<jallan> scribe: Jan Richards

<JR> slate: JR

<JR> scribe:JR

6.4 Programmatic access to information about rendered content (P1)

JA: Sent this earlier: http://lists.w3.org/Archives/Public/w3c-wai-ua/2006JulSep/0040.html

PP: Has some other links for Linux...
... Use AtkComponent as well as AtkText...

<parente> http://developer.gnome.org/doc/API/2.0/atk/AtkComponent.html

PP: Plus there is a broken link

<parente> http://developer.gnome.org/doc/API/2.0/atk/AtkText.html

CL: We don't generally do this in HPR
... Believe AT's are going directly to Windows APIs on windows not to DOM

JA: How is it on MacOS?

PP: MacOS has its own.
... Even on Linux - application uses ATK, AT's go through ATSPI(sp?)
... Usually 1 to 1 but ATSPI missing "write" calls

JA: WHat about AJAX?

CL: Think AJAX people are looking at adding location to AJAX events

PP: In Linux we do assume that the ATSPI will be kept up to dat by the browser.
... But for thinks like changing coordinates, not sure.
... New ATSPI will have doc events added - start loading, end loading.
... Also mentions ChildrenChange events

<parente> http://accessibility.freestandards.org/a11yspecs/atspi/adoc/atspi-events.html

PP: Javascript components have to tell browser when things change so browser can tell AT

CL: Someone at IBM (Gibson?) knows this stuff...

<scribe> ACTION: CL to Gather Gibson's presentations etc. for the group. [recorded in http://www.w3.org/2006/08/31-UA-minutes.html#action01]

<jallan> ACTION: fix links add refereces for Linux include previoius references to ATSPI [recorded in http://www.w3.org/2006/08/31-UA-minutes.html#action02]

<cklaws> http://www-306.ibm.com/able/resources/ajaxaccessibility.html

This is paper Gibson wrote.

Gibson = Becky Gibson

6.5 Programmatic operation of user agent user interface (P1)

JA: Has prob with def on user agent ui control...
... Def says this out of the box browser
... But may need to expand to extensions and toolbars etc.

CL: IE uses COM and ActiveX interfaces for plugins

JA: In firefox?

CL: Firefox supports XPcom and XUL(sp?)

JA: Wondering about accessibility to toolbars etc
... Are there guidelines, best practoces for these toolbars?

CL: Will ask Tom

JA: Will ask "Standards Project" about their accessibility toolbar

<scribe> ACTION: JA to Will ask "Standards Project" about their accessibility toolbar [recorded in http://www.w3.org/2006/08/31-UA-minutes.html#action03]

<cklaws> Firefo xa11y features: http://kb.mozillazine.org/Accessibility_features_of_Firefox

<scribe> ACTION: PP to Will look around for similar API mechanisms in Linux (for plugging in toolbars, etc.). [recorded in http://www.w3.org/2006/08/31-UA-minutes.html#action04]

CL, JR, PP: We should update the examples in Example Techniques 3.

CL, PP: Accessibility works best at "semantic" level - buttons etc, not a drawing level - draw rectangle etc

CL, PP: Use the desktop level as examples.

JA: Maybe add Linux to ex. tech 6?

CL: X windows is a weird example

PP: Window manager will be written to rely on X...
... Window managers are separte but at some point bundled

CL: In tha tCommunity peple think of Linux+Gnome or Solaris+Gnome

JA: Let's change X-Windows to Linux

<jallan> in techniques 3

<cklaws> http://msdn2.microsoft.com/en-us/library/727z646z.aspx

(Introduction to COM and ATL (goes in Refs for 6.5)

<jallan> ACTION: ja: added link to references for accessible Com objects [recorded in http://www.w3.org/2006/08/31-UA-minutes.html#action05]

JA: Ex. tech. 8 gets back in to low level stuff...
... So we don't need to add Linux here

PP: Will ook up another ref for 8.

6.6 Programmatic notification of changes (P1)

JR: Does first exclusion work with AJAX?

CL: There is the "unless" clause

PP: The was CSS used now...much can be done

CL: APIs still mapping down to the DOM

JR: Would this confuse a developer?

CL: Yeah we really mean the semantic content

PP: Imagine if CSS used to check/uncheck a box.

CL: Maybe WCAG prevents this?

JR: So what is 6.6 really getting at for deveopers?

CL: IN def for document model, CSS is required

JR: Does ex. tech. 1 belong here?

CL: Used to be common for AT's to hook display drivers...but now we want to use engineered APIs
... This way all of the samantic info comes along

JA: What do we need to do with 6.6?

CL: Did ask Rich if you can get where content changes occur as part of event notification, but answer was no.

<jallan> cl comments related to 6.4

CL: So change message happens but must then walk to tree to see where,

<jallan> jr: 6.6 says must be notified of change, but not WHERE the change occured.

PP: Technically if AtkComponent used could get location...
... eg newchild event, tells source..use source to get location

CL: What about text?

PP: In new ATSPI, there is an ObjectAttributesChange event that would be relant here
... For text...TextChangedEvents...but noit sureif any of these are fired by DHTML

<jallan> ACTION: ja: add reference to roadmap role/state information to references for 6.6 [recorded in http://www.w3.org/2006/08/31-UA-minutes.html#action06]

CL: ON Windows side, MSAA does not do thiz

JA: Next week we will review 6.6

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: Jan Richards
Found Scribe: JR
Inferring ScribeNick: JR
Scribes: Jan Richards, JR
Default Present: Jim_Allan, Parente, Jan_Richards, cklaws, DPoehlman
Present: Jim_Allan Parente Jan_Richards cklaws DPoehlman
Got date from IRC log name: 31 Aug 2006
Guessing minutes URL: http://www.w3.org/2006/08/31-UA-minutes.html
People with action items: add added cl fix for include ja link links linux pp previoius refereces reference references

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.