Thursday, December 15, 2005

Origins of Civilization

Fred Reed, long time curmudgeon and columnist, has a web site where he posts his periodic rantings. Recently he went all down on civilization, opining that we should never have given up the hunter-gatherer lifestyle.
While I have some sympathy with the point, I realized there was something important he'd missed, so I shipped a letter off to him:

The advantage to agriculture, and what I think is the reason that the hunter gatherers settled down, was that you get to make beer. Beer needs an abundance of grain, which you're not going to get bashing about the countryside. Then you have to hang around while it is fermenting, which means you start building more permanent structures. You need proper containers for the fermenting and storage, so you have to go invent clay pots and glass containers. Most beer is better you need to develop deep cellars or refridgeration. This leads naturally to pubs and bars, and all of their support structure. A few more iterations, and voila, civilization.

It's all about the beer

This got the one line reply:

Damn. You have destroyed my entire argument....

That made my day. :-)

Friday, October 28, 2005

Air-drop Programming

Back in 1992 or so, in a desperate effort to convey to prospective employers just what it is I do well, I came up with the term "Air-drop Programming". The idea is you get parachuted into the middle of a code base, then fight your way out until the problems are fixed. It implies a total effort at dealing with the unfamiliar, without timidity. It really describes much of the maintenance programming experience, just not as boring.

I haven't been shy about using the term, but it never really caught on. I used it recently at OOPSLA and it looks like it was noticed! And some othe bloggers have picked it up as well Cool!

Thursday, August 18, 2005

Estimating passengers

Via Danny Ayers and David Weinberger , Julian Bond asks how many people are flying on airplanes at any point in time. He figures a million, a commenter gets 750,000, Danny figures 50,000 +/- an order of magnitude. Nobody shows their reasoning, making it impossible to converge on a decent answer as nobody can check each others work. So here is my answer out in the open.

Let's just consider 747 passengers. There are roughly 1,000 passenger 747s, with an average passenger capacity of 500 passengers each. That's 500,000 right there. Most of them are flying at any given point in time as the economics of airlines doesn't permit them to keep their valuable resources sitting on the ground. Subtract some percentage for flights not being full (International fill rate averages 71%, see below) and call it 25% on the ground. This gives 266,250 passengers in the air on 747s at a time. Multiply by an off the top of my head estimate of 3 (all other airplane types) to gives total filled airplane seats of around 800,000.

Looking at it another way with the FAA stats (US domestic for May 2005), there were 56e6 passengers on US airlines domestic flights only in May. Divide by 31 to give us 1.8e6 passengers/day. Assuming the number of US flights matches the rest of the world, that's 3.6e6 passengers/day. (This is probably low as China has lots of flights these days.) I can't seem to find flight duration information, so lets assume an average of 3 hours. That gives us 450,000 as the average instantaneous value.

Unfortunately IATA sells it's statistics, so we can't get much from them (unless someone wants to give me USD 800), but it appears the load factor in international flights is about 71%.

So it looks like 500,000 +/- an order of magnitude. (50 thousand to 5 million) But if I'm wrong we'll all know where I miscalculated.

Technorati Tags: