IRC log of ua on 2004-11-16

Timestamps are in UTC.

01:00:42 [RRSAgent]
RRSAgent has joined #ua
01:00:48 [MattSEA]
rrsagent, make log public
01:00:59 [Zakim]
01:01:10 [MattSEA]
zakim, ??p6 is Jon
01:01:10 [Zakim]
+Jon; got it
01:01:27 [Zakim]
01:05:33 [MattSEA]
jg: Moving UAAG test suite to mozilla site.
01:05:39 [MattSEA]
jg: Work on keyboard interface?
01:05:50 [Zakim]
01:05:50 [MattSEA]
pk: Still working on it. Questions about configurability of interface.
01:06:01 [MattSEA]
zakim, ??p9 is Jessie_Li
01:06:01 [Zakim]
+Jessie_Li; got it
01:06:43 [JRG]
JRG has joined #UA
01:09:01 [JRG]
zakim, who is here?
01:09:01 [Zakim]
On the phone I see Jon, Matt, Peter_Korn, Susan_Crayne, Jessie_Li
01:09:02 [Zakim]
On IRC I see JRG, RRSAgent, Zakim, MattSEA
01:09:27 [korn]
korn has joined #ua
01:09:33 [korn]
01:10:09 [Zakim]
01:10:22 [MattSEA]
zakim, Rich is Aaron
01:10:22 [Zakim]
+Aaron; got it
01:12:10 [MattSEA]
kyle: We have fixed some bugs in Mozilla 1.7 tree, and are porting them to the trunk. Current feature is arrow keys in form control.
01:12:38 [kyle]
kyle has joined #ua
01:13:12 [kyle]
01:13:20 [MattSEA]
jg: What are you planning to do for the next version?
01:14:35 [MattSEA]
kyle: Keyboard nav for moving in and out of form control, and item-based navigation. Probably will be based on JavaScript components. No detail on second part; will discuss in the future.
01:15:15 [MattSEA]
jg: Is it caret-based, or based on document order?
01:17:35 [MattSEA]
kyle: In the keyboard nav model, we'll follow caret-based nav, but in item-based nav, we'll follow the document model.
01:17:52 [MattSEA]
jg: So if I'm on a line, and hit up arrow, how do you know where the caret goes?
01:18:11 [MattSEA]
pk: I believe it's visual, based roughly on where it is. Think it's covered in current keyboard proposal.
01:19:45 [MattSEA]
kyle: We use a special key binding for moving outside of text box.
01:20:25 [hopgrassor]
hopgrassor has joined #UA
01:20:53 [MattSEA]
kyle: We also need a special binding to move focus into form control.
01:21:19 [kyle]
no, I was wrong, no special key binding needed to go into form control
01:21:28 [kyle]
alt+[ and alt+]
01:21:48 [MattSEA]
(left and right bracket)
01:21:52 [korn]
[ = "right bracket" or "right square bracket"
01:22:04 [kyle]
alt+[ for move backward, alt+] for move forward
01:23:22 [MattSEA]
pk: There's a coding headache to this.
01:23:33 [MattSEA]
aaron: Also an issue of getting out of the form.
01:24:11 [MattSEA]
pk: in GNOME and Java, tab can insert characters into edit field, so need to press ctrl+tab to move out.
01:25:48 [MattSEA]
aaron: Any app can control any key at any time. Many things can be put into trees, etc. And some of this could be integrated into Web Forms 2.0 in the future.
01:27:22 [MattSEA]
aaron: WHAT WG is forking standards by adding common ideas, menus, forms, etc. Kind of important for Mozilla because they're working on it.
01:27:24 [MattSEA]
01:27:53 [MattSEA]
aaron: Working on Web Forms 2.0, Web Apps, which is for building Web applications, and one for making custom controls.
01:28:56 [MattSEA]
aaron: Ian Hickson is leading that.
01:29:12 [MattSEA]
jg: Anyone working on accessibility?
01:29:29 [MattSEA]
aaron: Political issues, no companies represented. They have accessibility addressed.
01:31:14 [MattSEA]
jg: Implication here is that there may be more complex things that the caret may have to recognize.
01:32:50 [MattSEA]
aaron: Starts by capturing input from the UI, then bubbles up through the chain.
01:33:12 [MattSEA]
aaron: Usually they set a prevent-default flag on it, which it passes up a level, but it can still be used at higher levels.
01:35:03 [MattSEA]
aaron: There can be some pretty awkward key combinations (incl. AltGr) that affect our keystroke choices.
01:35:13 [MattSEA]
jg: Could the keystroke be configurable?
01:35:53 [MattSEA]
susan: Can it be set to the locale?
