/irc-logs / mozilla / #accessibility / 2015-08-10 / end
Options:
Previous day, Next day
- # Session Start: Mon Aug 10 00:00:00 2015
- # Session Ident: #accessibility
- # [00:26] <@firebot> jamie@nvaccess.org cancelled needinfo?(jamie@nvaccess.org) on bug 1192108.
- # [00:27] <@firebot> https://bugzil.la/1192108 — NEW, nobody@mozilla.org — value of aria-labelledby not honored when element is focused
- # [00:59] * Quits: Justin_o (uid14648@moz-c6vipl.ealing.irccloud.com) (Quit: Connection closed for inactivity)
- # [01:07] * Joins: smaug (chatzilla@moz-qsiri0.elisa-mobile.fi)
- # [04:58] * Quits: satdav (uid15780@moz-sljh5r.tooting.irccloud.com) (Quit: Connection closed for inactivity)
- # [07:44] * Joins: ioanachiorean (ioanachiore@moz-q23.52n.2.5.IP)
- # [07:46] * Quits: MrMazda (fmcz@moz-ipbm7f.cable.mindspring.com) (Ping timeout: 121 seconds)
- # [07:47] * Quits: a-865 (fmcz@moz-ipbm7f.cable.mindspring.com) (Ping timeout: 121 seconds)
- # [08:46] * Joins: a-865 (fmcz@moz-ipbm7f.cable.mindspring.com)
- # [10:12] * Joins: Gijs (chatzilla@moz-v0lust.cable.virginm.net)
- # [10:41] <Gijs> MarcoZ: for your last comments in bug 1190366 - I don't suppose you have time to use mozregression (maybe with --inbound-branch fx-team) to narrow down exactly what caused the recent regression and if it is/isn't this bug?
- # [10:41] <@firebot> https://bugzil.la/1190366 — FIXED, adw@mozilla.com — Accessibility label for Search Engine AutoComplete item is wrong
- # [10:42] <&MarcoZ> Gijs: I could try. I've never used MozRegression, or if I tried, it had caused me problems and I hadn't bothered investigating. But I could certainly learn this tool and try if it helps.
- # [10:43] <Gijs> MarcoZ: I find it super useful, personally. I use the commandline version; don't know their GUI thing and if it's accessible or not (sorry) - but the cmdline version should be usable, I think...
- # [10:43] <&MarcoZ> Gijs: Where do I find info about it again? Need to install first.
- # [10:43] <Gijs> MarcoZ: http://mozilla.github.io/mozregression/
- # [10:48] <&MarcoZ> Gijs: Umm, yeah, typical Mozilla, it talks about what this thing does, and what command line options I can use, but not how I install the bloody thing.
- # [10:49] <&MarcoZ> I am on Windows, and MozillaBuild doesn't include it by default.
- # [10:49] <&MarcoZ> Why do we colleagues always make it so hard for each other?
- # [10:50] <&MarcoZ> Gijs: Not blaming you, but those who came up with this page...
- # [10:52] <Gijs> MarcoZ: err, doesn't it anymore? It used to.
- # [10:52] * Gijs sighs
- # [10:52] <Gijs> MarcoZ: easy_install mozregression
- # [10:52] <Gijs> should work
- # [10:52] <Gijs> even on Windows
- # [10:52] <Gijs> http://mozilla.github.io/mozregression/install.html
- # [10:52] <Gijs> MarcoZ: although that says to install pip and then install mozregression with that
- # [10:53] <Gijs> MarcoZ: that page is linked from the header, but I don't know if their page semantics are bad
- # [10:53] <Gijs> I can file a bug for that if necessary
- # [10:53] <Gijs> oh goodness
- # [10:54] <Gijs> <div id="site-title">mozregression</div>
- # [10:54] * Gijs bangs head into desk
- # [10:54] <&MarcoZ> Gijs: Heh, this is also typical Mozilla. Quickly hacked together web pages that don't give a shit about semantics. ;)
- # [10:55] <&MarcoZ> Anyway the command you gave me seems to work. I'll try it now.
- # [10:55] <Gijs> :(
- # [11:07] <&MarcoZ> Gijs: This is what it spits out: https://hg.mozilla.org/integration/fx-team/pushloghtml?fromchange=a4baa2a12eef&tochange=5b3ba9f596e3
- # [11:08] <Gijs> that's very surprising.
- # [11:08] <Gijs> MarcoZ: how did you start it?
- # [11:08] <&MarcoZ> It did, owever, tell me that it didn't find any inbound builds to bisect from. So the MozRegression 0.40 seems to still not be able to find the builds from the new locations.
- # [11:08] <Gijs> MarcoZ: but that's an fx-team pushlog...
- # [11:10] <&MarcoZ> Yes, I did mozregression --good 2015-08-06 --bad 2015-08-09 --inbound-branch fx-team
- # [11:11] <&MarcoZ> Gijs: That's what you suggested I use as the inbound branch.
- # [11:21] <Gijs> yes
- # [11:21] <Gijs> MarcoZ: did you have any "good" candidates ?
