Comments on 21 November draft

This latest version represents a marked improvement over earlier drafts.
Nevertheless, as usual, I have a number of detailed suggestions to make:

Checkpoint 2.3. The examples are well chosen and innovative, but they
could be clarified a little more by explaining what could appear in the
markup in each case, and how this differs from the presentation. Examples
2 and 3 discuss both the logical structure and the presentation (example 1
focuses on structure except for the mention of italics). The point is,
however, that the structures (sections, paragraphs, lists; acts, scenes;
and objects depicted in the graphic, respectively) can be indicated
abstractly in the markup, whereas the presentation is a separate matter:
it can be generated by a style sheet according to users' requirements,
stored in a separate "presentational" view of the content, for example a
graphic or page description language (SVG or PDF for instance), etc.

The purpose of checkpoint 2.3 is to ensure that the structures are marked
up and made available for further processing (navigation, presentation by
a variety of devices, transformations of various kinds, etc.). Clarifying
the examples under checkpoint 2.3 would facilitate comprehension of the
distinction between structure and semantics on the one hand, and
presentation on the other. Unfortunately, these important ideas are often
unknown to, or misunderstood by those who do not approach the web with a
background in SGML, and hence we need to explain the concepts as clearly
as possible.

Checkpoint 6.1: this text accurately captures the proposal which emerged
from last week's meeting. One item in the list of options below the
checkpoint concerns me: "a mechanism to enable the content to be processed
by the user". Shouldn't this read "by the user agent?" Now, if we only had
a device that would enable all content to be processed by the user (and I
will leave it to science fiction writers to imagine how this might work)
we probably wouldn't need accessibility guidelines in the first place.

In the text of checkpoint 6.1 the phrase "by the user" could be omitted
without changing the meaning appreciably.

Checkpoint 6.2 accords with my proposal, but it doesn't take account of
Charles' follow-up, in which he specifically suggested adding a note to
state that the W3C has incorporated considerations of accessibility into
the processes by which its specifications are developed and reviewed; and
hence they are expected to meet the requirements of the checkpoint. Other
organisations may also take steps to ensure that their technologies meet
the requirements. I am not insisting that this point be added, but merely
indicating that it has been raised in discussion and has not been included
in the current draft of the checkpoint.

As a more general comment, the guidelines use a variety of terms to
designate what we usually refer to as "the content": ("content",
"document", "page", etc.). If possible, we should use a single term (E.G.
"content") in all cases, unless we intend to restrict the application of
the requirement, definition or explanation to only a particular kind of
content (a user interface, a structured document, a web page, etc.). The
Web Characterisation Activity once proposed definitions of such terms as
"web site", "web page" etc., and we need to decide whether we want to use
these terms (and/or the suggested definitions) in WCAG deliverables. See
http://www.w3.org/1999/05/WCA-terms/ for the proposed definitions (this
document is apparently only a working draft).

Received on Wednesday, 22 November 2000 21:10:23 UTC