- From: Charles McCathieNevile <charlesn@sunrise.srl.rmit.edu.au>
- Date: Thu, 22 Jan 1998 11:09:48 +1100 (EST)
- To: Mike Burks <mburks952@worldnet.att.net>
- cc: "Kasday, Leonard" <kasday@att.com>, Chris Hasser <chasser@immerse.com>, "'wai-interest-group-post'" <w3c-wai-ig@w3.org>
The stats were gathered from the log files. Students using lynx accounted for (If I recall right) about 80% of those not retrieving the graphics. On the other hand proxy servers accounted for a number of retrievals, and it may be that they are in turn serving up pages without graphics. I assume the rest were graphic browsers. An interesting exercise I went through yesterday involved searching the catalogue of the Bibliotheque Nationale de France - the French National Library. They have a web-based system which is very poor for accessibility, and doesn't work properly anyway. They also have an older, telnet-style catalogue. Much easier to use, more efficient (It actually works) but difficult to find. They are not the only library to suffer from this problem. While I know that good graphics can be important, what I deplore on the web are meaningless or bad graphics. When I began working on the RMIT website, the main links were a small graphic icon, accompanied by a picture of a word. With no alt tagging. This is unfortunately not an uncommon experience. The presentation of authoring packages to the public relies heavily on their ability to provide visual control, at the expense of good HTML code (let alone accessibility). This gives people more time to spend on graphics, which should allow them to improve thier sites. Unfortunately the model of web design with which they are being presented does not lend itself to this. Charles McCathieNevile
Received on Wednesday, 21 January 1998 19:30:49 UTC