PDA

View Full Version : Colossus cracks codes once more



John Stevenson
11-15-2007, 05:30 AM
For the first time in more than 60 years a Colossus computer will be cracking codes at Bletchley Park. The machine is being put through its paces to mark the end of a project to rebuild the pioneering computer.
It will be used to crack messages enciphered using the same system employed by the German high command during World War II.
The Colossus will be pitted against modern PC technology which will also try to read the scrambled messages.

http://news.bbc.co.uk/1/hi/technology/7094881.stm


.

Your Old Dog
11-15-2007, 07:16 AM
Will it work on my wifes Check book? :D

I find most of these projects quite humbling. Coming up with the concept is one thing, being able to carry it out with hundreds of thousands of parts is quite another. I've been so impressed by this technology that I'm going to renew my efforts to raise my 700 pound ATV into the air so I can wire some safety lights on it for snowplowing the end of the drive :D

Can you imagine what the folks who originally built this thing could have done had they the benefit of modern computers and tooling.

Evan
11-15-2007, 08:54 AM
What I find remarkable is that they were able to keep it a secret for so long during the war. I understand that they were forced to allow some towns and cities to be bombed in order to not tip their hand.

As a side note, the Colossus isn't a true computer. It isn't capable of performing a "turing complete" instruction set. It is best described as a purpose built dechiphering machine with limited programmability. With around 2000 electronic switches it has about the same logical complexity as the average wrist watch today.

DickDastardly40
11-15-2007, 09:27 AM
What I find remarkable is that they were able to keep it a secret for so long during the war. I understand that they were forced to allow some towns and cities to be bombed in order to not tip their hand.


I think that this was much earlier in the war during the decyphering of the Enigma machines. According to Winston Churchill there was no secret greater than the fact the Allies could read Enigma. Later when this machine was in use I think the Allies were more in transmit bombers than receive.

I regret I haven't researched recently this but from what I have read a while ago, I am happy to be set straight.

Al

Norman Atkinson
11-15-2007, 09:33 AM
Come on. Let's get real.
If anyone thinks that Colossus numbers 1 to 10 in Colossus Mark One was not replaced by Colossus Mark Two or such, they must live in a CloudcuckooLand.
The War did not end the need for code breaking, it only began.

As for tipping the game, there was bugger all which would stop a Luftwaffe raid at night which was using co-ordinates. Ally Pally or Alexandra Palace was bombed. The Nazis knew all about us knowing about us, us knowing about-----! Google Swains Lane in Highgate, London. But there was something else in Highgate. Almost 60 years since I was there, The RAF and the powers that be, deny what was there. The Yanks don't.

I was injured in the RAF- like most people, but three of my comrades were killed. It took 50 years to get an admission. It is only in the past 6 months or so that the RAF has released the existence of something. They haven't said WHAT went on.

No, I wasn't playing with Colossus but I knew who was playing with 'something' I had two of the Geeks in my section.

Let's get real and stop playing with Biggles leather helmets and Mark 8 goggles with rose coloured lenses!

NickH
11-15-2007, 09:36 AM
It isn't a real computer because "It isn't capable of performing a "turing complete" instruction set."

I think it was a real computer because it was capable of complex iterative processing of data held on a storage media based on programmable instructions (even though these were set up on hardware they are still programming) and producing output data.

So there:D
Nick

John Stevenson
11-15-2007, 10:02 AM
What I find remarkable is that they were able to keep it a secret for so long during the war. I understand that they were forced to allow some towns and cities to be bombed in order to not tip their hand.

As a side note, the Colossus isn't a true computer. It isn't capable of performing a "turing complete" instruction set. It is best described as a purpose built dechiphering machine with limited programmability. With around 2000 electronic switches it has about the same logical complexity as the average wrist watch today.
The first real computer was the Baby.

http://www.computer50.org/

BTW Giles Parkes of gear cutting fame who contributes to MEW one one of the guys who worked on Colossus in WWII.

.

SGW
11-15-2007, 10:10 AM
The article says the paper tape was read at 5000 characters a second. That seems awfully fast for a paper tape reader. The paper tape readers I used at DEC in the late 1960s could read, I think, 240 chars/sec. I'd like to get verification of that 5000 chars/sec number.

thistle
11-15-2007, 10:24 AM
The article says the paper tape was read at 5000 characters a second. That seems awfully fast for a paper tape reader. The paper tape readers I used at DEC in the late 1960s could read, I think, 240 chars/sec. I'd like to get verification of that 5000 chars/sec number.


maybe the ones from the 60s were not afraid of getting a German bomb in the carapace!

lazlo
11-15-2007, 10:29 AM
The first real computer was the Baby.

That's a favorite bar argument of computer engineers: what was the first computer, the first electronic computer, the first programmable electronic computer,...

The details of Colossus have only recently been released, but here's a great summary of the various features of Zeus (German), the Atanasoff-Berry, Colossus, and Eniac from Wikipedia:

http://i164.photobucket.com/albums/u15/rtgeorge_album/computer.gif

From the descriptions I've read, Colossus was most definitely a programmable electronic computer. It also had the first ever implementation of shift registers and systolic arrays (a form of SIMD).


BTW Giles Parkes of gear cutting fame who contributes to MEW one one of the guys who worked on Colossus in WWII.

Wow, that's very cool! Did you ever meet Tommy Flowers (who designed Colossus)? He was completely overshadowed by Alan Turing, and never got much credit for the brilliant technical achievements at Bletchley Park.

John Stevenson
11-15-2007, 11:15 AM
After the war, all trace of what occurred at the park was burned or destroyed. Such was the culture of secrecy at Bletchley Park that no word of what happened began to emerge until the mid 1970’s. To this day, many people who worked there are still reluctant to talk about it.
During the war, many ingenious aids and machines were developed at Bletchley Park to aid the breaking of codes, one of these was Colossus — the World’s first electronic computer.


(For many years the honour of being the World’s first electronic computer was given to the American ENIAC. In recent years however, both the UK and US governments have declassified and released papers giving more information about Colossus. In the light of this historians have been forced to reconsider and most now agree that Colossus was in fact the World’s first electronic computer.)



.

NickH
11-15-2007, 11:18 AM
Ah, so now it's about whether it was a "Modern" computer, that's moving the goalposts a bit, theoretically the Antikythera Mechanism was a computer, a mechanical one, you have to state your terms of reference before making sweeping statements!
Regards,
Nick

lazlo
11-15-2007, 11:30 AM
Ah, so now it's about whether it was a "Modern" computer, that's moving the goalposts a bit

Agree Nick -- that was my point about the first computer, the first electronic computer, the first programmable electronic computer,...

The first electronic computer was the Atanasoff–Berry Computer from Iowa State University. It solved linear equations, and was abandoned by the research team because of the War.

