IRC log of UA on 2004-02-27

Timestamps are in UTC.

01:03:04 [RRSAgent]
RRSAgent has joined #UA
01:03:30 [m3mSEA_]
zakim, aabb is Marc_Mulcahy
01:03:30 [Zakim]
+Marc_Mulcahy; got it
01:03:44 [m3mSEA_]
zakim, unmute ??P2
01:03:44 [Zakim]
??P2 should no longer be muted
01:07:48 [Zakim]
01:08:21 [Zakim]
01:09:47 [MattSEA]
zakim, ??p2 is Beijing
01:09:47 [Zakim]
+Beijing; got it
01:09:55 [MattSEA]
zakim, ??p1 is Jon_Gunderson
01:09:55 [Zakim]
+Jon_Gunderson; got it
01:11:27 [sunbj]
sunbj has joined #ua
01:11:38 [MattSEA]
JRG: How much is related to making XUL accessible, and how much is related to Gecko?
01:12:30 [MattSEA]
Sun (Louie): Mozilla has some XUL accessibility pieces (event model).
01:12:46 [MattSEA]
JRG: Marc, is that what you're testing on?
01:12:56 [MattSEA]
Marc: I've been focusing on HTML content.
01:13:52 [MattSEA]
JRG: I have a student working on a Mozilla extension (link functions, header functions, access keys). If it uses XUL controls, it should be accessible?
01:13:55 [MattSEA]
Marc: yes
01:14:23 [MattSEA]
JRG: What do you in Beijing need help with that the user agent group can help with?
01:14:31 [MattSEA]
JRG: (to conform to UAAG)
01:16:25 [MattSEA]
Jessie: Test cases provided are very helpful. Peter Korn asked a question in email: 1.1, full keyboard access - specific or detailed test cases for that function?
01:16:47 [MattSEA]
JRG: The main tests we have for that in HTML 4.01 test suite are related to supporting keyboard in HTML (accesskey, tabs)
01:17:18 [MattSEA]
JRG: It's difficult for the working group to make tests for all browsers. We list the core functions in Guideline 10 - Orient the user
01:17:25 [MattSEA]
JRG: (maybe 11)
01:17:59 [MattSEA]
JRG: It's not required to implement these functions, but it's a priority 2 to make them accessible with a single key (stop, back, forward, add to favorites, view favorites, etc.)
01:18:08 [MattSEA]
JRG: (checkpoint 11.5)
01:18:29 [MattSEA]
JRG: In general, we don't say you have to support this, but in Priority 2. They're common to many technologies.
01:18:43 [MattSEA]
Cathy: There are other tests related to frames, accesskey, changing form controls.
01:19:18 [JRG]
01:19:53 [MattSEA]
JRG: We should have a report next week on Mozilla.
01:20:42 [MattSEA]
Louie: When I tested with keyboard, not sure what to do with arrow keys. Is there a guideline about those?
01:21:24 [MattSEA]
JRG: We don't define any certain keyboard binding for any function. You can use arrow keys to navigate links, etc., but we say to provide keyboard access, and document how to use them accessibly.
01:21:47 [MattSEA]
JRG: Access keys are an example. Mozilla, IE and Opera all implement accesskey differently.
01:22:29 [Zakim]
+ +1.510.655.aacc
01:22:33 [MattSEA]
JRG: Opera did it differently.
01:22:40 [MattSEA]
zakim, aacc is Peter_Korn
01:22:40 [Zakim]
+Peter_Korn; got it
01:23:22 [MattSEA]
JRG: Opera has a Shift-Esc "accesskey mode" that they designed not to interfere with OS shortcut keys.
01:24:43 [MattSEA]
Peter: In Gnopernicus, while it's programmatically accessible, the expectation is that the user will navigate the content using arrows and control arrows, etc. I looked through UAAG 1.0, and didn't see anything beyond "full keyboard access" to describe what you had in mind.
01:24:55 [MattSEA]
Peter: Also didn't see any tests other than form content.
01:25:15 [MattSEA]
JRG: We don't have specific keyboard bindings. Minimal number of navigational requirements. We encourage more.
01:26:48 [JRG]
JRG has joined #UA
01:28:14 [JRG]
01:29:14 [kor1]
kor1 has joined #ua
01:29:31 [MattSEA]
JRG: These deal with focus. Had to assume that there was some manner of keyboard focus.
01:29:45 [MattSEA]
JRG: 9.2 is to move keyboard focus.
01:30:03 [MattSEA]
JRG: 9.3 only requires moving forward at P1. Second requirement to move in reverse at P2 (9.7).
01:30:18 [MattSEA]
JRG: Also have a search requirement (9.8, P2)
01:31:29 [MattSEA]
JRG: 9.9 is structured navigation. The WG worked on this concept of headers, separation of form controls, etc. 9.9 and 9.10 were the result.
