/irc-logs / mozilla / #accessibility / 2013-05-24 / end
Options:
- # Session Start: Fri May 24 00:00:01 2013
- # Session Ident: #accessibility
- # [00:04] * Quits: victorporof (victorporo@7362939B.70F4578F.79933D60.IP) (Quit: victorporof)
- # [00:15] * Quits: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net) (Quit: scott_gonzalez)
- # [00:51] * Joins: Jesse (jruderman@2557E599.66715431.D25A875A.IP)
- # [00:53] * Joins: scott_gonzalez (scott_gonz@moz-8E5C24A9.hsd1.pa.comcast.net)
- # [01:01] * Quits: Jesse (jruderman@2557E599.66715431.D25A875A.IP) (Ping timeout)
- # [01:01] * Quits: jongund (chatzilla@moz-97EEA285.rehab.uiuc.edu) (Ping timeout)
- # [01:02] * Joins: Jesse (jruderman@2557E599.66715431.D25A875A.IP)
- # [01:06] * Quits: Jesse (jruderman@2557E599.66715431.D25A875A.IP) (Ping timeout)
- # [01:07] * Joins: Jesse (jruderman@2557E599.66715431.D25A875A.IP)
- # [01:09] * Quits: Jesse (jruderman@2557E599.66715431.D25A875A.IP) (Connection reset by peer)
- # [01:10] * Joins: Jesse (jruderman@2557E599.66715431.D25A875A.IP)
- # [01:21] * Quits: Jesse (jruderman@2557E599.66715431.D25A875A.IP) (Ping timeout)
- # [01:21] * Joins: Jesse (jruderman@2557E599.66715431.D25A875A.IP)
- # [01:24] * Joins: scottgonzalez (scott_gonz@moz-8E5C24A9.hsd1.pa.comcast.net)
- # [01:24] * Quits: scott_gonzalez (scott_gonz@moz-8E5C24A9.hsd1.pa.comcast.net) (Connection reset by peer)
- # [01:25] * Quits: Jesse (jruderman@2557E599.66715431.D25A875A.IP) (Ping timeout)
- # [02:26] <@firebot> jamie@nvaccess.org changed the Status on bug 786163 from RESOLVED to VERIFIED.
- # [02:26] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=786163 nor, --, mozilla21, surkov.alexander, VERI FIXED, Sort out name calculation for HTML input buttons
- # [02:52] * Quits: scottgonzalez (scott_gonz@moz-8E5C24A9.hsd1.pa.comcast.net) (Quit: scottgonzalez)
- # [03:10] * khuey|away is now known as khuey|tw
- # [04:05] * Joins: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net)
- # [04:16] * Quits: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net) (Quit: scott_gonzalez)
- # [05:04] * Quits: peteb-away (ptbrunet@moz-B51E1692.austin.res.rr.com) (Client exited)
- # [05:53] * Joins: Jesse (jruderman@moz-537BCF9.hsd1.ca.comcast.net)
- # [06:07] <@firebot> philringnalda@gmail.com changed the Resolution on bug 873412 from --- to FIXED.
- # [06:07] <@firebot> philringnalda@gmail.com changed the Status on bug 873412 from NEW to RESOLVED.
- # [06:07] <@firebot> philringnalda@gmail.com changed the Target Milestone on bug 873412 from --- to mozilla24.
- # [06:07] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=873412 nor, --, mozilla24, maxli, RESO FIXED, [AccessFu] Implement performAction for clicking
- # [06:07] <@firebot> philringnalda@gmail.com changed the Resolution on bug 873411 from --- to FIXED.
- # [06:07] <@firebot> philringnalda@gmail.com changed the Status on bug 873411 from NEW to RESOLVED.
- # [06:07] <@firebot> philringnalda@gmail.com changed the Target Milestone on bug 873411 from --- to mozilla24.
- # [06:07] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=873411 nor, --, mozilla24, maxli, RESO FIXED, [AccessFu] Provide extra node information to Brailleback
- # [06:15] * Joins: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net)
- # [06:48] * Joins: icaaq (Adium@moz-C35543CE.cust.bredbandsbolaget.se)
- # [07:10] * Quits: Jesse (jruderman@moz-537BCF9.hsd1.ca.comcast.net) (Quit: Jesse)
- # [07:19] * Quits: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net) (Quit: scott_gonzalez)
- # [07:21] * Quits: unicorn (fxa90id@moz-F3DFED37.dsl.dynamic.t-mobile.pl) (Ping timeout)
- # [07:47] * Quits: nhirata (anonymous@moz-1A50F7F8.hsd1.ca.comcast.net) (Ping timeout)
- # [08:01] * Joins: SteveF (chatzilla@moz-3F778890.cable.virginmedia.com)
- # [08:08] * Joins: Wheelman (Wheelman@87B8A04.29271E3B.360EF119.IP)
- # [08:15] * khuey|tw is now known as khuey|away
- # [08:16] * khuey|away is now known as khuey|tw
- # [08:31] <@firebot> New Core - Disability Access APIs bug 875665 filed by marco.zehe@googlemail.com.
