/irc-logs / mozilla / #accessibility / 2012-02-08 / end
Options:
- # Session Start: Wed Feb 08 00:00:00 2012
- # Session Ident: #accessibility
- # [00:06] * Quits: @jprmc (jprmc@F2D29657.F60B0462.67AC9B1.IP) (Ping timeout)
- # [00:11] * Joins: silvia (Adium@13A1E76C.497EC1B8.5EFB9497.IP)
- # [00:12] <@firebot> mak77@bonardo.net changed the Status on bug 672507 from NEW to RESOLVED.
- # [00:12] <@firebot> mak77@bonardo.net set the Resolution field on bug 672507 to FIXED.
- # [00:12] <@firebot> mak77@bonardo.net changed the Target Milestone on bug 672507 from --- to mozilla13.
- # [00:12] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=672507 nor, --, ---, trev.saunders, NEW, merge nsIAccessNode and nsIAccessible
- # [00:15] <@tbsaunde> \O/
- # [00:16] * Quits: skyler_brungardt (skyler@moz-9FA3C392.washdc.fios.verizon.net) (Quit: skyler_brungardt)
- # [00:20] <@hub> this one caused me some rebasing work
- # [00:20] <@hub> but it is over now
- # [00:21] <@tbsaunde> yeah, its a huge patch so I'm not suprised
- # [00:22] <@hub> neither am I
- # [00:22] <@hub> and hg is kinda incompetent with merges too
- # [00:22] <@hub> :-/
- # [00:22] * Joins: surkov (surkov@E3B837DE.44A4068D.222B27F0.IP)
- # [00:22] * ChanServ sets mode: +o surkov
- # [00:23] <@hub> how did we do before version control?
- # [00:24] <@tbsaunde> hub: that's why I try and only move patches to hg from git right before I push
- # [00:25] <@hub> tbsaunde: maybe I should just do that
- # [00:25] <@tbsaunde> hub: but you might be able to make mqueue use a merge tool or something slightly better than rej files
- # [00:26] <@tbsaunde> I've thought about keeping a public tree on github or a private server but never bothered
- # [00:27] <@hub> tbsaunde: there is one on github
- # [00:27] <@tbsaunde> hub: yeah, I know
- # [00:27] <@tbsaunde> I mean make the stuff I'm working on public
- # [00:27] <@tbsaunde> like a user repo on hg.m.o
- # [00:27] <satdav> does anyone know of a screen reader for linux
- # [00:27] <@hub> oh I see
- # [00:27] <@tbsaunde> satdav: orca
- # [00:27] <@hub> satdav: Orca
- # [00:28] <satdav> is it free
- # [00:28] <@hub> satdav: installed by default on most distro
- # [00:28] <satdav> as i am doing a kb on sumo
- # [00:28] <@hub> satdav: it is Free Software
- # [00:28] <satdav> i have added the mac default one
- # [00:28] <satdav> my college uses read and write pro or that
- # [00:28] <@hub> satdav: the current build of Firefox on Mac don't support VoiceOer
- # [00:28] <satdav> a paid one
- # [00:28] <satdav> it should support it
- # [00:28] <@hub> satdav: I'm fixing that, but it won't be before FF13
- # [00:29] <satdav> cool
- # [00:29] <satdav> and when will that be released
- # [00:29] <@hub> for now, it is not even scheduled
- # [00:29] <@hub> FF13 would be the earliest
- # [00:29] <@hub> and 18 weeks for now to hit release
- # [00:29] <@hub> at least
- # [00:30] <satdav> hub, would be better with a addon i would say
- # [00:30] <@hub> satdav: we have accessibility support in the other platforms. just that it was broken on Mac
- # [00:30] <satdav> http://www.screenreader.net/ has anyone tried http://www.screenreader.net/
- # [00:30] <satdav> oh
- # [00:32] <satdav> http://support.mozilla.org/en-US/kb/Accessibility
- # [00:32] <satdav> I have to keep that page updated reguler
- # [00:34] <satdav> if any of you guys wants to help then go ahead
- # [00:34] <satdav> does ken the guy who runs http://www.accessfirefox.org/ do irc
- # [00:35] <khuey> !seen ksaunders
- # [00:35] <@firebot> I've never seen a 'ksaunders', sorry.
- # [00:35] <khuey> hmm, not sure
- # [00:35] <khuey> thought I've seen him before
- # [00:36] <satdav> !seen ken
- # [00:36] <@firebot> ken was last seen 113 weeks, 4 days, 20 hours, 33 minutes and 44 seconds ago, saying 'Hi' in #firefox.
- # [00:36] <satdav> !seen kensaunders
- # [00:36] <@firebot> I've never seen a 'kensaunders', sorry.
- # [00:36] <satdav> khuey, would it be that ken nick
- # [00:36] <satdav> !seen kens
- # [00:37] <@firebot> kens was last seen 174 weeks, 2 hours, 52 minutes and 20 seconds ago, saying 'Not too hard to digest though' in #accessibility.
- # [00:37] * Quits: nhirata (nhirata.bu@moz-BBE3ABD.mv.mozilla.com) (Quit: nhirata)
- # [00:37] <satdav> or that one
- # [00:38] <@hub> that's a while ago
- # [00:38] <satdav> YES I KNOW
- # [00:39] <satdav> over a year
- # [00:39] * Quits: @aaronlev (aaronlev@moz-CDA191A6.c3-0.arl-ubr1.sbo-arl.ma.cable.rcn.com) (Quit: aaronlev)
- # [00:41] * Joins: nhirata (nhirata.bu@moz-BBE3ABD.mv.mozilla.com)
- # [00:43] <@hub> satdav: something like 3.5
- # [00:44] <satdav> true
- # [00:46] * Quits: silvia (Adium@13A1E76C.497EC1B8.5EFB9497.IP) (Ping timeout)
- # [00:47] * Joins: silvia (Adium@13A1E76C.497EC1B8.5EFB9497.IP)
- # [00:49] <@hub> and it never ends. this time it is Android that turns orange
- # [01:17] * Quits: mike5w3c (MikeS@moz-DAFE1A45.tokyo.ocn.ne.jp) (Quit: mike5w3c)
- # [01:19] <@tbsaunde> hub: I thought that was the normal color for android =p
- # [01:20] <@hub> now I have a leak in Debug....
- # [01:20] <@hub> :-/
- # [01:21] <@hub> tbsaunde: apparently it is
- # [01:22] <@tbsaunde> you actually star your inbound pushes yourself?
- # [01:23] * Quits: satdav (satdav@moz-1ECB932B.cable.virginmedia.com) (Quit: Leaving)
- # [01:24] <@hub> I didn't star anything
- # [01:26] <@tbsaunde> hub: sure, but you seem to be watching tbpl
- # [01:26] <@hub> tbsaunde: I was watching mine, yeah
- # [01:27] * Joins: davidb (davidb@B695C5DD.16851E93.3C16C231.IP)
- # [01:27] * ChanServ sets mode: +qo davidb davidb
- # [01:28] * Quits: ptheriault (ptheriault@moz-4BE034AB.ptr.us.xo.net) (Quit: ptheriault)
- # [01:31] * Quits: @davidb (davidb@B695C5DD.16851E93.3C16C231.IP) (Ping timeout)
- # [01:36] <@tbsaunde> surkov: I'll look into your comments in a day or two realy busy with school till then
- # [01:36] * @tbsaunde has lots of sml and a history paper to write asap
- # [01:37] <@surkov> sure
- # [01:37] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
- # [01:58] * Joins: skyler_brungardt (skyler@moz-9FA3C392.washdc.fios.verizon.net)
- # [02:09] * Joins: mike5w3c (MikeS@moz-DAFE1A45.tokyo.ocn.ne.jp)
- # [02:15] * Joins: ptheriault (ptheriault@moz-4BE034AB.ptr.us.xo.net)
- # [02:16] * Quits: skyler_brungardt (skyler@moz-9FA3C392.washdc.fios.verizon.net) (Quit: skyler_brungardt)
- # [02:17] * Quits: @hub (hub@21B7B9F2.B87E9213.6E712CE2.IP) (Input/output error)
- # [02:21] * Joins: aaronlev (aaronlev@moz-9058091D.bstnma.fios.verizon.net)
- # [02:24] * Quits: ptheriault (ptheriault@moz-4BE034AB.ptr.us.xo.net) (Quit: ptheriault)
- # [02:26] * Joins: skyler_brungardt (skyler@moz-9FA3C392.washdc.fios.verizon.net)
- # [02:27] * Quits: aaronlev (aaronlev@moz-9058091D.bstnma.fios.verizon.net) (Ping timeout)
- # [02:30] * Joins: aaronlev (aaronlev@moz-9058091D.bstnma.fios.verizon.net)
- # [02:30] <@firebot> New Core - Disability Access APIs bug 725178 filed by surkov.alexander@gmail.com.
- # [02:30] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=725178 nor, --, ---, nobody, NEW, get rid ensureAccessibleTree of common.js
- # [02:30] * Quits: aaronlev (aaronlev@moz-9058091D.bstnma.fios.verizon.net) (Quit: aaronlev)
- # [02:34] <@firebot> bmo@edmorley.co.uk changed the Target Milestone on bug 672504 from mozilla13 to ---.
- # [02:34] <@firebot> bmo@edmorley.co.uk changed the Target Milestone on bug 673405 from mozilla13 to ---.
- # [02:34] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=672504 nor, --, ---, hub, ASSI, Don't keep pointer to weak presshell in accessible
- # [02:34] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=673405 cri, --, ---, hub, NEW, Rename GetDocAccessible() to Document()
- # [02:34] * Quits: skyler_brungardt (skyler@moz-9FA3C392.washdc.fios.verizon.net) (Quit: skyler_brungardt)
- # [02:36] * Joins: ptheriault (ptheriault@moz-4BE034AB.ptr.us.xo.net)
- # [02:38] * Joins: hub (hub@83874EA1.EB7C1AF9.6F478678.IP)
- # [02:38] * ChanServ sets mode: +o hub
- # [02:43] <@firebot> surkov.alexander@gmail.com cancelled review?(surkov.alexander@gm ail.com) for attachment 595044 on bug 717506.