- # [11:22] <Gijs> ie was any of the builds you ran not broken
- # [11:31] <&MarcoZ> Gijs: I am pretty sure the 2015-08-06 nightly was still good, so I used that as a starting point.
- # [11:32] <&MarcoZ> Do you suspect a certain patch that might have landed earlier than that to be at fault?
- # [11:32] <&MarcoZ> Gijs: Note that I also had to struggle with the profile constantly getting reset, and me having to disable e10s again for each try because of a11y.
- # [11:33] <&MarcoZ> Gijs: And no, there was no non-broken build in the builds that ran.
- # [11:36] <Gijs> MarcoZ: sorry, multitasking. So...
- # [11:36] <Gijs> MarcoZ: if you had no non-broken builds then this is probably just the first commit in the range you specified
- # [11:37] <Gijs> MarcoZ: if you used --inbound-branch but mozregression still complains it has "no inbound builds to bisect from" after giving you a lot of builds that must be from fx-team, because that's the regression range you got, then something stupid is going wrong
- # [11:38] <Gijs> MarcoZ: regarding the profile thing... I should have thought about this before, sorry. You can use --profile to pass it the path to a profile, and it'll run the builds with that profile
- # [11:38] <Gijs> MarcoZ: because of the first thing I said, I expect the range you got is not correct. If the 06 nightly is good but you only got fx-team builds then maybe it's just an issue with when stuff merged...
- # [11:39] <Gijs> MarcoZ: is this just win8 and whether nvda announces suggestions from the awesomebar?
- # [11:39] <Gijs> MarcoZ: if you like I can try getting a range myself
- # [11:39] <Gijs> MarcoZ: or I can help you fight mozregression, but I dunno how interested you are :)
- # [11:40] <Gijs> MarcoZ: to be clear, when I say "something stupid is going wrong", I mean mozregression must have a bug of some sort, though I don't know exactly *what* is going wrong without seeing the build output
- # [11:50] <&MarcoZ> Gijs: It's just the awesome bar acting up. All the rest is fine.
- # [11:51] <&MarcoZ> Gijs: Perhaps it would be a good idea if you did this, since you can visually verify at the same time if something acts up or not.
- # [11:52] <&MarcoZ> And since you know how to use NVDA... ;)
- # [11:53] <&MarcoZ> Gijs: The funny thing with MozRegression was that it told me it was starting out with nightly builds first, and the URLs said mozilla-central. Then, when it had narrowed down to the Nightly range, it said it would now try to fetch inbound builds, starting 5 days earlier to make sure the range wasn't missed. And then it said "oops, no more Inbound builds
- # [11:53] <&MarcoZ> found", and gave me the range I pasted you.
- # [11:53] <Gijs> MarcoZ: most mysterious. :s
- # [11:53] <Gijs> MarcoZ: I will have a look in a bit. Ping me if the bug isn't updated in 2 hours or so
- # [11:55] <&MarcoZ> Will do! Just make sure you don't forget to have lunch some time. ;)
- # [12:03] * Quits: smaug (chatzilla@moz-qsiri0.elisa-mobile.fi) (Ping timeout: 121 seconds)
- # [12:19] <Gijs> MarcoZ: remind me if you can use etherpad?
- # [12:19] <Gijs> MarcoZ: or is that still completely inaccessible?
- # [12:19] * Gijs seems to recall it didn't work well
- # [12:20] <Gijs> and yes, I'll not forget lunch :)
- # [12:20] <&MarcoZ> Gijs: Dunno if it is accessible yet. The original etherpad.org now has some better accessibility, but I am not sure if our instance ever caught up.
- # [12:23] <Gijs> MarcoZ: OK. You have email (about something completely different)
- # [12:35] * Joins: smaug (chatzilla@moz-7dv8tf.bb.dnainternet.fi)
- # [13:13] <Gijs> MarcoZ: hrm, I'm testing current Nightly and I think I need more details. If I type in the urlbar, and then use the arrow keys to navigate to a search or tab switch icon, it seems to now correctly read things and no longer reads the moz-action URL
- # [13:13] <Gijs> s/icon/item/
- # [13:13] <Gijs> (this is with NVDA on)
- # [13:14] <Gijs> MarcoZ: comparing with devedition... maybe I need to manually get a previous nightly?
- # [13:54] <&MarcoZ> Gijs: Hm, let me check something else. I *may* have caused a false alarm, then. Because all I hear is silence when I arrow up and down once items appear in that list. :(
- # [13:55] <&MarcoZ> But I also got an NVDA update recently, and that introduced some changes to the way Mozilla applications are handled, so it *might* be a side effect of that. Let me check.
- # [13:55] <Gijs> MarcoZ: I got an NVDA update as well, but I think I'm running "normal" release, and I'm guessing you run something more bleeding-edge?
- # [13:56] <Gijs> NVDA says I'm running 2015.2
- # [13:57] <&MarcoZ> Gijs: Yes, you are on the release channel, I am on semi-nightly.
- # [14:05] <&MarcoZ> Gijs: Nope, seems to be us. I just ran 2015.2, too, and am seeing the problem. What OS are you on? I'm on Win 10 release.
