IRC log of wai-xtech on 2007-07-27

Timestamps are in UTC.

16:06:34 [RRSAgent]
RRSAgent has joined #wai-xtech
16:06:34 [RRSAgent]
logging to
16:06:58 [oedipus]
meeting: DHTML Sub-team Teleconference
16:07:08 [oedipus]
chair: Don Evans
16:07:28 [oedipus]
scribe: Gregory J. Rosmaita
16:07:33 [oedipus]
scribenick: oedipus
16:08:24 [oedipus]
Becky: if say right-left arrow with menu posted -- does that mean drop down? perhaps is covered (read from draft); with any level it would be same, right?
16:08:49 [oedipus]
Becky: spacebar opens actionable item -- exposes or executes?
16:09:15 [oedipus]
Becky: when have focus on menu item, pressing enter will execute the menu action
16:10:24 [oedipus]
GJR: XHTML2 group interested in 3 states: give focus, hang-fire (inspect), and fire
16:10:49 [oedipus]
Becky: nested menus -- not a drop-down menu -- or rather is but vertical instead of horizontal
16:11:06 [oedipus]
AG: how do you integrate abstract tree behavior
16:11:11 [oedipus]
Becky: right need both
16:11:25 [oedipus]
AlG: use menu or menubar consistently should be in best prActices
16:11:48 [oedipus]
AlG: if works like menubar, should follow menu keyboard conventions
16:11:51 [oedipus]
Becky: ok
16:12:13 [oedipus]
ACTION: Becky - post to xtech
16:12:53 [oedipus]
present: Don Evans, Al Gilman, Becky Gibson, Gregory J. Rosmaita
16:12:59 [oedipus]
Don: checkboxes?
16:13:26 [oedipus]
Becky: can do without proposal; should decide next one before telecon concludes so can discuss next meeting
16:13:51 [oedipus]
Chris: which keys are available by script events if running IE on Windows with JAWS or Window-Eyes?
16:15:33 [oedipus]
GJR: there is a passthrough key in most screen readers
16:15:58 [oedipus]
AlG: gregory explained how to get through screen-reader, but doesn't mean can get to application or script
16:17:14 [oedipus]
AlG: highly educated best-guess at keybindings that will work without conflict; some will require alternate methods; separate line of development around reengineering accesskey -- poster child for fuzzy implementatino and definition; need to be some kind of ARIA approach to getting raw materials in to bind to
16:18:02 [oedipus]
DonE: beginning to build javascript applications for applications; doing tree views, tab interfaces, etc.; not gracefully handled by at
16:18:23 [oedipus]
AlG: grep accessibility lists ( for accesskey
16:19:15 [oedipus]
GJR: start with existing keybindings defined by devs
16:22:09 [oedipus]
AlG: angle in action item to SAP to come up with working input tokenizer (text entry field with hidden grammar to check contents of text entry field) -- they have a validate action -- user can request it -- trick is hotkey in text edit mode also next to a button performing same fucntionality; style guide - anything give keybinding to somewhere there is an isolated action that could be pre-validated -- need right association between label on button and lab
16:22:24 [oedipus]
AlG: need second level in overall methodology
16:23:15 [oedipus]
Chris: attempted to take WinXP, MSIE7 and JAWS to ascertain what keys are already mapped -- short list; check against mozilla, even less; add window-eyes and you're left with nothing
16:23:49 [oedipus]
AlG: widget roles for keyboard access valuable because so much keybindings in AT
16:23:57 [oedipus]
Chris: handle elsewhere?
16:24:22 [oedipus]
AlG: perceive problem, use cases will get others to solve problem; need test interaction and usability
16:25:15 [oedipus]
AlG: part of the solution -- can machine follow from edit box with hotkey to the label that triggers the same event; stress in programming style level
16:25:27 [oedipus]
Chris: just wanted to ensure issue on table
16:25:52 [oedipus]
AlG: fact that have an issue is clear; what we need is a solution, which Becky is working on
16:26:47 [oedipus]
Becky: going to come into conflicts; how does user move focus from rich text editor to toolbar then back to rich text editor; can't make tabbable, will have to assign a key; will come up with proposal for dojo and submit to DHTML for consideration
16:27:26 [oedipus]
DonE: IE7 versus IE6 -- tabindex of positive number to DIV or SPAN to get focus -- any difference between IE6 and IE7?
16:27:40 [oedipus]
Becky: not had to do anything different for IE6 and IE7
16:28:02 [oedipus]
AlG: if have test case, bring up to mark silvey -- microsoft rep to PF;
16:28:16 [oedipus]
Don: some works in IE6 some in IE7, in our experience
16:28:20 [oedipus]
TOPIC: checkboxes
16:28:45 [oedipus]
Becky: checkboxes have checked, unchecked - space is trigger
16:28:54 [oedipus]
AlG: how many systems support that?
16:29:13 [oedipus]
Becky: have to check; for radio buttons, enter isn't supported; checkboxes enter and space
16:30:17 [oedipus]
GJR: tri-state for checkbox;
16:31:07 [oedipus]
Becky: isn't it part of group? what is use case
16:32:07 [oedipus]
AlG: software install -- summary install -- get checkboxes in advanced mode, can partially check options
16:32:38 [oedipus]
DonE: how do you trigger a partially checked check?
16:32:45 [oedipus]
Becky: only a programmatic state?