Since Colussus was not Turing Complete (i.e., it wasn't a fully general-purpose computer like Eniac), it doesn't hold the claim of the first "modern" computer. If you look at that Wikipedia chart I posted above, none of those machines is a binary, electronic, programmable, general purpose (Turning Complete) machine.

The Manchester I (circa 1948, well after the war), that John was calling "Baby," is really considered the first "modern" computer, but even that can be argued.


theoretically the Antikythera Mechanism was a computer, a mechanical one,

No one's entirely sure what the Antikythera is, but according to the Nature article that the international research team published last year, the leading theory is that it was an orrery (an astronomical clock). If you want to call that a mechanical computer, then sure :)

John Stevenson
11-15-2007, 11:38 AM
We could go on arguing about this for ever more but I posted the original link as I thought it was cool that they had restored this and allowed it to be seen by the general public as Bletchly Park is now open to visitors.

Lazlo,
Never met Tommy Flowers, in fact I had never heard of him until I read Station X a few years ago.

It's by coincidence that I'm re-reading it at the moment.

.

NickH
11-15-2007, 11:47 AM
John,
I agree that it was really cool and a hell of an achievement, I was just responding to the nit-picking, "terms of reference-free" comments following the original post, :D
Regards,
Nick

Evan
11-15-2007, 11:49 AM
I was taught that to be considered a computer rather than a special function machine (glorified calculator) the machine needed to be Turing Complete. This isn't a concept that came after Colossus or any other computer. The idea was well known to those that built it and ENIAC as well. Being electronic and binary doesn't make it a computer. Being "programmable" by changing the wiring doesn't make it a computer. Being able to change it's own instructions does. Turing completeness isn't a vague or poorly defined concept. To be Turing Complete a programming language only needs 8 instructions. The best example of that is the programming language called Brainf*ck.

http://www.muppetlabs.com/~breadbox/bf/ (http://www.muppetlabs.com/%7Ebreadbox/bf/)

The idea of Turing completeness defines what a computer is and that logic doesn't change regardless of how or when it was implemented. Modern computers share the exact same logic elements with every digital computer ever built. It only takes two logic elements to build a computer, AND and NOT. Every other logical function can be implemented using those two functions.

lazlo
11-15-2007, 12:03 PM
The problem with that definition Evan is that there are plenty of computers that aren't Turing Complete, like most analog computers -- i.e., the analog ballistics computers on the big WWII battleships.

"Turing Complete" just means that it's equivalent to a Universal Turing Machine -- the machine can perform any computational task that takes finite input and returns finite output in finitely-many steps.

So Charles Babbage's Analytical Engine was Turing Complete (circa 1830, but it was never completed).

Zeus, the German calculating machine, was the world's first Turing-Complete computer, but it was mechanical.

The Atanasoff–Berry machine from Iowa State was the world's first electronic Turing Complete computer, circa 1942, but it wasn't programmable (the program was hardwired).

The famous version of Eniac used at the Aberdeen Proving Grounds during the War to calculate ballistics tables was also not programmable (it was programmed with telephone plugs) and it wasn't Turing Complete.

The Mark II version of Eniac added firmware programmability, and was Turing complete, circa 1948.

NickH
11-15-2007, 12:09 PM
Evan,
You initially omitted the words "modern" and "electronic",
Quote-
"What I find remarkable is that they were able to keep it a secret for so long during the war. I understand that they were forced to allow some towns and cities to be bombed in order to not tip their hand.

As a side note, the Colossus isn't a true computer. It isn't capable of performing a "turing complete" instruction set. It is best described as a purpose built dechiphering machine with limited programmability. With around 2000 electronic switches it has about the same logical complexity as the average wrist watch today."

assuming something has to be modern and electronic to class as a computer is, well, you know the saying about the word assume, you have now added the words modern, electronic & even binary, so your stated premise has changed,
Regards,
Nick

lazlo
11-15-2007, 12:15 PM
We could go on arguing about this for ever more but I posted the original link as I thought it was cool that they had restored this and allowed it to be seen by the general public as Bletchly Park is now open to visitors.

John, I wasn't disputing the importance of Colussus -- just that it was one of a series of major milestones on the path to the modern computer.

I'm thrilled that the Colossus has not only been restored, but that it's up and running, and cracking codes again after 60 years! Sadly, ENIAC didn't fare as well -- there are pieces on display at the Smithsonian, but when I used to work up at the Aberdeen Proving Grounds in Maryland (the birthplace of Eniac) there were pieces of ENIAC ungraciously strewn all over the place.

I'm going to be in London for a technical conference next Spring -- I'm really looking forward to seeing Bletchley Park!

By the way, here's a picture of the newly restored Colossus:

Click on the link for a much bigger version:

http://upload.wikimedia.org/wikipedia/commons/e/e0/ColossusRebuild_11.jpg

http://upload.wikimedia.org/wikipedia/commons/thumb/e/e0/ColossusRebuild_11.jpg/320px-ColossusRebuild_11.jpg

Norman Atkinson
11-15-2007, 01:13 PM
Well, Robert. Your homework starts with 'Domino versus Y-Gerat'
I'll save you a bit of leg work. You are also looking for a story of an illigitimate offspring of a Norwegian who just happened to be 'connected' with a Swains Lane in Highgate, London N6. But there is also a something for P-47 pilots. That is the US version of some of the events.
The London Evening Standard on 24 Oct 1945 reports a radio mast accident- well in Highgate

I seem to be getting into radio- don't know why. the Polar Institute in Cambridge, England tells of Royal Air Force corporal radio fitter who was stuck out on chunk of Antarctica called Queen Maudland and a little radio station called Maudheim. Question- what was he really doing- and why did someone name an iceshelf Quarland after such an inconsequential lowly creature when there were Four others who outranked him?

You might get carried away- there is also something called Nazis in Antarctica.


The Vampire in Swains Lane? Nah, there are better ones than that.

Bletchley Park? There are more Enigmas than that

John Stevenson
11-15-2007, 01:32 PM
[quote=aviemoron]The London Evening Standard on 24 Oct 1945 reports a radio mast accident- well in Highgate

quote]

So was Karl Marks having a phone put in ?

.

Malc-Y
11-15-2007, 03:16 PM
:) I live only a mile from Bletchley Park and have visited it quite a few times though not for a few years. It's funny how you never seem to visit places of interest close to home. I shall have to make a point of going soon to see the rebuilt Colossus. I always know when there is a special open day on at Bletchley Park as we get the Battle Of Britain Memorial Flight Avro Lancaster fly over my house at very low altitude, and it is sometimes accompanied by the Spitfire and the Hurricane. It's very stirring to hear those Merlins close up.

Malc. :cool:

P.S. lazlo, contact me when you are coming to Bletchley Park, maybe we could meet up.

