/irc-logs / mozilla / #accessibility / 2014-09-29 / end
Options:
- # Session Start: Mon Sep 29 00:00:00 2014
- # Session Ident: #accessibility
- # [00:38] * Joins: Jamie (Instantbir@moz-CA26021.jantrid.net)
- # [01:21] * Quits: Jamie (Instantbir@moz-CA26021.jantrid.net) (Ping timeout)
- # [01:45] * Joins: surkov (surkov@moz-4AE5762E.cpe.teksavvy.com)
- # [01:45] * ChanServ sets mode: +o surkov
- # [02:15] * Quits: @surkov (surkov@moz-4AE5762E.cpe.teksavvy.com) (Quit: surkov)
- # [02:20] * Joins: Jamie (Instantbir@moz-CA26021.jantrid.net)
- # [03:37] * Joins: surkov (surkov@moz-4AE5762E.cpe.teksavvy.com)
- # [03:37] * ChanServ sets mode: +o surkov
- # [03:51] * Joins: yliao (yliao@moz-99690620.hinet-ip.hinet.net)
- # [04:34] * Joins: yzen (yzen@moz-F62769B5.cpe.pppoe.ca)
- # [04:34] * ChanServ sets mode: +o yzen
- # [04:37] * Quits: @yzen (yzen@moz-F62769B5.cpe.pppoe.ca) (Ping timeout)
- # [05:47] * Quits: @surkov (surkov@moz-4AE5762E.cpe.teksavvy.com) (Quit: surkov)
- # [06:09] <@firebot> maxli@maxli.ca changed the Assignee on bug 1011886 from nobody@mozilla.org to maxli@maxli.ca.
- # [06:09] <@firebot> https://bugzil.la/1011886 — NEW, maxli@maxli.ca — [AccessFu] Introduce key echo by character, word, and character and word
- # [07:35] * Joins: icaaq (icaaq@moz-683D578C.cust.bredbandsbolaget.se)
- # [07:58] * Quits: maxli (maxli@moz-B19F68ED.student.cs.uwaterloo.ca) (Quit: Leaving.)
- # [08:09] * Joins: ioanachiorean_ (ioanachior@6DC7A5F5.AA1FA0D2.6A4F8DA2.IP)
- # [09:48] * Quits: icaaq (icaaq@moz-683D578C.cust.bredbandsbolaget.se) (Connection reset by peer)
- # [09:48] * Joins: icaaq (icaaq@moz-683D578C.cust.bredbandsbolaget.se)
- # [09:58] * Quits: yliao (yliao@moz-99690620.hinet-ip.hinet.net) (Client exited)
- # [10:02] * Joins: yliao (yliao@moz-E37F359D.hinet-ip.hinet.net)
- # [10:21] * Joins: agibson (agibson@moz-2C643250.gate.cable.virginm.net)
- # [10:26] * Quits: Jamie (Instantbir@moz-CA26021.jantrid.net) (Quit: Instantbird 1.6a1pre -- http://www.instantbird.com)
- # [10:56] * Joins: slee (chatzilla@moz-31F23D35.range109-155.btcentralplus.com)
- # [11:20] * Joins: Gijs (gijs@moz-43FD6530.range86-164.btcentralplus.com)
- # [11:24] * Quits: yliao (yliao@moz-E37F359D.hinet-ip.hinet.net) (Client exited)
- # [11:28] * Joins: yliao (yliao@moz-E37F359D.hinet-ip.hinet.net)
- # [11:57] * Joins: SteveF (chatzilla@moz-6F24D0BD.cable.virginm.net)
- # [11:57] <slee> eeejay: ping
- # [12:08] * Quits: slee (chatzilla@moz-31F23D35.range109-155.btcentralplus.com) (Ping timeout)
- # [12:08] * Joins: slee_ (chatzilla@F9E55CD0.F555FE1C.51C6DE68.IP)
- # [12:08] * slee_ is now known as slee
- # [12:09] * Joins: slee_ (chatzilla@moz-31F23D35.range109-155.btcentralplus.com)
- # [12:10] * Quits: slee (chatzilla@F9E55CD0.F555FE1C.51C6DE68.IP) (Ping timeout)
- # [12:10] * slee_ is now known as slee
- # [12:31] * Quits: yliao (yliao@moz-E37F359D.hinet-ip.hinet.net) (Client exited)
- # [12:33] * Quits: slee (chatzilla@moz-31F23D35.range109-155.btcentralplus.com) (Ping timeout)
- # [12:33] * Joins: slee (chatzilla@moz-31F23D35.range109-155.btcentralplus.com)
- # [12:34] * Joins: yliao (yliao@moz-99690620.hinet-ip.hinet.net)
- # [12:37] * Quits: yliao (yliao@moz-99690620.hinet-ip.hinet.net) (Ping timeout)
- # [12:47] * Quits: slee (chatzilla@moz-31F23D35.range109-155.btcentralplus.com) (Ping timeout)
- # [13:03] * Quits: mdcurran1 (Instantbir@moz-F28D3F65.static.tpgi.com.au) (Ping timeout)
- # [13:04] * Joins: mdcurran (Instantbir@moz-F28D3F65.static.tpgi.com.au)
- # [13:23] * Joins: slee (chatzilla@moz-31F23D35.range109-155.btcentralplus.com)
- # [13:33] * Joins: slee_ (chatzilla@moz-31F23D35.range109-155.btcentralplus.com)
