Conference 7.286::digital

    I beleive the ENIAC was used during the war at the Aberdeen Proving
    Grounds in Maryland.  It was used to develop artillery projectal path
    information.  It would have taken thousands of man years to figure out
    all the trajectory computations if it had been done by hand.  This
    means that your 1945-1946 date would not be correct.
    
    

ENIAC was the first electronic digital computer that
was built as an electronic digital computer.

With the benefit of hindsight, we can see that Colossus
(the Bletchley code-breaking machine) was really an
early programmable digital computer, but that's
not what they were building.  (Or at least not what
they told management they were building :-)

Tommy Flowers (in charge of the team) is now
considered to be a computing pioneer.  Turing did,
I think, contribute ideas, but I'm not sure whether
he was a full time member; I'd have to look it up.
 
dave

Dave:

> (Or at least not what they told management they were building :-)

  A common M.O. in the business, isn't it! Pretty Darn Popular, that One!

                                   Atlant

    .3> A common M.O. in the business, isn't it! Pretty Darn Popular, that One!
                                                 P      D    P
    :-)
    

more like...


    .3> A common M.O. in the business, isn't it! Pretty Darn Popular, that One!
                                                 P      D    P             1
    :-)
    

re Note 3865.0 by MASS10::GERRY:

> Now i thought we had laid that myth to rest, and that the code-breaking
> machine used at Bletchly Park, England DURING the second-world war had been
> acknowledged as the first.

        Since the UK, with her allies, won the war, apparently even
        if you *win* the war you don't necessarily get to write the
        history!

        Bob

        P.S.  Wasn't it Konrad Zuse who had designed a workable
        electronic computer during the '30s, but the Nazis wouldn't
        fund its construction?

       > P.S.  Wasn't it Konrad Zuse who had designed a workable
       > electronic computer during the '30s, but the Nazis wouldn't
       > fund its construction?

       Actually, it wasn't a computer the Nazi's designed, it was a
       tape drive.  The "Death's Head" insignia was removed before it
       was marketed commercially in the '70's as the "TU45."  8^)


       bill

    RE: 7
    
    I knew there was a reason I hated working on those things.  They were
    cursed!!!   :-)
    
    
    

	I wouldn't say they were cursed, they just required a special
relationship to understand and maintain the  )(*&^%$#@ things...

	We had one customer who wrote their own device manipulation routines
for the TU45 on RSTS, and disabled error correction. This led to the scenario
(regularly) of me with a screwdriver adjusting gains to make their EBCDIC Social
Security Department transaction tapes balance correctly !! - those were 
simpler times...

	Eventually the drive caught fire when the fan seized and overheated;
when the young lady operator loaded a tape and hit the button the fan dropped
out of it's cage, down into the electronics and motors, spewing flames and
sparks out of the top of the cabinet. I believe she refused to ever touch the
drive again after that day.

	Much to the next account engineers horror, they wanted us to fix it -
which he did , of course (going the extra fifty miles in that case). Eventually
the whole system was replaced with Vaxes.

	The last we saw of it was when the system was decommissioned, (there 
was talk of a field service wide wake for that drive, to be sure and drop it
down the lift well)..but it was bought by a well meaning individual who was
setting up a bureau service - I wonder how it worked out ? 

	

>>Now i thought we had laid that myth to rest, and that the code-breaking
>>machine used at Bletchly Park, England DURING the second-world war had been
>>acknowledged as the first. This machine (whose name i have temporarily
>>mislaid) was i think designed (or at least worked on) by Alan Turing, and
>>would have been operational in 1942-43, long before ENIAC (1945-46) but was
>>considered to be so secret that news of its existance was not released until
>>the 1970's!! Hence the popular misconception!!

	the machine in question was called Enigma...it was used by the Allies
	to decode German encrypted communications....its success was largely
	due to the fact that Germany never discovered its existence and
	thus continued to use the same encoding methodology...this is also
	why it was kept secret until the '70s; follow-ons to the Enigma
	design were used to decode all sorts of encryped stuff...

	I have not seen it acknowledged as the first computer...IMHO it
	would be a stretch the think of it as a computer at all since it
	was largely electro-mechanical.....

	_kelley

    "first computer" ? Analog perhaps .....abacus
    
    "first computer" ? Digital perhaps ..... Charles Babbage - "Difference
    Engine" and "Analytical Machine". 

    I thought "enigma" was the coding scheme used by the Axis, not the
    machine that was used to break the code.. 
    