- # [02:43] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=717506 nor, --, ---, askalski, NEW, telemetry for xforms use
- # [02:44] * khuey is now known as khuey|away
- # [02:45] * Joins: skyler_brungardt (skyler@moz-9FA3C392.washdc.fios.verizon.net)
- # [02:45] * Joins: aaronlev (aaronlev@moz-9058091D.bstnma.fios.verizon.net)
- # [02:46] * Quits: skyler_brungardt (skyler@moz-9FA3C392.washdc.fios.verizon.net) (Quit: skyler_brungardt)
- # [02:55] * Quits: aaronlev (aaronlev@moz-9058091D.bstnma.fios.verizon.net) (Quit: aaronlev)
- # [02:58] <@firebot> surkov.alexander@gmail.com granted review for attachment 595165 on bug 723796.
- # [02:58] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=723796 nor, --, ---, nobody, NEW, rename nsAccessible::GetAllowsAnonChildAccessibles
- # [03:07] * Quits: nhirata (nhirata.bu@moz-BBE3ABD.mv.mozilla.com) (Quit: nhirata)
- # [03:21] <@firebot> surkov.alexander@gmail.com denied review for attachment 594986 on bug 706134.
- # [03:21] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=706134 nor, --, ---, askalski, NEW, ARIA listitem shouldn't expose selectable state and pick up aria-selected and aria-checked
- # [03:26] * khuey|away is now known as khuey
- # [03:28] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Connection reset by peer)
- # [03:37] * Joins: skyler_brungardt (skyler@moz-9FA3C392.washdc.fios.verizon.net)
- # [03:41] * Quits: skyler_brungardt (skyler@moz-9FA3C392.washdc.fios.verizon.net) (Quit: skyler_brungardt)
- # [03:52] * Joins: jprmc (jprmc@moz-7F2FF3EB.cpe.net.cable.rogers.com)
- # [03:52] * ChanServ sets mode: +o jprmc
- # [04:15] * Quits: mike5w3c (MikeS@moz-DAFE1A45.tokyo.ocn.ne.jp) (Quit: mike5w3c)
- # [04:35] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
- # [05:25] * khuey is now known as khuey|away
- # [05:28] * Joins: skyler_brungardt (skyler@moz-9FA3C392.washdc.fios.verizon.net)
- # [05:31] * Quits: silvia (Adium@13A1E76C.497EC1B8.5EFB9497.IP) (Quit: Leaving.)
- # [05:32] * Joins: silvia (Adium@13A1E76C.497EC1B8.5EFB9497.IP)
- # [05:38] * Quits: silvia (Adium@13A1E76C.497EC1B8.5EFB9497.IP) (Ping timeout)
- # [05:46] * Joins: silvia (Adium@13A1E76C.497EC1B8.5EFB9497.IP)
- # [05:48] * Quits: peteb-away (ptbrunet@moz-E9B02845.austin.res.rr.com) (Ping timeout)
- # [06:02] * Joins: mike5w3c (MikeS@moz-DAFE1A45.tokyo.ocn.ne.jp)
- # [06:04] * Quits: mike5w3c (MikeS@moz-DAFE1A45.tokyo.ocn.ne.jp) (Ping timeout)
- # [06:28] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Connection reset by peer)
- # [06:33] * Joins: mike5w3c (MikeS@moz-DAFE1A45.tokyo.ocn.ne.jp)
- # [06:46] * Quits: silvia (Adium@13A1E76C.497EC1B8.5EFB9497.IP) (Quit: Leaving.)
- # [06:49] * Quits: skyler_brungardt (skyler@moz-9FA3C392.washdc.fios.verizon.net) (Quit: skyler_brungardt)
- # [07:00] * Joins: Mark_Capella (chatzilla@moz-DD0C7E4F.twcny.res.rr.com)
- # [07:01] * Quits: ptheriault (ptheriault@moz-4BE034AB.ptr.us.xo.net) (Connection reset by peer)
- # [07:01] * Joins: ptheriault (ptheriault@moz-4BE034AB.ptr.us.xo.net)
- # [07:02] * Quits: ptheriault (ptheriault@moz-4BE034AB.ptr.us.xo.net) (Connection reset by peer)
- # [07:03] * Joins: ptheriault (ptheriault@moz-4BE034AB.ptr.us.xo.net)
- # [07:04] * Joins: ptheriault_ (ptheriault@moz-4BE034AB.ptr.us.xo.net)
- # [07:05] * Quits: ptheriault (ptheriault@moz-4BE034AB.ptr.us.xo.net) (Ping timeout)
- # [07:07] * Quits: ptheriault_ (ptheriault@moz-4BE034AB.ptr.us.xo.net) (Ping timeout)
- # [07:08] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
- # [07:27] * Quits: a-865 (fmcz@moz-A5D13CA.cable.mindspring.com) (Ping timeout)
- # [08:02] * Joins: silvia (Adium@moz-7033BBBB.dyn.iinet.net.au)
- # [08:08] * Joins: MarcoZ (marco.zehe@moz-20316C0D.dip.t-dialin.net)
- # [08:08] * ChanServ sets mode: +o MarcoZ
- # [08:09] * Quits: silvia (Adium@moz-7033BBBB.dyn.iinet.net.au) (Quit: Leaving.)
- # [08:29] <@firebot> marco.zehe@googlemail.com changed the Status on bug 558589 from NEW to ASSIGNED.
- # [08:29] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=558589 min, --, ---, marco.zehe, ASSI, A11y tests which depend on the "tabbrowser API" should explicitly check its availability.
- # [08:53] <@hub> rebase time
- # [08:53] <@hub> *sigh*
- # [08:59] * Quits: @surkov (surkov@E3B837DE.44A4068D.222B27F0.IP) (Quit: surkov)
- # [09:01] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Connection reset by peer)
- # [09:16] * Quits: sgimeno (santiago.g@A922400D.48362F18.BCC03455.IP) (Client exited)
- # [09:16] * Joins: sgimeno (santiago.g@A922400D.48362F18.BCC03455.IP)
- # [09:18] * Quits: @hub (hub@83874EA1.EB7C1AF9.6F478678.IP) (Ping timeout)
- # [09:22] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
- # [09:36] * Quits: Mark_Capella (chatzilla@moz-DD0C7E4F.twcny.res.rr.com) (Quit: ChatZilla 0.9.88 [Firefox 13.0a1/20120205013302])
- # [09:45] * Joins: victorporof (victorporo@F6D8F148.2E655143.79933D60.IP)
- # [09:45] * Joins: sgimeno|2 (santiago.g@A922400D.48362F18.BCC03455.IP)
- # [09:45] * Quits: sgimeno (santiago.g@A922400D.48362F18.BCC03455.IP) (Ping timeout)
- # [10:24] * Joins: surkov (surkov@E3B837DE.44A4068D.222B27F0.IP)
- # [10:24] * ChanServ sets mode: +o surkov
- # [10:31] * Joins: victorporo (victorporo@9098D3F7.602BD6C.79933D60.IP)
- # [10:32] * Quits: victorporof (victorporo@F6D8F148.2E655143.79933D60.IP) (Ping timeout)
- # [10:47] * victorporo is now known as victorporof
- # [11:13] * Quits: @MarcoZ (marco.zehe@moz-20316C0D.dip.t-dialin.net) (Ping timeout)
- # [11:14] * Joins: MarcoZ (marco.zehe@moz-20316C0D.dip.t-dialin.net)
- # [11:14] * ChanServ sets mode: +o MarcoZ
- # [11:16] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Connection reset by peer)
- # [11:25] * Quits: sgimeno|2 (santiago.g@A922400D.48362F18.BCC03455.IP) (Quit: KVIrc 4.0.2 Insomnia http://www.kvirc.net/)
- # [11:25] * Joins: sgimeno (santiago.g@A922400D.48362F18.BCC03455.IP)
- # [11:44] <@tbsaunde> surkov: I'm not sure we actually fire value change events for atk when the value of a combo box changes since we QI the accessible to nsIAccessibleValue before fireing the event
- # [11:44] <@tbsaunde> but perhaps we should fire value change for that case
- # [11:46] <@surkov> tbsaunde, it sounds we do since ARIA combo box should implement this interface
- # [11:46] <@tbsaunde> surkov: should it really? what is its numeric value?
- # [11:47] <@surkov> tbsaunde: that's very good question :)
- # [11:48] <@surkov> tbsaunde: yes, that's a bug
- # [11:49] <@tbsaunde> surkov: oh, and aria list box appears to be eNoValue so nsAccessible::QueryInterface will say its not a nsIAccessibleValue afaik
- # [11:49] <@surkov> right
- # [11:50] <@tbsaunde> surkov: but we shouldn't fire events in that case?
- # [11:50] <@surkov> maybe not, iirc orca is going to listen selection change events
- # [11:51] <@surkov> windows screen readers relies on value change event
- # [11:51] <@tbsaunde> surkov: ok
- # [11:51] <@tbsaunde> surkov: which part were you saying is a bug? I can try and file in a bit
- # [11:51] <@tbsaunde> unless you want to since you understand :)
- # [11:51] <@surkov> tbsaunde: role="combo box" shouldn't implement nsIAccessibleValue
- # [11:53] <@surkov> tbsaunde: Bug 390129
- # [11:53] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=390129 nor, --, ---, aaronlev, RESO FIXED, Support aaa:expanded="true" on ARIA combobox
- # [11:53] <@tbsaunde> surkov: ok
- # [11:54] <@surkov> oops not this one
- # [11:54] <@surkov> but change set refers to this bug - http://hg.mozilla.org/mozilla-central/rev/a55d93c82a17
- # [11:55] <@surkov> maybe we'll never know a reason
- # [11:55] * Quits: victorporof (victorporo@9098D3F7.602BD6C.79933D60.IP) (Ping timeout)
- # [11:55] <@tbsaunde> surkov: yeah, maybe :/
- # [11:55] <@surkov> anyway ARIA spec doesn't allow these attributes http://www.w3.org/TR/wai-aria/roles#combobox so I think we need to fix that on our side
- # [11:56] <@surkov> tbsaunde: are you going to file a bug? that should be a good first one
- # [11:57] <@tbsaunde> surkov: I can't for a while
- # [11:57] <@tbsaunde> I have several pages to write in the next couple hours
- # [11:57] <@surkov> I will then
- # [11:57] <@tbsaunde> I shouldn't be distracting my self now as much as I already am =p
- # [11:57] <@tbsaunde> thx
- # [12:01] * Joins: Stevef (chatzilla@moz-FBEAB712.croy.cable.virginmedia.com)
- # [12:01] <Stevef> surkov: re: HTML5 context menu https://bugzilla.mozilla.org/show_bug.cgi?id=617528#c169
- # [12:01] <@firebot> Bug 617528 enh, --, mozilla8, Jan.Varga, ASSI, implement the HTML5 "context menu" feature (contextmenu attribute)
- # [12:02] <Stevef> also any idea about whether mozilla are sticking with <menuitem> rather than as specced?
