14:15:42 RRSAgent has joined #aapi 14:15:42 logging to http://www.w3.org/2011/08/04-aapi-irc 14:15:56 RRSAgent, start meeting 14:15:56 I'm logging. I don't understand 'start meeting', davidb. Try /msg RRSAgent help 14:16:04 rrsagent, make logs public 14:16:10 rrsagent, make log world 14:16:21 cyns has joined #aapi 14:16:34 topic: issue-444 14:16:45 Do IE and FF modify the DOM? 14:17:07 DB: FF does modify the DOM in the case of aria-activedescendant and aria-selected. 14:17:22 RS: don't think ATs are going to use the API to modify selection 14:17:43 CS: seems like modifying the DOM is the right thing to do 14:18:07 I don't like modifying the DOM. 14:18:29 RS: the reason it's a problem is that authors are in complete control - track keyboard events 14:18:57 RS: but authors are not looking for changes to active-descendant 14:19:09 RS: are you modifying the DOM or doing styling? 14:19:21 DB: FF does modify the DOM for aria-selected 14:20:05 CS: seems like it is a bad idea to not have the DOM in sync with what the AT thinks 14:20:29 RS: ATs are not using API to change selection, web authors won't know it's changed 14:20:55 RS: to make it work end to end, we need the device independent events to notify the web app 14:21:04 CS: but that WG is not even off the ground yet 14:21:27 RS: are we going to take WAI-ARIA back to Last Call because of this? 14:22:04 DB: what about removal of features from ARIA? does that require another LC? 14:23:12 AS: we can only remove something if there is an Ed Note that says the feature is at risk and might be removed. 14:23:32 DB: so if we have Safari and IE that would be two releases 14:23:54 RS: would like to do a point release of WAI-ARIA right after this to include the device independent events strategy 14:24:10 CS: but the DI events will take two years 14:24:28 s/be two releases/be two implementations 14:25:36 RS: all agree we need DI events - reduces work for developers 14:26:01 CS: agree it is needed but if it went to LC tomorrow, it would be two years before it would be a Rec 14:26:12 RS: essential for mobile 14:26:15 CS: agree 14:27:52 RS: if we go back to LC, we would change the spec and remove the MUST which would raise a red flag with the WHATWG 14:29:16 action: Cynthia to find out what IE does 14:29:39 RS: voice reco technologies are not using APIs to change selection. They send keyboard and mouse events. 14:30:07 AS: what about using an API to change focus? 14:30:09 RS: not sure 14:30:43 AS: sounds like we need to poll the AT vendors to see if anyone is using the AAPI to change focus or selection 14:31:07 CS: even if they're not, this forces them to go to the DOM rather than using the API 14:31:29 CS: would like to find out how bad the issue is and publish an errata ASAP after WAI-ARIA goes to Rec 14:35:21 topic: skip to main content 14:35:32 RS: need a feature in the browsers to get to the main landmark 14:35:42 RS: can IE put this in IE 10? 14:36:01 CS: IE doesn't have document structure navigation feature - not sure where they would put it 14:36:13 CS: would love for somebody to do a plug-in 14:36:29 RS: if we could have a skip to main content keyboard function would be ideal 14:36:41 CS: changes to UI at this point have a low chance of getting in 14:39:19 topic: back to issue-444 14:39:36 AS: how does it work if the UA modifies the COM for aria-activedescendant? 14:39:51 RS: today can use DOMAttrModified event but that is going away 14:58:27 zakim, bye 14:58:27 Zakim has left #aapi 14:58:32 rrsagent, make minutes 14:58:32 I have made the request to generate http://www.w3.org/2011/08/04-aapi-minutes.html Andi 14:58:47 richardschwerdtfe has left #aapi 15:24:13 Andi has joined #aapi 15:25:03 AndiSnow has joined #aapi 15:25:12 AndiSnow has left #aapi