    On the original track: as with most things, what was "first" depends on
    how you define "what".. for example, I heard this morning that today is
    the anniversary of "GE starting the first scheduled Television
    broadcasts in New York" back in the mid-20's.. I suspect this is 
    partially true, but:
     - I'm pretty sure Jenkins was doing scheduled broadcasts before GE
       in the US (but not New York)
     - I'm pretty sure Baird and the BBC were doing scheduled 
       broadcasts before this data in the UK
     - What was being sent and received then bears little resemblence
       to "television" as we know it today.. (that honor probably 
       goes to the EMI/Marconi system that the BBC started using in
       the mid thirties, at least for "scheduled" broadcasts).
    
    So, depending on how you define the event, the "winner" may change.
    
    ENIAC was probably the first electronic, digital, general-purpose
    computer that did useful work outside of it's design center. (I think
    that has enough disclaimers to narrow the field..)
    
    ...tom

><<< Note 3865.12 by OTOOA::MOWBRAY "Wish I didn't know now what I didn't know then" >>>
>                            -< I've used them all >-
>
>    "first computer" ? Analog perhaps .....abacus
>    
>    "first computer" ? Digital perhaps ..... Charles Babbage - "Difference
>    Engine" and "Analytical Machine". 

An abacus is decidedly a digital device, both in operation and numerical
representation.
A better candidate for first analog arithmetic device (as opposed to 
a simulating computer) might be the Napier-style slide rule.
Neither an abacus nor a slide rule can be termed a "computer" in modern
parlance, since neither can be programmed.

- tom]

     re .13 Depends on what "first" means.
    
    ENIAC was the first computer to do general work. But, it wasn't the
    first to be built for general work. ENIAC was built to help build atom
    bombs and as such it was only a partial success as the atom bomb was
    largely built without it. ENIAC didn't start work until Nov '45 but it
    was used (and modified) extensively until the early '50s mainly solving
    ballistics problems.
    
    Re other notes
    
    Enigma was the encryption machine used by the axis forces during the
    war and is used as a general name for the encryption technique also.
    
    Colossus was the machine some clain to be the first electronic
    computer, and it was electronic. Its predecessors, Bombes, were
    electro-mechanical.
    
    Mike.
    
    PS Going back to the definition of "first", many people now refer to
    the first stored-program computer in an attempt to reduce ambiguity.
    This title goes to ENIAC which was upgraded to stored-program operation
    shortly before the Manchester University Mark 1 went live (1948). The
    MU1 was built mainly by the people working at Bletchly Park on
    Colossus.

    re. 14
    
    OK lets play it that way......
    
    1. Define Computer
    
    2. Define Programming
    
    Please provide above definitions in enough detail and clarity that they
    will not be disputed by other noters.
    
    Once we have the ultimate definitions, then we can decide that
    Babbage's machines were computers and programmable ... come to think of
    it how about Pascal's engine as a programmable analog machine ?

	I'd define computers as Turing machines.  By the way, although
	the Babbage machines could be thought of as the first computers,
	he never managed to build the second phase.  The first phase
	was built and they were used to calculate the Admirality tables
	that were used in navigation.   The second phase, which *would*
	have been a computer (albeit mechanical) was not fully completed
	for lack of government funds.  His (Charles Henry Babbage's) vision
	did spur on many advances in mechanical engineering.  I once met
	his granddaughter and have a cat named Babbage after him.

	Dave

Babbage definitely had the first computer project which
overran its schedule, exceeded its budget, and ultimately
failed to deliver on its promises.

    In the book "The Code Breakers" which is about the efforts at Bletchley
    Park, there is a photo of one of the German "Enigma" machines, the
    earlier version with three "rotors", and on the inside of the lower 
    front cover is the word "Enigma" inside of an oval. I believe it was
    also known as the Lorenz SZ 40 or SZ42. 
    
    A later version was increased to four rotors and the front plugboard
    had more available permutations.
    
    If I recall my Ancient Electronics History correctly, the "AC" at the
    end of various computer names (Like UNIVAC) stood for Analog Computer, 
    but the practice remained for some of the earlier digital
    tube-and-relay computers.

	uggg..yer all right of course; Enigma was the code breakee not
	the code breaker (whose name I don't remember as well)....none
	the less, it is this issue of analog vs digital that defines
	the basis for "first"; I prefer the latter...

	_kelley (I'll stick to software; my history s*cks)

>    If I recall my Ancient Electronics History correctly, the "AC" at the
>    end of various computer names (Like UNIVAC) stood for Analog Computer, 
>    but the practice remained for some of the earlier digital
>    tube-and-relay computers.

Not according to the official names... it was usually expanded
as "Automatic Computer" or "Automatic Calculator".  Of course,
the pioneers may have invented the now-common practice of inventing
an acronym first and then figuring out what it stood for later.

    I believe Enigma was a mechanical device.  The Allies snatched it with
    a raid on a German transport, replaced it with a typewriter that
    had been made to superficially resemble the machine, then burned the
    transport.
    
    The Germans were sufficiently convinced that the machine in question
    had been destroyed, not taken, and continued using the same encryption
    methods.  This deception had to be maintained at all costs.  Churchill
    had to decide whether to evacuate an English town scheduled for bombing
    the following day and risk German discovery of our secret, or ensure
    the secret was kept and let innocent people die.  He chose the latter
    and suffered the pain and hurt of this moral dilemma.
    
    As a side note, the Americans used Navajos in the Pacific for secrecy. 
    The Japanese weren't able to meet the daunting challenge imposed by
    learning a foreign language with no educational aids spoken by a
    relatively few people.
    
    								Tex

Thought I read somewhere that 

ENIAC = Electronic Numeric Integrator and ACcumulator

Mac

    I saw a special on PBS, called something like The Code Breakers?  Been
    awhile.  Anyway, on that show they pointed out that the first
    "computers" were people.  In particular, (and this is from memory)
    computers at one time were typically women who calculated trajectory
    tables for cannons.
    
    As for electronic computers, I used to service pinball machines when I
    was a lad.  This was back in the good old days when a pinball machine
    defined functions using relays and cams (counters, memory) stepped by 
    solenoids or motors.  I would definitely consider a relay-logic system 
    like a pinball machine to be a "computer."
    
    Also, if you think of DNA as a program that controls a living cell ...
    Gets mighty hard to really nail down when the first computer was made
    unless you narrow the definition down a lot.
    
    Steve

wasn't the "Enigma" breaker called "Ultra"?

- paul

    To further deepen the rathole there are some that claim the ABC as the
    first electronic computer. 

    >                          <<< Note 3865.25 by ARCANA::CONNELLY "Don't try this at home, kids!" >>>
    >
    >
    > wasn't the "Enigma" breaker called "Ultra"?


    I think that Ultra was the code name associated with the information
    gleaned from breaking Enigma transmissions.
    
    	Yaacov
    

>    I believe Enigma was a mechanical device.  The Allies snatched it with
>    a raid on a German transport, replaced it with a typewriter that
>    had been made to superficially resemble the machine, then burned the
>    transport.
 
      Enigma was electro-mechanical. The position and wiring of rotors in
    the machine determined the current path between the keyboard and the
    output display (letters with lights behind them, if I remember aright).
    The rotors could be shuffled around, and new ones, with different wiring,
    could be substituted; they advanced one position with each character
    encoded/decoded. One limitation (which was used in the breaking of the
    code) was that a letter could never be encoded as itself.
      The original Enigma breakthrough came shortly before the war, when
    Polish Intelligence intercepted a machine which was sent as ordinary
    freight (rather than in diplomatic baggage) to the German embassy in
    Warsaw. The principle of the machine, and indeed actual machines, were
    commercially available; it was the rotor wiring that was top secret.
    The Poles copied the wiring of the rotors, and this information was
    shared with British intelligence.
      The Germans knew well enough that machines would be captured; their
    fatal assumption was that there were so many possible combinations of
    rotor arrangement and starting positions, that decoding messages would
    take too long for the information obtained to be of any practical use.
    This was where the bombes, and later Colossus, came in; by using them
    it was possible to test millions of possible combinations in a
    reasonable time. The undecoded message would be fed in, and tried with
    each combination in turn; when recognisable German came out, you were
    getting somewhere. That is a very simplified version of course, but you
    get the basic idea. Lapses in procedure by German signallers
    considerably reduced the number of combinations which had to be tried.
       
    >The Germans were sufficiently convinced that the machine in question
    >had been destroyed, not taken, and continued using the same encryption
    >methods.  This deception had to be maintained at all costs.  Churchill
    >had to decide whether to evacuate an English town scheduled for bombing
    >the following day and risk German discovery of our secret, or ensure
    >the secret was kept and let innocent people die.  He chose the latter
    >and suffered the pain and hurt of this moral dilemma.
    
