W3C home > Mailing lists > Public > w3c-wai-er-ig@w3.org > January 2001

Re: example of site where text-only does not convey all info

From: Phill Jenkins <pjenkins@us.ibm.com>
Date: Tue, 23 Jan 2001 18:12:25 -0500
To: W3c-wai-er-ig@w3.org
Message-ID: <OFB4B5293F.4095DD12-ON862569DD.007E1CB7@raleigh.ibm.com>
>The question we asked this morning is, "how would you determine if these
>two sites (the text equivalent and the graphical with rollovers) are
>equivalent?"  They have the same links, but there is information presented
>on the graphical page that is not available on the text-only.

I think we have to be careful about the context in which we use the term
"equivalent". I give 3 types of contexts with examples:

For example some on the list have talked about the "site" - a whole
collection of pages and functions and features, and whether the user can
achieve the same result using the site, for example shopping for an item,
getting the information they are looking for, etc.  This is a general
design goal or principle.

Another context example is evaluating just the feature or tag, as in the
"equivalent" of the mouse over that Wendy started the thread with, or the
equivalent of an <image> is the alt attribute, etc. This could be more
machine checkable.

A third context is the "user experience or satisfaction".  This is the most
difficult since it may be physically impossible to have the same physical
experience without the appropriate physical senses or devices to render the
content in a particular modality.

So, in general, I look more to the first context, the site as a whole, and
not that each and every coding construct have an equivalent. There is a
finite subset of specific HTML tags that have equivalents that could be
listed and would be useful to ER-IG.  It would also be useful to list
popular real life examples of constructs and how they do, or could provide
"equivalents".  I do not think it is useful for a theoretical approach to
exhaustively list equivalents for each feature or tag.  So I agree, except
for the finite subset, that these are in the GL techniques domain.

Phill Jenkins
IBM Research Division - Accessibility Center
Received on Tuesday, 23 January 2001 18:12:41 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:01:31 UTC