- # [12:03] <@surkov> Stevef: you found a bug
- # [12:03] <Stevef> ok thought i might just be missing some magic
- # [12:04] <@surkov> Stevef: what platform is alt+f10 is used on?
- # [12:04] <Stevef> windows
- # [12:05] <Stevef> sorry not alt is shift+f10
- # [12:05] <@surkov> ok
- # [12:06] <Stevef> surkov:have updated bug
- # [12:07] <@surkov> thx
- # [12:07] * Joins: victorporof (victorporo@86FDEBD.6D2FB7A4.79933D60.IP)
- # [12:17] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
- # [12:21] * Quits: Stevef (chatzilla@moz-FBEAB712.croy.cable.virginmedia.com) (Connection reset by peer)
- # [12:24] <@firebot> New Core - Disability Access APIs bug 725259 filed by surkov.alexander@gmail.com.
- # [12:24] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=725259 nor, --, ---, nobody, NEW, ARIA role combobox shouldn't allow aria-valueXXX attributes
- # [12:26] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Ping timeout)
- # [12:29] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
- # [12:35] <@tbsaunde> surkov: what do you want to do about how that change will effect what atk fires value change events for
- # [12:35] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Ping timeout)
- # [12:36] <@surkov> tbsaunde: I hope they aren't used
- # [12:36] <@surkov> we need to ask joanie
- # [12:39] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
- # [12:42] <@MarcoZ> surkov: And NVDA.
- # [12:42] <@surkov> MarcoZ: NVDA shouldn't care
- # [12:43] <@MarcoZ> surkov: You may be right, I think they only use it for edit combos, not for combo boxes.
- # [12:43] <@surkov> MarcoZ: use what?
- # [12:43] <@MarcoZ> surkov: Valuechange.
- # [12:43] <@surkov> MarcoZ: ah, it doesn't affect on value change on windows
- # [12:44] <@MarcoZ> surkov: OK
- # [12:47] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Ping timeout)
- # [12:57] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
- # [13:41] * Quits: @jprmc (jprmc@moz-7F2FF3EB.cpe.net.cable.rogers.com) (Ping timeout)
- # [13:47] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Connection reset by peer)
- # [13:50] <@firebot> surkov.alexander@gmail.com changed the Status on bug 559749 from NEW to RESOLVED.
- # [13:50] <@firebot> surkov.alexander@gmail.com set the Resolution field on bug 559749 to WORKSFORME.
- # [13:50] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=559749 nor, --, ---, nobody, RESO WORKSFORME, add accessibility support of @autocomplete attribute on HTML input
- # [13:53] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
- # [14:30] * Joins: john_ (john@moz-20C4F32F.fbx.proxad.net)
- # [14:30] <john_> Ded
- # [14:30] <john_> Dd
- # [14:30] <john_> D
- # [14:30] <john_> D
- # [14:30] <john_> Dd
- # [14:30] <john_> D
- # [14:30] <john_> D
- # [14:31] * Parts: john_ (john@moz-20C4F32F.fbx.proxad.net)
- # [14:31] * khuey|away is now known as khuey
- # [14:47] * Joins: davidb (davidb@F2D29657.F60B0462.67AC9B1.IP)
- # [14:47] * ChanServ sets mode: +qo davidb davidb
- # [14:49] <@tbsaunde> mjorning davidb
- # [14:52] <@MarcoZ> Good morning davidb!
- # [14:52] <@davidb> hi all!
- # [14:52] <@davidb> hi tbsaunde MarcoZ
- # [14:53] <@davidb> surkov: you are filling my bug mail folder :)
- # [14:53] <@davidb> love it
- # [14:53] <@surkov> thx
- # [14:56] <@MarcoZ> davidb: Anything you'd like to have added to today's agenda? https://wiki.mozilla.org/Accessibility/Meetings/2012-02-08
- # [14:56] <@MarcoZ> surkov: Great blog post by the way!
- # [14:56] <@surkov> MarcoZ: thx!
- # [14:58] <@davidb> MarcoZ: maybe telemetry/memory
- # [15:09] <@MarcoZ> davidb: OK! Have you found something interesting or worrying?
- # [15:09] <@davidb> minor interesting thing
- # [15:09] <@MarcoZ> davidb: OK.
- # [15:10] <@MarcoZ> davidb: Speaking of telemetry: Has the screen reader picture changed at all since the uplift to Aurora happened?
- # [15:10] <@davidb> that's the little thing :)
- # [15:10] <@davidb> WE shows up
- # [15:10] <@MarcoZ> OK :)
- # [15:10] <@MarcoZ> davidb: Oh how nice!
- # [15:10] <@davidb> but also i want to comment briefly on what we are collecting, and what we might want to collect
- # [15:12] <@MarcoZ> davidb: Added.
- # [15:16] <@MarcoZ> surkov: Regarding bug 558589, I believe I'll have to split the tests up into two parts somehow: One that tests the manual autocomplete stuff and one that does the same with the searchBar autocomplete. Not sure how I have to do that yet, but I believe just putting it all into a separate browser window wouldn't work.
- # [15:16] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=558589 min, --, ---, marco.zehe, ASSI, A11y tests which depend on the "tabbrowser API" should explicitly check its availability.
- # [15:17] * Joins: peteb-away (ptbrunet@moz-E9B02845.austin.res.rr.com)
- # [15:26] * Joins: clown (clown@67828CC7.C1A51174.9D42CF23.IP)
- # [15:29] * Joins: jprmc (jprmc@F2D29657.F60B0462.67AC9B1.IP)
- # [15:29] * ChanServ sets mode: +o jprmc
- # [15:42] * Joins: skyler_brungardt (skyler@moz-9FA3C392.washdc.fios.verizon.net)
- # [15:42] * Quits: skyler_brungardt (skyler@moz-9FA3C392.washdc.fios.verizon.net) (Quit: skyler_brungardt)
- # [15:43] * Joins: skyler_brungardt (skyler@moz-9FA3C392.washdc.fios.verizon.net)
- # [15:49] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Connection reset by peer)
- # [15:49] <@surkov> MarcoZ: splitting is fine if we need to, I'm not sure what exactly you keep in mind though but I hope to read in bugzilla soon :)
- # [15:50] <@MarcoZ> surkov: I'm just thinking that it might be difficult to simply try to access the search bar within the current test file. But I can try that.
- # [15:50] <@surkov> yeah, right, separate test is fine I think
- # [15:51] <@surkov> we shouldn't mix UI and XUL/HTML things
- # [16:00] * Quits: @surkov (surkov@E3B837DE.44A4068D.222B27F0.IP) (Quit: surkov)
- # [16:04] * Joins: askalski (akuda@moz-104CC309.mv.mozilla.com)
- # [16:04] * ChanServ sets mode: +o askalski
- # [16:04] <@askalski> hi
- # [16:04] <@askalski> I have problem accessing Vidyo, anyone else?
- # [16:05] <@askalski> I always wanted to talk to firebot.
- # [16:05] <@askalski> firebot, hi!
- # [16:05] <@firebot> askalski
- # [16:06] <@tbsaunde> askalski: haven't tried today
- # [16:06] <@askalski> davidb, hi, is Vidyo working?
- # [16:06] <@davidb> i'll try now
- # [16:06] <@askalski> if it's not then it's probably IP settings here, I can either go via skype gateway (if it exists) or route the connection via VPN which will kill the performance
- # [16:07] <@davidb> WFM
- # [16:07] <@davidb> askalski: i'll hang in the room if you want to try (meeting is in 30 mins)
- # [16:08] <@tbsaunde> davidb: so, with video is there a way for people without ldap access to participate easily? I gues we can have them use the phone thing
- # [16:08] <@davidb> yes
- # [16:08] <@askalski> davidb, good, but it seems like my vidyodesktop died
- # [16:08] <@davidb> there is a link they can click
- # [16:08] <@tbsaunde> davidb: oh, I thought MarcoZ said that was just if you have an ldap account
- # [16:09] <@davidb> tbsaunde: there is a link here https://wiki.mozilla.org/Accessibility/Meetings
- # [16:09] <@davidb> not sure you need ldap for that
- # [16:09] <@davidb> here is the link: https://v.mozilla.com/flex.html?roomdirect.html&key=DsmJFkwXvWkY
- # [16:12] <@askalski> davidb, honestly, I don't think I will be able to connect
- # [16:12] <@askalski> the vidyodesktop is runned as always, but firefox does not see it
- # [16:12] <@davidb> odd
- # [16:12] <@askalski> it's the problem with application
- # [16:12] <@askalski> I guess it's ff update
- # [16:12] <@tbsaunde> davidb: ok, well, orca really doesn't like that page, put plugin-container doesn't seem to be running so I'm not sure what's up
- # [16:13] <@askalski> I have an idea, w8
- # [16:13] <@davidb> tbsaunde: oh, that link is for actual vidyo desktop client users
- # [16:13] <@davidb> (i think)
- # [16:13] * Joins: ptheriault (ptheriault@moz-4BE034AB.ptr.us.xo.net)
- # [16:13] <@tbsaunde> davidb: oic
- # [16:14] <@tbsaunde> davidb: I don't really care, but I think it should be possible for people with out ldap to join without having to ask for help :)
- # [16:14] * Joins: hub (hub@moz-E2FCA694.figuiere.net)
- # [16:14] * ChanServ sets mode: +o hub
- # [16:14] <@askalski> davidb, it's vidyodesktop. other browser doesn't change anything
- # [16:14] <@askalski> I'll reboot just for sure
- # [16:14] <@davidb> tbsaunde: i agree
- # [16:14] <@davidb> askalski: ok
- # [16:14] * Quits: @askalski (akuda@moz-104CC309.mv.mozilla.com) (Quit: Wychodzi)
- # [16:15] <@davidb> tbsaunde: so, as far as I know, people can freely dial in, and if they want to install vidyo they can then use that direct link.
