So who’s responsible for the torture and abuse in Iraq?

I watched Donald Rumsfeld’s Congressional performance and wondered where the buck really stops for the ill-treatment of Iraqi prisoners. Rumsfeld and Bush express horror about how the abuses are profoundly un-American. There’s a lot of talk about “bad apples”, etc. But two things published on the Net make one wonder.

The first is a remarkable New York Times account of the circumstances under which untrained reservists came to be in charge of a teeming prisoner population under unspeakable conditions. Why is this so revealing? Well, it vividly shows the extent to which Rumsfeld & Co did absolutely no planning for dealing with Iraq after the invasion and the defeat of Saddam’s regime.

As to the “bad apples” argument, well it doesn’t wash either. It’s a terrible truth about human beings that whenever some people have absolute power over other human beings, (whether in concentration camps, children’s homes, old people’s care centres, prisons) some of them behave badly, and some do truly terrible things. So any civilised country takes great pains to ensure that this kind of unaccountable power is exercised as little as possible by its forces. No pains were taken by Rumsfeld & Co.

The second salutary item on the Net is a sequence taken from the video-recording system of a US Apache helicopter operating in Iraq. If you have a strong stomach, you can find the MPEG here. If you don’t want to watch it, the gist is this: it’s night-time, so all the action is shot through night-vision lenses; there are three Iraqis on the ground, near a truck which may contain a Stinger anti-aircraft missile; as the chopper approaches, the Iraqis run in various directions. One by one they are picked off with the precision of a video game. The last survivor chooses to hide under the truck. The gunner fires at the truck, and vaporises it — and presumeably the Iraqi. But then a figure appears, on the ground, crawling away from the wreckage. The gunner is ordered to kill him. And he does — poof!

What’s interesting about this is not so much the growing similarity between real and virtual warfare, but the way overwhelming force is used to kill a wounded soldier, who was clearly beyond posing a threat to the helicopter. And what this says about the mindset of US forces in Iraq.

Through a glass brightly

Saw this hanging in the porch of a friend’s house yesterday.

Things to bear in mind when giving Commencement Addresses

One of the dubious pleasures of middle age is that your Alma Mater may decide you’re now sufficiently old/prominent/rich (delete as appropriate) to be invited back to give an Address on Graduation Day. I know — it’s happened to me. The temptation to give a pompous sermon to a captive audience is difficult to resist. (I’m not sure I succeeded.) In the meantime, here are some useful guidelines.

“Banana Republic has always been a store, not a puppet government in Latin America.

The statement ‘You sound like a broken record’ means nothing to them.

They do not have a clue how to use a typewriter.

They’ve never heard, ‘Where’s the beef?’

Paul Newman has always made salad dressing.

Michael Jackson has always been white.”

My favourite Commencement Address is still Woody Allen’s. It begins:

“Today we are at a crossroads. One road leads to hopelessness and despair; the other, to total extinction. Let us pray we choose wisely.”

Memo to CIOs: remember to count cost of security patches and repairing malware damage when computing TCO of Windows-based systems

One plank in the Microsoft hymn-sheet against Open Source software is that the ‘Total Cost of Ownership’ (TCO) is more important than the initial purchase price. The argument is that companies should not be distracted by the low initial cost of free software, but should add in the costs of conversion, support, etc. So indeed they should. But one thing that is consistently ignored in computing the TCO of a Microsoft system is the cost of coping with the security vulnerabilities of the software. All of which makes an item from Good Morning Silicon valley about Gartner Research’s views on Microsoft (in)security very interesting indeed:

“It’s never been cheap to run a Windows shop. Host intrusion detection. Scalable antivirus protection. Patch management. All these things are costly, especially given the amount of malware that finds its way into the wild these days. So it’s entirely likely that administrators around the world will respond to Gartner’s announcement that vulnerabilities in Windows raise the total cost of using the OS with a collective ‘no —-, Sherlock.’ But corporate types, who after all make up much of Gartner’s core audience, may sit up and take notice. And if we’re lucky, they might realize that turning on Windows’ automatic update feature doesn’t make you immune to worms like Sasser and that one can rarely budget too much for additional security technology….”.

The relevant quote from the Gartner source reads…:

“Dealing with widespread worms like Sasser raises the cost of using Windows, a research analyst said Wednesday.

Mark Nicolett, research director at Gartner, recommended that enterprises boost spending on patch management and intrusion prevention software to keep ahead of worms, which are appearing ever sooner after vulnerabilities in Windows are disclosed.

‘This is part of the carrying cost of using Windows,’ said Nicolett. ‘The cost of a Windows environment has gone up because enterprises have to install security patches very rapidly, deal with outages caused by secondary problems with these patches, and deploy additional layers of security technology.’

Although he placed some caveats on his numbers, Nicolett said that informal surveys with Gartner clients indicate that simply moving from a no rapid patch deployment capability to an ongoing process that can respond quickly to vulnerabilities raises the cost of using business by about 15 percent.

