Re: initial ACSS action item response

Here are a few comments on your draft response to the ACSS action item.

Firstly, I would suggest that you offer a list of specific recommendations
which the W3C might consider when revising its HTML Cougar proposal. Based
on your draft document, the most important of these recommendations would
appear to be that the user should be given the option of substituting
his/her own style sheets for those provided by the author of the document,
and this should include the ability to disallow the application of
persistent styles. As I read it, the current draft of the HTML styles
specification is ambiguous as to whether the user should be able to
override persistent style sheets, and if so, under what circumstances. This
point should be clarified. In the Cougar draft, under the discussion of
alternate styles, it is stated that the user should be permitted to choose
between alternate styles or to "switch off style sheets" altogether.
Later, however, a "persistent style" is defined as one which is always
applied, regardless of the user's preference. The document should at least
allow for persistent styles to be ignored whenever they are not
specifically intended for the medium in which the document will be
formatted. For example, audio-based user agents should be able to
disregard persistent style sheets which are not of media type "audio". For
maximum flexibility, it would be preferable that users be able to
deactivate persistent styles in all cases and substitute their own
preferences for those of the author. (I offer this as a suggestion for
further discussion).

Secondly, you could address the question of whether it is possible for
there to be a style sheet that can apply to all media. In the Cougar draft,
it is proposed that in the absence of an explicit "media" attribute, a
media type of "all" should be implied by default. Stated differently, the
style sheet is presumed to be relevant to all media types unless otherwise
specified. Consideration could be given to the question of whether this
arrangement is satisfactory, and if not, what alternative proposals might
be offered. For example, should the specification of a media type be made
obligatory? Presumably this would interfere with backward compatibility
and thus be unacceptable, but the general suggestion is that a range of
options be considered and discussed. It appears from the present proposal
that, in the absence of a "media" attribute, an audio-based user agent,
for example, would need to retrieve the style sheet and determine whether
any of its contents were relevant to the audio medium. If the style sheet
in question were written with print in mind, it is highly unlikely that it
would be relevant to the audio medium and would need to be discarded; but
assuming that audio and print-oriented parameters (as set out,
respectively in the CSS and ACSS specifications) can be given in a single
style sheet, there would be no alternative, in the absence of a media
attribute, but to parse the style sheet to determine whether any audio
parameters were contained therein. This issue is directly related to the
question, noted earlier, of whether there can be styles that are
applicable across all media types.

Please feel free to post this message to the WAI list or to forward it to
other interested parties if you consider such action appropriate.

Regards,

Jason White.

Received on Thursday, 26 June 1997 21:39:30 UTC