Al Hoang

File this under the I Lust One Dept.

The Register Link

The Inquirer reports that FEMA is only allowing users of the Microsoft Internet Explorer browser to apply for hurricane relief funds. “The now very much criticised US Federal Emergency Management Agency (FEMA) has stopped Mac and Linux victims of hurricane Katrina from applying for relief. The agency, which is already in hot water for its lack-lustre rescue efforts in New Orleans, has created a web-based service that only works for users of Windows and IE6.”

File this under the WTF Were They Thinking Dept.

Original Link

If you’ve not heard of Arc
from Paul Graham you probably don’t
follow programming language wanking that much. Philip Greenspun decided
to offer his thoughts on Arc:

No hopes for Arc, though Paul is a nice guy, unless someone writes a ground-breaking must-have open-source application in Arc that others want to modify. Lisp is more tasteful than C or Java but not that dramatic a help unless the programmer is very smart and tasteful him or herself. The average programmer seems to be getting progressively less devoted to taste and quality and therefore I would think that Lisp will continue to dwell in obscurity as things like PHP ascend to popularity.

To me
Arc sounds like an interesting
idea but if you take 100 years to design
the 100-year language
you’ll have to re-advertise again since all the people who first heard of
such a great idea will probably be in their graves. But if you have it
made financially what’s the
rush? Unless you’re trying to become ultra-famous…

Link to Phil Greenspun’s comments on Arc

Paolo Amoroso posed an interesting
question on the comp.ai newsgroup
(yes newsgroups still exist!) regarding what has happened to Expert Systems
since the AI Winter.
He got some very enlightening responses on what
has happened with these systems since their introduction a couple of decades
ago. Here is a choice quote from one of the replies:

In financial fraud detection, particularly the problem of finding
credit card transaction fraud, expert systems were an early candidate.
Based on experts’ experience, rules were posed which characterized
known fraud and the output on unseen data was used to figure out how
the rules would go wrong. These systems did as well as the state of
the art at the time which was considerably worse than the state of the
art now.

Go right to the discussion

Paolo’s blog post on Expert Systems

Paolo’s blog post on LISA and rule-based Programming