Evan
11-15-2007, 03:59 PM
I didn't add electronic, modern or binary, other posters did. Mechanical computers are every bit as valid as an electronic one if they meet the definition. As long as they perform the appropriate logic functions it doesn't matter how it is done. My point is that being electronic or binary or modern doesn't change anything. From my experience with the field of computing over the years since 1963 when I first learned how to program one it has been generally accepted that turing completeness is the essential difference between a true computer or computer language and any other sort of automaton. Computers are automatons but not all automatons are computers.

dp
11-15-2007, 04:53 PM
I wonder what the definition of a computer was at the time it was first applied to Colossus, and if perhaps that definition has changed over time thus leaving a legacy of confusion. Turing introduced the Turing test in 1950 or so, well after Colussus, and certainly before the concept of Turing completeness had been established. Such language changes create innocent errors in the historical record.

lazlo
11-15-2007, 06:21 PM
Alan Turing published his paper defining the Universal Turing Machine in 1936. If I can remember my undergraduate computer science classes, that was when he was still a graduate student at Princeton, studying under John Von Neumann. What's especially amusing about this discussion is that Alan Turing was the lead scientist at Bletchley Park, while Colussus was being designed and built.

The 1950 paper you're describing was his philosophical dissertation on ways to distinguish between a human and a computer, and proposed the "Turing Test."

But your point is quite apt: a lot of arguments over whether a particular machine was Turing Compete is largely semantic. The Zeus machine, for example, was for years considered not Turing Complete, because it didn't have conditional branches:


IF (X = Y) THEN (Foo)

ELSE (Bar)

But then in 1998, 60 years after the Zeus machine had been disassembled, a German academic published an article in IEEE Computer describing an obscure trick that could have been used to simulate a conditional branch on Zeus.

Then, suddenly, the history books were re-written, and Zeus went down as the World's First Turing Complete Machine.

dp
11-15-2007, 07:19 PM
Personally I still think Pluto is a planet, but that is the single reworked definition that stands out as having a lot of history to repress to be accepted.

But back to John's thread - that Colossus is being reactivated and even competing against PC's to solve Enigma encrypted data. That is simply astounding.

Don't know why but I'm reminded of the cute 80's cold war era song from Nena:

Neun und Neunzig Luftballons

(English)
Ninety-nine red balloons
Floating in the summer sky
Panic bells, it's red alert
There's something here
From somewhere else
The war machine springs to life
Opens up one eager eye
Focusing it on the sky
Where ninety-nine red balloons go by

kmp
11-15-2007, 07:47 PM
Here we go again, we won't have to wait long and one of those nutters will be along to tell us the world is not really flat!:D

History, has a way of being re-written regularly depending on the wind direction and price of eggs at the time of the latest inturpretation. Historians have to eat as well I suppose. Personally I just believe that they were all major steps in the development of "computers" and I don't really care in what precise order they came. I do however have a problem with re-defining a machine in light of later understanding - in my book it was what it did at the time, not what it was actually capable of in hindsight that is important.

Keith

lazlo
11-15-2007, 08:59 PM
:) I live only a mile from Bletchley Park and have visited it quite a few times though not for a few years.

P.S. lazlo, contact me when you are coming to Bletchley Park, maybe we could meet up.

Thanks Malc -- that would be great! I wouldn't mind meeting some of the British Model Engineers too :)

J Tiers
11-15-2007, 10:21 PM
Evan et al:

It really hardly matters whether "Colossus" was a full, complete modern computer or not.

The point, which all the academic fluff totally misses, is that:

"Colossus" WAS D**N GOOD AT EXACTLY WHAT IT WAS MADE FOR.

I put it to you; THAT is all you need to consider in evaluating it. I suspect that the complications required for "Turing completeness" would NOT have helped, and might well have hindered, or at least delayed deployment.

If any of those features would have helped, I think you can be pretty sure they would have been put in.

BTW, as for "nothing stopping a raid"..... read "The Wizard War". It may go by a different name in the UK.

It mentions, and NOBODY would have been in a better position to know than the author, Dr R. V. Jones, a number of "dirty tricks" used to break up and divert raids to the wrong place, or no place.

it seems that the Nazis used radio beam directions to set the locations to be hit. Depending on the exact system (there were several, each countered in turn) in use, manipulation of counter signals could, and did, send the raiders off in wrong directions, with some even ending up West off the coast of Wales..... and likely a bit too low on fuel.

The V-2 was completely unstoppable. But, later, manipulation of spy reports caused the V-2 impact points to be moved East, away from dense areas, saving an estimated 10,000 people.

The re-creation of a "Colossus" of whatever revision, is a great thing, and brings back to life the first widespread use of "Elint" and the modern forms of spying. It is important as the single most deciding factor in many aspects of WW2, from small beginnings in pre-Dunkirk operations, thru the early air war, N Africa, and even much later. You can read about it in the F. W. Winterbotham book. He was in charge of the output of B-P during much of the war.

Oh, yes, as for the dates, not only did the Nazis continue to use "Enigma" throughout the war, so did the Japanese. The first big breakthrough there was the battle of Midway Island, which was tilted in the US' favor by intercepted messages, then "won" by good preparation with some very good luck as the deciding factor.

dp
11-15-2007, 10:29 PM
I went to Google to see what I got confused and damn that was fun. Indeed it was 1936 when the idea of the universal machine was postulated and the Turing "game" of guess if it's real or a machine came in the 50's. So much for my aging gracefully.

And in re-reading it all I got distracted into tuple spaces which was a system I worked on while at Boeing using Java tools. Both IBM and Sun and introduced tuple spaces management tools that were clever and easy to work with. I can see some applications for machining with CNC and tuple spaces, but it would be very complex to implement for the average guy.

And I'd forgotten too what an awful death Turing suffered. Takes some of the shine off the "nifty fifties" for sure.

Evan
11-15-2007, 10:52 PM
Jerry,

I'm not trying to minimized the accomplishments of the code breakers. I'm quite sure they didn't give a lot of thought to the idea of building a machine that was Turing Complete. At that time the advantages wouldn't have been realized. It confers the ability to the machine to solve any problem that a computer can solve given the contraints of time and resources. A machine that isn't Turing Complete cannot be programmed to solve all problems that are solvable by a computer.

The nit I have to pick is with the reporting in the BBC story where they say "Colossus is widely recognised as being one of the first recognisably modern digital computers". Well, it isn't and it isn't. To be considered that it would have to be able to perform the same function set that other modern computers can and without Turing Completeness it cannot do that. There are numerous examples of machines that fall in the same category. The NSA uses purpose built deciphering machines that can outpace a regular computer by hundreds or maybe even millions of times in speed. Not being Turing Complete does not neccessarily confer some sort of handicap in performing the intended task, IF the intended task is very well defined to start with. It's when changes must be made that a fully programmable machine wins out.

