RE: ATAG2.0 - comments/questions on SC B.1.2.1

Comments marked "JR"

--
(Mr) Jan Richards, M.Sc.
jrichards@ocad.ca | 416-977-6000 ext. 3957 | fax: 416-977-9844
Inclusive Design Research Centre (IDRC) | http://idrc.ocad.ca/
Faculty of Design | OCAD University

From: w3c-wai-au-request@w3.org [mailto:w3c-wai-au-request@w3.org] On Behalf Of Boland Jr, Frederick E.
Sent: August 19, 2011 9:03 AM
To: w3c-wai-au@w3.org
Subject: ATAG2.0 - comments/questions on SC B.1.2.1

Going through spec a little at a time..

Thanks and best wishes
Tim Boland NIST

B.1.2.1 - What is a "restructuring transformation"?  Please define.. I see a definition for
"content transformations" but how is this different?  same with "recoding transformations" -
how is recoding different from restructuring?  What is being transformed?  Is there a defined
begin and end point of a transformation?

JR: The concepts come from here. http://www.w3.org/TR/2008/WD-ct-guidelines-20080801/#term-restructuring
We can grab more text if you like.

for (a) - what exactly constitutes - accessibility information - (I know we define it, but do
we need to change definition in light of recent discussions)? Does this relate to the
"WCAG-capable" discussion at the last teleconference?

JR: It relates, but so far we haven't decided on a resolution.

"All" accessibility information?  Some?

JR: Has to be all. Some can only really be tested to mean "at least one"

How much is sufficient?   What exactly does it mean to "preserve" -information- (or how can one
objectively test that -information- has been "preserved" (for example, is it the semantics of
the -information- in a different form, or does the syntax/representation of the -information-
have to be the same?

JR: The WCAG SCs are not technology-specific so it should be possible.

Is the -output- the output of the -transformation- (are we "comparing"
the output to the input at predetermined points in time?)

JR: Yes, the output is the output of the transformation. And yes, we are comparing before and after the transformation (which of course could be composed of multiple steps).

for (b) - the comments I made for the beginning of B.1.1.2 ("author", "default option", "authoring
tools provide a warning" - suggested wording, etc.)
apply here as well..  see comments for (a) above re: accessibility information (typo - should
have same term as part (a) for consistency - for example, (a) has (WCAG) here but (b) doesn't?)

JR: (WCAG) should be added.

"All" or "some" accessibility information may be lost (is it possible to "quantify")?

JR: If they want to.

Does the warning always occur before the transformation is initiated?

JR: Warnings work best before the things they warn of.

Is there always evidence of
the warning in some modality?  I think that there is always a possibility that information may be
lost when any change (transformation) is made, so I'm not sure what additional benefit the "may be lost"
gives (unless there's a much higher probability of loss when certain changes (transformations) occur?

JR: The accessibility of the warning is covered by Part A. The wording just gives developers a bit of flexibility in case they want to put up a standard warning on certain actions.

It depends on the nature of the transformation (who/what has control over the end
state of the transformation, how "drastic" the transformation is, how many transformations take place,
etc.).  For example, a minor change of one word or tag entirely within an authoring tool would seem to
be less drastic than when the entire contents are transferred to an entirely different tool or environment
under completely different control..

JR: Agreed. We try to take that into account with our different types.

for (c) how -soon- after the transformation (for example, before another transformation or before exiting the
authoring tool?)

JR: More or less immediately after.

How much accessibility checking is -automatically- performed (for example, on one element/
word, all elements/all content - how much is sufficient/enough?)  On what particularly is accessibility
checking performed (not specified).. Also see comments above for "transformation"..

JR: Right - these are the same issues as for auto-generated content.

Do we need to change our definition of "checking, accessibility" in light of recent WCAG-related discussions?

JR: We need to wait and see.

I assume that the accessibility checking is automatically performed by the authoring tool being tested by this
SC - is this always true (it's not explicitly specified here)?

JR: Right - remember the tool can be a conglomeration of tools.

for (d) , see comments above for (c) - is there always evidence of a prompt for testing purposes? - how much
accessibility checking?  On what?

NOTE - typo - change "criteria" to "criterion" - what exactly is an "output technology" (is this term defined?
- I know "technology, web content" is)

JR: The technology of the outputted content. I don't think it needs a definition.

  What specifically is an "-included?- technology for conformance?  Could you
please explain further and give examples?  Does this relate to the ATAG conformance section in some way?

JR: Yes, there should have been a link: http://www.w3.org/TR/ATAG20/#conf-techs-produced