- # [08:31] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=875665 nor, --, ---, nobody, NEW, [AccessFu] Keyboard does not come up when double-tapping the "What are you doing?" text area in Face
- # [08:34] * Joins: victorporof (victorporo@7362939B.70F4578F.79933D60.IP)
- # [08:49] * Joins: Wheelman_ (Wheelman@24A29F78.1B7FAD2C.7D0E51E.IP)
- # [08:50] * Quits: Wheelman (Wheelman@87B8A04.29271E3B.360EF119.IP) (Ping timeout)
- # [08:50] * Joins: Wheelman (Wheelman@32108533.FF82A40A.7D0E51E.IP)
- # [08:51] * Quits: Wheelman_ (Wheelman@24A29F78.1B7FAD2C.7D0E51E.IP) (Ping timeout)
- # [09:16] * Joins: nhirata (anonymous@moz-1A50F7F8.hsd1.ca.comcast.net)
- # [09:21] * Quits: nhirata (anonymous@moz-1A50F7F8.hsd1.ca.comcast.net) (Quit: nhirata)
- # [09:57] * Quits: Wheelman (Wheelman@32108533.FF82A40A.7D0E51E.IP) (Quit: Leaving)
- # [10:39] * Quits: brambles (xymox@moz-969AAE9B.barwen.ch) (Ping timeout)
- # [10:39] * Joins: brambles (xymox@moz-969AAE9B.barwen.ch)
- # [11:25] * Quits: SteveF (chatzilla@moz-3F778890.cable.virginmedia.com) (Ping timeout)
- # [11:26] * Joins: SteveF (chatzilla@moz-3F778890.cable.virginmedia.com)
- # [11:37] * Quits: victorporof (victorporo@7362939B.70F4578F.79933D60.IP) (Quit: victorporof)
- # [12:04] * Quits: SteveF (chatzilla@moz-3F778890.cable.virginmedia.com) (Connection reset by peer)
- # [12:20] * Quits: brambles (xymox@moz-969AAE9B.barwen.ch) (Ping timeout)
- # [12:20] * Joins: brambles (xymox@moz-969AAE9B.barwen.ch)
- # [12:33] * khuey|tw is now known as khuey|away
- # [12:47] * Joins: victorporof (victorporo@moz-23D6B301.info.uaic.ro)
- # [13:09] * Quits: brambles (xymox@moz-969AAE9B.barwen.ch) (Ping timeout)
- # [13:09] * Joins: brambles (xymox@moz-969AAE9B.barwen.ch)
- # [13:16] * Quits: victorporof (victorporo@moz-23D6B301.info.uaic.ro) (Quit: victorporof)
- # [13:22] * Joins: SteveF (chatzilla@moz-3F778890.cable.virginmedia.com)
- # [13:38] * Quits: fxa90id (fxa90id@moz-F3DFED37.dsl.dynamic.t-mobile.pl) (Ping timeout)
- # [13:38] * Joins: fxa90id_ (fxa90id@moz-BBC7E408.dsl.dynamic.t-mobile.pl)
- # [13:58] * Quits: fxa90id_ (fxa90id@moz-BBC7E408.dsl.dynamic.t-mobile.pl) (Quit: Leaving)
- # [13:58] * Joins: fxa90id (fxa90id@moz-BBC7E408.dsl.dynamic.t-mobile.pl)
- # [14:04] * Joins: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net)
- # [14:04] <@firebot> surkov.alexander@gmail.com granted review for attachment 753378 on bug 869280.
- # [14:05] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=869280 nor, --, ---, eitan, NEW, Add temporary modal subtree to pivot API
- # [14:13] * Joins: peteb-away (ptbrunet@moz-B51E1692.austin.res.rr.com)
- # [14:19] * Joins: marcoz (marco.zehe@moz-C1E9D8F1.dip0.t-ipconnect.de)
- # [14:19] * ChanServ sets mode: +o marcoz
- # [14:29] * Joins: victorporof (victorporo@moz-23D6B301.info.uaic.ro)
- # [14:39] <SteveF> marcoz; hello facebook buddy :-)
- # [14:43] <@marcoz> Hi SteveF! :)
- # [14:45] <@marcoz> SteveF: FB have really made some strides in improving accessibility. Things work much much better today than they did a year ago.
- # [14:46] <SteveF> marcoz: thats great!
