Re: Software Patent Concerns => New Black Markets?

["Robert I. Eachus" <rieachus@attbi.com>]

Warren W. Gay VE3WWG wrote:

> "They decreed that the stored-program idea rightfully
> belonged in the public domain."
> 
> And that is how it was decided, and we enjoy its  benefits today.

I knew many of the people involved in this.  (Pres Eckart was best man 
at my parent's wedding, and my father at Pres's first wedding.)  But 
more than that I knew a number of people who actually worked on Eniac in 
addition to the principle players.

Pres's father was actually the one who pushed him to patent everything 
that wasn't nailed down.  AFAIK, Pres never met Atsenoff until much 
later, and knew nothing about the ABC computer.  (In any case, the ABC 
computer was very different in design and implementation.)  Oh, and as 
far as I am concerned, the half-adder, which was the key circut in most 
early vacuum-tube computers was invented by Pres.  I don't know who 
originally came up with the flip-flop.

Now back to the stored program concept.  Originally, the ENIAC was 
designed with a bank of rotary switches which could be used to program 
it.  There were twenty instruction registers in the original design. 
There was a similar bank of registers implemented using 10 tubes to 
create a ring counter.  (These were 6SN7 tubes, dual triodes, and were 
connected in pairs so that if one burned out its partner would carry on. 
  So five tubes allowed for ten different on states, and each tube had a 
indentical tube in series.  The way that the tubes usually failed was 
that the filament burned out in the inrush when the machine was turned 
on.  Of course the real solution was to keep it always on.)

As soon as the machine was built, plans were made for a second bank of 
instruction switches, to be used for storing constants.  Until it was 
ready, the instruction switch bank could also be used for constants. 
But often programs needed more than 20 instructions and constants, so 
the programmers for the machine--those who actually set the 
switches--got pretty proficient at putting constants in registers. 
Basically, they had a program to read constants from cards into the 
registers.

Then someone accidently submitted a program that would need the new bank 
of instruction switches.  He was flabbergasted to get results back. 
When he went to find out how it had been run--was the new instruction 
bank finished?--he was told, "No problem, we just did like we always do, 
costants after the twentieth go into the registers."  Of course, those 
first three constants were actually program instructions...

It then became instantly obvious to the mathematicians who generated the 
programs that they could use as many instructions as they wanted to by 
using some of the registers for instructions and changing them.  To me 
that is the key insight of stored programs.  And no one who was there 
was willing to take particular credit or give it to anyone else.

Now understand that at the time this happened, the ENIAC was still at 
the Moore School, and the staff using it was joint between the 
University of Pennsylvania and Aberdeen Proving Ground, the actual US 
Army customer.  So yes, that discussion and decision that the stored 
program concept was unpatentable did occur.  But it occurred in the 
context that the key features and advantages of a stored program system 
had been instantly obvious to those present.  And they were literally 
all the people skilled in the art of programming computers at the time.

Johnny von then wrote his infamous paper, recently declassified.  The 
reason for the omitted references became obvious when that happened. 
The Johnny von paper was prepared with an annex that described some of 
the details of the implementation and credited Pres and John Mauchley. 
But the paper was distributed at a time when it was felt that the fact 
the machine actually existed at present should be very closely held. 
But still most people who saw the (classified) paper never saw the 
annex.  Why? Because the annex discussed how to use ENIAC to replace the 
bombes used to decrypt Enigma messages.  The Johnny Von paper was only 
Top Secret.  That annex was literally classified as Ultra Secret, and 
only available to those in the know on Ultra.  (And again, AFAIK, my 
father did get to see the full paper, but Pres Eckert never did. 
However, I was not able to discuss that with my father until much, much 
later when I could show him an unclassified copy of the annex, so we are 
talking memories thirty years later.)
--
                                            Robert I. Eachus

"As far as I'm concerned, war always means failure." -- Jacques Chirac, 
President of France
"As far as France is concerned, you're right." -- Rush Limbaugh