Doc Nickel
11-15-2007, 10:52 PM
I wonder what the definition of a computer was at the time it was first applied to Colossus[...]

-"Computer" was, at the time, basically "something that computes". As noted, the analog (clockwork) ballistics calculators for battleships were called "computers", because they 'computed' the trajectories.

We only have to add new modifiers to the definition today, because the term 'computer' now means something very, very different. We look at it from our current view of ubiquitous ultrapowerful desktop computers, but back in the day of Colossus, it was very much referred to as a "computer".

It may have been a glorified calculator, or little more than an improved Babbage Engine, but it was, indeed, a computer.

That doesn't help the definition for "first", of course, but as noted, it depends on how you define which kinds of computers.

Still and all, definitions and firsts aside, it's still very cool they've recreated a functional model.

Doc.

lazlo
11-15-2007, 11:00 PM
It really hardly matters whether "Colossus" was a full, complete modern computer or not.

"Colossus" WAS D**N GOOD AT EXACTLY WHAT IT WAS MADE FOR.

Agree 100% Jerry -- that was my point about Alan Turing being the Lead Scientist at Bletchley Park when Colossus was being designed and built. If Alan Turing, of all people, didn't think Turing Completeness was necessary for Colossus, then it probably wasn't.


The V-2 was completely unstoppable. But, later, manipulation of spy reports caused the V-2 impact points to be moved East, away from dense areas, saving an estimated 10,000 people.

I just watched this week's Nova episode about the spy story/ICBM race behind Sputnik. The Redstone rocket booster used by the US was an enhanced V-2 designed by Wernher von Braun, who also designed the V-2. They mentioned that of all the V-2 rocket attacks on France, England, and Belgium, only 6,000 people were killed, but 25,000 slave laborers died producing the rockets. Nova made the somber observation that it's the only weapon in history where more people died building it than were killed by it.


Oh, yes, as for the dates, not only did the Nazis continue to use "Enigma" throughout the war, so did the Japanese. The first big breakthrough there was the battle of Midway Island, which was tilted in the US' favor by intercepted messages, then "won" by good preparation with some very good luck as the deciding factor.

The Japanese Enigma was a different machine, of completely Japanese design, and it wasn't used very much. The Japanese code system that was broken by the US Navy (and was crucial to the victory at Midway) was "JN-25", which was a 5-digit substitution code. It consisted of a dictionary of 45,000 five-digit numbers. Each number represented a word or phrase. When the JN-25 codes were sent, the five-digit numbers were double-enciphered using a second substitution table.

The US Navy was unable to break JN-25, until the Japanese made the huge mistake of sending the same message in JN-25, and a previous, weaker code that the US had already broken.

The famous story of the Joint Chief's not being convinced that "AF" was really Midway Island is true: the Navy sent a bogus message saying that Midway's water distillation plant had broken, and sure enough, a JN-25 message was intercepted that the water distillation plant on "AF" was down. The rest is history :)

Evan
11-15-2007, 11:00 PM
We only have to add new modifiers to the definition today, because the term 'computer' now means something very, very different.

More different than you realize. The term "computer" is a lot older than the machines. It originally referred to a person who computed mathematical problems, particularly those who did so as a job. Richard Feynman was in charge of a group at Los Alamos Lab who did just that in computing the artillery firing tables used to compensate a gun for windage, elevation, charge etc. The people doing the calculations were employed as "computers". The term was transferred to the machine as was the task.

dp
11-15-2007, 11:05 PM
yah - I don't think anyone in this thread really gives a rip about what was the "first computer". I certainly don't. And that honor will remain a changing target so long as we keep changing the language. Same thing happened with television and radio. So for the same reason that I don't need some whacked scientific board to tell me what Pluto is, I'm going to consider that Colossus, Eniac, Brainiac, etc, were pioneer computers at a time before the modern definition of computer evolved. Turing be damned, and anyway, his is "a" definition, not, possibly "the only" definition. And the good news is they worked great.

fasto
11-15-2007, 11:12 PM
The article says the paper tape was read at 5000 characters a second. That seems awfully fast for a paper tape reader. The paper tape readers I used at DEC in the late 1960s could read, I think, 240 chars/sec. I'd like to get verification of that 5000 chars/sec number.

My reference, "Code Breakers, the inside story of Bletchley Park" (Oxford Univ. Press, ISBN 0-19-285304-X) indicates that Colossus used "ordinary teleprinter tape run on pulleys at nearly 30 miles per hour" (pg163).

This same section, authored by Jack Good, suggests that Colossus used boolean programming internally. He also says that several other uses were found, including (almost) multiplication by base 10.
Whether this machine fits the proper description of a 'computer' I cannot say, although I suspect that its flexibility may indeed be greater than we might expect.

Regards,
Aaron

lazlo
11-15-2007, 11:24 PM
Richard Feynman was in charge of a group at Los Alamos Lab who did just that in computing the artillery firing tables used to compensate a gun for windage, elevation, charge etc. The people doing the calculations were employed as "computers". The term was transferred to the machine as was the task.

Evan, you have your people and military sites confused.

Richard Feynman never worked on ballistics. He was hired while he was a Physics Student at Princeton to work on the Manhattan project.

Los Alamos was created by General Groves in 1942 as a super-secret site to isolate all the Manhattan Project researchers.

The ballistics calculations for windage, elevation, curvature of the earth, etc were all done at the Aberdeen Proving Grounds (the Ballistics Research Lab). Initially, the calculators were all women running manual calculators. At the height of the war, nearly two hundred young women, both civilian and military, worked as human "computers," performing ballistics computations. Later, ENIAC was built to automate the ballistics tables, and 6 women who were running the calculations manually on mechanical calculators were chosen to program ENIAC.

Feynman did at one point work with the "Hand Computing" group at Los Alamos, but on nuclear physics calculations for the Uranium Bomb. Stan Frankel, a friend of Richard Feynman, led the "Hand Computing" group, largely the Wives of the Nuclear Physicists working on the Manhattan Project:

http://www.hp9825.com/html/stan_frankel.html

"Los Alamos purchased a quantity of mechanical calculators (Marchants, Fridens, and Monroes) like the ones they used at Berkeley to help perform the many calculations required by the project physicists.

Although the Los Alamos physicists performed many calculations on their own, the sheer volume of calculations for the uranium bomb project proved too much, which led to the use of computers (the female kind). The wives of Los Alamos scientists—including Frankel’s wife Mary—were recruited to become, literally, human or “hand” computers that operated the mechanical calculators like industrial production equipment to crank out numbers. These human computers became the T-5 hand-computing group. Nick Metropolis and future Nobel Laureate and expert-on-pretty-much-everything Richard Feynman learned how to fix the mechanical calculators and became the T-5 handymen, working closely with Frankel and Nelson.

