"Anyone alive in 1958 could understand the programming languages we have today. Physicists would be surprised by what we know in physics today. That's bad."

A little exaggeration there. I think there would be at least a few that couldn't. But as programming languages become more abstract and expressive, they become easier to understand anyway, so I'm not sure what he's talking about.

I agree that the Internet is one of these ULSs - it also has all the characteristics that seem to strike this guy as so hard to achieve. It's too complex for any single person to understand, it's large, it's modular and it's fault tolerant.

IMHO, what we're likely to see is a lot of small devices in the future. It doesn't seem unlikely that these will be in constant or near constant contact with a wireless network; I suspect the real ultra-large systems will be more like (legitimate) botnets running on powerful mobile devices than data centres or mainframes.

Hmmm, I'm not convinced by this: where's the commercial imperative - how do people make money from all this investment.

What this guy seems to have done is take the rate of growth over the past 10 years and extrapolate - always a dodgy technique and usually comes up with ludicrous results. (E.g. extrapolate transport development in 1940's-50' and we'd all have hypersonic flying belts by now.)

He's also flat wrong about programming languages. While I admit I was taught FORTRAN at university and procedural languages are widely used ('cos they work! ), computer scientists from the 1958 would be completely bamboozled by SQL. It's like saying Old man "Benz" would recognise todays cars.

Finally, when did anyone ever "read" porn?

Could anyone care to give an example of what this unreliable and constantly failing system which is beyond human understanding would actually do? Provide highly unreliable distributed DNA cracking? Or a constantly failing high-speed communication network? Or generating the two pages of Java for every person on the planet? Work as a precompiler for the not-yet-invented programming language? Maybe even manage all of California's traffic lights? Work for the IRS?

Coor blimey, the possibilities are endless!

Not much crystal, but an awful lot of balls

amanfromMars's day job. Either that or they have the same dealer/ prescription.

"Anyone alive in 1958 could understand the programming languages we have today.

Oh yeah? Throw a few lines of Perl at them and see how they like it. Or even relatively simple C++. Or even do the same to 80-90% of people alive today and watch the results.

The plain fact is that *anyone* can understand *anything*, given sufficient training and experience. Physicists from 1958 would *not* be surprised at what we know today, because knowledge of facts does not equal improvement in intelligence. A physicist from 1958 would be perfectly able to operate in 2007, given training to cover the intervening 49 years.

And in fact software *has* advanced significantly. The primitive software of the 1950s was in an environment where operations and RAM were expensive compared to development time. Software today is in precisely the opposite environment, hence the development of OO, Java and more recently JIT-compiled semi-scripting languages like Python. Quite simply, this guy doesn't know crap about computing *or* physics.

You also have to ask what this network is actually going to *do*. He says it's going to be so much bigger that we can't grasp how big it'll be, but then he says that it's not going to be as functional as existing systems. So... errr... why build it then?

"computer scientists from the 1958 would be completely bamboozled by SQL."

...Which was developed in the early 70s.

And, I might add, is extremely simple. I was a procedural C programmer for ages and I got a decent handle on SQL in less than a week.

Computers are built on essentially the same concepts as they were in the 50s, with the same sort of data structures underneath. I think a developer from 1958 could be caught up with less than two years of study if he really wanted to go at it with an open mind, and he'd probably be better than you. Young programmers taught on PHP have no respect for silicon.

That said, I think the rest of the article is just rhetorical nonsense, including the notion that the lack of change is even a bad thing.

Please work in the following:

...Web X.0 (any number greater than 2)

... as a platform

... anthropomorphic

... open source

... orthogonal

... intrinsic

Nobody really knows what those mean, so they would be really useful in this article.