- # [16:15] * Joins: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP)
- # [16:17] <@tbsaunde> davidb: oh, I see MarcoZ did mention that way
- # [16:17] * Joins: muralisr92 (chatzilla@moz-A696D59A.pc.nus.edu.sg)
- # [16:17] <@davidb> hi muralisr92
- # [16:17] <muralisr92> hey
- # [16:18] <muralisr92> i'm going to look at our bug now
- # [16:18] <muralisr92> but i saw the comment from surkov
- # [16:19] <muralisr92> but i am unable to understand what he is trying to sayy :(
- # [16:19] <@davidb> muralisr92: basically you need to trigger the context menu like in the test he points to, then test the tree.
- # [16:20] <@davidb> unfortunately he's going to bed now
- # [16:20] * Joins: askalski (akuda@moz-104CC309.mv.mozilla.com)
- # [16:20] * ChanServ sets mode: +o askalski
- # [16:20] <@askalski> davidb, no progress
- # [16:21] <@davidb> askalski: can you dial in to the vidyo meeting?
- # [16:21] <@askalski> davidb, you mean by phone? double roaming would kill me
- # [16:21] <@davidb> can you use Skype to dial out?
- # [16:22] <@tbsaunde> askalski: you should be able to use skype I think if thats better
- # [16:22] <@davidb> maybe there is a sip bridge
- # [16:22] <@askalski> davidb, the only possible option I can think of is v... darn, virtualbox cannot bridge cameras, it's not vmware
- # [16:22] * @davidb multi-multitasks
- # [16:22] <@davidb> we don't need camera
- # [16:22] <@askalski> davidb, I can ask at #desktop if you
- # [16:22] <@askalski> *you're busy
- # [16:22] <@davidb> sure
- # [16:22] <@askalski> ok
- # [16:22] <@davidb> they know more than me anyway :)
- # [16:23] <@davidb> (always busy - always want to help)
- # [16:23] <muralisr92> davidb: so right now, im just testing the tree.. i need to do the stuff like opening the menu, selecting an item in it etc and then test the tree??
- # [16:23] <@askalski> davidb, darn, #desktop is yet asleep
- # [16:24] <@tbsaunde> muralisr92: I believe so
- # [16:26] <@davidb> muralisr92: you really need to actually be able to run the tests in order to fix this bug.
- # [16:26] <@davidb> I highly recommend that as your next step.
- # [16:26] <@davidb> details are here: https://wiki.mozilla.org/Accessibility/Contribute
- # [16:26] <muralisr92> i was just googling for that page!
- # [16:26] <@davidb> muralisr92: if you do it this way: python runtests.py --a11y
- # [16:27] <@davidb> then you can select the specific test to run (they won't all auto run)
- # [16:27] <@tbsaunde> davidb: you can do that with TEST_PATH=blah make mochitest-a11y too :)
- # [16:27] <muralisr92> okay, so i should write up the test, then run it and if it succeeds, i'll put it up??
- # [16:27] <@tbsaunde> fwiw
- # [16:28] <@davidb> tbsaunde: yep
- # [16:28] <@davidb> tbsaunde: feel free to add to the wiki
- # [16:28] <@tbsaunde> ok
- # [16:28] <@davidb> muralisr92: running the tests and seeing how they fail will help you iterate toward a good solution.
- # [16:29] * Joins: nhirata (nhirata.bu@moz-BBE3ABD.mv.mozilla.com)
- # [16:29] <@davidb> muralisr92: for example, you might find that you can't test the menu until it is invoked
- # [16:29] <@MarcoZ> tbsaunde: I don't understand why this convoluted make statement is necessary if you can just run runtests.py with the --a11y parameter. :)
- # [16:30] <@tbsaunde> MarcoZ: it's not really that conviluted, but it means once the thing starts you don't have to do anything
- # [16:30] <@davidb> convolution is in the eye of the beholder :)
- # [16:30] * @tbsaunde prefers to turn orca off when running tests
- # [16:30] <muralisr92> okay thanks a lot! i'll look at the context menu example, understand it and implement the same thing in my test!! and im really really sorry, but if i run into some piece of code that i dont understand i'll drop by here and nag u guys again... :P
- # [16:31] <@MarcoZ> tbsaunde: Ah, so that line tells the make -f client.mk build command to run the tests automatically once the build finishes?
- # [16:31] <@MarcoZ> tbsaunde: Yeah I turn off NVDA, too.
- # [16:32] <@davidb> askalski: can you use Skype to dial +1 800 707 2533 (pin 369) Conf# 99369
- # [16:32] <@tbsaunde> MarcoZ: no, it only runs the tests, but if you don't set TEST_PATH in the enviroment of make then it make mochitest-a11y does all the tests
- # [16:33] <@tbsaunde> anyway, I'm moving to someplace I can call from
- # [16:33] * Quits: mike5w3c (MikeS@moz-DAFE1A45.tokyo.ocn.ne.jp) (Quit: mike5w3c)
- # [16:34] * tbsaunde is now known as tbsaunde|mtg
- # [16:35] * davidb is now known as davidb|mtg
- # [16:37] * Joins: eeejay_mac (eitan@moz-6CE338E8.tukw.qwest.net)
- # [16:38] <@davidb|mtg> https://wiki.mozilla.org/Accessibility/Meetings/2012-02-08
- # [16:38] <@davidb|mtg> meeting is on!
- # [16:39] * Joins: mike5w3c (MikeS@moz-DAFE1A45.tokyo.ocn.ne.jp)
- # [16:43] * Quits: jhk (jiggy@8E6C34C1.A3F9767A.1C37C358.IP) (Ping timeout)
- # [16:45] <@davidb|mtg> askalski: can you mute
- # [16:45] <@askalski> I can mute
- # [16:45] <@askalski> OK
- # [16:45] <@askalski> davidb|mtg, done
- # [16:46] * Quits: victorporof (victorporo@86FDEBD.6D2FB7A4.79933D60.IP) (Ping timeout)
- # [16:47] <@askalski> davidb|mtg, I'm still waiting for confirmation from my promotor
- # [16:47] <@askalski> on adding a week
- # [16:47] <@davidb|mtg> askalski: thanks
- # [16:47] <@davidb|mtg> hub: not sure your mic is hooked up
- # [16:52] <@hub> davidb|mtg: it was hooked up, just low volume. i'm on a different machine this AM
- # [16:52] <@davidb|mtg> ok
- # [16:52] <@hub> so the setting were not the same
- # [17:00] <@askalski> number?
- # [17:01] <@davidb|mtg> bug 724672 - you don't need to worry too much about it.
- # [17:01] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=724672 nor, --, 1.7, bolterbugz, NEW, [One Mozilla clean-up] Accessibility Content
- # [17:01] <@davidb|mtg> mostly it is triage work for people senior to moz a11y
- # [17:02] <@firebot> tobbi.bugs@googlemail.com requested review from surkov.alexander@gm ail.com for attachment 595410 on bug 725178.
- # [17:02] <@firebot> tobbi.bugs@googlemail.com changed the Status on bug 725178 from NEW to ASSIGNED.
- # [17:02] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=725178 nor, --, ---, nobody, NEW, get rid ensureAccessibleTree of common.js
- # [17:03] * Joins: victorporof (victorporo@C092FEB2.1C233438.79933D60.IP)
- # [17:03] <@askalski> none offence taken :)
- # [17:04] * Quits: muralisr92 (chatzilla@moz-A696D59A.pc.nus.edu.sg) (Ping timeout)
- # [17:04] * Joins: muralisr92 (chatzilla@moz-A696D59A.pc.nus.edu.sg)
- # [17:04] <@askalski> you mean the desing of new standards?
- # [17:04] <@askalski> *design
- # [17:05] <@askalski> davidb|mtg, do you mean designing new standards?
- # [17:05] <@davidb|mtg> askalski: no, designing out web presence
- # [17:05] <@davidb|mtg> out/our
- # [17:05] <@askalski> ok
- # [17:09] <@askalski> davidb|mtg, can you describe more about thunderbird?
- # [17:09] <@askalski> davidb|mtg, what kind of experience is required?
- # [17:09] <@davidb|mtg> bug 763860
- # [17:09] <@davidb|mtg> askalski: will tell you later :)
- # [17:12] <@askalski> bye
- # [17:12] * Quits: muralisr92 (chatzilla@moz-A696D59A.pc.nus.edu.sg) (Quit: ChatZilla 0.9.88 [Firefox 10.0/20120129141257])
- # [17:12] <@davidb|mtg> MarcoZ: can you give me 5-10 mins?
- # [17:13] * @davidb|mtg has to preview a devtool blog
- # [17:15] <@MarcoZ> davidb|mtg: Sure!
- # [17:15] <@MarcoZ> davidb|mtg: The bug for TB that we talked about is bug 673860.
- # [17:15] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=673860 maj, --, ---, nobody, NEW, [Thunderbird] Adding more than one addressee in message composition, we don't create accessibles for
- # [17:15] <@davidb|mtg> thanks
- # [17:15] * tbsaunde|mtg is now known as tbsaunde
- # [17:19] <@davidb|mtg> who are our python hackers? eeejay, askalski … tbsaunde ?