Nicolett’s advice stemmed from the recent outbreak of the Sasser worm, which began striking Windows systems last Friday and has infected a large number of machines world-wide, with estimates ranging from 100,000 to well into the millions.”

Graphic beauty

I love Danny Gregory’s Blog, and envy him his talent. This is a beautiful image from an ingenious group project.

“Take a page, divide it into thirty squares, then do a drawing each day in one of the squares. After a month, it will be filled with a rich quilt of art. No matter how lame any one of the drawings is, the overall result will be beautiful. As the month ended, some of the participants have been uploading their work. It’s very interesting and inspired me.” Me too. I’m going to try this with thumbnail photographs.

iPod mania

Never ceases to amaze me. iPod Lounge has been encouraging people to send in photographs of their iPods. Some, like this, are deliciously wacky.

Intrinsic vulnerability of Linux?

Interesting paper by Dan O’Dowd arguing that objections to his assertion that Linux is unsafe for defense systems were based on (i) “dangerous misconceptions that it is equally easy for foreign intelligence agents or terrorists to infiltrate malicious code into any operating system” and (ii) “that the many eyes looking at the Linux source code will find any malicious code infiltrated into Linux”. In part, O’Dowd relies on the fact that UNIX co-author Ken Thompson showed many years ago that an open source process couldn’t find clever subversions, no matter how many people of whatever competence looked at the source code. O’Dowd is also claiming that the embedded Linux system sold by his company is not vulnerable in this way. He may be right (I hope he is), but embedded systems are of limited applicability.

Research and scholarship

Interesting conversation today with some academic colleagues. We were discussing the paradox that while it is an article of faith among academics that excellence in research is necessary if a university is to be excellent in teaching, the statistical evidence for this is thin (to put it mildly). So is this just an academic shibboleth (“a manner of speaking that is distinctive of a particular group of people” – www.cogsci.princeton.edu/cgi-bin/webwn)?

In thinking about it, it seemed to me that Thomas Kuhn’s concept of a ‘paradigm’ might be useful. A paradigm, in TK’s sense, is an agreed theoretical framework which characterises a mature academic discipline. Research (‘normal science’ in TK’s terms) takes place on the edges of the paradigm and consists of exploring anomalies between the paradigm and the real world. Teaching, in contrast, consists of articulating the paradigm in such a way that new generations of students can understand and absorb it.

Teaching and research require quite different abilities and skills. Many people who are good at research are poor at articulating the paradigm. (Of course there are some spectacular refutations of this — Richard Feynman is the one who comes immediately to mind, but I believe it to be a reasonable generalisation.) Conversely, many people who are excellent teachers are poor at ‘research’ of the peer-reviewed, pushing-back-the-frontiers kind.

So what is necessary for good teaching, if not research? The answer, I conjecture, might be scholarship, which I define as study which has the effect of deepening one’s understanding of a paradigm. To be a good teacher, one needs (i) a broad and deep understanding of one’s subject (the product of scholarship), together with (ii) sophisticated skills at imparting knowledge, (iii) empathy with learners (something talented researchers often lack) and (iv) the ability and inclination for self-reflection. My conclusion: to have excellent teaching it is more important for an institution to have scholarship than ‘research’. My question: what’s wrong with this conjecture?

Apostrophe rage

Even the New York Times is succumbing to SAS (Surplus Apostrophe Syndrome). One of its headlines yesterday read: “Worm Brings Down PC’s and Networks”. Growl.

Contrasting IPO styles

There’s a nice piece in today’s NYT ruminating on the coincidence that just as Google plans an auction as a way of launching its shares on the market, Frank Quattrone — a leading practitioner of the old, corrupt way of doing IPOs — was convicted of obstructing the course of justice. “Once, Wall Street considered it embarrassing to have a stock soar right after it went public”, says the Times, “because the underwriters had obviously left a lot of money on the table and deprived the issuing company of the best price for the shares sold.

But in the late 1990’s it became a badge of honor to have a new offering double or even quadruple the first day of trading. That meant the potential for phenomenal profits to those who could buy at the offering – profits that could be realized within minutes after the stock began trading, long before it became clear whether the company would prosper or fail.

Mr. Quattrone put himself at the center of that process. Moreover, despite nominal rules separating investment banking from research, Mr. Quattrone had analysts reporting to him, allowing him to influence their recommendations to clients.” Google wants to play by different rules. It proposes to sell its shares in a version of a Dutch auction. That means any investor – whether the best friend of the lead underwriter or a small investor should have an equal chance of buying shares at the offering price. The auction will set the price.

At the height of the Internet boom, Mr Quattrone’s cronies — those who had been allocated shares prior to the market launch — could make obscene profits on the first day of trading. The Times mentiones one of his triumphs — VA Linux which Quattrone’s group took public in December 1999. “The shares were priced at $30 in the offering, but traded for $320 the first day, ending that day at $239.25, for a gain of nearly 700 percent. By the end of 1999, the shares were trading for less than the first-day close, but were still viewed as a great success. (They later fell as low as 54 cents in 2002, and now sell for about $2.)”

The nice thing is that Google is not confining its innovation to technology.