- From: Charles McCathieNevile <charles@w3.org>
- Date: Tue, 15 Dec 1998 18:46:16 -0500 (EST)
- To: Lynn Alford <Lynn.Alford@jcu.edu.au>
- cc: "Christine A. Quinn" <cquinn@stanford.edu>, w3c-wai-ig@w3.org
About 18 months ago, in an old job, I did a little test to find out about use of images. We had a new front page, new image, new names, and after a week I did a quick check of the server access logs and some basic arithmetic. the results were as follows: hits on page: about 100 000 GET image requests: about 45 000 some users from the organisation (an unmeasurable quantity, but perhaps about half) were restricted to text only access. So I checked for identifiable requests from inside the organisation: 45 000 Assuming that all of those were text-only access, which is a bad assumption, then 10% of outside hits did not bother to take the image. This was one of Australia's largest websites, but had been recently rated as one of the least interesting unless you really wanted the information (as distinct to an exciting and visually stimulating browsing experience) It also counted hits to the server, so proxy servers only counted as one hit. I concluded that a substantial proportion of people accessing the fron page were not visually impaired, but were just not interested in the images - after all one of the most common efficieny tricks, and generally the most effective, since it cuts most download times by 25% - 95%, is to turn images off. Charles McCathieNevile
Received on Tuesday, 15 December 1998 18:46:29 UTC