- # [13:34] * Quits: slee (chatzilla@moz-31F23D35.range109-155.btcentralplus.com) (Ping timeout)
- # [13:34] * slee_ is now known as slee
- # [13:39] * Quits: slee (chatzilla@moz-31F23D35.range109-155.btcentralplus.com) (Quit: ChatZilla 0.9.90.1 [Firefox 33.0/20140923222114])
- # [13:42] * agibson is now known as agibson|afk
- # [13:48] * Joins: surkov (surkov@moz-4AE5762E.cpe.teksavvy.com)
- # [13:48] * ChanServ sets mode: +o surkov
- # [14:20] * Joins: Justin_o (uid14648@moz-E77DEB21.irccloud.com)
- # [14:38] * Quits: @surkov (surkov@moz-4AE5762E.cpe.teksavvy.com) (Quit: surkov)
- # [14:39] * agibson|afk is now known as agibson
- # [14:39] * Joins: surkov (surkov@moz-4AE5762E.cpe.teksavvy.com)
- # [14:39] * ChanServ sets mode: +o surkov
- # [14:48] * Quits: SteveF (chatzilla@moz-6F24D0BD.cable.virginm.net) (Ping timeout)
- # [15:00] * Joins: yzen (yzen@13F2CEC5.7672369.D8E68FF6.IP)
- # [15:00] * ChanServ sets mode: +o yzen
- # [15:01] * Joins: anvk (anovak@C141829F.3923648E.6468E038.IP)
- # [15:04] * Quits: ioanachiorean_ (ioanachior@6DC7A5F5.AA1FA0D2.6A4F8DA2.IP) (Quit: Leaving)
- # [15:04] * Joins: ioanachiorean (ioanachior@6DC7A5F5.AA1FA0D2.6A4F8DA2.IP)
- # [15:05] * Joins: davidb (davidb@13F2CEC5.7672369.D8E68FF6.IP)
- # [15:05] * ChanServ sets mode: +qo davidb davidb
- # [15:06] <@davidb> heyo
- # [15:07] * Joins: SteveF (chatzilla@moz-6F24D0BD.cable.virginm.net)
- # [15:11] * Joins: askalski (akuda@2BA4E5A7.CC7751E8.6712E96B.IP)
- # [15:12] * Quits: askalski (akuda@2BA4E5A7.CC7751E8.6712E96B.IP) (Quit: Wychodzi)
- # [15:35] * Joins: newtron (newtron@8E826829.173A0FBA.35182819.IP)
- # [15:35] * Quits: newtron (newtron@8E826829.173A0FBA.35182819.IP) (Client exited)
- # [15:35] * Joins: newtron (newtron@B24807A2.173A0FBA.35182819.IP)
- # [15:45] * Joins: API (api@moz-75739B3B.igalia.com)
- # [16:14] * Quits: scottgonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net) (Quit: scottgonzalez)
- # [16:23] <@firebot> gijskruitbosch+bugs@gmail.com changed the Component on bug 1070713 from Untriaged to Disability Access.
- # [16:23] <@firebot> https://bugzil.la/1070713 — UNCONFIRMED, nobody@mozilla.org — HTTP refresh header + accessibility.blockautorefresh causes refresh notification to disappear immedi
- # [16:29] * davidb is now known as davidb|afk
- # [16:31] * Joins: maxli (maxli@moz-EE42E0E.student.cs.uwaterloo.ca)
- # [16:43] * icaaq is now known as icaaq|afk
- # [16:45] * davidb|afk is now known as davidb
- # [16:51] <joanie> surkov: ping
- # [16:51] <@surkov> joanie: pong
- # [16:51] <joanie> https://www.w3.org/WAI/PF/Group/track/issues/397
- # [16:51] <joanie> I was chatting with API about this
- # [16:51] <joanie> and possible solutions
- # [16:51] <joanie> and here's what I am thinking
- # [16:52] <joanie> and API didn't say "NO!"
- # [16:52] <joanie> ;)
- # [16:52] <joanie> ATK_STATE_INDETERMINATE is far too specific ;)
- # [16:52] * davidb is now known as davidb|afk
- # [16:52] <joanie> perhaps it could be expanded -- and properly doc'ed with examples -- to include "something cannot be determined about this object"
- # [16:52] <joanie> in the case of checkboxes, it would be that intermediate/mixed state
- # [16:53] <joanie> in the case of an "indeterminate" (emphasis added) setsize,
- # [16:53] <joanie> it would mean the setsize could not be determined
- # [16:53] <joanie> this would mean that the value could still be an int
- # [16:53] <joanie> the "999+" thing just really bothers the heck out of me
- # [16:53] <joanie> if we cannot even count on size being a number, then....
- # [16:54] <joanie> thoughts in general, and then specifically about IA2?