- # [14:46] <@marcoz> SteveF: I'm actually right now collecting info to make into a blog post about this which I'll publish in a few days.
- # [14:47] <SteveF> marcoz: cool! look forward to it
- # [14:51] * Quits: victorporof (victorporo@moz-23D6B301.info.uaic.ro) (Quit: victorporof)
- # [14:55] <@marcoz> SteveF: It also makes for some interesting browser bug findings, too. For example NVDA + Firefox read the checked/unchecked states when you edit the friend and list info for a Facebook friend, Safari+VoiceOver do not.
- # [14:56] <@marcoz> SteveF: I mean the checkable menu items if somebody is a close friend, an acquaintance etc.
- # [14:56] <SteveF> right, so nvda/firefox get it right
- # [15:05] <@marcoz> SteveF: Yes. I just filed a bug with Apple.
- # [15:05] * Joins: victorporof (victorporo@moz-23D6B301.info.uaic.ro)
- # [15:25] * Joins: davidb (davidb@F2D29657.F60B0462.67AC9B1.IP)
- # [15:25] * ChanServ sets mode: +qo davidb davidb
- # [15:26] * Quits: @davidb (davidb@F2D29657.F60B0462.67AC9B1.IP) (Quit: blast off!)
- # [15:26] * Joins: davidb (davidb@F2D29657.F60B0462.67AC9B1.IP)
- # [15:26] * ChanServ sets mode: +qo davidb davidb
- # [15:27] <@davidb> heyo!
- # [15:29] <@marcoz> Hi davidb!
- # [15:31] <SteveF> oh its bolder time...
- # [15:31] <@davidb> is that like hammer time?
- # [15:32] <@davidb> can't touch this
- # [15:37] * Joins: habber (habber@moz-8654C0B7.nyc.res.rr.com)
- # [15:38] * Joins: jongund (chatzilla@moz-97EEA285.rehab.uiuc.edu)
- # [15:38] <SteveF> similar sub hammer though ;-)
- # [15:38] <SteveF> if i was ta school with you i would have called you bolder
- # [15:39] <SteveF> since i wasn't i have to do it now
- # [15:40] * Quits: jongund (chatzilla@moz-97EEA285.rehab.uiuc.edu) (Quit: ChatZilla 0.9.90 [Firefox 21.0/20130511120803])
- # [15:40] <@marcoz> LOL
- # [15:40] <SteveF> currently blown away after reading http://www.w3.org/WAI/history realisation that my boss why instrumental in WAI
- # [15:49] <@davidb> Yuri Rubinski was tight with Jutta
- # [15:49] <@davidb> and also, lol
- # [15:49] <@davidb> most people called me bolter
- # [15:50] <@firebot> eitan@monotonous.org requested review from surkov.alexander@gm ail.com for attachment 753767 on bug 775621.
- # [15:51] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=775621 nor, --, ---, nobody, NEW, Add pivot traversal flag for ignoring aria-hidden
- # [15:54] * @marcoz grits his teeth. aria-hidden. Grrrrrrrrrrrrr *bang bang bang bang* :)
- # [16:00] <@tbsaunde> marcoz: talk to eeejay I'd happily wontfix any bugs to support it more
- # [16:00] * @eeejay hides
- # [16:00] * Joins: yzen (Adium@AA2C085E.10C8AEF8.9A8C35B4.IP)
- # [16:01] <@marcoz> eeejay: :) Serious question: Why do you come back to this bug now? Is there a use case that prompted this?
- # [16:01] <@marcoz> Hi yzen!
- # [16:01] <@eeejay> marcoz, i would love to be a purist about it. but there are some cases where i don't see alternatives
- # [16:02] * Joins: maxli (maxli@F2D29657.F60B0462.67AC9B1.IP)
- # [16:02] <@marcoz> eeejay: Like?
- # [16:02] <@eeejay> marcoz, the drawer in gaia building blocks
- # [16:02] <yzen> marcoz: morning/afternoon :)
- # [16:02] <@eeejay> yzen, morning
- # [16:02] <yzen> eeejay: morning
- # [16:03] <@firebot> surkov.alexander@gmail.com granted review for attachment 753767 on bug 775621.
- # [16:03] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=775621 nor, --, ---, nobody, NEW, Add pivot traversal flag for ignoring aria-hidden
- # [16:03] <@marcoz> eeejay: Is the drawer a full set of items, or just one item? if just one, role="presentation" should do the trick just fine.
- # [16:04] <@eeejay> marcoz, it is a container with a big hierarchy
- # [16:04] <@marcoz> eeejay: And what does that drawer actually contain?
- # [16:04] <@marcoz> Remember that aria-hidden nukes a whole subtree.
- # [16:04] <@marcoz> I'm just trying to understand this thing.