Frankel and Nelson did a superb job of organizing the calculations into an assembly line. Frankel’s deep understanding of the mathematical simulation of physical phenomena allowed him to decompose complex calculations into many simpler calculations that were distributed among the human computers. These simpler calculations incur fewer errors and the system worked well, until the humans started to tire of the endless, repetitive calculations.

Before 1943 ended, the endless bomb calculations proved more than the T-5 hand-computing group could handle and the T division decided to automate. There are no electronic computers to be had in 1943 (they didn’t exist yet, except in Iowa) so Los Alamos ordered IBM punched-card tabulation equipment including the IBM model 601 multiplier to help with the calculations. The tabulation equipment could add, subtract, and multiply using IBM punched cards for data input. The IBM punched-card tabulation group was designated T-6 and the tabulating machines were run by enlisted soldiers from the Army’s Special Engineering Detachment."

J Tiers
11-15-2007, 11:35 PM
The Japanese Enigma was a different machine, of completely Japanese design, and it wasn't used very much. The Japanese code system that was broken by the US Navy (and was crucial to the victory at Midway) was "JN-25", which was a 5-digit substitution code. It consisted of a dictionary of 45,000 five-digit numbers. Each number represented a word or phrase. When the JN-25 codes were sent, the five-digit numbers were double-enciphered using a second substitution table.

The US Navy was unable to break JN-25, until the Japanese made the huge mistake of sending the same message in JN-25, and a previous, weaker code that the US had already broken.

The famous story of the Joint Chief's not being convinced that "AF" was really Midway Island is true: the Navy sent a bogus message saying that Midway's water distillation plant had broken, and sure enough, a JN-25 message was intercepted that the water distillation plant on "AF" was down. The rest is history :)

The basis of the "enigma" machine was a "rolling substitution table", with various enhancements. So it can be classed with other substitution encypherments, if you want to be a little loose about it. The same type of attack can break any substitution, so long as you have some idea how it is made. Obviously some will take longer.

Most substitutions used to depend on assumptions, such as using non-intuitive substitutions, or using a different language prior to or in the process of substituting. but those are things a machine is uninterested in. The programmer handles that, and can afford to try some variations if the process is faster.

I had understood that the code geeks generally classed the German and Japanese versions as similar, although not identical, and that similar methods would work on the Japanese version.

Evan
11-15-2007, 11:39 PM
Evan, you have your people and military sites confused.
It doesn't sound like it from what you just posted. Feynman told the story a bit differently in his biography "Surely You're Joking Mr. Feynman". I don't have it to hand right now as when I read it I borrowed it from the library but it covers his days at Los Alamos and the computing group.

lazlo
11-15-2007, 11:46 PM
I had understood that the code geeks generally classed the German and Japanese versions as similar, although not identical, and that similar methods would work on the Japanese version.

I think you're right Jerry -- that the Japanese Enigma (called "Green" by the Allies) was probably reverse engineered from the commerical version of the 3 rotor enigma that was freely sold in Japan before World War II. That version had 3 rotors, and didn't have the switchboard to scramble the letters on the typewriter before it hit the first rotor.

The German Military Enigma machine had 4 - 8 rotors mounted coaxially, and the switchboard.

http://upload.wikimedia.org/wikipedia/commons/a/ae/Enigma.jpg

The Japanese "Green" Enigma had 4 rotors that were mounted flat, and no plug-board, so I don't know if it used the same cross-wiring scheme in the rotors:

http://upload.wikimedia.org/wikipedia/commons/thumb/9/93/Japanese-enigma.jpg/306px-Japanese-enigma.jpg

Enigmas show up on Ebay from time to time. I'd love to have one - fascinating machines! :D

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=320179963059

J Tiers
11-16-2007, 12:51 AM
Because of the flat mounting, I would assume that the rotors were built entirely differently, with, say, two circles of contacts on one side.

The latter would allow "face-mounted" rotors to function identically to the "through" type, at the cost of more wiring, but the advantage of easy swapping for more convolutions.

tony ennis
11-16-2007, 01:25 AM
The German Military Enigma machine had 4 - 8 rotors mounted coaxially, and the switchboard.

The German naval version had 4 rotors, the 4th being weaker than the rest - it didn't rotate like the other rotors though it could be set into any of 26 positions. The German army Enigma had 3 rotors. There were a few oddball Enigmas with more rotors but these weren't in common use, if at all.

Norman Atkinson
11-16-2007, 04:59 AM
JT,
I am quite adamant about the inability to stop air raids.
I am one of those who was at the receiving end too many times.

The big raid on Coventry was an illustration about 'Domino versus Y-Gerat'

What the 'Foreskin Fusiliers' forget is that an air raid on Southern England was a matter of less than 30 minutes warning from take off from Occupied France to 8000 lbs of bombings tipping out of a He111 on Southern England.

To get any sort of defence, a night fighter had to be airborne above the attack. The next question is just how many night fighters had the Brits defending 1000 miles of coastline. Was the legendary Spitfire and Hurricane up in the skies waiting for a He 111 or Ju88 in the dark? The answer is 'No'

I once saw a night fighter in attack in 6 years. The 'Foreskin Fusiliers' saw none- they were not there! They had their hands in their pockets and f*** all to do. Had we enough anti aircraft guns to respond? 'No'

Out of three raiding aircraft, one will always get through. Read the Augsburg Raid when Lancs were first used. Most of the new Lancs got the chop but Nettleton and his now burning wingman got through. My neighbour 'got through' but didn't get home. He escaped out of Luft 3, was recaptured but a couple got home. Remember that the Germans were just as determined as our lads. Today, a Tornado bomber will always get through. There is no change! When you have faced the guns of a He111 and then its bombs- the facts are painfully obvious. Oh, the crew in question got the chop but-- they got through.

J Tiers
11-16-2007, 09:24 AM
That is naturally true. Whoever said different?

I think you see a windmill where there is only laundry on a line.

A 100% blocking is very difficult, although the NAZIS attained that figure, or a figure so close as to make no difference, quite a number of times to the discomfiture of the allies.

They had much more warning, but very little in the way of technical assistance in terms of dirty tricks, and had to do it the hard way. Apparently it worked for them, even if the means included the mistakes of the allies (Leave the IFF on, fail to form up properly, navigate improperly, etc) .

The items I mentioned were "blunting" not prevention. It is quite a different thing if you can send 2/3 or more of the attackers off "into the weeds", vs having them concentrate on the target you want to save.

Not very heartening to the folks who DO get a "gift from above", but much better overall for the nation if the factory gets only 10 hits instead of 100, or the death toll can be cut by 10,000.