- # [17:19] * @davidb|mtg is ready when MarcoZ is
- # [17:20] <@askalski> davidb|mtg, oh, actually there is something I would like you to ask
- # [17:20] <@tbsaunde> davidb|mtg: I wouldn't really say I know python, but if I need to I can hit it with a hammer till it works
- # [17:20] <@askalski> davidb|mtg, for a long time I was thinking about becoming a scala-hacker :), but that would consume some more time
- # [17:20] <@davidb|mtg> heh
- # [17:20] * @davidb|mtg really in meeting now
- # [17:20] <@askalski> davidb|mtg, OK
- # [17:24] * Quits: skyler_brungardt (skyler@moz-9FA3C392.washdc.fios.verizon.net) (Quit: skyler_brungardt)
- # [17:25] <@tbsaunde> MarcoZ: isn't that bug probably a dup of one gin filed a while ago? 65xxxx about the profile manager thing and creating new profiles
- # [17:26] * Quits: ptheriault (ptheriault@moz-4BE034AB.ptr.us.xo.net) (Connection reset by peer)
- # [17:29] <@tbsaunde> ok, /me -> class
- # [17:29] * tbsaunde is now known as tbsaunde|afk
- # [17:34] * davidb|mtg is now known as davidb
- # [17:34] <@davidb> askalski: a question?
- # [17:35] <@askalski> davidb, about whether I can waste some time learning scala for next scripts, or should I stick with Python for performance and reason that people are more familiar with it
- # [17:36] <@askalski> davidb, I have some basics, but I guess I would be writing slowly. Still, scripts does not consume much of the time. I spend it mostly compiling and mochitesting :D
- # [17:36] <@askalski> (and reading)
- # [17:37] <@askalski> davidb, anyway, if you don't see future applications of scala, then I'll just do it in my spare time some day later
- # [17:40] * Joins: skyler_brungardt (skyler@3459447D.62BEB748.75EC3910.IP)
- # [17:40] <@davidb> sorry stepped away to look at a popcorn app
- # [17:40] <@davidb> back now
- # [17:40] <@askalski> davidb, popcorn app?
- # [17:40] <@davidb> askalski: i wouldn't spend time on scala
- # [17:41] <@askalski> davidb, OK
- # [17:41] <@askalski> what is popcorn app?
- # [17:41] <@davidb> http://popcornjs.org/
- # [17:41] <@davidb> somebody at a local school is building a slideshow using popcorn.js and we'll make it accessible.
- # [17:42] <@davidb> i'll pour an hour or two into it when he has the next rev done
- # [17:43] <@MarcoZ> davidb: Very nice!
- # [17:43] <@MarcoZ> I'd love to try it out then.
- # [17:43] <@davidb> it is very promising
- # [17:43] <@davidb> great
- # [17:43] <@askalski> davidb, I am going to relocate now, it got cold in this room, be back in 15
- # [17:43] <@davidb> ok
- # [17:43] * Quits: @askalski (akuda@moz-104CC309.mv.mozilla.com) (Quit: Wychodzi)
- # [17:50] <@firebot> hub@mozilla.com requested review from surkov.alexander@gm ail.com for attachment 595422 on bug 721947.
- # [17:50] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=721947 nor, --, ---, hub, NEW, don't use nsIWeakShell
- # [17:52] * khuey is now known as khuey|away
- # [18:00] * Joins: a-865 (fmcz@moz-A5D13CA.cable.mindspring.com)
- # [18:06] <@firebot> bmo@edmorley.co.uk changed the Status on bug 717506 from NEW to RESOLVED.
- # [18:06] <@firebot> bmo@edmorley.co.uk set the Resolution field on bug 717506 to FIXED.
- # [18:07] <@firebot> bmo@edmorley.co.uk changed the Target Milestone on bug 717506 from --- to mozilla13.
- # [18:07] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=717506 nor, --, mozilla13, askalski, RESO FIXED, telemetry for xforms use
- # [18:07] <@firebot> bmo@edmorley.co.uk changed the Status on bug 723833 from ASSIGNED to RESOLVED.
- # [18:07] <@firebot> bmo@edmorley.co.uk set the Resolution field on bug 723833 to FIXED.
- # [18:07] <@firebot> bmo@edmorley.co.uk changed the Target Milestone on bug 723833 from --- to mozilla13.
- # [18:07] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=723833 nor, --, mozilla13, surkov.alexander, RESO FIXED, IAccessibleText::setCaretOffset on location or search bar causes focus to jump
- # [18:16] <@firebot> bmo@edmorley.co.uk changed the Status on bug 723796 from NEW to RESOLVED.
- # [18:16] <@firebot> bmo@edmorley.co.uk set the Resolution field on bug 723796 to FIXED.
- # [18:16] <@firebot> bmo@edmorley.co.uk changed the Target Milestone on bug 723796 from --- to mozilla13.
- # [18:16] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=723796 nor, --, mozilla13, joey.blacksmith, RESO FIXED, rename nsAccessible::GetAllowsAnonChildAccessibles
- # [18:24] * Joins: askalski (akuda@moz-104CC309.mv.mozilla.com)
- # [18:24] * ChanServ sets mode: +o askalski
- # [18:26] * khuey|away is now known as khuey
- # [18:29] * Joins: ptheriault (ptheriault@moz-4BE034AB.ptr.us.xo.net)
- # [18:32] <@askalski> MarcoZ, is there a way to tell hg to revert a changes for a single file, and by saying revert-the-changes I mean "unapply current patch to this single file"?
- # [18:32] <@hub> hg revert path/to/file
- # [18:33] <@hub> oh wait, from the patch queue?
- # [18:45] * Joins: richardschwerdtfeger (richardsch@moz-130F1494.lightspeed.austtx.sbcglobal.net)
- # [18:52] <@askalski> hub, I wanted to revert all changes induced by queue, and just re-download head version
- # [18:53] <@askalski> hub, did it manually this time :(
- # [18:53] <@askalski> davidb, how often do you checkout (update) repository? or - how long one can work without updating? recompile still takes me a long time
- # [18:55] <@davidb> on a development day i usually do an update
- # [18:56] <@askalski> davidb, ok. same here
- # [18:56] <@davidb> askalski: if compiling is slow for you i'd like to solve that
- # [18:56] <@askalski> davidb, I use ccache and all the stuff, and talked with #desktop
- # [18:57] <@davidb> (i7 + SSD + >=8 GB RAM)
- # [18:57] <@askalski> they offered me a laptop to pick up while I'll be in Toronto
- # [18:57] <@davidb> good
- # [18:57] <@askalski> yes. that I think. lenovo w500 is my type now, or dell m4600 (it has a lot of ram option)
- # [18:57] <@davidb> askalski: cc me on any related bug please
- # [18:57] <@askalski> OK
- # [18:57] <@askalski> *w520
- # [18:58] <@askalski> ok, i need to relocate again, be back in an hour. I am finishing with ARIA bug right now
- # [18:58] <@askalski> it just compiled, gonna test at home
- # [18:58] <@askalski> see you
- # [18:58] <@davidb> ciao
- # [18:58] * Quits: @askalski (akuda@moz-104CC309.mv.mozilla.com) (Quit: Wychodzi)
- # [19:02] <@MarcoZ> davidb: I'm *really* curious if we will ever see XForms use show up in telemetry.
- # [19:02] <@davidb> i don't see how we can
- # [19:03] <@davidb> but yes
- # [19:03] <@MarcoZ> davidb: Have you looked at other data that we collect, like the depricated table stuff and iSimpleDOM usage statistics?
- # [19:03] <@davidb> that's the one we can't get clarity of the data on yet, except that it is used.
- # [19:04] <@davidb> (ISimpleDOM)
- # [19:10] <@davidb> I see IAccessibleTable usage as well
- # [19:10] <@davidb> i can't correlate by screen reader though
- # [19:10] * @davidb needs coffee
- # [19:20] <@davidb> ok espresso, do your thing
- # [19:22] <@davidb> oh 11% WE
- # [19:24] <@MarcoZ> davidb: It's probably older versions of WE and JAWS. Can't imagine much older NVDA still floating around.
- # [19:25] <@MarcoZ> OK, I'm going off for the evening. Hub, good luck with getting a review and your third landing attempt! :)
- # [19:25] <@davidb> g'night
- # [19:25] * Quits: @MarcoZ (marco.zehe@moz-20316C0D.dip.t-dialin.net) (Quit: Good night!)
- # [19:31] * Quits: @hub (hub@moz-E2FCA694.figuiere.net) (Ping timeout)
- # [19:34] <@firebot> jigneshhk1992@gmail.com requested review from surkov.alexander@gm ail.com for attachment 595446 on bug 539694.
- # [19:34] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=539694 nor, --, mozilla11, jigneshhk1992, ASSI, accessible objects should have private copy constructor
- # [19:46] * Joins: askalski (akuda@moz-4C8A107E.pool85-48-91.dynamic.orange.es)
- # [19:46] * ChanServ sets mode: +o askalski
- # [19:47] <@askalski> back!
- # [19:48] * Joins: hub (hub@21B7B9F2.B87E9213.6E712CE2.IP)
- # [19:48] * ChanServ sets mode: +o hub
- # [19:53] <@tbsaunde|afk> askalski: I usually update every so often but probably not every day
- # [19:54] <@askalski> tbsaunde|afk, thanks. so far I usually did every morning while doing first coffee. you know, I/O operations :)
- # [19:55] * tbsaunde|afk is now known as tbsaunde
- # [19:55] <@askalski> tbsaunde, today I just wanted to avoid that while fixing yesterday patches
- # [19:55] <@tbsaunde> askalski: I gues that works, but it often will mean you'll have to start the day with a full build
- # [19:55] <@tbsaunde> and I have several work dirs
- # [19:56] <@askalski> I have two repos, cause it never happened to me to get clean mochitest results, so I need a clean build as a reference of what is OK right now :)
- # [19:56] <@tbsaunde> I ... see
- # [19:57] <@tbsaunde> one of the focus auto complete foo tests pretty reproducably fails for me, but I just sort of remember that and ignore it
- # [19:57] <@tbsaunde> I should probably debug it, but ;efort
- # [19:58] <@tbsaunde> anyway unless I think stuff has landed that will substantially break I'll go a week without pulling easily
- # [19:58] <@tbsaunde> breka me I mean
- # [19:58] <@tbsaunde> err, break me
- # [19:59] <@tbsaunde> and then I basically always just rebuild accessible/ and libxul which only takes me a minute or two
- # [19:59] <@askalski> tbsaunde, I posted a python script that does some nice job differing results of mochitests
- # [19:59] <@tbsaunde> I don't think it possibly took more than 10 minutes on my old core 2 thinkpad
- # [20:00] <@tbsaunde> askalski: thats nice, but I've never wanted to do that :)
- # [20:01] <@askalski> tbsaunde, your old core 2 thinkpad sound pretty much as my present main machine :D
- # [20:03] <@tbsaunde> really? that machine only had 2 gb of ram
- # [20:03] <@tbsaunde> I thought you ad an i5 with 6gb?