- # [16:07] <@eeejay> marcoz, in the example in the building blocks. they show an email inbox.
- # [16:08] <@eeejay> marcoz, when you press a button on the upper left corner, it pans to the right, almost out of the screen, but with maybe 15% still showing
- # [16:08] * Quits: habber (habber@moz-8654C0B7.nyc.res.rr.com) (Quit: habber)
- # [16:08] <@eeejay> and under it is revealed a list of email folders
- # [16:10] <@eeejay> so when we show the list of email folders, there is still a visual hint that the inbox is waiting on the right, but functionally and perceptibly we are not in that screen.
- # [16:12] <@eeejay> marcoz, the browser app uses drawers as the main interface
- # [16:12] <@eeejay> marcoz, in that instance, a button with the tab count is on the right top corner
- # [16:13] <yzen> eeejay: re 2 outstanding patches for the utterance, should they go in soon, i know you mentioned you wanted another one opened that deals with subtrees..
- # [16:13] <@tbsaunde> if its 15% isn't that enough that you can read the start of lines or something?
- # [16:13] <@eeejay> pressing it brings up the tabs panel by panning the web view left, almost all the way (but not entirely)
- # [16:13] <@eeejay> tbsaunde, you could maybe read half a word on each line
- # [16:15] <@tbsaunde> so it seems like the real problem is arguably then that an accessible has to be either on screen or off screen and can't be partially onscreen?
- # [16:16] <@eeejay> tbsaunde, not sure what you mean
- # [16:16] <@eeejay> yzen, oh right..
- # [16:17] <@tbsaunde> eeejay: I mean an accessible either has the onscreen state or it doesn't
- # [16:17] <yzen> eeejay: not sure if they still apply though
- # [16:17] <@tbsaunde> you can't say the bounds of this accessible are this and its partially onscreen but only this part of its on screen
- # [16:18] <@eeejay> tbsaunde, ah. so you are saying that we should change the state to offscreen if a certain proportion is clipped?
- # [16:19] <@eeejay> yzen, seems like there is one more change needed. accorsing to the last comment
- # [16:20] <@tbsaunde> eeejay: not exactly I'm saying we can't really give a good answer to the question is this accessible offscreen given the current api
- # [16:20] <@eeejay> tbsaunde, i know. that is why i am resorting to aria-hidden. unless somebody comes up with something else
- # [16:21] <@tbsaunde> eeejay: I'd say you should do that since you're the person who wants it ;)
- # [16:21] <@tbsaunde> but aria-hidden there is sort of wrong
- # [16:22] <@eeejay> tbsaunde, also, offscreen state is not usable in this case. since many sites put special screen reader test in negative offsets
- # [16:23] <@tbsaunde> sure, I'm mostly argueing that 1 aria-hidden is broken and 2 you should be the one to come up with something better since you want it
- # [16:26] <@eeejay> tbsaunde, for our case, aria-hidden works fine. but i am open to alternatives if people so badly want to see it die
- # [16:27] <@tbsaunde> I'd say it only mostly works
- # [16:30] * Quits: victorporof (victorporo@moz-23D6B301.info.uaic.ro) (Ping timeout)
- # [16:34] <@marcoz> yzen: eeejay: I'll be curious to see what the solution to bug 875665 is. Really curious why keyboard won't come up for these text fields.
- # [16:34] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=875665 nor, --, ---, nobody, NEW, [AccessFu] Keyboard does not come up when double-tapping the "What are you doing?" text area in Face
- # [16:35] <@eeejay> marcoz, yeah. that sucks that we never saw that before
- # [16:35] <@eeejay> i am pretty sure twitter used to work
- # [16:35] <@marcoz> eeejay: You may have found a test case where aria-hidden is currently indeed the only solution. Like tbsaunde said, there is no current way for the API to specify that something is partially clipped and should therefore be ignored.
- # [16:36] * Joins: surkov (surkov@4F778CCB.95871A73.E17943EE.IP)
- # [16:36] * ChanServ sets mode: +o surkov
- # [16:36] <@eeejay> right
- # [16:36] <@marcoz> eeejay: Yes, Twitter worked at some point, but when I checked Facebook today, I also checked Mobile Twitter and found that the same problem exists there.
- # [16:36] <@eeejay> surkov, you missed a whole aria-hidden conversation, your favorite :)
- # [16:36] <@surkov> :)
- # [16:37] <yzen> eeejay: oh ya i missed that one :) , ill fix it up
- # [16:37] <fxa90id> :)
- # [16:37] <fxa90id> surkov, :D
- # [16:44] <@tbsaunde> marcoz: so the thing I don't think things should be ignored if they're only partially visible, which means aria-hidden is wrong
- # [16:46] <@tbsaunde> surkov: so, I was looking at protecting WillRefresh() from reentering but it looks like we set mObservingState back to mRefreshObserving before firing events
- # [16:48] <@surkov> tbsaunde: it seems we need to add extra state for mRefreshObserving
- # [16:48] <@marcoz> tbsaunde: The thing is: This partially visible thing on the right cannot be interacted with if I understood eeejay correctly. This means that presenting it on-screen would confuse the screen reader user. This would be the same as if a hidden side bar was leaking through to the front when it is not supposed to, because it is currently out of sight.