01:32:09 [MattSEA]
Peter: How do you turn that on?
01:32:35 [MattSEA]
JRG: That's a standard feature in Opera.
01:33:06 [MattSEA]
Peter: Copy text to the clipboard?
01:33:26 [MattSEA]
Peter: No way to mark text to cut and copy/
01:33:46 [MattSEA]
JRG: That's ambiguous. You can select the whole resource and cut and paste with the keyboard, but not a subsection.
01:33:57 [MattSEA]
JRG: (in Internet Explorer)
01:34:18 [MattSEA]
Peter: So arguably they could comply, but don't conform to Section 508 1194.21(a).
01:34:35 [MattSEA]
Peter: What guideline would that fall under?
01:34:44 [MattSEA]
JRG: UAAG 1.1, full keyboard access.
01:35:37 [MattSEA]
Peter: Wasn't clear when I went through UAAG which one that applied to. I thought 1.1 as well.
01:36:06 [MattSEA]
Peter: You cite Opera as an example of mapping things. I understand that W3C is method-agnostic. Curious about thoughts/advice/suggestions since we don't have a lot of this implemented.
01:36:13 [MattSEA]
Peter: What keystrokes should we use and why?
01:36:39 [MattSEA]
JRG: We have one student project working on a Moz extension that will solve some of these problems.
01:37:15 [MattSEA]
JRG: Some of these functions may be available soon as extensions to Mozilla.
01:37:31 [MattSEA]
Peter: We can share our work in open source and make it better.
01:38:28 [MattSEA]
JRG: Not sure how well Opera model would work with Moz extensions.
01:38:41 [MattSEA]
JRG: Opera has a modal keyboard model.
01:39:34 [MattSEA]
Peter: We have F7 to turn caret navigation on.
01:40:52 [MattSEA]
Peter: There are some bugs with this. If you look at the Free Standards Group page ( you'll see a number of problems. Table navigation, etc.
01:41:26 [MattSEA]
Peter: That issue comes up once you have keyboard navigation. We don't have good answers yet.
01:42:31 [MattSEA]
Cathy: With HPR, we made some changes to the reading mode. Could read line by line or char by char, but inside a table, arrows are reading up and down a column or across the row.
01:43:39 [MattSEA]
Cathy: Had to deal with keyboard conflicts.
01:43:49 [MattSEA]
Peter: What about Alt, Ctrl, Shift?
01:44:11 [MattSEA]
Cathy: We didn't use the Alt key because it's used in so many dialogs. We used Alt-Shift for the accesskey.
01:45:03 [MattSEA]
Peter: Implementation is not doing the right thing, but we have to determine what the right thing is.
01:45:26 [MattSEA]
JRG: Did you not pick a document order for some reason?
01:45:40 [MattSEA]
Peter: I don't think the Sun Beijing team had a lot to do with the implementation.
01:46:20 [MattSEA]
JRG: Could add some additional navigation for table cells with Alt-arrow, for example.
01:47:43 [MattSEA]
JRG: The map element can be used in HTML as a container for related links. If we use that with nav bars, people can identify these collections of links and skip over them, or link to them.
01:48:56 [MattSEA]
JRG: Another way is to move in document order, and use different keyboard commands. Maybe the caret function becomes a structure-navigation mode.
01:50:22 [MattSEA]
Peter: We have a few models to work with: HPR, Opera, JAWS+IE, WindowEyes+IE. They provide their own navigation system of Web content from the DOM.
01:50:56 [MattSEA]
JRG: WindowEyes relies more on MSAA, which is DOM-like, but not as rich. I think the Mozilla extension will help.
01:51:46 [MattSEA]
JRG: List of links function has the advantage of allowing manipulation of nav features.
01:52:20 [MattSEA]
Peter: Any other models to look at?
01:52:41 [MattSEA]
JRG: Not sure about lynx...
01:52:50 [MattSEA]
Peter: pwWebSpeak?
01:52:54 [MattSEA]
JRG: They died 2 years ago.
01:53:00 [MattSEA]
Cathy: Author refers to HPR a lot.
01:53:32 [MattSEA]
Peter: stats on how many people use which?
01:53:35 [MattSEA]
01:55:15 [MattSEA]
MattSEA: So many different implementations, not sure whether to follow just one.
01:55:39 [MattSEA]
Peter: If 3 implementations use one key, and the other uses a different one, could help us decide.
01:56:08 [MattSEA]
Cathy: some apps are going to have to design around the browser, others are the browser.
01:56:33 [MattSEA]
Peter: Not arguing that because JAWS does it, we should do it. Just curious what they do because I think it would be useful info.
01:56:50 [MattSEA]
Cathy: When you look at their keys, they have unusual shift states, and different keys for laptop vs. desktop.
