This page is a collection of links to various bits and pieces of text that I found interesting. I don't think there's any grand scheme here, unless it is the study of what I find interesting.

Some of the documents may have imperfect formatting; they may have been saved from a web site, but saved without external image files or style files needed to complete their appearance.

Web testing

A simple test of using character reference codes (#DDD;) for represting non-ASCII characters in HTML.
I'd like to learn how browsers treat favicon. This is part of unfinished research into the topic.
Dealing with flash
This is a document describing how to suppress Flash in Mozilla (and Firefox).
font tests
A nuisance of the web is that fonts are not portable across platforms. (A secondary nuisance is that Microsoft keeps releasing new fonts for Windows, which half the world adopts, and then shrugs at you when you report that their pages look bad. e.g: first Arial, then Verdana/Tahoma.) This page tests the appearance of some fonts.
Harrison Bergeron
Vonnegut wrote a good story called "Harrison Bergeron". I use it here as part of a test of curly-quotes and cascading style sheets.
Monitor gamma information
Information about monitor gamma, which can affect why web page images sometimes look brighter or darker as you switch machines. Not my text—this is copied from the web.
Finding your monitor's gamma
Another chart that you can use to estimate your monitor's gamma setting. Not my text—this is copied from the web.
Greek Alphabet
The magic of Unicode lets you make web pages that contain true Greek text, instead of containing English text that is displayed with the Symbol font. These characters are hard to type on an American keyboard, so I have a web page that allows me to copy them from the page and paste them into documents where I need them. As a bonus, I can look at their appearance in various fonts.
ISO 8601 dates
Dates written in the format "NN/NN/NN" (for example, "01/02/03") are ambiguous, because different countries have different interpretations about the meaning of each of the N's. ISO 8601 dates avoid this problem by writing dates in an unambiguous fashion.
Unfriendly JavaScript and email
In 2001 I got upset about web pages that could send email to a site, based only the fact that I had visited a web page. This exposed your email address to that company, which led to extra spam.
The battle against spam is lost, and the spammers have won, and I don't think I've worried about this issue for a while.
Old-time computer fans have a love-hate-hate relationship with Real, the company that makes the RealPlayer video and audio players. In early 2004, the radio show Car Talk dumped RealPlayer for Microsoft Windows Media Player, because of their poor experience with Real. I empathize, and wrote a small blurb.
server-side includes
MIT's web servers now support server-side includes.
If the question is "can you type a nice TeX logo in HTML?", the answer is "it depends".
Typesetting math on the web is tough. One way that was invented to do it used a Java applet as a helper. I have an older page testing this method, and also a slightly newer page. Sadly, bit rot appears to be setting in.
Palatino fonts
Ick. Most computers have a Palatino font (or equivalent), and it is one of the core fonts for Postscript printers. However, on Unix using FontConfig, Palatino looks wretched.
Web-safe fonts
This document is incomplete, but I'm testing web fonts to see what recommended practice is for using web fonts that work on various viewing platforms.
A List Apart (CSS)
I found an article on ALA that described setting up lists using CSS, and I thought it was interesting enough to try to keep it around.
xhtml, and why not to use it
Mark Pilgrim is a big fan of standards... and he's decided that xhtml isn't worth working with. After flirting with it (twice), he converted his site back to html 4.01. These articles show some of his reasoning.

Calvin and Hobbes

Watterson (of Calvin&Hobbes) commencement speech
A speech Bill Watterson gave at Kenyon College, Gambier Ohio, to the 1990 graduating class.


Sometime March 2000, I read an article about voting methods in which the author argued that simple voting techniques do a poor job of choosing winners that the majority of the voters are really happy with. I thought it was interesting, and wrote a small blurb about it.

Since then, I've become less convinved that there is a magic bullet. There is a website at that advocates other methods of voting, but in doing so give a good summary of the problems that can be encountered in traditional voting. If you're interested in this type of thing, you're probably already heard of Instand Runoff Voting (IRV), Approval Voting, Borda's method, and other terms.


When I was shopping for computers for myself and for my lab, I put together a list of advice to myself. The advice is from 1997, and is dated by now.


In 1999, Don Knuth gave some lectures at MIT. I wrote a small page, recording session titles, and linking to some lecture archives. Links on the web decay over time, so I'm not sure how many of the page's links still work.

During Knuth's talk, he mentioned an odd parable by Smullyan. I'm not sure I agree with it.

Legal: Wen Ho Lee

Wen Ho Lee of Los Alamos was accused of espionage. The government appears to have overplayed its hand, and was accused of being desperate to find a scapegoat in order to please public opinion. In the end, Lee was convicted of not following proper procedures when handling classified data, and all other charges were dismissed. The judge made the remarkable step of apologizing to Lee for the behavior of the government. Here is the transcript of the final hearing of the case.


A parallel computing class I took at MIT required that we do a final project. Part of my project involved learning how to use a program that would process three-dimensional data and turn it into images. My writeup is here.