- # [16:49] * Quits: brambles (xymox@moz-969AAE9B.barwen.ch) (Ping timeout)
- # [16:50] <@tbsaunde> marcoz: well, as I understand it you can interact with it a little just not nearly completely otherwise why is it visible at all
- # [16:50] * Joins: brambles (xymox@moz-969AAE9B.barwen.ch)
- # [16:50] <@eeejay> tbsaunde, you can't interact with it at all
- # [16:50] <@marcoz> eeejay: Can you try to explain again what this partially visible thing is supposed to do/indicate, and how you get it back into full view as a sighted user?
- # [16:50] <@tbsaunde> surkov: like mObservingDuringEvents you mean?
- # [16:51] <@marcoz> eeejay: Ah thanks, i thought so.
- # [16:51] <@tbsaunde> eeejay: I thought you said you could read parts of words which sounds like interacting to me
- # [16:51] <@tbsaunde> not much I'll grant of course, but > 0
- # [16:51] <@eeejay> marcoz, , it is a visual queue to tell the user that they are not on the main screen
- # [16:52] <@surkov> tbsaunde, like eRefreshProcessing to prevent reentering
- # [16:54] <@eeejay> marcoz, tbsaunde, it is not meant to be read.
- # [16:55] <@eeejay> marcoz, this might be a stretched metaphor: imagine a scene in a play that takes place in a crowded space. there is a lot of indiscernible murmuring that is not meant to be understood, just to give the effect of a crowded scene
- # [16:56] <@marcoz> eeejay: Wonder why they chose to do it this way. Sounds like quite a bit of screen estate of the 3 inch screen is taken away just by this visual indicator.
- # [16:56] <@eeejay> in this case the screen slides out of view, but just enough is visible to know it is there - but not to read or interact with
- # [16:57] <@eeejay> marcoz, that might be true. aesthetically it looks good
- # [16:58] <@marcoz> eeejay: Yeah…..Funny what looks good sometimes. xD
- # [16:58] <@eeejay> in a windowed environment, imagine if you were tabbing in a window, and it just switched to the next window
- # [16:58] <@marcoz> #eeejayAnyway, you're right that we don't want to interact with this bit, only the bit that allows us to slide the hidden part back into view.
- # [16:58] <@eeejay> yup
- # [17:02] <@firebot> emorley@mozilla.com changed the Resolution on bug 875201 from --- to FIXED.
- # [17:02] <@firebot> emorley@mozilla.com changed the Status on bug 875201 from NEW to RESOLVED.
- # [17:02] <@firebot> emorley@mozilla.com changed the Target Milestone on bug 875201 from --- to mozilla24.
- # [17:02] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=875201 nor, --, mozilla24, surkov.alexander, RESO FIXED, move getText at word boundary tests into test_wordboundery.html
- # [17:07] <@tbsaunde> surkov: it seems I forgot how shuting down accessibles is linked to firing events which makes delaying xpcom events kind of tricky
- # [17:08] <@surkov> tbsaunde: iirc there was a problem on atk since tree changes results in events
- # [17:08] <@tbsaunde> I guess we culd do that in the runnable to, but I wonder if we should just move the observer code before ProcessEventQueue()
- # [17:09] <@tbsaunde> surkov: that's not what I was thinking about, look at the stuff in EventQueue::ProcessEvents()
- # [17:09] <@marcoz> surkov: Bug 873358 causes NVDA to not show a blank line on a braille display when pressing Enter to insert one at the end of a textarea.
- # [17:09] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=873358 nor, --, mozilla24, surkov.alexander, RESO FIXED, textarea claims bogus trailing \n (HTML br)
- # [17:10] <@marcoz> surkov: To reproduce, one just needs to type something in a textarea and press enter. The expected behavior is that the cursor goes onto a new blank line. However, it stays at the end of the previous line and blinks after the last character.
- # [17:11] <@surkov> marcoz: could you file a bug please?
- # [17:12] <@surkov> tbsaunde: not sure I get you
- # [17:12] <@marcoz> surkov: Sure!
- # [17:15] <@tbsaunde> surkov: I mean moving firing xpcom events to after ProcessEventQueue() would cause them to happen after we've unbound the accessibles from the document if it was a hide event
- # [17:15] <@surkov> true
- # [17:16] <@tbsaunde> surkov: which I think e probably can't do?