01:57:13 [MattSEA]
JRG: Both JAWS and WindowEyes use h key to navigate headers, while opera uses s and w.
01:57:24 [MattSEA]
Peter: Any info/studies/anecdotes about what users like and why?
01:57:49 [MattSEA]
JRG: From my experience, so few headers on Web pages that most people have never heard of header nav. They go back to search.
01:58:07 [MattSEA]
JRG: Part of the battle is that there's not a lot of good markup.
01:58:39 [MattSEA]
JRG: For table nav, JAWS uses Ctrl-Alt for nav
02:00:22 [MattSEA]
Cathy: Our users like the reading modes. We don't like to change the keys that users are accustomed to.
02:00:44 [MattSEA]
Peter: Any plans for HPR outside of Windows?
02:00:51 [MattSEA]
Cathy: We don't have any announced plans.
02:01:21 [MattSEA]
JRG: HPR and Opera apply as models, since they are the browser.
02:01:48 [MattSEA]
Peter: Is there a doc that outlines the modes in HPR?
02:01:58 [MattSEA]
Cathy: Getting Started page, and the keys help file.
02:05:28 [MattSEA]
Peter: We will have to prioritize, and that priority will have to focus around blind and motor disabilities with our screen reader and OSK, because of the goals of Sun Java Desktop.
02:06:39 [MattSEA]
Peter: Your work on the Mozilla extension may be the easiest way to get keyboard nav into Mozilla.
02:08:18 [MattSEA]
Peter: Also Swing HTML and JPane, gtkhtml and gtkhtml2, Java help HTML, HTML renderer in Evolution mail client, and would make a lot of sense for users to have the same keyboard idioms for doing the same things in all of these.
02:10:39 [MattSEA]
Peter: Could also include IE, Konqueror...
02:11:33 [MattSEA]
Peter: We've had a lot of cooperation between GNOME and KDE communities. Konqueror folks may be amenable to standardizing on keystrokes.
02:12:30 [MattSEA]
ACTION: Jon work on cross-browser keystroke matrix
02:12:46 [MattSEA]
Peter: What are the important dates for current Mozilla schedule?
02:13:45 [MattSEA]
Kyle: 3 months for each release.
02:14:18 [MattSEA]
02:14:42 [sunbj]
02:15:43 [MattSEA]
MattSEA: 1.7beta is 10 March, 1.8alpha is 12 May, 1.8beta is 9 June
02:16:02 [MattSEA]
Peter: We've missed 1.7 for most things. Our more sophisticated navigation is probably 1.8.
02:17:03 [MattSEA]
Kyle: Since Netscape is over, nobody working on keyboard nav.
02:20:42 [MattSEA]
Peter: We're interested in user feedback on Mozilla with Gnopernicus.
02:21:08 [MattSEA]
JRG: We could have a couple people working on it, if we can get a system working.
02:22:00 [kor1]
Kyle - would it be reasonable for Beijing to make builds of accessible Mozilla available to Jon's students?
02:25:39 [MattSEA]
JRG: Next meeting 11 March?
02:25:58 [MattSEA]
Peter: close to CSUN
02:26:04 [MattSEA]
MattSEA: I'll be traveling
02:26:12 [MattSEA]
JRG: During CSUN?
02:26:36 [MattSEA]
Peter: Maybe. CSUN is crazy.
02:26:45 [MattSEA]
JRG: 25 March?
02:27:32 [MattSEA]
JRG: I'm out of town.
02:28:52 [MattSEA]
Peter: Status of
02:29:32 [MattSEA]
JRG: Not working on new version of UAAG. May be revisions to the current, but trying to get people to implement 1.0. Some discussion of mobile/voice apps.
02:30:04 [sunbj]
what's the final decision of next meeting? 3/11 or 3/25?
02:30:09 [MattSEA]
JRG: Would like to see conformance claims.
02:30:22 [kor1]
I believe 3/11 (in the U.S.; 3/12 for Beijing)
02:31:06 [Zakim]
02:31:38 [Zakim]
02:31:39 [Zakim]
02:31:40 [Zakim]
02:31:41 [Zakim]
02:31:44 [Zakim]
02:31:45 [Zakim]
WAI_UAWG(beijing)8:00PM has ended
02:31:46 [Zakim]
Attendees were +1.512.246.aaaa, +1.303.554.aabb, Matt, Cathy_Laws, Marc_Mulcahy, Beijing, Jon_Gunderson, +1.510.655.aacc, Peter_Korn
02:34:34 [MattSEA]
rrsagent, bye
02:34:34 [RRSAgent]
I see 1 open action item:
02:34:34 [RRSAgent]
ACTION: Jon work on cross-browser keystroke matrix [1]
02:34:34 [RRSAgent]
recorded in