16:45:40 RRSAgent has joined #ua 16:45:40 logging to http://www.w3.org/2007/11/01-ua-irc 16:45:49 Zakim, this will be UAWG 16:45:49 ok, JR; I see WAI_UAWG()1:00PM scheduled to start in 15 minutes 16:45:58 Meeting: WAI UA 16:46:07 Chair: Jim Allen 16:46:17 Scribe: JR 16:46:49 Regrets: Kelly Ford, David Poehlman, Peter Parente, Gregory Rosmaita 16:49:00 Agenda: http://lists.w3.org/Archives/Public/w3c-wai-ua/2007OctDec/0025.html 16:54:29 AllanJ has joined #ua 16:56:34 WAI_UAWG()1:00PM has now started 16:56:41 +AllanJ 16:57:03 +??P5 16:57:25 zakim, ??P5 is really JR 16:57:26 +JR; got it 17:02:17 +[IBM] 17:02:35 zakim, [IBM] is really cklaws 17:02:35 +cklaws; got it 17:02:41 cklaws has joined #ua 17:04:45 JA, JR: OK with putting "Until User Agents" stuff on the Wiki 17:06:19 JA, CL: Discussing status of Forms, XForms, ... 17:15:36 JA: Option labels not generally read by Scr Readers 17:15:55 CL: only way to get it from Accessibility API is to get it from the tree 17:16:11 CL: But not considered an attribute like groupbox would be 17:16:28 JA: So it's animplicit association 17:16:54 C: Tried to get Firefox to establish a "relationship" 17:17:10 CL: "Labeeled by", "Described by" 17:17:30 JA: Easier if more in HTML code? 17:17:58 CL: Not really...until lots of people implement optgroup they won't put in support and vice versa 17:21:46 CL: So if you're in a given option...don't know what optgroup you're in unless you traverse back up the tree to the parent optgroup 17:23:04 CL: EVERYTHING in the DOM, just some stuff isn't in the API 17:23:46 JA: Asks about header, ids in tables 17:24:06 CL: If browser implements, then yes 17:24:14 CL: Inspect32 17:24:53 http://www.microsoft.com/downloads/details.aspx?familyid=3755582a-a707-460a-bf21-1373316e13f0&displaylang=en 17:25:22 CL: IBM has new tool 17:26:29 http://www.mail-archive.com/accessibility-ia2@lists.linux-foundation.org/msg00042.html 17:26:45 http://www.eclipse.org/actf/ 17:27:11 CL: THat's essentially an enhanced version of Inspect32 that includes support for iAccessible2 17:28:19 CL: Web developers do nothing with APIs...the browser converts markup into accessibility APIs...ATs use them 17:28:59 JR: Asks about WebAPI 17:29:29 JA: Doing communication with OS... 17:30:03 CL: It's about OS tool features 17:30:28 CL: Browser takes care of mapping to the specific OS 17:30:45 CL: Kind of like SAPI 17:30:49 WebAPI - http://www.w3.org/2006/webapi/ 17:32:43 JR:WebAPI called by Web developer... 17:32:59 JR: to call OS functions (e.g. clipboard) 17:34:39 JA: So in UAAG we have "6.10 Timely exchanges through APIs (P2) " 17:35:52 CL: e.g. state of control changes.....state in accessibility API must change ASAP 17:36:46 CL: ...events needed too because Scr Readers caching the screen and need to know if something has changed 17:37:44 JA: During PF 1.5 hours... 17:38:03 CL: Who's leading - do we need list of questions? 17:38:15 JA: Prob need to prepare... 17:38:49 JR: Question list? 17:39:37 JA: With AJAX lots of functionning that browser doesn't know about - functions, keybindings until event happens. 17:39:59 JA: Javascript might grab keyd before browser? 17:40:14 CL: Browser still handling no processing 17:41:15 CL: THere are parsers helping....but browser does run the javascript 17:41:39 CL: Write to DOM command in Javascript goes via browser 17:41:54 JA: Key binding? 17:42:15 CL: THinks key gets key first? 17:42:33 JA: Access key wins over menus in IE 17:42:55 CL: OK what about keystroke handled by javascript? 17:43:54 CL: So a QUESTION is how to resolve all the different keybindings. 17:45:02 CL: AT, OS,...browser... 17:46:15 CL: Actually OS lets AT get them first... 17:46:31 JR: Sounds like we could be writing guidance to browser about what to do... 17:46:48 CL: Yes - between its chrome, access keys, javascript key events 17:47:59 JR: QUESTION 1: Out of all browser-level and below processes, what kind of guidance can we give. 17:48:40 JA: Event handlers... 17:49:26 JA: Programmatic exposure may not be enough... 17:51:32 CL: When key comes in Browser doesn't know if the javascript really wants a key 17:53:22 JR: So wuation about viability of activate event handlers? 17:54:11 JA: Another question is about exposing bindings to users who don't have ATs. 17:54:54 JA: Other big issue - GL6....major concern is out of process APIs are read only 17:55:14 JA: ARIA folks working on DOM API 17:55:36 CL: User agent issue 17:57:01 CL: There is a perfomance issue...if you are not in process you have to cross process boundary... 17:57:39 CL: Quicker if you are in process 18:00:07 CL: Client-server architecture 18:00:25 CL: Using COM transport in Windows 18:00:38 CL: CORBA in Linux 18:01:39 CL: Complicated by fact that browser's interface is different than AT 18:03:22 JA: Other issue is programmatic notification of changes....and not all changes are in DOM...caret changes, CSS changes 18:03:52 -cklaws 18:05:30 Action JR: Change Agenda to set PF-ARIA questions just after lunch on Monday. 18:06:15 -AllanJ 18:06:19 -JR 18:06:20 WAI_UAWG()1:00PM has ended 18:06:21 Attendees were AllanJ, JR, cklaws 18:06:33 RRSAgent, make minutes 18:06:33 I have made the request to generate http://www.w3.org/2007/11/01-ua-minutes.html JR 18:06:37 RRSAgent, set logs public 18:06:41 Zakim, bye 18:06:41 Zakim has left #ua 18:06:45 RRSAgent, bye 18:06:45 I see 1 open action item saved in http://www.w3.org/2007/11/01-ua-actions.rdf : 18:06:45 ACTION: JR to Change Agenda to set PF-ARIA questions just after lunch on Monday. [1] 18:06:45 recorded in http://www.w3.org/2007/11/01-ua-irc#T18-05-30