- # [17:17] <@surkov> it won't be good I think
- # [17:18] <@tbsaunde> surkov: so I'm thinking we move the stuff to remove the refresh observer before firing events so that isn't a concern
- # [17:19] <@surkov> tbsaunde: why would we need to remove it before events processing?
- # [17:19] <@surkov> it may make us to remove and then add it
- # [17:19] <@tbsaunde> surkov: so we can be sure funny things don't happen if we renter
- # [17:20] <@surkov> didn't we wanted to add a guard to make sure we never reenter
- # [17:21] <@tbsaunde> surkov: yeah, ignore me there
- # [17:21] <@tbsaunde> surkov: but that doesn't solve the out of order events problem, but I'm not sure we can solve that :(
- # [17:21] <@surkov> maybe it's not big deal
- # [17:22] <@marcoz> Ouch, being bitten hard by bug 875757. Can hardly do anything with today's nightly.
- # [17:22] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=875757 blo, --, ---, general, NEW, crash in js::ion::DoTypeUpdateFallback
- # [17:22] <@tbsaunde> surkov: yeah that's what I'm thinking, but what do we do about the assert then?
- # [17:23] <@surkov> assert on reentrance? probably we shouldn't add it since reentrance can happen
- # [17:24] <@tbsaunde> surkov: no, I mean the assert in AccessibleWrap.cpp that fires because the doc accessible that's the event target is defunct
- # [17:25] <@surkov> tbsaunde: that assert is about events typing (if you mean atk file) so it makes sense to keep it
- # [17:26] <@surkov> I wouldn't process an event if IsDefunct() triggers
- # [17:27] <@tbsaunde> surkov: so only assert if event target isn't defunct?
- # [17:27] <@firebot> New Core - Disability Access APIs bug 875794 filed by marco.zehe@googlemail.com.
- # [17:27] <@surkov> yep
- # [17:27] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=875794 maj, --, ---, nobody, NEW, After bug 873358, when inserting a newline at the end of a textarea, the braille display does no lon
- # [17:29] <@tbsaunde> surkov: actually I think maybe checking if WillRefresh() is reentering may fix the assert because we won't fire the hide event until the js finishes and we get back to the first WillRefresh()
- # [17:31] <@tbsaunde> yeah, concurrency is hard and I'm not sure but I thinking bailing out of WillRefresh() if its reentering will fix the issue
- # [17:31] <@firebot> surkov.alexander@gmail.com requested needinfo from jamie@nvaccess.or g on bug 875794.
- # [17:32] <@surkov> it could be
- # [17:44] <yzen> eeejay: so if the pref cache supports Ci.nsISupportsWeakReference removing it will remove the observer as well ?
- # [17:45] <@eeejay> yzen, i'm assuming lazily..
- # [17:45] <@eeejay> yzen, when the observer will be called next time a pref is changed, the service will see it as dead and remove it
- # [17:47] <@firebot> eitan@monotonous.org requested review from surkov.alexander@gm ail.com for attachment 753818 on bug 775621.
- # [17:47] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=775621 nor, --, ---, nobody, NEW, Add pivot traversal flag for ignoring aria-hidden
- # [17:47] <yzen> eeejay: right, hence the aHoldWeak set to true in addObserver
- # [17:47] <@eeejay> yzen, yes
- # [17:50] <@firebot> surkov.alexander@gmail.com granted review for attachment 753818 on bug 775621.
- # [17:52] <@tbsaunde> eeejay: what is your use case for observing pref changes at all instead of just getting it each time? (just out of curosity)
- # [17:53] <@eeejay> tbsaunde, changing behavior real-time
- # [17:53] <@eeejay> tbsaunde, plus, caching. although i don't know if we really need that
- # [17:53] <@tbsaunde> eeejay: I really doubt a cache is a good thing
- # [17:54] <@tbsaunde> getting prefs is just a virtual call and a hash table look or so
- # [17:54] <@tbsaunde> I don't see why you can't change behaviour by just getting the pref each time
- # [17:55] <@eeejay> tbsaunde, hm does js/xpconnect have any cost?
- # [17:55] <@eeejay> tbsaunde, the behavior is more like activate/deactivate
- # [17:55] <@tbsaunde> eeejay: some, not sure exactly how much
- # [17:56] <@tbsaunde> but I'm pretty sure I'd want to see data showing a cahce is a good idea
- # [17:56] <@eeejay> tbsaunde, yeah. since we set up observers anyway, it is easy to cache. but if we didn't it might not be worth it
- # [17:57] <@eeejay> tbsaunde, my assumption is that a cache is faster or equal to getting it each time
- # [17:57] <@tbsaunde> probably true if you really need an observer anyway
- # [18:05] * Quits: @surkov (surkov@4F778CCB.95871A73.E17943EE.IP) (Quit: surkov)
- # [18:09] <@marcoz> Have a good weekend everyone!
