- From: Charles McCathieNevile <charles@w3.org>
- Date: Tue, 1 Feb 2000 18:12:59 -0500 (EST)
- To: Scott Luebking <phoenixl@netcom.com>
- cc: w3c-wai-gl@w3.org
This seems like a pretty sound approach in general. However I am not sure
that having preference settings like "upper arm disability" is going to be as
well received as having the resulting features be what the user selects on -
fast navigation, no images, or whatever.
Charles Munat has also worked on this approach a fair bit, but doing browser
sniffing to determine whether to include thins like accesskeys (based on the
idea that if the browser doesn't handle them they are a wase of
bandwidth). As I understand it, CC/PP does not deal specifically with
disabilities, but instead with functional requirements/limitations. Another
area this is being used is with system atributes in SMIL and SVG.
Cheers
Charles McCN
On Tue, 1 Feb 2000, Scott Luebking wrote:
Hi,
In a previous posting, Wendy asked me to look at CC/PP amd see
if it is in sync with my ideas.
Before going into that discussion, I thought it might be helpful to
talk a little about attributes/preferences and software architecture.
An arrangement that I've found helpful in software architecture is to
provide for external attributes/preferences and internal attributes/preferences.
The external attributes/preferences are known outside of the program
while internal attributes/preferences are known only inside of the program.
There are several advantages to this arrangement. One is that the names
used for the external attributes/preferences can be renamed, aliased, etc
and the internal attributes/preferences can be left the same. Only
the mapping of the external attributes/preferences to internal
attributes/preferences need be changes. This arrangement also allows
for external attributes/preferences to be specified from different sources
using names relevant to the source without needing to change the
internal attributes/preferences.
In this arrangement, values from CC/PP would be seen as external
attributes/preferences from the CC/PP source. Also, a web page
could be provided for specifying user preferences using attributes/preferences
which need not have the same names as CC/PP. The names would
be more relevant to the purpose of the web page where the names
from CC/PP would be more generic.
Assuming that a web site provides an ability to store a user's
preferences from a previous visit, a user preference page and the
handling for CC/PP, the following steps would probably occur
when a user visits the web site:
1. set internal attributes/preferences from defaults
2. set internal attributes/preferences from stored user preferences
3. query user agent for CC/PP information
4.a. if no CC/PP information is returned, display a greeting web page
with a link to user preferences web page which has a list
of external attributes/preferences relevant to the web site
4.b. if CC/PP information is returned, set the internal attributes/preferences
from the external attributes/preferences provided by the
CC/PP. then display a greeting web page with a link
to user preferences web page which has a list of external
attributes/preferences relevant to the web site. (This web
page will let a user over-ride values provided by CC/PP
as might be needed for the particular session. It is also
highly useful for debugging.)
In this organization, CC/PP should identify major classes of disability,
e.g. "blind", "upper arm limitations". The application would then translate
these external attributes/preferences into possible internal
attributes/preferences like "use simple format" or "include fast
section navigation". (Ideally, the user preference web page would allow
the user to turn off all features relevant to a disability or
let the user pick and choose which disability features to include.)
As near as I can tell, CC/PP isn't providing information about
a person's disability.
Scott
--
Charles McCathieNevile mailto:charles@w3.org phone: +61 (0) 409 134 136
W3C Web Accessibility Initiative http://www.w3.org/WAI
21 Mitchell Street, Footscray, VIC 3011, Australia
Received on Tuesday, 1 February 2000 18:13:09 UTC