See also: IRC log
<jallan> Jim and Cathy work on action item http://lists.w3.org/Archives/Public/w3c-wai-ua/2006AprJun/0052.html
JA: reports on conversation with
... "programmatically determined" seems to take author away from technology
Al: two parts: 1-on prog det as
author responsbility: think of 2 views of content - rendered
content and part that can go to automated ua
... Al interprets as convey in notation that has a formal semantics
... Should follow specs and fill in fields.
JA: Just doesn't seem like this would be understandable by others who read.
Al: we should be paying attention
to what comes over wire from author's automation to the user's
... THey didn't want to say "in attribute"...to tech specific
... But in the end this is too indirect for authors to understand.
JA: CL and JA just think that
this boils down to use well formed, semantic, etc. markup
... So are you saying you think things in WCAG should change?
Al: Yes we should write some
... By not separating user experience from data from server...the only things left are what's in user experience.
<jallan> JR: its multipart, its a complex concept
JA: For example wcag
... When you drill down...techs are use <stong> etc. instead of span bold.
<jallan> Al: wcag is at a different level
<jallan> success criteria is standalone - normative
<jallan> don't need techniques, understanding, etc.
<jallan> to meet the success criteria
Al: Success criteria are only thing normative - but techs really reqd in some cases to understand...
<jallan> JR: in AU had to invent another layer.
<jallan> technology specific benchmark document
<jallan> tool maker must go through wcag and state what their tool does to meet wcag
<jallan> then the Authoring tool conformance statement becomes binding on themselves
<jallan> they must match their (autoring tools) interpretation of what wcag checkpoints/success criteria mean
<jallan> Al: very much what EARL has done.
Al: shadi separates test objective from what was specifically tested
Shadi = EARL
Al: Years ago we had mtg with fed stats people about accessible stats...
they found JAWS inspect command - it would read of all headers if header attribute used...otherwise not.
Al: So some prog deter. things do
reach user experience.
... New idea: Has to do with what the user agent calls configurability...
there are 2 kinds of controls in addition to what comes in over wire - view adjustments (eg view size) and the other is navigation.
Al: WCAG seems to have ignored
... ex, they ask you to make structure and presentation seperable, then they ask for sufficient contrast...even when ua is a ble to correct if struct and pres are seprated
... SUggest using HTML version of the mail-in form to draft text.
<jallan> ACTION: Jim to rewrite prog det items to conform to wcag model [recorded in http://www.w3.org/2006/05/18-ua-minutes.html#action01]
<parente> there's also http://www.w3.org/WAI/WCAG20/comments/form.html
author that really should be done by the user agent.
JA: Decided to drop this.
... Other thing that came up on CG call was issue from compound documents...
... Essentially Al revised some comments by JA to compund docs group...
... Issue was what if in my UA I set 18pt font, what would happen with an SVG inside in 100x100 container in an xhtml
... Something needs to be done (scrolling) etc.
... They said not our prob.,,its the xhtml object.
... But the browser may not know what's inside object so ewhat's inside needs to porovide.
This is scribe.perl Revision: 1.127 of Date: 2005/08/16 15:12:03 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Found Scribe: JR Inferring ScribeNick: JR Default Present: Jim_Allan, [IBM], Jan_Richards, AlGilman Present: Jim_Allan [IBM] Jan_Richards AlGilman Regrets: Dave P. Got date from IRC log name: 18 May 2006 Guessing minutes URL: http://www.w3.org/2006/05/18-ua-minutes.html People with action items: jim WARNING: Input appears to use implicit continuation lines. You may need the "-implicitContinuations" option.[End of scribe.perl diagnostic output]