- # [16:54] <joanie> (done)
- # [16:57] * joanie waits for surkov's thoughts
- # [17:00] <@surkov> joanie: I don’t see anything bad in state, not sure which state is most suitable though
- # [17:01] <joanie> STATE_INDETERMINATE
- # [17:01] <@surkov> yeah, perhaps this one good
- # [17:01] <joanie> :)
- # [17:01] <@firebot> gijskruitbosch+bugs@gmail.com changed the Component on bug 1071553 from DOM: Events to Keyboard: Navigation.
- # [17:01] <joanie> ok, would that work in IA2?
- # [17:01] <@firebot> https://bugzil.la/1071553 — UNCONFIRMED, nobody@mozilla.org — Tab key should skip elements with "pointer-events: none"
- # [17:01] <@surkov> so I dont’ say no too :)
- # [17:01] <joanie> hehehe
- # [17:01] <@surkov> I think it should work, at least I don’t see possible conflicts
- # [17:01] <joanie> ok, perhaps we should triple-check with jamie when he is around
- # [17:02] <joanie> but this can be solved by a doc patch to ATK
- # [17:02] <@surkov> right and it’d be a good idea to file suggesation to the list
- # [17:02] <joanie> and size remains an int, so cats and dogs don't sleep together
- # [17:02] <joanie> which list?
- # [17:02] <@surkov> IA2 list I meant
- # [17:02] <joanie> ok
- # [17:02] <joanie> I can do that :)
- # [17:02] <joanie> thanks!
- # [17:03] <@surkov> however MIXED state comes from MSAA
- # [17:03] <joanie> oh, you have MIXED?
- # [17:03] <@surkov> perhaps that’s ok if IA2 will override it
- # [17:03] <@surkov> yeah, it’s used for tri state checkboxes
- # [17:03] <joanie> argh
- # [17:03] <joanie> it doesn't fit as well there
- # [17:03] <joanie> could you add a state to IA2?
- # [17:04] <@surkov> I think so
- # [17:04] <joanie> then you'd have just one implementation. ;)
- # [17:04] <tbsaunde> yeah, it seems pretty unlikely ia2 is out of biyts
- # [17:04] <@surkov> :)
- # [17:04] <tbsaunde> it was a typo even ;)
- # [17:05] <joanie> ok, I'm going to do a patch for ATK
- # [17:05] <@surkov> it’s still readable
- # [17:05] <joanie> then I'll email the ia2 list
- # [17:05] <@surkov> thank you, please cc me
- # [17:05] <joanie> thanks!
- # [17:05] <joanie> on the atk patch or the ia2 list email?
- # [17:05] <@surkov> both
- # [17:05] <joanie> heh
- # [17:05] <joanie> ok
- # [17:06] <@surkov> alos it’d be good to refer to ATK expierence
- # [17:06] <joanie> ?
- # [17:11] * Joins: scott_gonzalez (scott_gonz@93F79037.623559CB.6038A538.IP)
- # [17:20] * davidb|afk is now known as davidb
- # [17:38] * Quits: scott_gonzalez (scott_gonz@93F79037.623559CB.6038A538.IP) (Ping timeout)
- # [17:39] * @davidb shakes his fist at windows high contrast mode
- # [17:40] * Quits: API (api@moz-75739B3B.igalia.com) (Ping timeout)
- # [17:42] * Joins: API (api@moz-75739B3B.igalia.com)
- # [17:47] * Quits: ioanachiorean (ioanachior@6DC7A5F5.AA1FA0D2.6A4F8DA2.IP) (Ping timeout)
- # [17:47] * davidb is now known as davidb|afk
- # [18:07] * Joins: slee (chatzilla@moz-31F23D35.range109-155.btcentralplus.com)
- # [18:32] * icaaq|afk is now known as icaaq
- # [18:35] <@firebot> bugs@pettay.fi changed the Resolution on bug 1071553 from --- to INVALID.
- # [18:35] <@firebot> bugs@pettay.fi changed the Status on bug 1071553 from UNCONFIRMED to RESOLVED.
- # [18:36] <@firebot> https://bugzil.la/1071553 — INVALID, nobody@mozilla.org — Tab key should skip elements with "pointer-events: none"
- # [18:38] * Quits: SteveF (chatzilla@moz-6F24D0BD.cable.virginm.net) (Ping timeout)
- # [18:42] <joanie> surkov: so as you probably saw, I sent the email to the ia2 list
- # [18:42] <joanie> but it's not showing up in the archives
- # [18:42] <@surkov> joanie: yeah, thank you
- # [18:42] <joanie> http://lists.linuxfoundation.org/pipermail/accessibility-ia2/
- # [18:42] <joanie> so... I dunno
- # [18:43] <joanie> I was rather hoping to point to it in my action item
- # [18:43] <joanie> :)
- # [18:43] <joanie> also API already blessed the doc change so I pushed that
- # [18:44] <tbsaunde> joanie: I didn't get it, so presumably stuck in moderation or something?
- # [18:44] <joanie> tbsaunde: that's my guess
- # [18:44] <joanie> and I'm subscribed to the list in theory
- # [18:44] <joanie> surkov: mind seeing if I'm in the moderation queue? says your an admin of the list
- # [18:45] <@surkov> joanie: it’s there
- # [18:45] <@surkov> let me fix it
- # [18:45] <joanie> thanks
- # [18:46] <@surkov> done
- # [18:47] <joanie> tyvm! :)
- # [18:47] * davidb|afk is now known as davidb
- # [18:53] * icaaq is now known as icaaq|afk
- # [18:56] * Quits: maxli (maxli@moz-EE42E0E.student.cs.uwaterloo.ca) (Quit: Leaving.)
- # [18:58] * icaaq|afk is now known as icaaq
- # [18:59] * Joins: clown_mtg (clown@67828CC7.C1A51174.9D42CF23.IP)
- # [19:00] * Joins: jongund (chatzilla@moz-FE12876D.dres.uiuc.edu)
- # [19:01] * Quits: slee (chatzilla@moz-31F23D35.range109-155.btcentralplus.com) (Quit: ChatZilla 0.9.90.1 [Firefox 33.0/20140923222114])
- # [19:20] * Quits: API (api@moz-75739B3B.igalia.com) (Quit: Leaving)
- # [19:25] * khuey|away is now known as khuey
- # [19:36] * Quits: @surkov (surkov@moz-4AE5762E.cpe.teksavvy.com) (Quit: surkov)
- # [19:47] * Joins: jamesn (jnurthen@moz-34C566A2.oracle.com)
- # [19:57] * Quits: @davidb (davidb@13F2CEC5.7672369.D8E68FF6.IP) (Ping timeout)
- # [19:59] * Joins: davidb (davidb@13F2CEC5.7672369.D8E68FF6.IP)
- # [19:59] * ChanServ sets mode: +qo davidb davidb
- # [20:00] * Quits: @davidb (davidb@13F2CEC5.7672369.D8E68FF6.IP) (Connection reset by peer)
- # [20:00] * Joins: davidb (davidb@13F2CEC5.7672369.D8E68FF6.IP)
- # [20:00] * ChanServ sets mode: +qo davidb davidb
- # [20:01] * agibson is now known as agibson|afk
- # [20:21] * Quits: agibson|afk (agibson@moz-2C643250.gate.cable.virginm.net) (Quit: )
- # [20:30] <@firebot> ehsan.akhgari@gmail.com changed the Assignee on bug 1073820 from nobody@mozilla.org to ehsan.akhgari@gmail.com.
- # [20:30] <@firebot> https://bugzil.la/1073820 — NEW, ehsan.akhgari@gmail.com — Remove an unused variable from accessible
- # [20:34] * Joins: surkov (surkov@moz-4AE5762E.cpe.teksavvy.com)
- # [20:34] * ChanServ sets mode: +o surkov
- # [20:36] * Joins: SteveF (chatzilla@moz-6F24D0BD.cable.virginm.net)
- # [20:37] * Quits: icaaq (icaaq@moz-683D578C.cust.bredbandsbolaget.se) (Ping timeout)
- # [20:42] <joanie> surkov: ping
- # [20:42] <@surkov> joanie: pong
- # [20:42] <joanie> So at today's meeting
- # [20:42] <joanie> we concluded that there would be some backwards compatibility issues with "999+" as a setsize
- # [20:42] <@davidb> tbsaunde, heh, r=themaid ?
- # [20:43] <joanie> and rich didn't want to introduce new api if it isn't needed
- # [20:43] <@surkov> agree
- # [20:43] <joanie> so the group felt like "-1" to mean "unknown setsize" was a good option
- # [20:43] * joanie looks for the action
- # [20:43] <@surkov> that sounds restrictive
- # [20:43] <@surkov> since you don’t know number of loaded items
- # [20:44] <joanie> action 1514
- # [20:44] <@surkov> link pls?
- # [20:44] <joanie> https://www.w3.org/WAI/PF/Group/track/actions/1514
- # [20:44] <joanie> do you need to know the number of loaded items necessarily?
- # [20:45] <tbsaunde> davidb: I stole it from waldo
- # [20:45] <joanie> i.e. it's often on the screen (in the case of gmail)
- # [20:45] <@surkov> joanie: you are AT vendor, but I guess saying something like “1 of 4, items are loading” might be nice
- # [20:45] <joanie> but they are not necessarily loading
- # [20:46] <@surkov> then, “items are missed” :) dunno
- # [20:46] <joanie> if you see "-1" you can say "items are missed"
- # [20:46] <joanie> -1 means "who the heck knows?"
- # [20:46] <joanie> in this proposal
- # [20:46] <@surkov> yeah but you never can say how many items are presented
- # [20:46] <joanie> i was fine with the numeric and state indeterminate
- # [20:46] <joanie> and mentioned this to the group
- # [20:47] <@surkov> after all the user has to work with what he has
- # [20:47] <joanie> :)
- # [20:47] <@surkov> yeah, that’d be working
- # [20:47] <joanie> so I was incorrect with my "Surkov was willing to consider a whole new state for IA@, surely he'll be ok with -1"
- # [20:47] <joanie> oops
- # [20:47] <@surkov> but I shouldn’t convience you that you need to know number of items when not all items are presented since you’re AT
- # [20:48] <joanie> s/IA@/IA2/
- # [20:48] <@surkov> I would ask Jamie also though
- # [20:48] <tbsaunde> what is the backwards compat problem? before that 999+ was invalid code so people shouldn't expect anything
- # [20:48] <joanie> at least in the case of WebKit, it assumes an int
- # [20:48] <joanie> and converts the value provided by the web div into an int
- # [20:48] <joanie> so even if we change WebCore, which we can
- # [20:49] <@surkov> tbsaunde: they couldn’t read it if they previously were able to do so
- # [20:49] <tbsaunde> surkov: huh?
- # [20:49] <joanie> web devs doing it now will presumably break older versions of webkit based browsers
- # [20:49] <joanie> what I haven't tested yet is how bad the spit up will be
- # [20:49] <joanie> maybe it will just shrug and do nothing
- # [20:49] <tbsaunde> ok, sure, so if you care about old things you cna't use new features, or have to feature detect
- # [20:49] <joanie> I suppose that's worth verifying
- # [20:50] <@surkov> tbsaunde: if the author did something like <div aria-setize=“999”> and he didn’t provided this is unfinished list then 999+ will be not backward compatible
- # [20:50] <tbsaunde> surkov: no? that'll just be 99
- # [20:50] * joanie fires up her mac to see what safari does in the face of this
- # [20:50] * davidb is now known as davidb|afk
- # [20:51] <@surkov> why? because int parsing will fail
- # [20:51] <joanie> btw Cynthia (MS) was also not a fan of the numbers turning into strings in need of parsing
- # [20:52] <@surkov> anyway another problem of 999+ is we cannot expose it via groupPosition in IA2
- # [20:52] <tbsaunde> surkov: how does it fail on 999 ? or what are you saying
- # [20:52] <joanie> surkov: that is what they pointed out in the action I think
- # [20:52] <@surkov> I meant 999+
- # [20:52] * joanie looks
- # [20:52] <tbsaunde> surkov: well, you can't really expose x of -1 either so...
- # [20:53] * Quits: @davidb|afk (davidb@13F2CEC5.7672369.D8E68FF6.IP) (Connection reset by peer)
- # [20:53] <@surkov> tbsaunde: IA2 groupPositon takes signed int
- # [20:53] <@surkov> but not sure I like -1 approach
- # [20:53] <tbsaunde> surkov: if your just saying aria-setsize="999+" with something that doesn't support that well, that's the authors problem
- # [20:54] <@surkov> anyway, I guess ARIA can do 999+ approach, AT mapping is different thing
- # [20:54] <tbsaunde> its like saying we should never add Window.foo because people can't use it in old browsers
- # [20:54] <joanie> surkov: I'm not sure sure
- # [20:55] * Quits: clown_mtg (clown@67828CC7.C1A51174.9D42CF23.IP) (Quit: Leaving.)
- # [20:55] <@surkov> joanie: why? markup and mapping is not the same
- # [20:55] <@surkov> I see backward compatibility issue but no other problems
- # [20:56] <joanie> i'm talking about the backward compatibility
- # [20:56] * Quits: jamesn (jnurthen@moz-34C566A2.oracle.com) (Quit: Leaving)
- # [20:56] <@surkov> it shouldn’t be huge issue, do we have many apps that use aria-setize on list of undefined size?
- # [20:57] <@surkov> but if we do then it means -1 doesn’t work either :)
- # [20:57] <@surkov> agree?
- # [20:57] <tbsaunde> the idea this is a backwards compat issue makes no sense IMO
- # [20:57] <joanie> http://trac.webkit.org/browser/trunk/Source/WebCore/accessibility/AccessibilityObject.cpp#L2100
- # [20:57] <joanie> I don't know the string code much
- # [20:58] <joanie> I assume it fails gracefully
- # [20:58] <@surkov> joanie: yeah, webkit has to make something with it
- # [20:58] <joanie> and right now it assumes an in
- # [20:58] <joanie> int even
- # [20:58] <@surkov> that’s ok
- # [20:58] <joanie> -1 fine 1+ not find
- # [20:58] * Joins: icaaq (icaaq@moz-683D578C.cust.bredbandsbolaget.se)
- # [20:58] <tbsaunde> joanie: so what if new things don't work in old browsers, you can't possibly refuse to add anything new defining undefined behavior
- # [20:58] <joanie> fine
- # [20:58] <@surkov> new app is not supposed to work in old browser
- # [20:59] <joanie> surkov: you should come to meetings then :P
- # [20:59] <@surkov> :)
- # [20:59] <@surkov> I’m quite ok to put my thoughts on the list if needed
- # [21:00] <joanie> that might be helpful
- # [21:00] <@surkov> if the point is webkit doesn’t want to change their code then it’s also the reason but it’s not backward compatibility
- # [21:00] <joanie> no
- # [21:00] <joanie> we can change webkit
- # [21:01] <joanie> I just want to see what happens when it's given a non-int
- # [21:01] * Joins: davidb (davidb@13F2CEC5.7672369.D8E68FF6.IP)
- # [21:01] * ChanServ sets mode: +qo davidb davidb
- # [21:01] <@surkov> ok, then my statement is if 999+ has backward compatibilty issue then -1 has it also
- # [21:01] <joanie> I don't follow
- # [21:01] <joanie> -1 is an int
- # [21:02] <@surkov> let me explain
- # [21:02] <joanie> 999+ is not an int
- # [21:02] <@surkov> if we have backward compatibility issue then we have the app that has <div aria-setsize=“999”> on undefined size list
- # [21:03] <@surkov> the app wants to expose info that the list is undefined
- # [21:03] <@surkov> and it wants to expose number of items
- # [21:03] <@surkov> so you cannot do 999+ becuase it runs into int issue
- # [21:03] <@surkov> but you cannot use -1 becuase you cannot expose number of preloaded items anymore
- # [21:04] <@surkov> otherwise we don’t have backward compatibility issue and we can do both 999+ and -1
- # [21:04] <@surkov> in case of 999+ webkit should fix the code
- # [21:04] <joanie> yeah, we can fix webkit
- # [21:04] <@surkov> does it sound correct?
- # [21:05] <joanie> I have a migraine that's making me dumb(er)
- # [21:05] <@surkov> ok, what is backward compatibility issue in this case, I mgiht miss something
- # [21:06] <joanie> lemme see what happens in safari with "999+"
- # [21:07] <@surkov> if it returns int then I guess that’s supposed to be some negative value
- # [21:07] <tbsaunde> joanie: so, I'm argueing it doesn't matter, if you write page that does that in browser that doesn't implement 999+ then your inoking undefined behavior
- # [21:08] <@surkov> agree, the solution doesn’t have to work in old browser
- # [21:09] <@surkov> it’d be nice if it worked though of course
- # [21:10] <@surkov> a key question for me if AT need ever know actual set size of the list
- # [21:11] <tbsaunde> Well, if nobody knows and somebody must know then there's a unsolvable problem ;)
- # [21:11] <joanie> if the web app doesn't know the actual set size, then....
- # [21:11] <joanie> hahaha
- # [21:11] <joanie> exactly tbsaunde
- # [21:12] <@surkov> I referred to actual size as amount of elemnets in the list
- # [21:12] <joanie> i seriously need to get myself voiceover chops
- # [21:13] <@davidb> I would like morgan freeman to do my voiceovers
- # [21:13] <joanie> other kind of voiceover silly
- # [21:13] <@davidb> pffft
- # [21:14] <@surkov> it’s like somebody has one kid but its familty size is undefined because new kids may appear later, so we know actual familty size and can expose this info and also we can expose intederminty flag
- # [21:14] <tbsaunde> I think we need a remote control fish at davidb's desk he makes stoo many bad jokes
- # [21:14] <@davidb> surkov, this is why family planning is important
- # [21:14] <khuey> lol
- # [21:14] <@surkov> davidb: say the to web
- # [21:15] <@davidb> ok khuey laughed, i'm done here :)
- # [21:16] <@surkov> davidb: I wish you’re not since you’re part of ARIA effort :)
- # [21:16] <@davidb> yeah, 999+ problems
- # [21:16] <@davidb> lots of backscroll
- # [21:16] <@davidb> is it solved?
- # [21:17] <@davidb> I'd say not to worry too much about 999+ backwards compat.
- # [21:18] <@davidb> surkov, earlier you wondered if AT cares about set size... like "3 of 15" is the 15 important to announce?
- # [21:18] <@surkov> screen readers do this afaik
- # [21:18] <@davidb> yeah were you wondering if it is important?
- # [21:19] <@davidb> (i'd guess sometimes)
- # [21:19] <@surkov> since the do that then I assumed it’s important so I was curious if this is alos important when you don’t know future list size
- # [21:19] <@davidb> oh
- # [21:20] * Joins: clown (clown@67828CC7.C1A51174.9D42CF23.IP)
- # [21:20] <tbsaunde> yeah, my mail client shows x / y in the status bar so I assume people like it
- # [21:20] <@surkov> joanie, btw, aria-setsize=“-1” is too techy I guess aria-setsize=“undefined” could look better for web developer
- # [21:20] <@davidb> tbsaunde, yeah
- # [21:21] <joanie> sure
- # [21:21] <tbsaunde> on the other hand I might take not having that or something if it got me faster loading of giant mailboxen
- # [21:21] <@davidb> tbsaunde, i'd bet if it was "3 of 3 million 2 hundred and 36 thousand, four hundred and thiry six" it could get annoying
- # [21:21] <joanie> btw, I cannot get voiceover to present my (bogus) setsize of 1000
- # [21:21] <joanie> so I don't know what it would do with 1000+ yet
- # [21:21] <@davidb> joanie, is it "busy" or just silent
- # [21:21] <@davidb> ?
- # [21:21] * Quits: SteveF (chatzilla@moz-6F24D0BD.cable.virginm.net) (Ping timeout)
- # [21:22] <joanie> neither. it tells the truth
- # [21:22] <joanie> 2 of 4
- # [21:22] <joanie> :)
- # [21:22] <@davidb> oh heh
- # [21:22] <joanie> mind you I am not a webdev
- # [21:22] <@davidb> so aria does not win in your csae
- # [21:22] <tbsaunde> surkov: well, -1 seems just generally worse than x+ or I guess x- maybe?
- # [21:22] <joanie> so I did a quick test with an unordered list roles, and setsize and posinset (based on an example I found from searching)
- # [21:22] <@surkov> tbsaunde: x+ is nice approach if we want to expose number of presented items
- # [21:23] <joanie> to be honest, I was fond of the user agents turning "999+" into "999" with STATE_INDETERMINATE
- # [21:23] <@surkov> if not then -1 or undefined works well
- # [21:23] <joanie> but I guess that's "new API"
- # [21:24] <joanie> i'm also fine with "-1" or "undefined" turned into "-1" by user agents
- # [21:24] <tbsaunde> well, so is -1 or undefined
- # [21:24] <@surkov> as long as we have this state it’s not new API
- # [21:24] <@surkov> state constant I mean
- # [21:25] <@surkov> joanie: whch list should I use to email about this?
- # [21:25] <joanie> hmmmm
- # [21:25] <joanie> came up in the aria group
- # [21:25] <joanie> so....
- # [21:25] <joanie> the public pf one?
- # [21:25] <clown> surkov: I suggest public-pfwg.
- # [21:25] <clown> :-)
- # [21:25] <@surkov> ok, htanks
- # [21:25] <joanie> clown: !
- # [21:25] <joanie> you have voiceover chops?
- # [21:25] <clown> reference the action too.
- # [21:25] * clown looking it up.
- # [21:26] <clown> there is this one, surkov: https://www.w3.org/WAI/PF/Group/track/actions/1514
- # [21:26] <clown> But, it's not the original one that caused the discussion...
- # [21:27] <joanie> https://www.w3.org/WAI/PF/Group/track/actions/1485
- # [21:28] <joanie> https://www.w3.org/WAI/PF/Group/track/actions/1486
- # [21:28] <clown> and this issue, joanie, surkov: https://www.w3.org/WAI/PF/Group/track/issues/397
- # [21:28] <joanie> yup
- # [21:29] <clown> joanie, I have not screen reader chops. I fumble at orca, even. It's amusing to watch, I expect.
- # [21:30] <clown> s/no/not
- # [21:30] * khuey is now known as khuey|away
- # [21:31] <joanie> does firefox + voiceover work fairly well?
- # [21:31] * joanie downloads firefox on her mac
- # [21:31] <@davidb> not great
- # [21:31] <@davidb> we don't really officially support it
- # [21:32] * Quits: @yzen (yzen@13F2CEC5.7672369.D8E68FF6.IP) (Ping timeout)
- # [21:33] <clown> Yeah, I gave up trying FF+VoiceOver a long time ago.
- # [21:34] <clown> Does FF do AXAPI, davidb?
- # [21:35] <@davidb> yeah AIUI
- # [21:35] <@davidb> clown, http://mxr.mozilla.org/mozilla-central/search?find=%2Faccessible%2Fmac%2F&string=ax
- # [21:36] <clown> Perhaps I should try FF+VO again, sometime.
- # [21:36] <@davidb> i'd love us to be awesome on mac but.
- # [21:37] <tbsaunde> patches welcome ;)
- # [21:37] <clown> joanie, if it helps, if you go to the "Accessibility" preferences in "System Preferences", and choose VoiceOver, there is a button labelled "Open VoiceOver Training…". That might help you.
- # [21:38] <clown> joanie: nit: that's using OS X Mountain Lion. I don't know if that's what you are running.
- # [21:43] <joanie> ok, so it seems that VoiceOver does not yet speak it in OS X
- # [21:43] <joanie> and the inspector just drops the "+" so webcore is fine
- # [21:44] <joanie> so if we don't care about older browsers knowing the set size is indeterminate
- # [21:44] <joanie> "999+" seems to work
- # [21:44] <joanie> tbsaunde: it wasn't a complaint. it was more of a "just want to have voiceover speak this"
- # [21:45] <joanie> my main purpose for a mac is to answer the question, how exactly did my webkitgtk+ patch break safari a11y?
- # [21:45] <joanie> ;)
- # [21:46] <tbsaunde> joanie: heh, I actually intended that at clown but welcome from however ;-)
- # [21:47] * Joins: SteveF (chatzilla@moz-6F24D0BD.cable.virginm.net)
- # [21:48] * clown is thick-headed, tbsaunde. What did you intend for me?
- # [21:49] <tbsaunde> clown: the patches welcome comment
- # [21:49] <tbsaunde> so not much of value ;)
- # [21:49] <clown> tbsaunde: , will this do? http://www.patchesrus.com/
- # [21:50] * joanie grins
- # [21:50] <clown> also, not of much value...
- # [21:50] <joanie> surkov: fwiw, I just updated https://www.w3.org/WAI/PF/Group/track/actions/1486
- # [21:52] <clown> joanie, another nit: action-1486 is "closed" (it can be reopened).
- # [21:52] <joanie> it was closed deliberately
- # [21:52] <joanie> but yes, it can
- # [21:53] <joanie> at the meeting, as I understood it, the consensus was to not do the "+" thang
- # [21:53] <clown> yes, that was the consensus. Everyone was fine with using "-1", except for my caveat about how that would work with IA2 groupPosition.
- # [21:54] <joanie> and what else would I do on that item regardless?
- # [21:54] <joanie> i.e. i found the code where it assumes an int, I've since verified it doesn't spit up, but just drops the "+"
- # [21:54] <joanie> investigation: complete
- # [21:54] <joanie> :)
- # [21:55] <clown> ah, okay.
- # [21:58] * Quits: newtron (newtron@B24807A2.173A0FBA.35182819.IP) (Quit: Leaving...)
- # [22:00] * Quits: SteveF (chatzilla@moz-6F24D0BD.cable.virginm.net) (Ping timeout)
- # [22:00] <clown> joanie, fyi. I just discovered that JavaScript's parseInt ("999+") returns the integer value 999.
- # [22:00] <joanie> cool
- # [22:01] <clown> i thought it might throw a parsing exception, but nope.
- # [22:01] <joanie> yeah, I was concerned about what webkit's string code would do was well
- # [22:01] <joanie> and I looked in webkitgtk and even the object attribute wasn't there
- # [22:01] <clown> now, what does ATK do with it?
- # [22:01] <joanie> contrary to what the code suggests
- # [22:02] <joanie> as an object attribute it's just a stupid string
- # [22:02] <joanie> Orca assumes int
- # [22:02] * Joins: yzen (yzen@moz-F62769B5.cpe.pppoe.ca)
- # [22:02] * ChanServ sets mode: +o yzen
- # [22:02] <joanie> but other things that are not exposed via object attribute (aka everything non-aria)
- # [22:02] <joanie> we have real ints
- # [22:02] <clown> oh yeah. aria-setsize is mapped to "object property setsize:n"
- # [22:03] <joanie> until I do my gnome action item of new API for this
- # [22:03] * clown wants mythical ints. that would be cool.
- # [22:04] <joanie> what would be cool is to not have to play "guess where the set size is" ;)
- # [22:04] <joanie> i seriously need to prioritize that ATK/AT-SPI2 API
- # [22:04] <joanie> since it appears I am blocking myself
- # [22:04] <clown> so, time to port IA2 groupPosition() to ATK/AT-SPI2 ?
- # [22:04] <joanie> ;)
- # [22:04] <joanie> possibly
- # [22:05] <tbsaunde> bleh glib why you feel the need to have special types
- # [22:05] <clown> yes, maybe.
- # [22:21] <tbsaunde> davidb: I kind of sort of have simple pages working in e10s you can atleast read links when tabbed to on apache's dir index page now :)
- # [22:21] <@davidb> \o/
- # [22:21] <@davidb> exciting
- # [22:22] <tbsaunde> yeah, now to fix the 999 things that still don't work
- # [22:22] <@davidb> ITYM 999+
- # [22:22] <tbsaunde> ;)
- # [22:22] <joanie> you beat me to that one
- # [22:22] <tbsaunde> btw you still in the office?
- # [22:22] * joanie is getting old and slow
- # [22:23] <tbsaunde> davidb: ?
- # [22:23] <@davidb> tbsaunde i am here until 6pm today
- # [22:23] <@davidb> joanie heh
- # [22:23] <tbsaunde> davidb: want to drop by at some point and tell me what some mail is?
- # [22:24] <@davidb> tbsaunde, now wfm you?
- # [22:24] <tbsaunde> davidb: yes
- # [22:24] <@davidb> omw
- # [22:25] * Quits: @surkov (surkov@moz-4AE5762E.cpe.teksavvy.com) (Quit: surkov)
- # [22:29] * Quits: anvk (anovak@C141829F.3923648E.6468E038.IP) (Quit: Leaving.)
- # [22:29] * khuey|away is now known as khuey
- # [22:57] * davidb is now known as davidb|afk
- # [22:59] * davidb|afk is now known as davidb
- # [23:04] * Parts: clown (clown@67828CC7.C1A51174.9D42CF23.IP)
- # [23:28] * Quits: @yzen (yzen@moz-F62769B5.cpe.pppoe.ca) (Ping timeout)
- # [23:29] * Quits: jongund (chatzilla@moz-FE12876D.dres.uiuc.edu) (Quit: ChatZilla 0.9.90.1 [Firefox 32.0.3/20140923175406])
- # [23:32] * Joins: yzen (yzen@moz-F62769B5.cpe.pppoe.ca)
- # [23:32] * ChanServ sets mode: +o yzen
- # [23:34] * Quits: @yzen (yzen@moz-F62769B5.cpe.pppoe.ca) (Ping timeout)
- # [23:38] * Quits: icaaq (icaaq@moz-683D578C.cust.bredbandsbolaget.se) (Quit: 10-79 Notify coroner, over and out)
- # [23:48] * Quits: @davidb (davidb@13F2CEC5.7672369.D8E68FF6.IP) (Quit: Blah blah blah)
- # Session Close: Tue Sep 30 00:00:00 2014
The end :)