15:01:07 RRSAgent has joined #aapi 15:01:07 logging to http://www.w3.org/2010/01/12-aapi-irc 15:01:22 hi andi 15:01:28 i'm having trouble passcoding in 15:01:29 zakim, this will be aapi 15:01:31 Zakim has joined #aapi 15:01:38 okay - I haven't tried yet 15:01:50 rrsagent, make logs public 15:01:54 zakim, this will be 2119 15:01:58 I do not see a conference matching that name scheduled within the next hour, MichaelC 15:02:03 rrsagent, make log world 15:02:13 scribe: Andi 15:02:48 Meeting: AAPI 15:04:17 I'm in 15:04:21 WAI_PFWG(AAPI)10:00AM has now started 15:04:36 +Andi_Snow-Weaver 15:04:41 s/I'm in// 15:05:02 +Cynthia_Shelly 15:06:35 davidb? 15:06:43 1 sec 15:06:48 s/davidb// 15:07:18 s/1 sec// 15:07:30 +[Mozilla] 15:07:38 Zakim, Mozilla is David_Bolter 15:07:42 +David_Bolter; got it 15:08:54 topic: bug 7066 http://www.w3.org/Bugs/Public/show_bug.cgi?id=7066 15:10:42 +??P38 15:11:08 Is this an issue for any API? Apple? 15:11:25 zakim, ??P38 is Stevef 15:11:25 +Stevef; got it 15:11:29 action: Andi to contact Apple, reassign 7066 to Andi for now 15:11:39 topic: HTML mapping table 15:12:30 SF: not a lot more has been done. Have filed a few bugs 15:13:16 http://spreadsheets.google.com/ccc?key=0AlVP5_A996c5dG9RSE9GMy1JaVlBQ2dIWDliczJHckE&hl=en_GB 15:13:42 cshelly has joined #aapi 15:14:23 http://spreadsheets.google.com/ccc?key=0AlVP5_A996c5dG9RSE9GMy1JaVlBQ2dIWDliczJHckE&hl=en_GB 15:19:25 SF: will we be able to work out categorization that will be acceptable? 15:19:40 CS: impression from TPAC is that we will 15:19:54 CS: made a report to larger group about where we were - no strong objections 15:20:07 CS: some may have been taking wait and see attitude 15:20:42 SF: some feel you should be able to turn semantic roles into controls, some disagree 15:22:54 SF: not a clear categorization which makes it difficult to argue 15:23:08 CS: that's what this exercise is about 15:23:33 CS: think we got through the first pass - maybe we should look at the types of categories/groupings we came up with 15:23:45 CS: might be able to build a taxonomy which would make them easier to argue 15:24:42 CS: line 3, hyperlink - why did we pick this particular subset 15:24:49 CS: are they "commands"? 15:25:12 CS: if all the things that are commands can replace each other, that seems clear 15:27:55 SF: example of checkbox - if nothing else can have checked/uncheced state, seems to preclude a checkbox being anything else 15:28:09 s/uncheced/unchecked/ 15:28:42 SF adds Properties column 15:29:11 CS: anchor - special - in absence of ARIA, actuallly sends MSAA click event 15:29:17 s/actuallly/actually/ 15:33:27 scribe notes that discussion is being recorded in the spreadsheet 15:35:11 CS: divisions we have seem "pragmatic" but not "pure" which may be an issue 15:35:17 SF: examples would help 15:37:14 SF: we need more help to get this done 15:37:36 action: CS and SF to set up an hour to discuss offline 15:38:17 SF: go through and pick off the ones we think we could get agreement on and submit the bugs 15:39:58 agreed on Friday at 9 PT 15:40:12 -Stevef 15:40:48 topic: 7100 http://www.w3.org/Bugs/Public/show_bug.cgi?id=7100 15:44:31 add note before role mapping table 15:44:42 topic: 7101 http://www.w3.org/Bugs/Public/show_bug.cgi?id=7101 15:45:06 disagreement between browser vendors on how aria-owns should be implemented 15:48:53 DB: goes against the idea of aria-owns. Purpose was to define a relationship that is NOT in the tree - for things that only have visual relationships. 15:49:03 CS: on desktop, containment relationships, buttons in dialog - what is the extra relationship? 15:49:11 DB: some of newer APIs have notion of relation. At abstract level, aria-owns allows you to define relations between one accessible object and another regardless of where they are in the accessible tree. Emerged from those who are aware of these APIs as a way to define relationships that are not apparent in the DOM. 15:49:19 CS: IE doesn't change the DOM, just the MSAA tree. 15:49:28 DB: if request children, pass back all children and the ones defined by aria-owns? 15:49:35 CS: think so. 15:49:55 DB: need to research what each browser is doing, develop some test cases, and determine how big this issue is. 15:50:47 +Cooper 15:52:23 topic: Don't want a child to be a parent of itself - makes tree walking take forever. 15:52:23 IE changes the accessible tree based on ARIA-OWNS. 15:52:23 FF and Apple disagree. 15:52:23 Probably have diverging implementations - need to research what each browser is 15:52:23 doing, develop some test cases, and determine how big this issue is 15:52:36 topic: http://www.w3.org/Bugs/Public/show_bug.cgi?id=7115 15:52:52 Cynthia reviewed the table but don't think it's complete 15:53:46 http://www.w3.org/WAI/PF/aria-implementation/#mapping_actions 15:55:41 CS: got some feedback - need to see if relevant and send to Andi 15:58:36 topic: bug 7119 http://www.w3.org/Bugs/Public/show_bug.cgi?id=7119 and 7120 http://www.w3.org/Bugs/Public/show_bug.cgi?id=7120 15:59:28 CS: got feedback from IE - need to send to Andi 16:00:16 topic: bug 7129 http://www.w3.org/Bugs/Public/show_bug.cgi?id=7129 16:00:26 DB: doesn't think there is a role for heading 16:00:33 CS: may have been thinking of table headers 16:04:11 -Andi_Snow-Weaver 16:04:12 -David_Bolter 16:04:12 -Cynthia_Shelly 16:04:20 -Cooper 16:04:23 WAI_PFWG(AAPI)10:00AM has ended 16:04:26 Attendees were Andi_Snow-Weaver, Cynthia_Shelly, David_Bolter, Stevef, Cooper 16:12:22 Andi: what did we decide about for when js is turned off? map only the landmarks? 17:44:10 Andi has joined #aapi 17:44:19 zakim, bye 17:44:19 Zakim has left #aapi 17:44:23 rrsagent, make minutes 17:44:23 I have made the request to generate http://www.w3.org/2010/01/12-aapi-minutes.html Andi 17:46:11 rrsagent, off