(A Javascript-enabled browser is required to email me.)

TBTF for 1995-09-10: What did the server know and when did it know it?

Keith Dawson (dawson dot tbtf at gmail dot com)
Sun, 10 Sep 1995 20:25:30 -0400

>>From Edupage (1995-09-07):

> Nielsen Media Research will work together with Internet Profiles to offer
> a service that will track the number of visits to Internet Web sites. The
> new service, called Nielsen I/Pro I/Count and Nielsen I/Pro I/Audit, will
> be able to tell subscribers how many times an individual visited the site,
> and the company name and location, assuming the individual is using a
> corporate account. (Wall Street Journal 6 Sep 95 B2)

Sorry, but you just can't count Web visitors as accurately as Nielson counts TV viewers (however accurately you believe that to be).

Internet Profiles (see <<http://www.ipro.com/>) can't really know about "individuals" and their "visits." Ipro works from data in Web server log files. How much does a server know about its visitors? We can find the answer from a Web service provided by Popco at <http://www.popco.com/cgi-bin/env-testers>.

Note added 1996-04-28: Another demonstration of the same data is available at <http://www.patents.com/status.cgi>; thanks to Carl Oppedahl <carl at oppedahl dot com> for the example.
Below is part of the output from the Popco URL -- the data that are in any way personal to me, the requestor.

HTTP_USER_AGENT:Mozilla/1.1N (Macintosh; I; 68K)
HTTP_ACCEPT:*/*, image/gif, image/x-xbitmap, image/jpeg

My Web client program tells the Popco server that I'm running Netscape ("Mozilla") 1.1N on a Macintosh 68K-based machine. The server knows my IP address ( and from that can look up my hostname (ip176.boston.ma.interramp.com). My client reports what types of files it is prepared to accept.

The next time I dial my Internet service provider, Performance Systems International, my IP address will be different and my hostname will be different. This is true of many, but not all, ISPs. This circumstance will lead Ipro to count me as a different individual each time I dial up. But it would summarize all of my visits in its geographical report as visits by people inside PSI from the state of Virginia (where PSI is headquartered).

On the other hand, visits by all the users of Compuserve, AOL, and Prodigy will be homogenized -- these users don't get individual IP addresses and so are indistinguishable. Ipro's log analysis can't discriminate whether a flurry of "hits" on your server from aol.com is the result of one or of many AOL users. My reading is that all hits inside a window of 30 minutes will be counted as a single visitor. Now imagine that numerous AOL users were hitting your Web site steadily throughout the day: Ipro would report to you a single long "visit." The same homogenization happens for corporations that funnel all their Internet traffic through a single "gateway" machine, as many do for security reasons.

> A presidential task force has recommended that electronic transmission
> of books, magazine articles and software should be classified as copies
> subject to existing copyright laws. The task force also recommended that
> it should be illegal to make or distribute products aimed at decoding
> encrypted software without the consent of the copyright owner. (Wall
> Street Journal 6 Sep 95 B3)

This task force was headed by Bruce Lehman, Commissioner of Patents and Trademarks. It's fair to say that Mr. Lehman does not subscribe to the credo espoused by Tom Parmenter: "Digital technology is the universal solvent of intellectual property rights."


>>Edupage -- mail listproc@educom.edu without subject
> and with message: subscribe edupage <your name> .

TBTF alerts you twice a week to bellwethers in computer and communications
technology, with special attention to commerce on the Internet. See the
archive at <http://www.tbtf.com/>. To subscribe send the
message "subscribe" to tbtf-request@world.std.com.
Keith Dawson dawson dot tbtf at gmail dot com dawson@atria.com
Layer of ash separates morning and evening milk.