16:33:29 [oedipus]
AlG: if activate custom, enter dialog; if use to check and uncheck, will push a partial?
16:33:41 [oedipus]
AlG: if using WAI ARIA states on checkbox, can have partial value
16:34:04 [oedipus]
DonE: ARIA states checked unchecked partially checked -- question is, which is child of other
16:34:27 [oedipus]
Becky: IE7 only supports 2 options
16:34:42 [oedipus]
Becky: don't use key if grouped - go from one to another
16:34:59 [oedipus]
AlG: caret in edit box, hit enter and submitted form, rather than making newline in text field
16:35:13 [oedipus]
Becky: how to do that without a save or cancel button is still unknown
16:35:35 [oedipus]
Chris: checked or unchecked - partial derived state -- programmatically set
16:36:13 [oedipus]
Becky: checked and unchecked by mouse click or spacebar -- a toggle
16:36:33 [oedipus]
DonE: partial checkbox is programmatically determined
16:36:43 [oedipus]
Becky: specify that all checkboxes have labels?
16:36:59 [oedipus]
AlG: spec will require that -- can help people in coaching on good practice
16:37:40 [oedipus]
Chris: inverse - if on partially checked checkbox and hit space should all children be affected? typical GUI behavior
16:38:26 [oedipus]
AlG: chasing a will'o'the wisp -- if light, full and custom are 3 choices, hitting spacebar should get you into finer grain
16:39:02 [oedipus]
Becky: is a case where use GUI app, have nested checkboxes; pick one, top-level checkbox shows partial
16:39:23 [oedipus]
AlG: does spacebar select all or clear all -- what is sequence
16:40:30 [oedipus]
DonE: partial to off to on -- can't get back to partial; if hit spacebar, turns it off, if go into details all children checkboxes have been cleared
16:40:53 [oedipus]
Chris: font list - if turn on half of them, should first click on partially checked checkbox
16:41:15 [oedipus]
AlG: need to check other apps to find out if uniformity applied
16:42:48 [oedipus]
DonE: reflected in options
16:43:50 [oedipus]
GJR: often when use multiple select option, often loose focus when inspecting options; could substitute checkboxes with partial states (lynx model: comobo boxes as radio buttons; multiple select combo boxes as checkboxes)
16:45:31 [oedipus]
Becky: mapped browser behavior, stick a style rule over it, for performance reasons decided to let UA do most of the work;
16:46:27 [oedipus]
AlG: pragmatic decision need to spell out clearly in Best Practices; clear conflict between browser behavior between browsers -- there is an HTML coding for radio button functionality, best practice is don't script, use HTML/XHTML -- feel will be different on different UAs,
16:47:21 [oedipus]
DonE: checkboxes: 2 state and 3 state; 2 state select or deselect with space; 3 state if partially checked, spacebar fully checked and so are child elements
16:48:13 [oedipus]
LisaP: from SAP, volunteered to be integrator editor for best practices document which include keybinding decisions
16:48:39 [oedipus]
DonE: valueable skill
16:49:34 [oedipus]
LisaP: about a month away
16:50:00 [oedipus]
DonE: recommendations for agenda?
16:51:17 [oedipus]
Becky: Earl has a lot to say about issues posted to list; not sure ready to share -- brings up issues for ARIA team -- dojo accordian container just a tab container, earl working on much more complex accordion -- have to figure out how to distinguish through UI; will ping Earl offlist
16:51:30 [oedipus]
Becky: need to add a few more use cases for menus to ensure they are covered
16:51:57 [oedipus]
AlG: request for ChrisB - how to construct context menus
16:52:09 [oedipus]
ChrisB: how to move focus to create a dom event that at notices
16:52:17 [oedipus]
AlG: agenda items for this group or no?
16:52:33 [oedipus]
ChrisB: generalized technique -- how some of these things can be implemented
16:53:09 [oedipus]
AlG: input tokenizer needs menus; RichS working on drag-and-drop which includes menus
16:54:12 [oedipus]
AlG: idea of a prologue -- authored introduction; is there markup for author to indicate "this is what you need to pay attention to" -- blew off aw custom rule, but still a use case hanging out there;
16:54:35 [oedipus]
AlG: context help and prologue -- need use cases
16:56:00 [oedipus]
Becky: need to address tabindex behavior and focus; one thing in control taht has tabindex=0 -- do we agree? should we make it official? from forum questions, think need to address
16:56:48 [oedipus]
AlG: may need 2 techniques: one that relies on ARIA support, one that shous how built-in portions of ML can be effectively used
16:57:18 [oedipus]
AlG: RichS has put work into talking about use cases on PF/ARIA wiki
16:57:36 [oedipus]
16:58:36 [oedipus]
GJR: will send IRC log to DonE
16:59:26 [becka11y]
becka11y has left #wai-xtech
16:59:53 [oedipus]
present: Al Gilman, Becky Gibson, Lisa Papas, Becky Gibson, Chris B, Gregory J. Rosmaita
17:00:10 [oedipus]
rrsagent, set logs world-visible
17:00:17 [oedipus]
rrsagent, create minutes
17:00:17 [RRSAgent]
I have made the request to generate oedipus
17:01:39 [oedipus]
rrsagent, format minutes
17:01:39 [RRSAgent]
I have made the request to generate oedipus
17:01:47 [oedipus]
rrsagent, publish minutes
17:01:47 [RRSAgent]
I have made the request to generate oedipus