Chair: Jon Gunderson
Date: Wednesday, July 21st
Time: 12:00 noon to 1:30 pm Eastern Standard Time
Call-in: W3C Tobin Bridge (+1) 617-252-7000
Chair: Jon Gunderson
Scribe: Jim Allan
Present: Ian Jacobs
Glen Gordon
Rich Schwerdtfeger
Regrets: Charles McCathieNevile
Gregory J. Rosmaita
RS: tech question, some items don't make sense. bullet item- how to specify in html
IJ: right, to be filled in later, techniques still in process
JG: move to proposed REC in next month or two, in proposed for 6 weeks, need f2f to discuss issues, looking at sites on west coast in early October
GG: ATIA conference (1st week October-Orlando 6-9) and CTG Oct 18-25
RS: out last week of Oct 14-16 and Oct 31-Nov 1
IJ: arrange calendars, has open calendar
JA: not available Sept 25 - Oct 3, CTG Oct 18-25
RS: perhaps piggy back on ATIA
JG: what about west coast
RS: ATIA convenient for GG
JG: never had meeting in SE USA, what about before conference
GG: after is better
JG: need a host group, with some funding from host group
IJ: host coordinates, funds a dinner
GG: talk to Wilson about this, St. Pete is far from
JG: talk to Judy about this, perhaps 10-12
IJ: need 8 weeks notice
ACTION IJ: set up call with Judy JG, IJ to discuss f2f
JG: need f2f comments
IJ: YES, more efficient, get more done
IJ: section on conformance in some revision, either conform or don't , then changed to conform by desktop or dependent, then WCAG, then use WCAG conformance, second clause - native implementation or interaction to other software, no check points for interaction with other software. if we don't talk about it then drop clause. propose dropping because no checkpoints
JG: communication with other software was murky, MS was not happy, if we provide it but nobody uses it is that still conformance, don't need clause
IJ: lots of evolution in GL, have dropped lots of language, shifted to dependent UA
JG: both clauses for GUI and dependent agents would be removed.
GG: makes life easier, conform by doing it natively, the native thing will be to support this API
JG: conformance, one statement for AT or dependent, broke it up to Desktop UA and AT, previous version said one conformance - if GUI with AT can meet check point then you have conformance. Now, only do what UAGL says, with out dependence on other software or technology
IJ: do it through interface, 3) it needs to be done (i.e. table access)
RS: why not require cell navigation
JG: desktop not braille, not speech, but provide access to this information, must make browser accessible to text to speech software,
RS: browser is designed for the blind, need to allow other AT to connect to your browser
JG: yes, must allow interoperability with other software
IJ: people may have other disabilities, may not be just blind, user may be using other software in addition to talking browser, must support standard API in answer to HB question-can I enter text from mouse...
RS: special case browser for mobility, blind,
IJ: yes, if they are stand alone, and may be used with other technology then yes, if in a kiosk and no ability to connect other software then now.
RS: special API like IE, help facility add helper object and access to DOM
IJ: expect timely access to DOM, manipulate user interface controls, read and write information about controls, selection, focus, may need to review guideline-
JG: what is required of desktop or AT
IJ: GL aimed at desktop and AT, RS sited another class, if new class want to conform must choose one of the classes, if WebSpeak claim conformance as AT then don't have to work with or allow other AT access, if conform as a desktop then must work with AT
JG: want cooperation between AT, need cross disability AT cooperation and working together
ACTION RS: review conformance statement, and classes of browsers (HPR) see where it fits into classes, present proposal to list if needed.
JG: questions about IJ conformance suggestion
RESOLVED: checkpoint 1.3 review, clause removal--all ok
IJ: if still questions, another class of browser- make a proposal if we finish document, and make a profile of class of browser and conformance list, intentionally choose these two classes,
JG: RS made earlier comment, not much there, JG: to add to techniques what strategies can we use to get people to write techniques
IJ: techniques history...less vital than guidelines, storehouse of information, hard to keep it interesting, structure keeps changing, current structure seems to work, hard to contribute in past, should be easier not, wide open to contributions how to get them? tricky, think about it in parts, review existing browsers and make suggestions based on what's needed. need developer input
RS: like structure
IJ: in order to go to REC, Tim BL says must have solid techniques before last call and REC, techniques is reflection on how real the guidelines are, use techniques to strengthen the GL
RS: id sections that need work, not clear what's needed
IJ: like RS contribution on Java
RS: problems with speed accessing DOM or COM, reality is different from making statements and test case
JG: what would help, make a list of what's needed, identify others to focus on a section
IJ: assigned tasks in AU group, brute force, is one approach, each person has expertise in specific areas, as editor created a usable structure
JG: people have contributed work
RS: id sections, who has expertise
ACTION JG and IJ: to id who has contributed in the past, to contribute more or review existing materials, then contact people
JG: other ideas
IJ: some material is mildly complete, sketch form, bullets, need to make paragraphs
JG: describe how something happens based on some tool with screen shots,
IJ: looking for resources, how to XXX?
RS: table stuff would be good for Kathy Laws (IBM) to look at
RS: nobody from Microsoft in attendance lately, Why?
IJ: Judy been talking, questions about MS and Netscape participation. Don't know.
ACTION IJ: Review member participation for next week
JG: need a laundry list of needed features and CMN: has queried disability listservs to get important issues- what users need orientation to, currently 25 checkpoints and need more list versus succinct list is good for conformance, lots of things to check off at different levels succinct easier for integration and creativity, i.e. for navigation, give general concept or direction, then expand in techniques.
IJ: no feedback from blinux people, what to shorten list, but no feedback shorten to 10 checkpoints and move other stuff to techniques. where is line drawn for inaccessibility
JG: schism between GL 8 and GL 9, 8 is general, 9 is specific, problem editorially, may be confusing to end user of GL
IJ: option, more abstract principle, or combine, or chop out what is not needed
JG: question, discrepancy between 8 and 9, should 9 be more similar to 8 (less checkpoints)
GG: afraid to say yes, fear of task, lean toward brevity
RS: move some to checkpoints
ACTION JG and JA: review 9 and propose consolidation of items deadline tomorrow afternoon
RS: would make document less ominous
GG: examples in techniques, put blindfold on try to use, or have page with 1000 links and get to link 800 without a mouse
IJ: scenario
GG: allow nav without mouse, have page with 1000 links and get to link 800 without a mouse rhetorical to make a point
IJ: Great! need more of this, giving rationale use scenario to illustrate, Strong YES
GG: developers should be able to get check point right if they have the concept
JG: other issues
no all around