Bill Kelly

[Jan. 2001] My feelings about dealings with XSLT. (With apologies to MontyPython)

Strange languages, lying about in XML, distributing tags, is no basis for a system of document conversion. Supreme executive power derives from a mandate from Larry Wall, not from some farcical W3 consortium ceremony. I mean, if I went round, saying I was Emperor of document conversion, just 'cause some moistened bint threw a recursively scoped syntactically self-embedded translation language at me, they'd put me away!

[Dec. 2000] I'm learning Smalltalk, finally! My primary languages at this point are, Forth, C, C++, Perl, and Java (in no particular order . . . well, that's the order I learned them actually :) Here are some observations or thoughts I've had about Smalltalk so far:

And my most-recurring thought about Smalltalk so far:

Long ago, on a computer terminal while visiting my Dad at work...

	You are standing at the end of a road before a small brick building.
	Around you is a forest.  A small stream flows out of the building and
	down a gully.

Somehow it's remained one of my favorites. I still recall that Saturday afternoon, heat waves rising off the parking lot, and the transition through smoky glassed doors into the air conditioned office building. A hardware R&D lab and more-like hangar, inside it had been partitioned with high walls stopping only half way to the double high ceiling; a dangling network of wires and fluorescent lighting occupied the spaces above. Corridors and surfaces lay inhabited by electronic equipment or piled with bare circuitry. And my Dad, clearing a few obstacles and seating me at a terminal, typed an inscrutable sequence of commands: then;

	Welcome to Adventure!

Would you like instructions? ?

My eight- or nine-year-old mind, making the leap into an alternate universe... I'm pretty sure a part of me's been there ever since. :)

I'm currently programming for Encyclopaedia Britannica working on our CD-ROM products; for about nine years preceding I did games programming; and various projects before that, to my first steps in programming in about 1982.

In my spare time I'm working on QwikWiki and recently hacked up a for-fun Forth compiler in Perl (

My current methodological leanings are more toward WorseIsBetter than the RightThing, although I think not-100% as correctness is still a closely held tenet for me - especially where incorrectness would lead to crashing or flaky software behavior.

I have a recurring obsession--and when it occurs it carries a feeling of some certainty--the feasibility of devising some form of a robust, general machine learning intelligence.

Update, Sep. 2000: I feel like I could be nearing an actual implementable concept for the AI. It's weird to be getting closer to something, something that is becoming sufficiently defined it would seem I could nearly implement it, and that it would even exceed my original expectations; and yet honestly have no idea whether it's even possible, let alone workable and emergent out of the particular configuration I'm converging on. Torture. Beautiful revelations. --- . . .or pipe dreams?

Wiki pages I've initiated: AutoVivification, LatticeOfCoincidence, QwikWiki

Wiki pages I've enjoyed: many, many, many! But to name a few offhand: DoTheSimplestThingThatCouldPossiblyWork, YouArentGonnaNeedIt, RefactorMercilessly, BigOmega

used to be: (foosenblat wiki)


View edit of February 25, 2005 or FindPage with title or text search