W3C home > Mailing lists > Public > w3c-wai-ua@w3.org > October to December 1998

Re: Please no browser assembly

From: Paul Adelson <paul.adelson@citicorp.com>
Date: Tue, 15 Dec 1998 14:16:30 -0600
Message-Id: <199812152017.PAA22000@egate2.citicorp.com>
To: Charles McCathieNevile <charles@w3.org>
Cc: Bryan Campbell <bryany@pathcom.com>, w3c-wai-ua@w3.org
In response to Charles McCathieNevile:
>>> I would be surprised if any major manufacturers did not provide simple
interfaces to their products. <<<

Rather than expecting / hoping developers will do what we think they should,
we should provide clear guidelines. For the conscientious developers much of
this will be 'low hanging fruit', but for others it will be a wake-up call.

The group has discussed CSS for alternative table rendering and other complex
processes that go well beyond fonts / colors. If we state that enabling CSS
support for an alternate rendering style is what a UA needs to do, we are
_not_ making a clear statement that the user shouldn't need to research and
reinvent arcane stylesheet code if they want to implement that alternative.

An extreme example: can a vendor who makes their source code available claim
that their product is therefore 100% accessible and 100% flexible, because all
it takes to meet specific needs is writing your own code? A developer once
made a similar claim to me (that his product was compatible with all
peripherals and major operating systems because you could buy the source code
and re-write it however you wanted.)

Charles McCathieNevile wrote:

> 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

  -- Paul Adelson
* The views expressed are those of the
* author and do not necessarily reflect the
* position of Citibank or its affiliates.
Received on Tuesday, 15 December 1998 15:16:16 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:49:21 UTC