These days, we'd have a government commission to investigate the wrongdoing of those who knowingly sacrificed some for the greater good, with the goal of jailing them. But then it made a lot of sense, Herbert Morrison notwithstanding.

lazlo
11-16-2007, 09:43 AM
The German naval version had 4 rotors, the 4th being weaker than the rest - it didn't rotate like the other rotors though it could be set into any of 26 positions. The German army Enigma had 3 rotors. There were a few oddball Enigmas with more rotors but these weren't in common use, if at all.

The original Wehrmacht Enigma (Enigma I) issued to the German army in 1928 was the commercially available 3 rotor Enigma with the addition of the front plug-board to pre-scramble the keyboard. The German Navy added a fourth rotor for U-Boat traffic in 1942. In order to fit the 4th rotor in the same space, the early version had to make a thin rotor and shave the telegraph transmitter (the "reflector ring", so the 4th rotor looks odd, but it works the same as the other three. When the 4-rotor machines hit full production, they were back to 4 identical rotors:

http://cnm.open.ac.uk/projects/stationx/enigma/images/details/topview.jpg

There's a great description of the internal mechanism, and a cryptanalysis here:

Enigma Keys
http://www.cromwell-intl.com/security/history/enigma.html

An Enigma key is made up of three setting components:

1. A selection of rotors in a specific order, what the Germans called Walzenlage. The mathematics of this varied with service during the war. The Army and Air Force selected three rotors out of a set of five, while in 1 February 1942 the German Navy introduced a system which selected four rotors out of a set of eight.

2. The "roll-over" ring setting, or Ringstellung, for each rotor. The rotors generally changed position for each input character, moving as odometer wheels move. But the ring setting specified where the rollover happened.
* Army / Air Force 3-rotor possible ring settings = 26^3 = 17,576
* Navy 4-rotor possible ring settings = 26^4 = 456,976




The 8 rotor "Enigma II" was used for high-level military communications:

http://upload.wikimedia.org/wikipedia/commons/thumb/b/be/Enigma-8-rotor.jpg/399px-Enigma-8-rotor.jpg

... and the Swiss made a 10 rotor Enigma variant towards the end of the war:

http://i5.ebayimg.com/05/i/000/b7/88/b43f_12.JPGhttp://i17.ebayimg.com/02/i/000/b7/88/b695_12.JPG

Evan
11-16-2007, 12:23 PM
Navy 4-rotor possible ring settings = 26^4 = 456,976

That should take a modern desktop CPU a few milliseconds to break by a brute force attack. The current fastest production CPUs are capable of around 30 gigaflops (30 billion floating point operations) per second. Graphics card GPUs are even faster although integer only.

lazlo
11-16-2007, 12:47 PM
That should take a modern desktop CPU a few milliseconds to break by a brute force attack.

Fortunately, the 4-rotor Enigma was built 65 years ago :)


The current fastest production CPUs are capable of around 30 gigaflops (30 billion floating point operations) per second.

Try 50 GFlops: A 3 Ghz Clovertown (quad-core) is 4 cores @ 4 Flops/cycle = 48 GFlops.


Graphics card GPUs are even faster although integer only.

All modern GPU's (Nvidia G80, G85, ATI R600,...) are single-precision floating-point.

Evan
11-16-2007, 12:55 PM
It's hard to keep up with it all and since I closed the computer store I'm not trying very hard to either. Most people don't appreciate how powerful modern computers really are. Of course Windows does a great deal to conceal that fact. A single average laptop computer (with a good modern programmer) could have changed the outcome of WWII depending on whose hands it was in. That could make an interesting science fiction story. Of course you would have to stop the military from trying to take it apart to see what makes it go.

lazlo
11-16-2007, 01:12 PM
I agree Evan, especially with respect to the power of GPU's. I've been involved with a project at work where part of my job has been to assess the difficulty in programming GPU's for non graphics workloads, such as (stock) Option Pricing Models, Oil and Gas (seismic FFT analysis), etc.

Although the GPU's are currently a major pain in the ass to program non-graphics applications, with the latest crop from ATI and Nvidia, they've made it a whole lot easier, and they have (free) integrated development systems you can download.

So if you have a G80 (Geforce 8800) or an R600 graphics card, you actually have 320 GFlops of real, programmable single-precision float available to you.
To put that in perspective, the Stanford GPGPU guys have ported Folding@Home to run on the Nvidia G80 cards, and the PS3. It's a free download at Stanford.

I was really surprised on customer visits that many eminent financial services companies, oil companies, etc were building GPU farms as a poor man's supercomputer.

A couple of GPU cards would make a great code breaking machine ;)

BobWarfield
11-16-2007, 01:26 PM
Lovely clockwork on those enigmas. Some enterprising soul should undertake to build one as a model engineering project. Perhaps an old Smith Corona could give it's life in the effort. The task would be a lot easier with CNC, so whoever does it should publish the g-codes. Such a machine would look slick alongside the model machine guns others have been building.

Speaking of lovely clockwork, I've been fascinated by the Difference Engines since I first saw pictures:

http://www.sciencemuseum.org.uk/images/I033/10303307.aspx

Now that's a computer suitable for this crowd--all mechanical!

BTW, the Turing Machine is a mathematical abstraction no real piece of hardware has ever implemented, although we like to talk about modern computers as Turing Machines. The reason? The mathematical model calls for a tape of infinite length. No machines have ever been built with infinite memory. Here is the Stanford University definition, for example:

http://plato.stanford.edu/entries/turing-machine/

What does it all mean? Well, it's quite interesting. I annoyed the profs at grad school by showing that in fact problems Turing Machines can't solve, such as the famous "halting problem", can be solved for real computers. When you create a Turing Machine with finite tape (e.g. a computer with finite memory), it boils down to a simple (easy for me to say) Finite State Automaton.

Whether such nonsense is practical in any real sense remains to be seen, but it seemed entertaining to launch the "there ain't no such thing as a real Turing Machine" hand grenade into this fray. I'll try to restrain myself from further Digital Deviances in future!

Cheers,

BW

PS And speaking of Antikythera, it's been a project I've been interested in doing on CNC:

http://www.thewarfields.com/cnccookbook/CCOrreryNotes.html

Spin Doctor
11-16-2007, 05:51 PM
Speaking of lovely clockwork, I've been fascinated by the Difference Engines since I first saw pictures:

http://www.sciencemuseum.org.uk/images/I033/10303307.aspx

Now that's a computer suitable for this crowd--all mechanical!
http://www.thewarfields.com/cnccookbook/CCOrreryNotes.html

Now drive the crankshaft with a steam turbine to stay Victorian or at least Edwardian and watch the solution time really come down :D

Now if I can just find a link to that clock that was being built to go into a time vault in the Rockies.

BobWarfield
11-16-2007, 06:05 PM
Now if I can just find a link to that clock that was being built to go into a time vault in the Rockies.

