Sick Transit

michkap@MS links to an interesting rant by famed technical author Charles Petzold. Putting aside the main thrust of Petzold’s argument, about the pluses and minuses of RAD and IntelliSense, I found his concluding anecdote–about writing a program to solve a math puzzle–very revealing:

I think that those are the moments that programmers live for–at least, that we code for. The joy of pure problem-solving, unburdened by the messy constraints of the real world.

Unfortunately, most programming work is not like that. Most programming work involves the implementation of simple, well-known concepts; the complexity comes from the need to interface to existing systems, map real-world data to logical concepts, etc. And a lot of that is, quite simply, bull work.

Which is precisely why we have a plethora of APIs, and RAD and frameworks: to do as much of this bull work as possible for us. And so, despite Petzold’s nostalgia, the frameworks allow us to spend more time on “real programming”–analysis and problem-solving–than if we had to reimplement the wheel, over and over again.

Am I the only one who saw the name Americans for Better Justice and thought of “For Great Justice”? All your nominee are belong to us!

(There’s no meme like an old meme.)

Tyler Cown discusses possibilities for speeding up airplane boarding:

In my experience, Southwest actually takes longer to board, or at least it felt that way to me. The first people to board take the seats up front, which is the exact opposite of the most efficient boarding order (last row first). Instead of aisle traffic moving in a single direction, people move back and forth, looking for the best seat they can find. And then people who are already seated have to move to allow families with children to sit together. Personally, I refuse to fly Southwest unless it’s a major savings, just because the boarding process is so atrocious.

The charge per carry-on sounds like a reasonable idea, although it would have to be phrased as a discount rather than a charge. People hate being charged for what they’re accustomed to receiving for free.

And the “race for the rebate” would create an utter disaster. At least right now, most people are polite and orderly when boarding. If boarding were a race with a prize, that detente could easily break down.

Aside from the window-middle-aisle system which Tyler mentions, the best idea I can think of is to implement last-row-first order more granularly. Instead of ordering by large blocks of rows, as is currently done, order by small blocks or even single rows. This would require more co-ordination, however, especially since not all passengers are at the gate when boarding starts. So I’m not sure if it would actually be a gain.

…when you’re smiling in your mug shot.

(Via Slate’s Chatterbox)

Well I’m a bit late to the party, but for any readers who aren’t personal friends whom I’ve already told, it seems I got a job at Microsoft. I’m going to be an SDET in the Real-Time Communications Group (part of the Office division, responsible for Live Meeting, Communicator, and other fun stuff), developing tools for internal use.

I start at Microsoft on November 14th, and will be relocating to the Seattle area on some as-yet undetermined date before then.

My complete assimilation approaches.

A couple of months ago, I wrote about custodians at NorthWest airlines striking to protest wage cuts. I had to side with the employers in that dispute, because the custodians were making $40k a year.

This week in Slate, Daniel Gross tells a harrowing union-busting tale of workers being asked to take a 2/3 pay cut. Which sounds unbelievable, until Gross explains that union workers at auto-parts manufacturer Delphi were making $65 an hour.

Gross mourns that the suggested new wage of $20/hour will drop the workers out of the middle class, and calls this the “re-proletarianization of industrial work”. But an annual income of $41,600 (not counting overtime) is fairly close to the national median household income ($43,318 in 2003). It is a little above the median income for a male earner, and well above the median for a female.

Lest anyone think me unsympathetic to the plight of laborers, I should say that something similar might equally well happen to programmers in my lifetime. Outsourcing has already struck the profession, of course; and sometime in the next few decades, AI may well replace most ordinary programming positions, just as robotics took over industrial labor.

But as painful as a drastic cut in income undoubtedly is, the Delphi workers are being brought down to the national median. They are not falling out of the middle class.

Via Drug WarRant, a libertarian judge and former prosecutor James Gray makes a strong point:

Emphasis on our drug money. The Colombian mafia’s billions ultimately come from American junkies. So drug prohibition not only harms us, but exports our problem to countries even less able to handle it.

The Cold War is long over, and (pace Iraq) so is its policy of using developing nations as pawns in the superpowers’ games. But we are still in reckless disregard of the effect our policies have on poorer nations.

I’m a little late to the post, but I just have to link to Publius’ discussion of why the hawks vs doves dichotomy is bad framing:

Jan Haugland writes about Darfur, and I can’t do better than to quote him:

Of course, the United States is not responsible for the whole world’s failures. But one of the greatest tragedies of the Iraq war is the lost opportunity of Darfur.

A couple of new studies are out on pot and driving:

In other words, this doesn’t mean it’s safe to drive while stoned. But it does show that pot-smokers tend to be responsible about not driving while impaired. It also suggest that marijuana use does not significantly impact day-to-day motor skills.

To investigate the issue of impairment, the authors suggest a follow-up study based on blood rather than urine samples. However, another panel (not identified in the article) found the following:

In other words, if you’re going to smoke pot, wait for a couple of hours before driving, just as you would if you were drinking moderately.

I’m sure it will take further research to establish these conclusions more solidly, but it’s nice to see the question being approached scientifically, instead of through the lens of anti-drug hysteria.

(link via Drug War Rant)

Tyler Cowen makes an important point about immigration:

It follows that new immigrants are more likely to be illiterate and to have an even harder time integrating, and that more open borders might actually help alleviate the problem. If that seems counterintuitive, read the whole post.