HAPPY 60th BIRTHDAY

[Bogstandard]

I thought that I would remind everyone on here that 60 years ago today, the first genuine modern day working computer started up in Manchester, UK. It was called 'The Baby'

This was the first computer that could store and use a program, and every other computer since was all due to this great invention.

There were other computers before this, but none could store, as well as use, at the same time.

http://www.bbc.co.uk/manchester/content/articles/2005/11/07/baby_computer_40_interview_feature.shtml

So many thanks to those people in Manchester in allowing us to have the communications we have today.

John

 

[mklotz]

What we really need to do is hold a bang-up celebration next Monday, the birthday of Alan Turing. Without his theoretical and practical work on Colossus, the development of the computer as we know it would have been grossly delayed - not to mention we might have lost WWII. He was treated rather shabbily during his life but that shouldn't stand in the way of honoring his achievements - many of which are central to computer theory today.

 

[Bogstandard]

As you must already know Marv, he died under very suspicious circumstances and there was a lot of speculation. But since the liberalisation of homosexuality, he has been honoured by everyone and his dog. A real tribute to a person, in my view was one of the most brilliant people of his time, and must have been a major force behind the above mentioned computer.

What really hits me about people like Turing, to put it bluntly, where the hell did they get their ideas and inspiration from. They must have been using their minds for 36 hours every day, not for personal glory, but just to get the job in hand done.

Lowly John

 

[mklotz]

I can't speak for Turing and I imagine every genius' mind works somewhat differently.

In the early days of Bletchley Park, Turing was charged with organizing the efforts of many human "computors" (i.e., people who performed some computation) who sat in ranks at long tables in one of the buildings. The idea was to break the task (decryption) down into tiny operations in order to speed things up and also to reduce errors. IIRC, Turing organized the computations on color-coded cards where the color was a primeval "program" that guided the card through the maze of human computors.

It's not difficult to imagine Turing mentally picturing each of these people replaced with a machine that performed some simple task (without error). From there it's not a big leap to picturing himself as a "super machine", handing out computational orders to these drone machines according to a list of instructions - his color coding cheat sheet.

There's a lot to be said for immersion in the problem. We've all had the experience of working feverishly for days on a knotty problem only to have the solution spring, full-blown, into our mind during the morning shower. Turing was about as immersed in his problem as one could get.

Before coming to Bletchley, Turing had already invented the concept of the "Turing machine", which defined the absolute minimum computation needed to solve a host of problems. That already in his mind and his immersion in what constituted a crude, real world Turing machine would certainly have helped point the way to a mechanism to replace the people.

Also, he wasn't unaware of the immediacy of the problem he was working. Decoding an "attack at dawn" message two days after it was written is pretty useless. Certainly he would appreciate that a mechanized solution would be faster and the payoff of "fast" was supply ships actually reaching Britain.


Please note that I'm not suggesting that I actually know how Turing's mind worked. It's just that, with my own personal experience with immersion, I can make what I think is a fairly logical extrapolation.

Immersion isn't everything. While Einstein claimed that Special Relativity came from his imagining "what it would be like to ride on a beam of light", most of his breakthroughs involved concepts where immersion was clearly impossible. All of which makes what he did all the more incredible. Nevertheless, Einstein did immerse himself in a problem in a sense. His favorite trick was to perform what he called Gedankenexperimenten (thought experiments) in which he imagined things like what would happen if we bent space and made time a relative parameter.

 

[John S]

That's all well and good Marv but even with bent space and time a relative parameter why does British rail not run trains on the same calender that others use ?

As an aside Dr Giles Parkes who is quite well known in UK model engineering for his articles on gears was one of the technicians at Bletchley Park during WWII working on the bombe machines that collected the decoded information.

Really nice guy if you ever get to meet him.

There is also I good book called Station X on Bletchley Park, ironically enough I'm re-reading it at the moment.

.

 

[rake60]

Amazing history!

If it hadn't been for those people and machines none of this could exist today.

I was 13 years old when I saw a computer for the very first time.
It filled a 20 X 20 foot AIR CONDITIONED room!
(Air conditioning wasn't a common thing here in those days.)
It's removable memory consisted of open reels of 9 track, 1/2" wide tape.
Each reel could hold a stagering 250MB of data.
Today a $10 thumb drive holds 40,000 times that but it will never change the
way I felt seeing a living, thinking machine at work the very first time.

Rick

 

[Circlip]

Yes John S the 3rd, Station X is a very good read, Quite an eye opener if everything in there is factual, but not very "Friendly"
Regards Ian.

 

[mklotz]

Hey, Rick, don't get carried away here.

40,000 * 250 MB = 10 TERAbytes

I don't think you can buy a 10 TB thumb drive, much less one for $10

 

[Bogstandard]

In the mid 80's I was involved in the manufacture of the Maxtor 5 1/4" full height hard disc drives. It had 165 MB of storage and the cost was approx. £2,000 ($4,000). Everyone then thought that this would be the maximum size of a hard drive, and you wouldn't be able to fill it. How wrong they were.

When the IBM 'pixie dust' magnetic coatings were used in later years, the capacity of hard drives went thru the roof, and have carried on at an alarming rate.

John

 

[rake60]

Hey, Rick, don't get carried away here.

40,000 * 250 MB = 10 TERAbytes

I don't think you can buy a 10 TB thumb drive, much less one for $10

There shows my math skills without a caculator.
See how dependant I've become on a computer chip? :-\

Rick

 

[Loose nut]

Turing was charged with organizing the efforts of many human "computors" (i.e., people who performed some computation) who sat in ranks at long tables in one of the buildings. The idea was to break the task (decryption) down into tiny operations in order to speed things up and also to reduce errors.

.

If computers were originally humans doing repetitive math calculations, this must have been parallel processing???

 

[mklotz]

No, the processing was serial. Person 1 would perform some small calculation then pass the work to person 2 who would perform another small part and so on.

There were several advantages to this system...

Each person had to learn only one small task so training them was simplified and they could easily be replaced should they fall ill or be killed in the nightly bombings.

The thought was that, with only a simple task to perform, the error rate for each task would drop and thus the error rate for the overall decrypt would fall. Whether this was actually true is hard to evaluate.

Several decrypts could be in the flow at once. After person 1 finishes his small task on decrypt A, he can be handed decrypt B and begin work immediately. This speeded up the decrypt rate and, of course, speed is essential in this work. See my example of decrypting a "we attack on Monday" message on Thursday.

Most (but not all) decryption is a serial process. One can't begin work on task 2 until the results from task 1 are available. It's not like adding a long column of numbers where each of N people can be given some of the numbers to add and then their subsums are summed to find the final answer. That would be true parallel processing.

Subdividing a complex task in such a way that it can be parallel processed remains one of the most challenging problems of modern computational theory.

 

[shred]

IIRC, some of the Manhattan Project "Computors" were used in a parallel-processing manner. Some fascinating reading there.

 

[shred]

What about 'not', the unary operator? NAND and NOR were merely high-level coding conveniences that worked well in electronics.. ;D ;D

 

[rake60]

LOL
Perhaps I was the NOT or Knot Head in those days...

Rick