I'm proposing a SIG session at CHI 99 on:

"Universal Web Access: Delivering Services to Everyone"
http://www.acm.org/perlman/accessig.html
for a combination of reasons. On its surface, it is about internationalization, disabled access, access for kids, the elderly, and more. Each of these target groups have had their own SIG sessions, panels, and papers both at CHI and elsewhere. They are each different, but at some level, they are the same: developing services for people who are "different" from the canonical "English-speaking normal-sensory/motor/cognitive males aged 35-45", or more generally, developing flexible, highly customizable systems. The SIG is focused on the Web, mostly because the Web is so universally accessible.

I work at OCLC, which provides libraries with FirstSearch®, a bibliographic and fulltext online retrieval system. I worked on the multilingual version of FirstSearch, http://www.oclc.org/cgi-oclc/fs-sample.scr the installation of which required the installation of a new mechanism for building our Web pages. This mechanism, and others we are working on, will support new versions of FirstSearch:

All these have become related in that we use many of the same methods for development (although less so for analysis and evaluation).

I'm having a similar experience with Web-based UI evaluation with questionnaires:

http://www.acm.org/perlman/question.html
I just completed, completely over the Web, a French version of the CSUQ, using an internationalized version of the administration system:
http://www.acm.org/perlman/question.cgi?language=fr
As I now work on a WAI version, the same issues come up: Each new target group gets easier, and "minor" enhancements (e.g., Spanish version) should take a day at most. Often, as I have add a new feature, like preferred-language detection or browser-type detection, I have commented to myself that I wish I had been thinking about all these issues earlier, and that I wish I had access to resources and methods.

It's this "Big Picture" accessibility that I am especially interested in. Ideally, this SIG could begin to integrate the efforts of people working on addressing the needs of specific groups. At least people could broaden their horizons and share information.

The SIG session can't possibly address specific groups in much detail, so I would propose to organize the SIG around the steps of analysis, design, implementation, and evaluation. So, for example, a group of people working on services for kids, multilingual, visually handicapped, ..., would discuss methods and resources for implementing Web-based services. Another group would try to cover evaluation issues. By the end of the SIG, I would expect that people would be motivated to combine their efforts in these areas. I would measure success by the amount of discussion after the session, possibly resulting in a workshop for CHI2K.

Gary PERLMAN, OCLC Online Computer Library Center
6565 Frantz Road, Dublin, Ohio 43017 USA
Voice: +1-614-761-5058 Fax: +1-614-793-0915
perlman@acm.org   http://www.acm.org/perlman/