That wouldn't be this clock, would it?

http://www.longnow.org/projects/clock/prototype1/

Best,

BW

J Tiers
11-17-2007, 12:02 AM
One could in principle "build" an N-rotor machine as a PC program.

2 or 3 hundred rotors could make breaking the code a matter of long enough time to make the "news" worthless.

Or, by simply flooding the available communication means with thousands of messages, each differently encoded, by virtue of your programmable N-rotor machine, you could keep the NSA so busy that you could slip your messages in at random and probably never even get noticed.

A sort of "denial of service" attack, aimed at the spy folks.

Since you wouldn't want to be identified as the only source, it could be done as a trojan worm, that infected machines, but only spewed a certain number of messages at specific times, in order to avoid truly denying service. Zombie message generators emailing encoded messages all over would probably make things rather hot. The NSA would probably get the internet put in lockdown to stop it.

Someone has surely already thought of it before me, so this is nothing new or too hot to let out into the wild.

And the truly 100% unbreakable code is already known. An N-rotor machine is an approach to it, but the real deal has existed for some time. All the NSA geeks in the world cannot break it except by the merest chance, and then for one message only.

But, the Enigma machines are nice pieces of machine work.

BobWarfield
11-17-2007, 01:04 AM
One could in principle "build" an N-rotor machine as a PC program.



Here's an Enigma simulator:

http://users.telenet.be/d.rijmenants/en/enigmasim.htm

Google quantum cryptography for some gnarly crypto.

Cheers,

BW

Evan
11-17-2007, 03:13 AM
BTW, the Turing Machine is a mathematical abstraction no real piece of hardware has ever implemented, although we like to talk about modern computers as Turing Machines. The reason? The mathematical model calls for a tape of infinite length. No machines have ever been built with infinite memory.
There is a difference between a Turing Machine and a machine that is Turing Complete. We aren't talking about a Turing Machine when we speak of a Turing Complete computer since it cannot of course have infinite capacity. However, if infinite capacity were magically granted then a Turing Complete computer can be a Turing Machine. While the Turing Machine is a mathematical abstraction it isn't very abstract. It can be easily implemented in simulation on a Turing Complete computer.

As the link you provided points out, there has been no better definition of a computer (computability) proposed as yet than the notion of Turing Completeness. The requirement for infinite memory and infinite time to solve a problem isn't a logical requirement but a merely practical one. It ensures that a problem won't be rejected because there isn't enough time to compute it. Making a real world computer that is Turing Complete into a true Turing Machine is only a matter of granting time and memory which doesn't change the logic of the machine at all.

As for problems that a Turing Machine cannot compute, that in no way invalidates the concept. There is an entire class of problems in mathematics known as "intractable problems" as they do not admit of an answer. That can be for various reasons but often it is simply because the known algorithms do not converge on an answer fast enough. A good example is the general case of the n-body problem where n>2. A method of calculating a solution is known but cannot be computed for all values of n as the universe will be in heat death before the algorithm converges sufficiently.

lazlo
11-17-2007, 09:29 AM
Here's an Enigma simulator:

http://users.telenet.be/d.rijmenants/en/enigmasim.htm

Hey, that's pretty cool Bob! I'm off to 9 hours of TIG welding, but I'll play with that when I get home.

BobWarfield
11-17-2007, 03:47 PM
There is a difference between a Turing Machine and a machine that is Turing Complete. We aren't talking about a Turing Machine when we speak of a Turing Complete computer since it cannot of course have infinite capacity. However, if infinite capacity were magically granted then a Turing Complete computer can be a Turing Machine. While the Turing Machine is a mathematical abstraction it isn't very abstract. It can be easily implemented in simulation on a Turing Complete computer.

As the link you provided points out, there has been no better definition of a computer (computability) proposed as yet than the notion of Turing Completeness. The requirement for infinite memory and infinite time to solve a problem isn't a logical requirement but a merely practical one. It ensures that a problem won't be rejected because there isn't enough time to compute it. Making a real world computer that is Turing Complete into a true Turing Machine is only a matter of granting time and memory which doesn't change the logic of the machine at all.

As for problems that a Turing Machine cannot compute, that in no way invalidates the concept. There is an entire class of problems in mathematics known as "intractable problems" as they do not admit of an answer. That can be for various reasons but often it is simply because the known algorithms do not converge on an answer fast enough. A good example is the general case of the n-body problem where n>2. A method of calculating a solution is known but cannot be computed for all values of n as the universe will be in heat death before the algorithm converges sufficiently.

Evan, I'm not sure what point you're trying to make, but with due respect, "Turing Complete" simply means that whatever language you're considering can solve any problem a Turing machine can. In other words, the language is isomorphic to a Turing machine and there is no difference between a Turing Machine and a Turing Complete computer. Because of that, such languages (or theoretical computers claimed to be Turing Complete) are subject to the same criterion that the tape or memory must be infinite for the automata to be Turing Complete.

Machines that can perform all of the Turing operations but with finite memory are not Turing Complete nor Turing Machines, they are Finite State Automata. You can satisfy yourself on that very easily by looking at any computer science text on automata theory.

Leaving aside that any "Turing Complete" computer must have infinite memory, when you say:

"Making a real world computer that is Turing Complete into a true Turing Machine is only a matter of granting time and memory which doesn't change the logic of the machine at all."

You are incorrect. There is a world of difference when the finiteness constraint comes into play. You should know from elementary calculus alone that taking things to their limit at infinity produces interesting and often counterintuitive results. Here, for example, are a list of problems that are decideable by finite state machines, but not Turing machines:

http://www.cse.ohio-state.edu/~gurari/theory-bk/theory-bk-twose6.html

In any case, this is decidedly off the track even for CNC enthusiasts let alone folks more concerned with making chips of the metal kind than with 1's, 0's, and the decideability of abstract computational automata, so I won't press it further. I wouldn't have brought it up in the first place except that there seemed rather a lot of pissing about on what the first computer was based on which one was a Turing Machine. I say again, none of them are for the reasons I've mentioned.

Cheers,

BW

dp
11-17-2007, 04:10 PM
There is a difference between a Turing Machine and a machine that is Turing Complete. We aren't talking about a Turing Machine when we speak of a Turing Complete computer since it cannot of course have infinite capacity. However, if infinite capacity were magically granted then a Turing Complete computer can be a Turing Machine. While the Turing Machine is a mathematical abstraction it isn't very abstract. It can be easily implemented in simulation on a Turing Complete computer.

The ability of ring lasers to hold boundless amounts of data could solve the infinity problem. Building arrays of ring lasers allows for scaling required by the limitations of the electronics involved. Pragmatically, any data source that is infinite in size represents a lot of data at risk should there be catastrophic failure. Makes you wonder how much data is contained in all the fiber interconnects at any one time. Or the amount of data contained in the miles between Earth and Mars. There's quite a large universe out there to put things in.

