Wednesday, September 30, 2009

Clojure

Clojure is an incredible achievement.

I've been using it for three weeks now, and as far as I'm concerned it's the new lisp that we've been promised for so long, only much better. Lisp as McCarthy, Steele, and Sussman would have done it if they'd had the computers and the perspective we have.

A lisp where vectors and sets and maps are as important and as usable and behave in the same way as lists, and in fact where if you're not concerned about speed you actually may not be able to remember
which you're using.

A lisp where the code looks like python with brackets.

As well as that, it has a whole new paradigm for concurrency, which I've started to play with but wouldn't claim to understand, but which looks really good.

The third big feature is its interoperability with Java, and the huge libraries that this brings along for free.

At the moment, you can keep all that. Last time I tried to use a Java library from Java itself, for xml parsing, I gave up in disgust and went and rewrote my program in python because learning how python does it, and rewriting the whole program, was easier than learning how Java does it.

But I am told that this is a common reaction from people coming at clojure from lisp-y directions, and that actually after a few months, they regard it as a major strength.

And I have a friend who is a programmer doing an in-house system, who already has a great heap of Java to work with, who has taken up clojure entirely because it's a much easier way to write Java, and who doesn't really get the whole macros and code-is-data and functional programming thing yet.

And we have him now....

Well done Rich Hickey!

Beyond Satire

This is a very thought provoking article about the benefits to society from the collapse of large
software projects:

http://www.berglas.org/Articles/ImportantThatSoftwareFails/ImportantThatSoftwareFails.html

I literally can't tell whether it's meant seriously or not. And obviously I am too lazy to check his facts in even the most cursory way.

The thesis is that the size of the Australian Tax Office is politically determined at about 1% of GDP,
and that the vast office productivity gains that have occurred since the 1960s have been entirely swallowed up by expanding regulation.

There's a lovely statistic about productivity in kilopages of tax regulations administered per unit GDP which shows that the ability of the tax office to administer regulation has gone through the roof while the number of people working there has stayed constant.

The article ties in quite nicely with my theory that the bewildering complexity of tax laws may be the result of an arms race between the Government and the population.

If we can also assume that computerization makes private accountants more productive, which is to say that more real citizens can use their services, then there will be more money flowing through the loopholes that accountants will find, hence more regulations and complexity to close the loopholes.

There is a monster here that feeds on itself, absorbing a reasonably fixed percentage of everybody's effort in activities that nobody enjoys.

I wonder if there is a way to tame it?

Or perhaps we should just marvel at the beauty of it all, as we do with lions and antelopes.

Tuesday, September 29, 2009

I like walls.

Just now I saw an advert for Windows in the Times. Leaving apart the question of why on earth Microsoft has to advertise, I'm not sure that "Windows. Life without walls." is the best strapline.

I can't remember who first said it, but:

In a world without walls, who needs windows and gates?

Besides, what's wrong with walls? I like walls. The walls of my house, for instance. Without them I would be cold, and I would have no things. And probably lots of viruses.

Tuesday, September 22, 2009

Extraordinary Foresight

This is one of the most impressive articles I've ever read:

http://www.joelonsoftware.com/items/2007/09/18.html

Admittedly it's an argument by analogy, but it's a damned good analogy.
Lisp/Paul Graham fans will find themselves laughing towards the end, but it's not really about either.

However, it's two years old now. I can't see what's gone wrong with the argument. Where's NewSDK?

Tuesday, September 8, 2009

Mavis Batey at Bletchley

From the Times of Monday September 7th 2009

At Bletchley she met her future husband, Keith, a Cambridge mathematician. Their romance did not get off to a good start. "I dropped my pencil to see if he would pick it up. He said, 'You've dropped your pencil'."

Followers