- # [20:03] * Joins: ehsan (ehsan@F2D29657.F60B0462.67AC9B1.IP)
- # [20:04] <@askalski> tbsaunde, I bought more
- # [20:05] <@tbsaunde> anyway a bit better machine
- # [20:06] * Joins: hub_ (hub@21B7B9F2.B87E9213.6E712CE2.IP)
- # [20:06] * Quits: @hub (hub@21B7B9F2.B87E9213.6E712CE2.IP) (Input/output error)
- # [20:06] <@tbsaunde> any way if your spending a lot of time waiting for compiles I would tend to expect your somehow doing something wrong
- # [20:07] * Quits: hub_ (hub@21B7B9F2.B87E9213.6E712CE2.IP) (Quit: Leaving)
- # [20:07] * Joins: hub_ (hub@21B7B9F2.B87E9213.6E712CE2.IP)
- # [20:07] * hub_ is now known as hub
- # [20:10] * Quits: eeejay_mac (eitan@moz-6CE338E8.tukw.qwest.net) (Quit: eeejay_mac)
- # [20:11] <@askalski> tbsaunde, around 5-12 minutes, 5 is when I recompile after a patch an 12 is after update
- # [20:12] <@tbsaunde> askalski: how are you compiling?\
- # [20:13] <@askalski> tbsaunde, --ccache, I guess that's all
- # [20:13] <@askalski> tbsaunde, I can send you my .mozbuild
- # [20:14] <@askalski> tbsaunde, it's not a scientific data though
- # [20:14] <@tbsaunde> askalski: sure, but how do you invoke make?
- # [20:14] <@askalski> make -f client.mk
- # [20:14] <@askalski> tbsaunde, or more specifically "time make -f client.mk"
- # [20:15] <@tbsaunde> askalski: unless you really need a full build for some reason that's dumb
- # [20:16] <@askalski> tbsaunde, so how?
- # [20:16] <@tbsaunde> askalski: you can generally get away with make -C objdir/accessible/ -<n> && make -C objdir/toolkit/library/ -j<n>
- # [20:16] <@askalski> tbsaunde, -<n>?
- # [20:16] <@tbsaunde> and of course y objdir/accessible/ as appropriate if you changed other dirs
- # [20:17] <@tbsaunde> askalski: -j<n>
- # [20:17] <@tbsaunde> read the man page for make
- # [20:17] <@askalski> tbsaunde, multithreading, right?
- # [20:17] <@askalski> tbsaunde, you just forgot the first "j"
- # [20:17] <@askalski> tbsaunde, I am gentoo user, I read the gcc/make man :)
- # [20:19] <hub> it is even slower on Mac
- # [20:23] <@firebot> tobbi.bugs@googlemail.com cancelled review?(surkov.alexander@gm ail.com) for attachment 595410 on bug 725178.
- # [20:23] <@firebot> tobbi.bugs@googlemail.com requested review from surkov.alexander@gm ail.com for attachment 595463 on bug 725178.
- # [20:23] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=725178 nor, --, ---, tobbi.bugs, ASSI, get rid ensureAccessibleTree of common.js
- # [20:23] * Joins: WeirdAl (chatzilla@moz-D461843.ask.info)
- # [20:24] * Joins: jimm (jmathies@moz-7F164CA1.pn.at.cox.net)
- # [20:24] <@tbsaunde> hub: ... sucks to suck :/
- # [20:24] <@davidb> hi WeirdAl, jimm
- # [20:24] <WeirdAl> hello
- # [20:24] * Joins: Mook_as (mook@moz-1FCC0032.activestate.com)
- # [20:24] <@davidb> all, we're going to be discussing bug 719561 for a bit - sorry for noise
- # [20:24] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=719561 cri, --, ---, nobody, NEW, Updater.exe from Ask Toolbar causes non-English Firefox to hang when viewing Flash videos in YouTube
- # [20:24] <@davidb> WeirdAl: are you Eric?
- # [20:24] <@tbsaunde> davidb: honestly I'd rather just fix our code to the point random people choosing to use a11y was fine
- # [20:25] <@davidb> tbsaunde: yes, but in the meantime.
- # [20:25] <WeirdAl> no, he's a product VP. I'm a Mozilla dev, but not familiar with accessibility code
- # [20:25] <@davidb> WeirdAl: ok
- # [20:25] <@tbsaunde> davidb: true
- # [20:25] <WeirdAl> ideally, we'd do both: fix the hang in FF code, and fix Ask code to not do stupid stuff
- # [20:26] <@davidb> yep
- # [20:26] <jimm> do we support both MSAA and UIA?
- # [20:26] <@davidb> jimm: not UIA.
- # [20:26] <@davidb> (MSAA + IA2)
- # [20:26] <@davidb> WeirdAl: so to catch me up, what do you think is the best short term action?
- # [20:26] <jimm> hmm, ok, anyway
- # [20:27] <WeirdAl> davidb: first, build this XMSAALib tool and run it against FF to reproduce the hang
- # [20:27] * @davidb is swamped
- # [20:27] <jimm> do we have any stacks from those hangs?
- # [20:28] <WeirdAl> unfortunately not because I haven't had much practice at isolating hangs
- # [20:28] * Quits: mike5w3c (MikeS@moz-DAFE1A45.tokyo.ocn.ne.jp) (Quit: mike5w3c)
- # [20:28] <jimm> who is :Cww?
- # [20:28] <jimm> sounds like he can reproduce
- # [20:28] <@davidb> i don't know
- # [20:28] <WeirdAl> this can be reproduced w/o the Ask toolbar installed
- # [20:28] <khuey> that's Cheng
- # [20:28] <khuey> the SUMO guy
- # [20:28] <WeirdAl> w/o any Ask code at all
- # [20:29] <@davidb> khuey: ty
- # [20:30] <jimm> hmm, maybe a regular old plugin hang related to ipc.
- # [20:30] <WeirdAl> good news is I do have a symbols build of FF9 with a11y enabled
- # [20:30] <khuey> davidb: we can invite him to the channel if it will help
- # [20:30] <khuey> or I can track him down in person
- # [20:30] <jimm> it would be great if we can get str (on english would be wonderful) and some stacks from a debugger.
- # [20:30] <@davidb> khuey: go ahead with both please
- # [20:30] <WeirdAl> jimm: str on english is easy, once you build that XMSAALib tool
- # [20:30] <khuey> actually, he may be in the product meeting
- # [20:31] <khuey> so I'm not going to chase him down just yet
- # [20:31] <WeirdAl> I have a standup at 11:30, but I can be back here about 11:50
- # [20:32] <@davidb> ok
- # [20:34] <@davidb> what is Updater.exe?
- # [20:34] <jimm> trying to get this XMSAALib downloaded..
- # [20:34] <@davidb> jimm++
- # [20:35] <WeirdAl> Updater.exe is a program that's supposed to update our toolbar installation from one version to another, including in IE and Chrome
- # [20:35] <WeirdAl> wait
- # [20:35] <WeirdAl> unfortunately, I'm not the guy who wrote that code, so I might be misspeaking a little
- # [20:36] <@davidb> this sounds… not quite kosher
- # [20:36] <WeirdAl> preaching to the choir
- # [20:36] <@davidb> :)
- # [20:37] <WeirdAl> we're working on a complete rewrite of this platform, where we actually do follow best practices, but it's not ready yet
- # [20:37] <@tbsaunde> davidb: what does it really matter to us? its strange, but its still a test case for a bug in our code
- # [20:37] <@davidb> WeirdAl: that's good to hear. will you require a11y api?
- # [20:37] <@davidb> tbsaunde: depends how you define we.
- # [20:37] <@davidb> or us
- # [20:38] <@davidb> the immediate problem is all kinds of people hanging.
- # [20:38] <WeirdAl> davidb: doubtful, since I'm working on a separate proposal that changes this whole third-party apps discussion.
- # [20:38] <@davidb> WeirdAl: good
- # [20:39] <@davidb> tbsaunde: there are basically two things I want to do in these situations:
- # [20:39] <@davidb> 1. stop the bleeding.
- # [20:39] <@davidb> 2. fix the problem.
- # [20:39] <@davidb> If 2 is going to take a while… well the metaphor holds.
- # [20:39] <WeirdAl> sorry, I have my standup now, back asap
- # [20:39] <@davidb> later
- # [20:39] <WeirdAl> I'll stay on IRC
- # [20:39] * @davidb nods
- # [20:39] <@davidb> thanks WeirdAl
- # [20:40] * Joins: aaronlev (chatzilla@moz-CDA191A6.c3-0.arl-ubr1.sbo-arl.ma.cable.rcn.com)
- # [20:40] <aaronlev> davidb, mind if I email bugs until I have more time to use bugzilla again?
- # [20:41] <@davidb> aaronlev: i'll take em.
- # [20:41] <@davidb> aaronlev: please use my mozilla email
- # [20:41] <jimm> hmm, so this test app runs, and it can query fx for various fields info. I'm not seeing any lockups, but maybe there is specific STR here as well.
- # [20:42] <@davidb> jimm: is dom.ipc.plugin.enabled true?
- # [20:43] <jimm> yep, oh, I didn't try youtube, lemmie try that
- # [20:43] <WeirdAl> jimm: yeah, somehow the plugin container is key: kill plugin-container and it comes back to life
- # [20:43] <WeirdAl> sorry, I guess I forgot that
- # [20:44] <jimm> hmm, so clicking around youtube videos - no problems thus far.
- # [20:45] <WeirdAl> a video has to be playing to see the hang
- # [20:45] <WeirdAl> Display Object Tree is the button in question to cause it
- # [20:45] <jimm> hmm, that is generating an error in the test app.