      I think in this, and the comment about the steamer being captured,
    you have confused the two wars. Code books were captured from a steamer
    (the name Magdeburg? comes to mind) in WWI (I mean 1914-18, for the Johnnie
    come latelies over there, but that's another rathole 8*)). These were used
    successfully during that war to decode German naval signals, by room 40 in
    the Admiralty. Much of the intelligence gleaned was wasted - the Battle of
    Jutland might have ended very differently had more information been passed
    to Jellicoe in time - but it was there.
    
      Further Enigmas were captured in WWII, as the Germans had of course
    known that they would be. Mechanical hardware is much more difficult to
    destroy completely, especially by burning, than code books. A particularly
    useful one was from a U-boat captured (initially by an aircraft!) off
    Iceland early in the war; the Navy machines had an extra rotor. I think
    this loss was unknown to the Germans for a long time. The U-boat,
    incidentally, was repaired and served as HMS Graph until 1944.
    
      I'm not sure which town you are referring to; could this be Coventry?
    If so, then the target was not known for sure in advance, only that it
    was probably somewhere in the Midlands. There was a serious cockup over
    the jamming of the German radio navigation beams, which resulted in the
    bombing being very accurate, but this was nothing to do with
    Enigma/Ultra. Enigma intelligence was not always timely; the decrypt might
    come out within an hour or two on a good day, sometimes it might take
    several days. There were sometimes long periods (months) when no decrypts
    could be made at all for at least a part of the traffic, due to changes in
    procedures, new rotors etc. Obviously rotor changes were not frequent, as
    all machines on a network had to be changed simultaneously, but they
    occurred. When they did, decryption was slow or stopped until the new
    rotor's wiring had been figured out.
    
      And now, back to your regular programme:
                                                                 
      Another candidate for the earliest programmable computer; the Jacquard
    loom. Programmed using punched cards, the output was patterned fabric.
    Not sure when it was invented, but it pre-dated, and possibly helped
    inspire, Babbage's Difference Engine. I believe they were in widespread
    use by the 1820's. Purely mechanical of course.
    
    Ken

     
    To deepen the "fascinating" rathole a bit further:
             
    The enigma machine stolen by the Poles was an early version of the
    coding machine but gave the Blethley guys enough to guess the rest. The
    Polish machine had, I believe, only three wheels and were relatively
    easily broken without much mechanical/electro-mechanical help. Usually,
    because the operators were too lazy to change the wheel positions
    between consecutive messages: ( message A + code A ) - ( message B +
    code A ) = Message A - Message B -- easy if you can do baudot
    arithmetic in your head :-). Oversimplified, of course.
    
    The Bombes, and later the Robinsons, were machines set up to try just
    this exercise on every consecutive pair of intercepts. This took care
    of most of the German signals intelligence including army, air force
    and diplomatic channels. The b****r was the navy who didn't trust the
    three and five wheeled enigmas and used an enhanced 8-wheel model which
    wasn't so easy. Although the machines were still called enigma, the
    codes were called the "fish" codes and named: tunny, silverfish ...
    Colossus was built (actually it now seems like "it" was a "they" -
    at least three) to handle the fish codes. The first success was just
    before D-day 1944.
    
    It is true that Bletchley warned Churchill about the bombing of
    Coventry before it happened but its very unlikely that he failed to
    evacuate the town to protect Ultra. This speculation appeared in a book
    on the subject and added spice. The reason its unlikely is that
    Churchill ordered all the spare mobile anti-aircraft batteries to the
    town - a broadcast to opposition intelligence that he knew something
    was up (no pun intended). In all probability, Coventry couldn't be
    evacuated: 250,000 people, with little transport and even less fuel in
    wartime conditions, couldn't be moved in the time available.
    
    Now before the moderators jump on us for talking about things that
    should be in the History notes conference, there is a very tenuous link
    to Digital in all this. Following the war, the US government decided
    that the future of intelligence services was largely in SigInt (Signals
    Intelligence) mainly because of the perceived success of Bletchly. This
    led to a very large budget for a good customer of ours - the NSA - and
    lots of Digital revenue. 
    
    Mike.

Here's another tie-in to DEC... a few years ago, when the
computer museum was still in Marlboro, Tommy Flowers gave
a short talk on his work on Colossue.

    Amazing what you learn reading this conference!  My entries were based
    mainly on "A Man Called Intrepid", which detailed the birth and
    development of America and Britain's intelligence services. 
    Fascinating reading, if you have a bent towards historical
    cloak-and-dagger.
    
    							Brent Eyster

> As a side note, the Americans used Navajos in the Pacific for secrecy.
> The Japanese weren't able to meet the daunting challenge imposed by
> learning a foreign language with no educational aids spoken by a
> relatively few people.

And it was known that there no Navajo speakers in Japan.