- # [18:09] <fxa90id> !
- # [18:10] <fxa90id> you too :D
- # [18:10] * Quits: @marcoz (marco.zehe@moz-C1E9D8F1.dip0.t-ipconnect.de) (Quit: Leaving.)
- # [18:13] <@firebot> yura.zenevich@gmail.com granted review for attachment 752976 on bug 865023.
- # [18:13] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=865023 nor, --, ---, eitan, NEW, [AccessFu] Streamline preference getting and observing
- # [18:15] * Joins: nhirata (anonymous@moz-BE33DA21.fw1.sfo1.mozilla.net)
- # [18:15] * Quits: nhirata (anonymous@moz-BE33DA21.fw1.sfo1.mozilla.net) (Quit: nhirata)
- # [18:15] * Joins: nhirata (anonymous@moz-BE33DA21.fw1.sfo1.mozilla.net)
- # [18:33] * Joins: webatou (Instantbir@455F2973.12C6E01D.BF9F5BB8.IP)
- # [18:43] * Joins: victorporof (victorporo@DE2B5DEF.67CDC207.79933D60.IP)
- # [18:49] * Quits: timeless (uid4015@moz-D8B63698.irccloud.com) (Ping timeout)
- # [18:53] * Quits: yzen (Adium@AA2C085E.10C8AEF8.9A8C35B4.IP) (Quit: Leaving.)
- # [19:08] <fxa90id> :D
- # [19:08] <fxa90id> 58:42.25
- # [19:46] * Joins: yzen (Adium@AA2C085E.10C8AEF8.9A8C35B4.IP)
- # [19:52] * Quits: webatou (Instantbir@455F2973.12C6E01D.BF9F5BB8.IP) (Connection reset by peer)
- # [19:52] * Joins: webatou (Instantbir@455F2973.12C6E01D.BF9F5BB8.IP)
- # [20:16] <@firebot> eitan@monotonous.org changed the Assignee on bug 775621 from nobody@mozilla.org to eitan@monotonous.org.
- # [20:16] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=775621 nor, --, ---, eitan, NEW, Add pivot traversal flag for ignoring aria-hidden
- # [20:26] <@firebot> New Core - Disability Access APIs bug 875874 filed by eitan@monotonous.org.
- # [20:26] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=875874 nor, --, ---, nobody, NEW, [AccessFu] Ignore aria-hidden nodes and their descendants
- # [20:30] <@firebot> eitan@monotonous.org requested review from marco.zehe@googlema il.com for attachment 753898 on bug 875874.
- # [20:46] * Joins: Jesse (jruderman@moz-537BCF9.hsd1.ca.comcast.net)
- # [20:56] <yzen> eeejay: hey
- # [20:56] <@eeejay> yzen, yo
- # [20:57] <yzen> line 44 in accessfu.jsm i dont think you even need that bind(this) there
- # [20:57] <yzen> i might be wrong though
- # [20:57] <@eeejay> yzen, i tested.
- # [20:57] <@eeejay> we do
- # [20:59] <yzen> interesting
- # [21:02] <yzen> eeejay: when i m merging verbosityRoleMap and nameRuleMap, what would be a good name for it ?
- # [21:03] <@eeejay> ruleRoleMap?
- # [21:03] <@eeejay> dunno :)
- # [21:09] <yzen> ok
- # [21:21] <@firebot> marco.zehe@googlemail.com granted review for attachment 753898 on bug 875874.
- # [21:21] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=875874 nor, --, ---, nobody, NEW, [AccessFu] Ignore aria-hidden nodes and their descendants
- # [21:23] <@firebot> trev.saunders@gmail.com requested review from surkov.alexander@gm ail.com for attachment 753917 on bug 869806.
- # [21:23] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=869806 nor, --, ---, trev.saunders, NEW, fix asserts in actions/test_link.html
- # [21:28] <@tbsaunde> davidb: have a chance to try that xhtml thing?
- # [21:28] <@davidb> not yet - will soon
- # [21:29] <@davidb> i'm trying to listen to something
- # [21:29] <@davidb> tbsaunde: does it matter xhtml transitional or strict?
- # [21:30] <@tbsaunde> davidb: sure, no idea
- # [21:30] <@davidb> ok
- # [21:30] <@tbsaunde> I'd probably mimick the test case just to be safe
- # [21:30] <@davidb> ditto
- # [21:33] <@davidb> and there it is (a table row)
- # [21:33] * Quits: webatou (Instantbir@455F2973.12C6E01D.BF9F5BB8.IP) (Ping timeout)
- # [21:33] * Joins: webatou (Instantbir@455F2973.12C6E01D.BF9F5BB8.IP)
- # [21:35] <@tbsaunde> hah
- # [21:36] <fxa90id> tbsaunde, hello you there ?
