A little computer science

There's been lots of talk about synchronisation around DoC recently. First, Benjamin Pierce talked about Harmony, a framework for bi-directional programming. Take any source tree or relation - maybe an address book - and compute a simplified view of it - maybe just names and phone numbers. Edit the data in the view. Now synchronise the view's data with the original tree or relation - that's the bi-directional bit. It's one-part schema integration, one-part regular typing and one-part formal reasoning about the properties of the synchronisation. Then, Gilad Bracha talked about "Objects as Software Services". This extremely interesting idea observes that hotswapping of components / services is made possible in modern, network-aware applications by synchronising code at the same time as user data. (Think of that address book again.) Lotus Notes was famous for (amongst, ahem, other things[*]) its replication of databases that contained code and data; it made distributed content management as easy as pi(e).

[*] I'm playing devil's advocate. That site is way out of date. There have been more recent attacks of Notes, entirely wrong-headed as always. The answer is Notes; now what was your question?

Moving on ... Erik Meijer has an excellent walk-through of how VB9 is going to be the world's most popular functional programming language in "Confessions of a Used Programming Language Salesman (Getting the Masses Hooked on Haskell)". It says:

To appreciate the difficulties of XML let us now make a short excursion to the world of XML schema, perhaps the best example of a south-pointing chariot in computer science. XSD must be one of the most complex artifacts invented by mankind, where simply using DTDs (or a compass instead of the chariot) would be a perfectly fine solution.

What on earth is a south-pointing chariot, we wondered ... the answer is here. Now it makes sense!

The RISKS digest today has this interesting snippet:

From: ches@cheswick.com (ches)
Subject: Re: Triple DES Upgrades (Redspin, RISKS-24.26)
> In stereotypical and steadfastly arrogant fashion,
> USA banks are refusing to move to chip-and-PIN
I have heard briefings from highly-placed people at both MC and Visa discussing this. They are steadfast and firm: chips will not be implemented in the US credit cards. There is insufficient justification for the expense, given the cheap modems and phone system available to retail outlets.

And finally ... a TIME article about Bill Clinton's work on removing unhealthy food and drink from US schools contains a remarkable factoid: "Five years removed from the Oval Office, he is 10 years younger than Ronald Reagan was when he entered it."