- # [20:46] <WeirdAl> what's the error?
- # [20:47] <jimm> something about an xml config file missing
- # [20:47] <WeirdAl> try to run the "enum all browser objects" button first
- # [20:47] <WeirdAl> and if that fails, it's because XMSAALib can't write to the folder it's running in
- # [20:48] <WeirdAl> (the folder you launched it from)
- # [20:48] <jimm> ah, it says, "that browser is not running"
- # [20:48] <jimm> but fx is running. seems like there's some sort of browser detection problem.
- # [20:48] <WeirdAl> I hit the hang by deliberately closing my stock FF and running the symbols-built FF w/ a11y
- # [20:49] <@tbsaunde> davidb: yes, I'm saying we should fix hangs
- # [20:49] <jimm> just a regular old release build of the latest fx should work right?
- # [20:49] <@davidb> WeirdAl: can you attach a debugger and break?
- # [20:49] <WeirdAl> oh, damn, I forgot
- # [20:49] <@tbsaunde> and maybe take hacky patches if we have to
- # [20:49] <WeirdAl> the eng who found the hang said there was a patch to XMSAALib that he had to apply
- # [20:50] <@tbsaunde> my point was that I don't really care that my response to somebody saying I'm doing these str is "don't do that!"
- # [20:51] <WeirdAl> sorry - maybe I should get the eng in question on IRC
- # [20:51] <WeirdAl> For XMSAALib, there's this code change:
- # [20:51] <WeirdAl> Define the correct window class in the file MSAALib.cpp line #51 as follows:
- # [20:52] <WeirdAl> #define MOZILLA_FIREFOX_APP_WINDOW _T("MozillaWindowClass")
- # [20:52] <@davidb> WeirdAl: can you add that to the bug?
- # [20:52] <WeirdAl> will do
- # [20:52] <@davidb> thanks
- # [20:52] <@davidb> jimm: ^
- # [20:53] <jimm> heh, from when we removed sub widgets
- # [20:53] <jimm> trying
- # [20:53] <WeirdAl> ?
- # [20:54] <jimm> the change of the window class - we broke a bunch of accessibility apps in 4.0 with that
- # [20:54] * @davidb nods solemnly
- # [20:54] <jimm> enum is working now
- # [20:55] <WeirdAl> ah, so we're not the only ones with obsolete, stupid code ;)
- # [20:56] <@davidb> all code over 2 years old is legacy code :)
- # [20:57] <WeirdAl> so, enum has to run first, then display object tree
- # [20:57] <WeirdAl> it may not happen on your first run while the video is playing, so you may have to try a few times while the video is playing
- # [20:57] <jimm> what should I expect to see - a freeze of fx proper, or just something weird with the video playing?
- # [20:58] * Quits: a-865 (fmcz@moz-A5D13CA.cable.mindspring.com) (Ping timeout)
- # [20:58] <WeirdAl> FF freezes, as does the XMSAALib app
- # [20:58] <jimm> ok
- # [20:58] <WeirdAl> basically XMSAALib is asking for info that FF doesn't provide and they both deadlock
- # [20:59] * Joins: a-865 (fmcz@moz-A5D13CA.cable.mindspring.com)
- # [20:59] <WeirdAl> that was the word used, "deadlock"
- # [20:59] <jimm> it
- # [20:59] <jimm> er..
- # [20:59] <jimm> "got it"
- # [20:59] <jimm> the plugin crashed
- # [21:00] <WeirdAl> we didn't see that crash. we only saw the hang and had to kill plugin-container (which of course reports plugin crash), so we didn't know for sure
- # [21:01] <WeirdAl> a plugin crash is most definitely a possibility I thought about, but I had no way to prove.
- # [21:01] <WeirdAl> or disprove, for that matter
- # [21:02] <@firebot> New Core - Disability Access APIs bug 725432 filed by bolterbugz@gmail.com.
- # [21:02] <@firebot> Bug 725432 was not found.
- # [21:04] * WeirdAl waits
- # [21:05] <WeirdAl> davidb: on your earlier question about a debugger: yes, I can
- # [21:05] <@davidb> WeirdAl: great, would be good to know the stack
- # [21:06] <WeirdAl> dammit
- # [21:06] <WeirdAl> this thing's talking to my stock FF
- # [21:06] <WeirdAl> stand by one, I have to disconnect
- # [21:06] * Quits: WeirdAl (chatzilla@moz-D461843.ask.info) (Quit: ChatZilla 0.9.88 [Firefox 10.0/20120129021758])
- # [21:07] <jimm> rats, I don't have ms's latest symbols loaded.
- # [21:07] <jimm> I have call stacks for all three though.
- # [21:07] <jimm> let me restart with new symbols and see if I can get it again, I'll post the stacks to the bug
- # [21:07] <@davidb> excellent!
- # [21:08] * Joins: WeirdAl (chatzilla@moz-D461843.ask.info)
- # [21:08] <WeirdAl> ok, I have FF9 in the debugger with symbols
- # [21:09] <WeirdAl> on the hang
- # [21:10] <@davidb> WeirdAl: jimm is posting to the bug (his stacks) so maybe compare?
- # [21:10] <WeirdAl> uh, I don't know how to get a stack on this one
- # [21:11] <WeirdAl> I can't say for sure I'm at the breakpoint of the exact hang
- # [21:11] <@davidb> WeirdAl: which debugger? you should be able to force a break
- # [21:11] <@davidb> ok
- # [21:11] <WeirdAl> VS 2008 Pro, and all I hit was the pause button
- # [21:11] <@davidb> i don't recall the UI off hand
- # [21:11] <WeirdAl> in the debugger, not the video :)
- # [21:11] <@davidb> lol
- # [21:12] <WeirdAl> it's labeled "break all"
- # [21:12] <@davidb> yeah
- # [21:12] <@davidb> what does that stack look like?
- # [21:12] <WeirdAl> pastebin coming up
- # [21:12] <@davidb> then continue and try again
- # [21:12] <WeirdAl> http://pastebin.mozilla.org/1480015
- # [21:13] <@davidb> yeah this is as i suspected
- # [21:13] <WeirdAl> stack looks the same when the next pause appears (or close enough that I can't spot a difference)
- # [21:13] <WeirdAl> you have a theory? :)
- # [21:14] <@davidb> well, the request is hanging (like was already suggested)
- # [21:16] * Quits: ptheriault (ptheriault@moz-4BE034AB.ptr.us.xo.net) (Quit: ptheriault)
- # [21:16] <@davidb> WeirdAl: it wouldn't hurt to attach your stack to the bug
- # [21:16] <@davidb> describing how you got it
- # [21:16] <jimm> posted
- # [21:17] <@davidb> ok looking
- # [21:17] <@davidb> WeirdAl: nevermind :)
- # [21:17] <WeirdAl> oh, you hit the hang?
- # [21:17] <@davidb> he did
- # [21:18] <WeirdAl> ah
- # [21:18] <jimm> both fx and the test app are bound up in blocking ole loops
- # [21:19] <WeirdAl> :?
- # [21:19] <WeirdAl> OLE?
- # [21:19] <jimm> and the container is trying to call NPN_GetValue_NPNVWindowNPObject, and never gets a response
- # [21:19] <@davidb> pre-COM :)
- # [21:19] <jimm> ole/com, that's out the interprocess stuff communicates
- # [21:19] <jimm> out -> how
- # [21:19] <jimm> for microsoft's accessibility
- # [21:20] <@davidb> jimm: what can we do?
- # [21:20] <jimm> a lot of debugging to try and find the cause, and hopfully a work around? :)
- # [21:21] * @davidb cries a little
- # [21:22] <WeirdAl> well, if it is a blocking loop, might that explain why the plugin container was never killed by FF code?
- # [21:22] <jimm> davidb: during a test like this, work XMSAALibTest be communicating directly with the flash process, or trying to go through fx?
- # [21:22] <jimm> work -> would
- # [21:23] <@davidb> i'm fuzzy on that
- # [21:23] <@davidb> for windowless plugins i think we connect the parent and child
- # [21:23] <@davidb> but for windowed plugins i think it is direct
- # [21:24] <jimm> it's got an IAccessible * that it's invoking - pAcc->accLocation(..)
- # [21:25] <jimm> hmm, best guess -
- # [21:28] <jimm> XMSAALibTest -> IAccessible (blocks) -> routes through fx (during a DispatchMessage win32 call). About the same time flash tries to send an event to fx (GetValue_NPNVWindowNPObject) which also blocks waiting for the reply. Fx can't reply, plugin-container is blocks, and the test app is blocked.
- # [21:28] <@firebot> askalski@mozilla.com requested review from surkov.alexander@gm ail.com for attachment 595486 on bug 706134.
- # [21:28] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=706134 nor, --, ---, askalski, NEW, ARIA listitem shouldn't expose selectable state and pick up aria-selected and aria-checked
- # [21:28] <@davidb> jimm: that's amazing
- # [21:28] <@davidb> please get that in the bug
- # [21:29] <WeirdAl> I think I understand that... :)
- # [21:29] <@askalski> Ok, I have submitted the ARIA patch with all Alex's comments included, I am finishing for today. bye everyone!
- # [21:29] <@davidb> jimm: why does flash send GetValue_NPNVWindowNPObject?
- # [21:29] <@davidb> askalski: splendid! g'night
- # [21:30] <WeirdAl> btw, a blacklist will _not_ help
- # [21:30] <jimm> ah, so I bet that blocked call in fx is trying to send a windows message to the plugin container, but pc is blocked waiting on fx to reply.
- # [21:30] <WeirdAl> blacklist would apply to the extension code, but updater.exe is launched in other ways (in fact, the extension doesn't launch updater at all)
- # [21:30] <@davidb> WeirdAl: ok, I probably misunderstood a comment
- # [21:30] * Quits: @askalski (akuda@moz-4C8A107E.pool85-48-91.dynamic.orange.es) (Quit: Wychodzi)
- # [21:30] <WeirdAl> we do have some other things we can do internally to try and prevent this
- # [21:31] <jimm> davidb: it's running it's script internally, and the flash script wants something flash has to go to the browser for.
