Re: Please no browser assembly

I think your fears of users being required to learn twenty languages a
week are misplaced. I would be surprised if any major manufacturers did
not provide simple interfaces to their products. What concerns me is if
there is no possibility of a major product being reconfigured extensively
by a user who has some need not anticipated by the designer of the 'simple
interface'. This problem is even greater if manufacturers of 'assistive
technology' do not have the ability to reconfigure the interface.

Do address your example of CSS, it should be possible to create a CSS
stylesheet. Actually there are a number of simple tools used to specify
styles already - the fonts/colours section of IE and Netscape, or the more
extensive controls offered by Opera, none of which require any
particular knowledge, can all be used to generate style sheets without the
user having to know what underlies the technology. But the ability of the
user to control those features is very important.

As I see it we are not telling anyone how to make their user interface.
Whether they use a command line interface that requires knowledge of HTML,
CSS, HTTP, GNU, and Swahili, or whether they use a series of plain english
questions and multiple choice answers is how manufacturers provide product
differentiation. In this personal view, the role of the UA group is to
determine a set of functions which a User Agent should be (somehow) able
to perform in order to ensure that it facilitates access to the content
which it renders. This is also, of course, not a total specification -
there are many functions which could be provided which are not necessary,
but again enhance product differentiation. And there are many functions
which are taken for granted - the most obvious in the case of browsers
being the ability to render the text and follow links.

--Charles McCathieNevile
W3C Web Accessibility Initiative (visiting)
email: charles@w3.org telephone: +1 (617) 258 8143
mail: LCS, 545 Technology sq., Cambridge MA, USA
http://purl.oclc.org/net/charles

Received on Monday, 14 December 1998 10:55:13 UTC