W3C home > Mailing lists > Public > w3c-wai-eo@w3.org > July to September 1999

Re: Comments on SMIL note

From: Madeleine Rothberg <Madeleine_Rothberg@wgbh.org>
Date: 30 Jul 1999 09:38:31 +0000
Message-ID: <n1278794988.5182@wgbh.org>
To: "Ian Jacobs" <ij@w3.org>, "Marja-Riitta Koivunen" <marja@w3.org>, "Geoff Freed" <Geoff_Freed@wgbh.org>
Cc: ij@w3.org, "Judy Brewer" <jbrewer@w3.org>, "WAI-EO" <w3c-wai-eo@w3.org>
        Reply to:   RE>>Comments on SMIL note

What Geoff and I are concerned about in section 4.1 is not whether
authors test for system-captions="on" or "off" but rather the
suggestion that authors use the system-captions attribute as
a proxy for whether the user needs text links in addition to
links in an image map or video map. Using captions generally 
implies lack of audio input (whether because of disability,
environmental sounds, or technical limits). Using text links
generally implies lack of visual input or difficulty using images.
The current example encourages authors to confuse these two needs.


Date: 7/29/99 10:08 PM
To: Madeleine Rothberg
From: Marja-Riitta Koivunen
>> Section 4.1:
>> It looks like you're recommending that authors use SYSTEM-CAPTONS to
turn on >text links corresponding to a video map.  Aren't you just supposed
to use this >element for captions only?
>Good catch. I'll have to talk this over with Marja. 
The text with the systeem-captions test attribute is not very clear. The
attribute tests whether the user has preferred to see captions or not. And
you can test both ways. So the attribute is not exactly metadata telling
that something is a caption. However, I agree that I would like to use it
as much as possible for that purpose (system-caption="on"). If we test
system-caption="off" it is always attached to some other elements than
captions anyway as it does not make sense to show captions when user has
preferred not to see them. Therefore it can be easily used e.g. to define
different layouts with and without captions. 

Sometimes it might be easier to test system-caption="on" and the authors
can do that too, but is there a borderline for what the authors are allowed
to use the test and how to define that?

Received on Friday, 30 July 1999 09:36:37 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:55:46 UTC