with electronic word processing and networking. Usually, the memo was written in longhand by the engineer (or perhaps dictated by a manager), then typed by a secretary, then proof read and corrected by the engineer, then retyped, etc. ( I noticed several corrections using "white-out" in Fran's memos.)
The ensuing interactions between secretary and engineer required patience on both sides, particularly if the engineer had poor penmanship or if the secretary was a poor typist. (I recall letting a minor typo slide rather than go back to the secretary for a third typing.) It amazes me that the group was so conscientious about such memos. I was because everyone else was. I guess it was in the culture.
> I always thought that Fran Underwood was a type of genius and these memos help verify that. He was a Renaissance man in his everyday life from his art to his technical savvy, to his social life. He and I became somewhat close in the time between when we re-met in 2009 in California (due to your instigation) and his death. We corresponded by email often, on everything from puzzles and conundrums on the internet to 1401 memories, to advice to me to get on with my life after I mentioned I missed my late wife.
But his 1957-58 memos show in pretty much detail how the 1401 evolved. I am amazed how he did this in one year. I assume it was by himself and wonder if he ever bounced ideas off of others. The memos show the big picture of business needs and how they relate to the system he proposes.
Yet he shows great detail in his logic descriptions, printer design, programming, etc. Few people could do this. (I still chuckle at his hatred of plug boards. I recall him getting disgusted when a peripheral proposed to be attached to the 1401 contained a plug board. "GET IT OUT!") Note that the "blueprint" drawings are drafted by him and are impeccable. Similarly, the punch card layout and computer program are his own work down to the last detail, even his beautiful signature and initials. He was one of a kind.
> John Pokoski
> To: Carl.Claunch@gartner.com
> From: robgarn@us.ibm.com
> Date: Fri, 21 Aug 2015 15:23:21 -0700
> CC: ...
> Subject: [1401_founders] Re: Fran's SPACE design memos (was RE: [1401_software] How did the "d-character" get its name?
> Carl,
>
> > First concept for the architecture http://ibm-1401.info/SPACE-1stMemo-Aug1957--.pdf
> > Second concept for the architecture, closer to what it became but still having substantial differences
> > By 1958, the design has shaped up and the usage of d-character is consistent
>
> It's admirable that you dug into Fran's early design documents - they're an interesting read!
>
> - Robert
>
> p.s. I had serendipitously discovered Fran's early design memos while sifting through random boxes
in the "Endicott section" of the IBM archives storage shelves in 2011 (out of 100's).
I suspect the boxes were from IP/patent offices.
>
> There's still one more of Fran's SPACE design memos that I need to scan and upload.
> (Plus a dozen more period memos relevant to the 1401...)
>
> We get to benefit from the earlier practice that engineers and managers typically
authored (commonly via dictation) and disseminated their partially baked deliberations
in well thought out memos that have survived for fifty years as paper artifacts.
> From: "Claunch,Carl"
> To: Ken Shirriff
> Date: 08/21/2015 12:35 PM
> Subject: RE: [1401_software] How did the "d-character" get its name?
> Sent by: 1401_software-bounces@computerhistory.org
> Hi Ken > > I have never found a direct etymology but through some reading of the evolution of the 1401 architecture, you can see the introduction of this term as the digit (d) field for branch instructions, but also there is an argument for simple sequential naming, at a time when the mutating 1401 architecture had A, B and C address fields. > > It was in the second memo from Fran Underwood about the SPACE architecture that I first encountered the d-char field, on page 3 where he discusses branching instructions. > > Reading the sequence of memos is fascinating � as the design shifted from fixed instruction length to variable, as ideas like chaining begin to arise, and they discard the idea of four-address instructions or two-character binary only addresses. > > First concept for the architecture http://ibm-1401.info/SPACE-1stMemo-Aug1957--.pdf > Second concept for the architecture, closer to what it became but still having substantial differences http://ibm-1401.info/SPACE-2ndMemo-Oct1957l--.pdf > By 1958, the design has shaped up and the usage of d-character is consistent http://ibm-1401.info/SPACE-OpsManual-June1958lq--.pdf > > Carl > > From: On Behalf Of Ken Shirriff > Sent: Tuesday, August 18, 2015 5:52 PM > To: 1401 Software Team > Subject: [1401_software] How did the "d-character" get its name? > > This is sort of a random question, but an opcode can optionally end with a d-character (or d modifier) that changes the action of the opcode. What does the "d" stand for? > > Ken
return to main page