01:36:24 [MattSEA]
aaron: It's not that simple. We have one set of punctuation keys, Ctrl- and Ctrl+ for font size... you never know where the key is on the keyboard, for example.
01:36:35 [MattSEA]
aaron: left and right bracket are sometimes the same key on a keyboard.
01:37:06 [MattSEA]
aaron: There were so many bugs related to Ctrl-/+. May want to consider Alt-up/down
01:37:16 [MattSEA]
01:37:27 [MattSEA]
kyle: We can try.
01:38:32 [MattSEA]
jg: Since this is hard-coded into the browser, browser would get to respond before the document.
01:39:09 [MattSEA]
aaron: I think it might be possible for the document to get to it first. I think you can register to get keystrokes first. Also security considerations.
01:40:13 [MattSEA]
kyle: All key bindings are in an XBL file?
01:40:21 [MattSEA]
aaron: Anything related to tabbing is in C++, but that should be fixed.
01:42:19 [MattSEA]
aaron: Letting people select their keybindings is a slippery slope because then that becomes the default answer for developers.
01:42:55 [MattSEA]
aaron: i18n comes for the mnemonics, but for the default keystrokes, we haven't had to deal with that. We should allow power users to configure, but should have a default.
01:43:37 [MattSEA]
jg: Agree, should be designed to work out of the box in the most situations.
01:44:07 [MattSEA]
aaron: Somebody's going to write a keyboard config extension for Firefox. But when we talk about a preference panel, etc., we have to think whether it's a design flaw or not.
01:45:13 [MattSEA]
aaron: Someone can write an overlay for Emacspeak, etc., but in terms of writing an interface for keyboard config, that's off the radar.
01:45:29 [MattSEA]
jg: Think the only place it'd be an issue is with screen reader conflicts.
01:46:04 [MattSEA]
pk: Gnopernicus uses only numeric keyboard by default.
01:46:31 [MattSEA]
aaron: Maybe should make Ctrl-Up/Down. Could move the caret to previous form control.
01:46:41 [MattSEA]
01:48:01 [aaronlev]
aaronlev has joined #ua
01:48:09 [MattSEA]
jg: Planning to update status line so metainformation about a particular element would be there?
01:49:10 [MattSEA]
jg: Want to know if this would be useful. Does this document help?
01:49:28 [MattSEA]
jessie: Yes, it's useful, but need more tests.
01:49:59 [MattSEA]
kyle: We haven't implemented status info for keyboard nav.
01:50:08 [MattSEA]
aaron: What are you proposing the status line says?
01:50:15 [MattSEA]
jg: What the current label is?
01:50:25 [MattSEA]
pk: What happens when you click on it?
01:50:33 [MattSEA]
aaron: Only links update the status.
01:50:48 [MattSEA]
aaron: I don't see a lot of use for status in form controls.
01:50:58 [MattSEA]
jg: Could be used for authors.
01:51:55 [MattSEA]
pk: I'm not aware of it on the radar.
01:52:04 [MattSEA]
pk: Could be implemented without keyboard nav.
01:52:48 [MattSEA]
aaron: Should expect a lot of pushback from Firefox developers, because they wanted all developer features outside of Firefox.
01:53:37 [MattSEA]
susan: There's an option in the install for DOM Inspector.
01:55:16 [MattSEA]
aaron: If you go to a radio button, it should focus on the label and not just the button itself.
01:55:56 [MattSEA]
jg: What happens when in a form control? Form control gets keyboard focus?
01:57:21 [MattSEA]
jg: What if you find an iframe?
01:57:33 [MattSEA]
kyle: tab key can go into iframe.
01:59:23 [MattSEA]
kyle: Need special key to get out of text field.
02:02:24 [JRG]
zakim, who is here?
02:02:24 [Zakim]
On the phone I see Jon, Matt, Peter_Korn, Susan_Crayne, Jessie_Li, Aaron
02:02:26 [Zakim]
On IRC I see aaronlev, hopgrassor, kyle, korn, JRG, RRSAgent, Zakim, MattSEA
02:03:09 [MattSEA]
kyle: Keyboard nav patch may not be accepted by the community, because it's huge.
02:10:00 [MattSEA]
jg: Next meeting Dec 6
02:10:24 [Zakim]
02:10:25 [Zakim]
02:10:26 [Zakim]
02:10:28 [Zakim]
02:10:35 [MattSEA]
rrsagent, pointer?
02:10:35 [RRSAgent]
02:10:35 [Zakim]
02:10:36 [Zakim]
02:10:37 [Zakim]
WAI_UAWG()8:00PM has ended
02:10:38 [Zakim]
Attendees were Matt, Peter_Korn, Jon, Susan_Crayne, Jessie_Li, Aaron
02:11:01 [aaronlev]
kyle, still there?