- # [21:40] * Quits: Jesse (jruderman@moz-537BCF9.hsd1.ca.comcast.net) (Quit: Jesse)
- # [21:44] <fxa90id> is this file is what I should be looking for ? http://mxr.mozilla.org/mozilla-central/source/accessible/src/html/HTMLTableAccessible.h
- # [21:45] <@tbsaunde> fxa90id: what are you looking for?
- # [21:45] <fxa90id> https://bugzilla.mozilla.org/show_bug.cgi?id=789245
- # [21:45] <@firebot> Bug 789245 nor, --, ---, michaljev, NEW, add TableCellAccessible::Row()
- # [21:47] <@tbsaunde> fxa90id: then that is one place you should look yes
- # [21:47] <fxa90id> great :D
- # [21:48] <fxa90id> and I should look for code which is using Row() and change it to RowFor(something)
- # [21:48] <fxa90id> ?
- # [21:51] <@tbsaunde> fxa90id: well code that gets the row somehow
- # [21:52] <fxa90id> :D
- # [21:56] <@davidb> tbsaunde: what is is that we want to test has a parent? which node?
- # [21:56] <@davidb> the row?
- # [21:56] <fxa90id> ;>
- # [21:57] <@tbsaunde> davidb: hm, probably the most importnat thing is no asserts
- # [21:57] <@davidb> that's not an explicit test then
- # [21:58] <@davidb> correct?
- # [21:58] <@tbsaunde> davidb: correct
- # [21:58] <@tbsaunde> davidb: we could also test the table and the tr don't have accessibles
- # [21:58] <@davidb> right
- # [21:59] <@davidb> that seems like the best option
- # [21:59] <@tbsaunde> davidb: yeah
- # [21:59] <@davidb> ok doing that
- # [21:59] <@tbsaunde> I guess you could test the document just has no kids but I'm not sure that's that useful given the other tests
- # [22:02] * Quits: SteveF (chatzilla@moz-3F778890.cable.virginmedia.com) (Ping timeout)
- # [22:11] <@davidb> hmm not sure how to test this nicely…
- # [22:12] <@davidb> tbsaunde: i at least have the mochitest set up to actually run and to do a bogus test. can i hand it back to you?
- # [22:12] <@davidb> it captures jesse's fuzz
- # [22:13] <@tbsaunde> davidb: yeah, I think I can figure out how to test that stuff quickly
- # [22:13] <@davidb> groovy… how do you want it? email?
- # [22:13] <@tbsaunde> sure
- # [22:16] * Quits: icaaq (Adium@moz-C35543CE.cust.bredbandsbolaget.se) (Quit: Leaving.)
- # [22:18] <@tbsaunde> davidb: got it thanks
- # [22:18] <@davidb> np
- # [22:19] <@davidb> tbsaunde: there was something in the test makefile… some opt group test addition that didn't seem right… so i removed it… please add it back if you need it
- # [22:19] <@davidb> perhaps it was cruft from another patch or something
- # [22:20] * Joins: Jesse (jruderman@2557E599.66715431.D25A875A.IP)
- # [22:23] <@tbsaunde> possible, I'm not sure
- # [22:27] * Joins: fxa90id_ (fxa90id@moz-BBC7E408.dsl.dynamic.t-mobile.pl)
- # [22:27] * fxa90id_ is now known as unicorn
- # [22:39] * Quits: @davidb (davidb@F2D29657.F60B0462.67AC9B1.IP) (Ping timeout)
- # [23:32] * Quits: maxli (maxli@F2D29657.F60B0462.67AC9B1.IP) (Quit: Leaving.)
- # [23:35] * Quits: unicorn (fxa90id@moz-BBC7E408.dsl.dynamic.t-mobile.pl) (Quit: Wychodzi)
- # [23:40] * Quits: victorporof (victorporo@DE2B5DEF.67CDC207.79933D60.IP) (Quit: victorporof)
- # [23:43] * Quits: yzen (Adium@AA2C085E.10C8AEF8.9A8C35B4.IP) (Quit: Leaving.)
- # [23:43] * Joins: yzen (Adium@AA2C085E.10C8AEF8.9A8C35B4.IP)
- # [23:50] * Quits: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net) (Quit: scott_gonzalez)
- # [23:53] * Quits: webatou (Instantbir@455F2973.12C6E01D.BF9F5BB8.IP) (Quit: Instantbird 1.3 -- http://www.instantbird.com)
- # Session Close: Sat May 25 00:00:00 2013
The end :)