Copyright, Terms of Use, and Privacy

Original work made freely accessible on the web is still automatically copyrighted. But, at least within the U.S., you have the right of fair use: you don't need permission to make reasonable quotation, with attribution, for purposes of discussion. Fair use is a right, not a privilege, and you shouldn't give up that right by asking for permission if your proposed use really is fair, involving mere snippets or reasonable quotation, particularly for noncommercial purposes.

Fair use does not include harvesting addresses for purposes of sending unsolicited advertising or bulk messages. This practice is expressly forbidden.

We do not give, sell, or otherwise transfer addresses maintained by this website to any other party for the purposes of sending, or enabling others to send, unsolicited bulk messages. (Can-Spam Act language)

By using this website you agree that you will not add any of the posted addresses, whether email, postal, telephone, or any other type of address to any type of unsolicited bulk message list. You agree that the addresses have a value not less than US $50 each derived in part from not being receptacles for unsolicited bulk messages and that this is substantially reduced each time an address is added to any list maintained for such purposes. Compilation of these addresses into an unsolicited bulk sending list of any type is a violation of this agreement and is expressly prohibited. Robot visitors are agents of those who control them and are bound by these terms. To help robots comply, the site may obfuscate some contact information, while attempting to maintain accessibility to human visitors by providing normal versions in images or with javascript.

Privacy, or the Lack Thereof

"Now here's a tail of simple fool. Just glance at a page or two. You laugh ha-ha but you blush a bit for you realize while you're reading it that it's also reading you." -- from Hans Christian Anderson

The web server and page counter automatically keep logs of readers. The server logs may include, by way of example and not by way of limitation, the URL of the page you read; the date and time you read it, the URL of the page you were reading before, which for search engine referrals usually includes the search terms you used; and a fingerprint of the computer you are using, including such details as the browser version, screen resolution, operating system version, geographic location, and current internet address. I also use Statcounter, which makes statistical information easily available. Statcounter stores a cookie on your computer so it can distinguish between repeat visits and new visits. They have their own privacy policy (www.statcounter.com/ privacy.html).

The logfile information itself does not identify users; it identifies computers. But the header of every e-mail you send contains your computer's current internet address, and the body contains whatever personal information you disclose. So if I receive a copy of an e-mail you sent, whether forwarded or directly, I may possess all of the information needed to unmask your otherwise anonymous reading on this site. Furthermore, the web server logs are only overwritten when space runs low, and the page counter log remembers the last 100-500 visits. Thus the nominally anonymous reading records may exist for months or even years, and then a single e-mail could be used to compromise the privacy of everything you read previously.

Almost all web servers collect such information, whether the page author knows it or not. Horrifying, isn't it? I hadn't really considered the implications of what the servers and counters do automatically until reading that every website should have a privacy policy. The obvious solution is to scrub the logs, but right now I would have to do that manually, and since I would certainly forget, I can't very well promise to do that. Perhaps some automatic process can be set up to deal with this, but I haven't looked into it.

I use the logfiles for administrative purposes, not to invade reader's privacy. Routinely, I will check the statistics for such things as finding out what browser, screen resolution, and operating systems visitors are actually using. This helps with accessibility testing. The data also tells how often each page is viewed, which can help for determing if there is actually anything interesting on the site. Search terms are useful for telling whether appropropriate pages are turning up in response to those terms. I make no attempt to associate this data with individual users.

It is my policy to not associate logfile data with individual readers and not to disclose it to third parties. However, there may be exceptions of necessity. The most obvious is if there is a denial of service or spam attack originating with a specific address in the logfiles. In this case, it may be necessary to unmask and contact the reader who has the infected computer or to share logfile data with investigators. There might also be cases where disclosure of logfiles would be required by law. Additionally, should a given address be a spam source, I reserve the right to report that address to the government and to blacklisting services.

Other Leaks. It is distressing to discover that nearly every destination web server routinely logs everything you read there. Page authors frequently aren't even aware of this. But that's not the half of it. In a typical case, your reading might be logged by your employer; their ISP; a proxy server; and the destination website. The first three have a complete transcript of your session; the latter a complete transcript of your visit to that site. To avoid creating such records, it may be necessary to use a service such as the-cloak.com or behidden.com. If you select SSL (https:) mode, this means the intermediate logs will only show a connection to or from the anonymizing service provider. This at least reduces the number of transcripts to one. But you have to trust the provider. Do not conduct electronic banking this way; it opens you to a man-in-the-middle attack, and no one deserves that level of trust.

Disclaimer