02:11:15 [kyle]
02:11:31 [aaronlev]
I rewrote a lot of the Mozilla tab navigation code in the trunk
02:11:48 [aaronlev]
There are new methods, nsIFrame::IsFocusable() and nsIContent::IsFocusable()
02:12:00 [aaronlev]
which also return whether something is tab navigable
02:12:17 [aaronlev]
nsEventStateManager::GetNextTabbableContent() is much shorter
02:12:22 [kyle]
all these are not in caret browing mode?
02:12:34 [aaronlev]
They only affect tabbing
02:13:09 [kyle]
how can these help us?
02:13:33 [aaronlev]
I wanted to tell you about it, because it may make merging code more difficult for you
02:13:47 [kyle]
oh, I see.
02:13:58 [aaronlev]
I think it's going to be difficult for you to make patches that apply to the current trunk
02:14:05 [aaronlev]
1.7 was branched a long time ago
02:14:20 [aaronlev]
I think the ATK stuff will be a lot easier than the keyboard stuff
02:14:32 [kyle]
yes, we also realized that porting patches is very diffucult
02:14:42 [kyle]
yes, agree
02:14:50 [aaronlev]
I'm looking forward to working together though
02:15:15 [aaronlev]
I plan to upgrade to Fedora 3 as soon as your team starts putting up patches
02:15:17 [kyle]
btw, did you mean (on the phone) that we should send some engineer to Boston to work with you together? (I probably got you wrong)
02:15:32 [aaronlev]
It was an idea I discusses with Peter at the last meeting
02:15:51 [aaronlev]
He said we could all talk this week
02:16:13 [kyle]
Do you think that is the most efficient way to port patch?
02:16:21 [aaronlev]
02:16:30 [aaronlev]
But expensive
02:16:35 [kyle]
and how long do you think it will take?
02:16:44 [aaronlev]
I haven't seen the code yet
02:16:47 [aaronlev]
so i don't know
02:17:06 [aaronlev]
I think it's probably better to get started without doing that
02:17:10 [aaronlev]
and see how well that goes
02:17:19 [kyle]
I was told some code they changed are yours
02:17:31 [aaronlev]
i'm sure, my code is not perfect :)
02:18:08 [aaronlev]
i had to change some of your code too :P
02:18:20 [kyle]
I didn't mean that :) I meant that your help will more appriciated.
02:18:47 [aaronlev]
Ok. Well if it touches keyboard navigation or accessibility api stuff I want to see it
02:19:06 [kyle]
sorry for the last word, I always can not spell it correctly :(
02:19:22 [aaronlev]
I can't spell either
02:19:29 [aaronlev]
Don't worry, English is hard
02:19:38 [aaronlev]
The only easy language is C++
02:19:51 [kyle]
Ginn is on vacation this week, I'll push other team member to post their patches to bugzilla from this week.
02:19:53 [aaronlev]
Actually, you spell better than most of the people on #mozilla
02:20:13 [aaronlev]
Unfortunately I am mostly on vacation from November 20-30
02:20:18 [kyle]
what is the event you mentioned on Feb?
02:20:20 [aaronlev]
but I will try to do some reviews
02:20:56 [kyle]
sorry, I meant the event on Feb, not you mentioned on Feb
02:21:12 [aaronlev]
looking for a link
02:21:15 [aaronlev]
W3C technical plenary
02:21:26 [aaronlev]
MattSEA, you have a link to that?
02:21:30 [kyle]
what's that?
02:21:39 [kyle]
I can try google
02:22:58 [MattSEA]
tech plenary? 1 sec.
02:23:09 [kyle]
I found the link for the 2004
02:23:35 [MattSEA]
No page for it yet.
02:23:52 [MattSEA]
It's 28 Feb-5 Mar in Boston.
02:24:15 [kyle]
aaronlev, both you and Peter will be there?
02:25:17 [aaronlev]
I will
02:25:24 [aaronlev]
Don't know about Peter
02:27:17 [aaronlev]
Do you think Sun will work on a separate branch for future accessibility in Mozilla?
02:27:30 [kyle]
aaronlev, I'm going to back to my seat. If you want to discuss more, I can login from there.
02:27:47 [aaronlev]
ok, i will see you later
02:28:00 [aaronlev]
join #developers sometimes to chat
02:28:27 [kyle]
ok, see you in #developers of
02:30:06 [MattSEA]
MattSEA has left #ua
02:31:17 [kyle]
kyle has joined #ua
02:31:37 [kyle]
aaronlev, are you still there?
03:49:10 [kyle]
kyle has left #ua
06:34:25 [korn]
korn has left #ua