Dave,
I believe there are two, separate issues here. They are:
1. What are the appropriate units to use for table refresh in ODF?
-> what I've heard suggested from TC members is the ISO
standard for this, which allows time to be expressed in milliseconds as
well as larger increments
2. Where is/are the appropriate place/s to ensure that user
interactions with an ODF document won't cause a seizure?
-> what I suggest is the appropriate place is in the ODF
application, not the document format specification
The reasons I suggest the appropriate place is in the ODF applications
are:
1. The app is where the rendering & user interaction occur
2. Not in all cases does a table refresh have the potential to trigger
a seizure (only if enough of the field of view is making a sufficient
luminosity change in the triggering frequency range).
3. My belief that the purview of the ODF accessibility subcommittee is
accessibility issues (and not anything beyond that). I personally feel
free to offer my opinions on all manner of things ODF-related, but I do
so as a member of OASIS interested in ODF; not with my "accessibility
hat" on.
4. My own sense that user interface considerations should not dictate
encoding schemes, they should simply place requirements on what is
needed (and thus I wouldn't say that the only valid table refresh
number must explicitly be outside of the range of 3-50Hz).
Thus whether or not I believe a vendor or any application author does
or does not have good reason to update some portion of the screen at
greater than >3Hz, my sole accessibility concern is whether that
update has the other attributes needed to trigger a seizure. If not,
then my aesthetic or other considerations are only those, and not an
accessibility statement.
And I further believe that only the rendering application is in a
position to make the determination as to whether the other
seizure-triggering characteristics are at play or not (e.g. a two cell
table in 9 point font where only black pixel text on a white background
is updating at potentially 5Hz when displayed on a 1280x1024 screen at
72dpi should be well below the seizure threshold, while a table update
that includes changing the background color of the cell from black to
white in a 50 cell table at 24 point font at 5Hz is certainly over the
threshold).
Whether or not something is a good idea from a design or aesthetics
point of view, if it doesn't cause a true accessibility problem then it
isn't my job to make sure it is fixed.
Regards,
Peter Korn
Accessibility Architect,
Sun Microsystems, Inc.
711a73df0807200634k1c2321a8n27ec1b45076929a1@mail.gmail.com" type="cite">
2008/7/20 Richard Schwerdtfeger <schwer@us.ibm.com>:
My suggestion would be to be to take an excerpt from our ODF accessibility
Guidelines for 1.2 and place it in the ODF guidelines in the section on
Table Refresh Delay. Something on the order that Office applications SHOULD
provide a facility to enable the user to limit table refresh delays to no
more than three times per second. Failure to do so may cause seizures in
some ODF users.
We could make this a MUST but I don't know of all uses for ODF documents.
Additionally, we should provide guidance to ODF content authors which would
be in line with this W3C WCAG requirement.
Tell me a vendor who justifiably refreshes at >3Hz in a human facing app?
Yet again we bow to the vendors Rich?
Pretty please instead of do it because its right?
I'll ride with it, but it's wrong IMO.
regards