___Pete

For those of you cloak and dagger, wartime intelligence buffs who want to know 
the full story of Enigma and Ultra (the codename given to the intelligence 
wrought from the Enigma decodes carried out initially by the SIS [Secret 
Intelligence Service/MI6] at Bletchley Park), "The Ultra Secret" by F W 
Winterbotham, Chief of the Air Department SIS, is an excellent read. Originally 
published in 1974 by Weidenfield and Nicholson, it should still be available in 
paperback/through Public Libraries.

In "The Ultra Secret", Winterbotham gives a fairly complete account of how the 
SIS got their hands on a wooden mock-up of the Enigma machine in 1938, before 
war broke out in Europe. This was courtesy of the Polish Secret Service and the   
Deuxieme Bureau in France (who never seemed to have received much credit for 
their part in what was to become the biggest intelligence break of the war until 
Winterbotham's book). The book goes onto deal with the organisation set up to 
distribute Ultra intelligence throughout all theatres during the war, including 
the Pacific - the Japanese having thoughfully taken the Enigma encoder from the 
Germans and then using it without substantial modification. By the end of the 
war there were Ultra centres operating in Washington, Brisbane and Delhi in 
addition to Bletchley. You'll find the complete lowdown on Coventry and why 
nothing was done to evacuate the City before the Luftwaffe blitzed it.

There are no specific details of "Colossus" in this book - it was still 
considered secret under the 30 year rule when the book was written. "Colossus" 
was constructed by Post Office Telecommunications at the Dollis Hill Research 
Centre in London around 1943. As already noted elsewhere in this conference, 
"Colossus" was used to speed-up the decoding process as more and more Enigma 
traffic was intercepted and recorded. "Colossus" had many of the features of 
later, all-electronic computers; an input device (7 track paper tape reader), a  
processing unit made-up of relays where compare operations were carried out in 
order to determine the Enigma encoding sequence used for the particular message 
being decoded and an output device (a Teletype printer).

One of the most interesting bits about Ultra was that the allies apparently knew   
the exact location of all German U-boat supply submarines (called "Milchcows") 
in the North Atlantic on an almost daily basis by late 1942. However, they 
couldn't all be sunk without compromising Ultra.

    AFAIR Coventry was known to be on a list of towns to be bombed, but
    they did not decode the message in time to know which one was to have
    been the target on the actual night. Churchill headed back to London,
    as that was believed to be the target. I think it was around this time
    that the first attempts to jam the guidance equipment were being made,
    but these were unsuccessful, as the frequency had been measured
    wrongly.
    
    Re .34
>traffic was intercepted and recorded. "Colossus" had many of the features of 
>later, all-electronic computers; an input device (7 track paper tape reader), a  
    
    The paper tape reader was *very* fast (12,000 characters per second).
    It would destroy the tape as it read it.
    
    Peter

    The Navaho Code Talkers not only spoke in Navaho, but used a highly
    cryptic slang that they'd invented during training.
    
    Think of trying to understand Cockney Rhyming Slang rendered in
    Japanese.
    
    \dave

Andrew, et. al.:

> All previous computers did not have stored program capability. ENIAC is 
> discussed but not in detail. The original ENIAC was built for ballistics 
> calculations, then rebuilt during 1948 and 1949 to include a mercury delay 
> line and was then used to solve other problems. ENIAC spanned the boundary 
> of machines with and without srored programs.

  Didn't Eniac sport Williams tube memory at some point in its life?
  Or am I thinking of an early Univac?

                                   Atlant

Corrected date - sorry
    
Taking as source "Early British Computers" by Simon Lavington, Digital 
Press.

The first stored-program computer was a prototype of the Manchester Mark 1 
which ran a 52 minute factoring program on the morning of June 21st 1948.

All previous computers did not have stored program capability. ENIAC is 
discussed but not in detail. The original ENIAC was built for ballistics 
calculations, then rebuilt during 1948 and 1949 to include a mercury delay 
line and was then used to solve other problems. ENIAC spanned the boundary 
of machines with and without srored programs.

EDSAC first ran a program on 6 May 1949, the Manchester Mark 1 in early 
April 1949.

As an aside...
Enigma was largely solved/deciphered using electromagnetic relay machines 
and hundreds of people, plus clever mathematics to reduce the initial 
possibilities. The COLOSSUS was built to decipher the Geheimschreibers 
which had ten rotors and used 5-bit teleprinter code. COLOSSUS began work 
in December 1943. The 5-bit paper tape reader ran at 5000cps. 

Andrew