Evan
11-17-2007, 05:17 PM
"Boundless" and infinity aren't the same. The universe may be boundless but not infinite if it is closed.

As I stated Bob, the reason for infinite memory in a Turing Machine is to deal with infinite problems so that they are not excluded from the limit case of what the machine can compute. The fact is that no such machine is possible but the logic used to make a "limited" Turing Machine is the same as one with infinite memory. So, in theory a real Turing Machine can compute a solution to the n=3 n-body orbital problem but it will take infinite memory to do so using the Sundman solution since the number of terms increases exponentially.


"Turing Complete" simply means that whatever language you're considering can solve any problem a Turing machine can. In other words, the language is isomorphic to a Turing machine and there is no difference between a Turing Machine and a Turing Complete computer.

Sure there is. The abstract Turing Machine is considered to have infinite memory (and time). That's a rather significant difference. However, logically, if you insist that there is no difference then we have a consistency problem in computer science where machines and languages are described as being or not being Turing Complete.

BobWarfield
11-17-2007, 07:54 PM
Sure there is. The abstract Turing Machine is considered to have infinite memory (and time). That's a rather significant difference. However, logically, if you insist that there is no difference then we have a consistency problem in computer science where machines and languages are described as being or not being Turing Complete.

Evan, you are still laboring under the delusion that something can be Turing Complete with finite memory. It isn't so. There is no definition of Turing Complete that describes it that way. "Turing Complete" is used to describe other formal languages that are as powerful as Turing Machines but that have a different "instruction set".

Languages are abstract and don't have a notion of memory. The particular implementation of a language has the memory, but implementations are not described as Turing Complete, only languages. Hence it is fine and well to describe a language as Turing Complete and have an implementation that is not Turing Complete.

Similarly, machines are seldom spoken about as Turing Complete unless they are abstract architectures too. It is understood by convention that one can refer to a machine as Turing Complete with finite memory, but that it isn't really Turing Complete, it only would be if it had infinite memory. The sticklers among computer scientists just don't like to call real machines Turing Complete. Even the Wikipedia calls this out very clearly.

If you do in fact understand what a finite state automaton is, and can write one in the language of your choice (they're very easy), it shouldn't be hard for you to see why they are wonderful simulators of the instruction set of any real machine.

I hope that helps enhance your understanding of these terms. If not, you shouldn't worry about it. Computability theory is unlikely to be critical for a home machine shop in the nether regions of Canada. It's not awfully clear how critical it is right here in sunny Silly-con Valley where I am.

Cheers,

BW

Evan
11-17-2007, 08:52 PM
Evan, you are still laboring under the delusion that something can be Turing Complete with finite memory.
No, I am not. The machine may be described as Turing complete if it could function as a Turing Machine given infinite memory. And, in fact, that is how the computer industry describes computers that are "Turing Complete". The issue of infinite memory does not change what it is to be a Turing Machine, it only makes the problem set more inclusive.



A Mathematical Definition of Computation

...

The label Turing Complete is used to describe a computer or computing system which is as powerful but no more powerful than a Turing Machine. It's easy to prove that all modern computers are Turing complete. This means not only that you can use a Turing machine to do simple arithmetic or calculate pi to n places but that you could simulate a word processor, spreadsheet or video game on a Turing machine. The input would need to be converted to 0's and 1's on the tape and the output would need to be converted to letters or graphics but the "computational essence" of the problem could be done by the Turing Machine.

http://www.cs.unm.edu/~saia/computability.html (http://www.cs.unm.edu/%7Esaia/computability.html)

There are numerous similar references available.

JRouche
11-17-2007, 11:28 PM
Ahh? Oh, Im lost.. lol But is fun to watch the various computer geeks (again, a compliment, not a neg. description) we have here bantering back and forth about a subject that will have most folks baffled..

Computers? Hmmm? We all have them and they preform flawlessly, yeah right... It is absolutely amazing where we have come though.

Simple folks such as myself have processing abilities at our finger tips that could have ruled the world fifty years ago. Fifty years ago is in most of our living history.

We have graphical user interfaces to "talk" with our boxes. The GUI is the most important design to come around, in all of our lives.

The GUI has impacted the world more than anything else in our time, anything... JRouche

tony ennis
11-18-2007, 12:21 AM
The GUI has impacted the world more than anything else in our time, anything

I'd sooner say that the internet holds that honor. Though it is much easier to enjoy if you have a GUI. I've done a lot of webbing on pure text, however. HTTP and HTML made the internet more accessible to the masses. I've used a pure text web browser. Not as much fun as Firefox to be sure, and less useful as time has passed.

dp
11-18-2007, 12:29 AM
"Boundless" and infinity aren't the same. The universe may be boundless but not infinite if it is closed.

Customer: Jack - We're building a little Turing Complete system over here to do some word processing and maybe solve the mystery of the universe and I'm going to need infinite storage. Wha' d'ya have I can use?

Jack: I don't have infinite storage. I have boundless storage.

Customer: What's the difference?

Jack: I don't know - it's different. I read it on the Internet.

Customer: Must be true then. Will I ever run out of storage with your boundless stuff?

Jack: Nope. It's boundless.

Customer: Close enough - send one over.

Jack: You're going to need two of them.

Customer: Two? What the hell, you said I'd never run out of storage!

Jack: You need a second one for a backup. You can't put an infinite amount of data at risk by running without a good backup.

Customer: That's why I call you first, Jack.

Evan
11-18-2007, 03:44 AM
:D

Really, boundless and infinite are different. If the universe is closed it has no boundaries but has finite volume even though it is expanding.

[rdit]

Simple example: The surface of a sphere is boundless. It is also finite.

dp
11-18-2007, 03:50 AM
:D

Really, boundless and infinite are different. If the universe is closed it has no boundaries but has finite volume even though it is expanding.

[rdit]

Simple example: The surface of a sphere is boundless. It is also finite.

Yup - I know all that. But I was caught up in the humor of it :). Constrained infinities exist all around us. For example, a guy with an infinite (or boundless) number of bb's and a bb gun can hit an infinite number of targets, but none further way than the range of the bb gun allows.

wierdscience
11-18-2007, 01:16 PM
Yup - I know all that. But I was caught up in the humor of it :). Constrained infinities exist all around us. For example, a guy with an infinite (or boundless) number of bb's and a bb gun can hit an infinite number of targets, but none further way than the range of the bb gun allows.

When I was in grade school we had a science teacher who siad that if a person took a BB gun and shot at opposite sides of the Empire state building and timed each shot to the buildings recoil if enough shots were made the building would be pushed beyond it's maximum recovery and fall over.

I said it was BS and no way would the building fall,he asked why I thought that,"simple" I said,"how is it going to fall over with that big freaking pile of BB's on either side"?:D