- # [21:31] <@davidb> jimm: ok
- # [21:32] <jimm> lemmie write this up more clearly and post to the bug
- # [21:32] <@davidb> jimm: the blocking you've described rings a bell, is there a name for this kind of standoff blocking?
- # [21:32] <@davidb> ah ok yes sounds good
- # [21:33] <Mook_as> fwiw, I'm guessing Ask uses MSAA because that's the closest thing Mozilla supports as a stable ABI.
- # [21:33] <WeirdAl> Mook - I'd have to ask my colleagues sitting 30 feet away :)
- # [21:34] <jimm> davidb: we should ask the Ask folks to confirm this is the same problem they are seeing, since they may not be the same.
- # [21:34] <Mook_as> that's what #extdev promoted for a while, anyway, given that everything else kept breaking.
- # [21:34] <WeirdAl> if you'll give me a few minutes, I can get one of the more relevant engineers on IRC
- # [21:34] <jimm> cool
- # [21:34] <jimm> I'd hate to be posting all of this in the wrong bug!
- # [21:35] * @davidb nods
- # [21:35] <jimm> WeirdAl: you work for Ask?
- # [21:36] <WeirdAl> yes, and my main job is maintenance.
- # [21:37] <WeirdAl> of our code
- # [21:37] * Quits: nhirata (nhirata.bu@moz-BBE3ABD.mv.mozilla.com) (Quit: nhirata)
- # [21:37] <jimm> WeirdAl: ah sweet, that helps!
- # [21:37] * Joins: ptheriault (ptheriault@moz-4BE034AB.ptr.us.xo.net)
- # [21:37] * Joins: levantineb (chatzilla@moz-D461843.ask.info)
- # [21:38] <levantineb> Hi Brian here to talk about accessibility code
- # [21:38] <WeirdAl> Brian's one of our engineers too
- # [21:38] <@davidb> welcome levantineb!
- # [21:38] <levantineb> thanks
- # [21:39] <@davidb> levantineb: jimm wants to make sure he is seeing the same problem
- # [21:39] <@davidb> he posted stacks
- # [21:39] <@davidb> https://bugzilla.mozilla.org/attachment.cgi?id=595482
- # [21:39] * @davidb steps aside
- # [21:40] <jimm> we would like to try and confirm the hang we've reproduced with this test app is the same you all are seeing with your Ask software.
- # [21:40] <levantineb> Got it. Checking...
- # [21:42] <levantineb> Similar. we're actually calling to get the object name before we try for location.
- # [21:43] <jimm> ok, the actual IAccessible call isn't really important
- # [21:43] <jimm> is ask bound up in the same ole code?
- # [21:43] <levantineb> Agreed.
- # [21:44] <levantineb> I can't tell what code FF is in.
- # [21:44] <levantineb> The call is in the call HRESULT hr = pChild->get_accName(varChild, &bstrValue);
- # [21:45] <jimm> hmm, so you can attach to both firefox.exe and our plugin-container.exe but first you would need to configure access to our symbol server.
- # [21:45] <WeirdAl> I have the symbols build, Brian doesn't. I'm at the breakpoint I mentioned earlier, http://pastebin.mozilla.org/1480015
- # [21:45] <jimm> are you in visual studio or windbg?
- # [21:45] <WeirdAl> oh, but I'm using that MSAA tool
- # [21:45] * Joins: nhirata (nhirata.bu@moz-BBE3ABD.mv.mozilla.com)
- # [21:46] <jimm> mozilla symbol server setup FAQ: https://developer.mozilla.org/en/Using_the_Mozilla_symbol_server
- # [21:46] <jimm> WeirdAl: looking..
- # [21:47] <jimm> yep, same here for the fx process
- # [21:48] <WeirdAl> jimm: again, that stack wasn't with our app running, just MSAA, but Brian believes that it fundamentally is the same bug
- # [21:48] <jimm> ok
- # [21:49] <levantineb> Both client programs are traversing the MSAA tree when the hang occurs.
- # [21:55] <jimm> levantineb: curious, which client problems are you referring to?
- # [21:55] <jimm> problems -> programs
- # [21:56] <levantineb> our own app and MSAALibTest
- # [21:56] <jimm> ah, ok
- # [21:56] <WeirdAl> I need to step out for lunch for about fifteen minutes... back asap
- # [21:57] <jimm> so MSAALibTest is querying. what is at the top of your program's stack?
- # [21:57] <jimm> levantineb: ^
- # [21:58] <levantineb> In both cases (us and MSAALibTest) we're nested down about 15 or 16 levels - enumerating children.
- # [21:58] <jimm> levantineb: what's the last call on your stack? is it in Win system code, or in your code doing something else?
- # [21:59] <jimm> (can you post a stack to to pastebin maybe?)
- # [21:59] <jimm> http://pastebin.mozilla.org/
- # [22:02] <levantineb> It'll take me awhile to do this. MSAA crashes because it always traverses the entire tree. Our code is conditional so I'll need to set myself up to make the hang happen. It may take a few minutes.
- # [22:03] <levantineb> I can tell you this, it will hang in a COM call on the IAccessible object.
- # [22:09] * Quits: levantineb (chatzilla@moz-D461843.ask.info) (Ping timeout)
- # [22:09] * Joins: levantineb (chatzilla@moz-D461843.ask.info)
- # [22:10] <levantineb> stack trace is posted
- # [22:11] <levantineb> http://pastebin.mozilla.org/1480065
- # [22:13] * @davidb nudges jimm
- # [22:13] * @davidb knows he's busy though
- # [22:14] * Quits: levantineb (chatzilla@moz-D461843.ask.info) (Ping timeout)
- # [22:14] * Joins: levantineb (chatzilla@moz-D461843.ask.info)
- # [22:18] <WeirdAl> davidb: given that this hang did not happen in FF8, would an hg bisect session be useful?
- # [22:19] <jimm> sorry, talos regression on inblound I had to deal with
- # [22:19] <@davidb> jimm: yeah saw that
- # [22:20] <jimm> WeirdAl: really? yes definitely.
- # [22:21] <jimm> that would rule a lot of stuff out if it were true
- # [22:21] * WeirdAl sighs - it takes a loooooong time to build FF
- # [22:21] <jimm> I do it every day, countless times!
- # [22:21] <jimm> :)
- # [22:21] <@davidb> max your hardware
- # [22:21] <WeirdAl> building normally isn't part of my job, so I don't have one of those fancy SSD's ;)
- # [22:22] <jimm> WeirdAl: maybe just use nightlies to nail down a particular patch range?
- # [22:22] <@davidb> good idea
- # [22:22] <jimm> I'd test a few times on ff8 though, I have a feeling you'll get it there too.
- # [22:22] <WeirdAl> that might be better, if I can keep the installation from taking over my settings and autoupdating to FF10
- # [22:22] <jimm> WeirdAl: just use the zips
- # [22:22] <WeirdAl> zip's don't auto-update?
- # [22:23] <jimm> never had that problem while regression testing
- # [22:24] <WeirdAl> nuts, my local FF8 build didn't have a11y on either
- # [22:34] <WeirdAl> hm, I can't find zips of FF8 release... looks like when it goes to beta, they turn off zip building
- # [22:38] * WeirdAl kicks off his FF8 acc build
- # [22:40] * Quits: peteb-away (ptbrunet@moz-E9B02845.austin.res.rr.com) (Ping timeout)
- # [22:41] <@davidb> i have to go AFK for 3.5 hours.
- # [22:41] * Quits: @davidb (davidb@F2D29657.F60B0462.67AC9B1.IP) (Quit: davidb)
- # [22:44] * Quits: levantineb (chatzilla@moz-D461843.ask.info) (Quit: ChatZilla 0.9.88 [Firefox 10.0/20120123235200])
- # [22:49] * Parts: clown (clown@67828CC7.C1A51174.9D42CF23.IP)
- # [23:02] * Joins: peteb-away (ptbrunet@moz-E9B02845.austin.res.rr.com)
- # [23:05] * Quits: @jprmc (jprmc@F2D29657.F60B0462.67AC9B1.IP) (Ping timeout)
- # [23:07] * Quits: peteb-away (ptbrunet@moz-E9B02845.austin.res.rr.com) (Ping timeout)
- # [23:14] * Joins: peteb-away (ptbrunet@moz-E9B02845.austin.res.rr.com)
- # [23:15] * Quits: WeirdAl (chatzilla@moz-D461843.ask.info) (Quit: ChatZilla 0.9.88 [Firefox 10.0/20120129021758])
- # [23:18] * Joins: WeirdAl (chatzilla@moz-D461843.ask.info)
- # [23:31] * Joins: drexler (chatzilla@moz-2C2B7D1F.hsd1.vt.comcast.net)
- # [23:37] * Joins: davidb (davidb@moz-A4A01B28.eng.wind.ca)
- # [23:37] * ChanServ sets mode: +qo davidb davidb
- # [23:38] * Quits: aaronlev (chatzilla@moz-CDA191A6.c3-0.arl-ubr1.sbo-arl.ma.cable.rcn.com) (Ping timeout)
- # [23:42] <jimm> WeirdAl: so this is related to our ipc mechanism. we'll have to search for a work around of some sort.
- # [23:43] <WeirdAl> yeah. Is it worth me bisecting, or no?
- # [23:44] <WeirdAl> also, do you believe Mozilla has enough to reproduce this bug without Ask involved?
- # [23:44] <jimm> bisecting will likely take you back to the 3.7 release
- # [23:44] <WeirdAl> Uh, I don't remember there being a 3.7 release
- # [23:44] <jimm> yeah, we have that test app and a good understanding of what's going on.
- # [23:45] <jimm> it was a major minor update to the previous release
- # [23:45] <jimm> :)
- # [23:46] <jimm> http://www.mozilla.org/en-US/firefox/3.6.3plugin1/releasenotes/
- # [23:48] * Quits: hub (hub@21B7B9F2.B87E9213.6E712CE2.IP) (Ping timeout)
- # [23:50] * Joins: Jamie (jamie@moz-CA26021.jantrid.net)
- # Session Close: Thu Feb 09 00:00:00 2012
The end :)