- # [14:05] <Gijs> MarcoZ: win8.1
- # [14:06] <Gijs> MarcoZ: I can test win10 though, but I'll need to install NVDA...
- # [14:06] <Gijs> MarcoZ: is it easy for you to make a video (with the NVDA audio) of what you're doing + hearing ?
- # [14:06] <Gijs> s/video/screencast/
- # [14:06] <&MarcoZ> Gijs: Hm, don't have that here. Let me see if I can get a few builds from last week and run on my regular profile. I *think* the 2015-08-06 build was still OK. At least that's what I was using all Friday.
- # [14:07] <&MarcoZ> Gijs: I don't even know how to do a screen cast on Windows. :) I know how to do one on Mac using QuickTime, but for Windows, there's nothing that's accessible, I believe. Or at least I don't know of anything.
- # [14:10] <Gijs> MarcoZ: I only know of jing and that's flash-based, I think, so probably horrible. :\
- # [14:18] * Quits: logbot (logbot@moz-4j4idd.glob.com.au) (Ping timeout: 121 seconds)
- # [14:22] * Joins: logbot (logbot@moz-4j4idd.glob.com.au)
- # [14:26] * Joins: alex_mayorga (uid4422@moz-lffp2u.ealing.irccloud.com)
- # [14:29] <&MarcoZ> Gijs: Manually investigating. Confirmed that the 2015-08-06 build was still good.
- # [14:41] <&MarcoZ> Gijs: OK, this is getting weird. A nightly build from yesterday I downloaded and run in a separate folder with the same profile does not exhibit the problem for me. The new labels are correct there.
- # [14:41] <&MarcoZ> My regular installed nightly, updated via regular software update, seems to be borked.
- # [14:41] <Gijs> MarcoZ: is there some kind of app-specific rule in NVDA that might depend on the path of the app you're running?
- # [14:42] <Gijs> MarcoZ: alternatively, you could do a diff of the app folder for your broken Firefox and the non-broken one and see if there are files that differ (assuming they should otherwise be the same nightly build) ?
- # [14:43] <&MarcoZ> Gijs: No. NVDA doesn't work by such rules at all.
- # [14:43] <&MarcoZ> Gijs: OK, and now my regular nightly has repaired itself, too. WTF?
- # [14:43] <Gijs> MarcoZ: :s
- # [14:43] <Gijs> yeah, err, I have no idea anymore
- # [14:44] <&MarcoZ> OK, sorry for all that noise, then! Don't know what that was! Commenting in the bug now.
- # [14:52] * Joins: anvk (anovak@moz-871.mvk.11.204.IP)
- # [15:24] * Joins: yzen (yzen@moz-ql1r6k.cpe.pppoe.ca)
- # [15:24] * ChanServ sets mode: +o yzen
- # [16:53] * Joins: davidb (davidb@moz-i5m.05u.207.66.IP)
- # [16:53] * ChanServ sets mode: +qo davidb davidb
- # [16:53] <~davidb> heyo
- # [16:53] <~davidb> MarcoZ, shall we?
- # [16:53] <~davidb> tbsaunde, I'll be a bit late for our 1:1
- # [16:54] <~davidb> had dentist torture this morning
- # [16:57] <&MarcoZ> davidb: Yes, Sir! :D
- # [16:57] <~davidb> great
- # [16:59] * Quits: ioanachiorean (ioanachiore@moz-q23.52n.2.5.IP) (Ping timeout: 121 seconds)
- # [17:06] <@tbsaunde> davidb: ack
- # [17:17] * wlach|afk is now known as wlach
- # [17:45] <~davidb> tbsaunde, just finished up, want to meet after lunch?
- # [17:47] <@tbsaunde> davidb: up to you
- # [17:53] * Joins: kkus (kkus@moz-oqbbml.fios.verizon.net)
- # [18:17] * Quits: ~davidb (davidb@moz-i5m.05u.207.66.IP) (Connection closed)
- # [18:33] * Joins: davidb (davidb@moz-i5m.05u.207.66.IP)
- # [18:33] * ChanServ sets mode: +qo davidb davidb
- # [18:40] * Joins: MrMazda (fmcz@moz-ipbm7f.cable.mindspring.com)
- # [18:40] <@tbsaunde> davidb: sup?
- # [18:41] <~davidb> tbsaunde, hey let's meet
- # [19:05] * Quits: @yzen (yzen@moz-ql1r6k.cpe.pppoe.ca) (Connection closed)
- # [19:05] * Joins: yzen (yzen@moz-ql1r6k.cpe.pppoe.ca)
- # [19:05] * ChanServ sets mode: +o yzen
- # [19:28] * Joins: satdav (uid15780@moz-sljh5r.tooting.irccloud.com)
- # [19:46] * Joins: surkov (surkov@moz-i5m.05u.207.66.IP)
- # [19:46] * ChanServ sets mode: +o surkov
- # [19:50] * Quits: ~davidb (davidb@moz-i5m.05u.207.66.IP) (Connection closed)
- # [19:53] * Joins: davidb (davidb@moz-i5m.05u.207.66.IP)
- # [19:53] * ChanServ sets mode: +qo davidb davidb
- # [20:05] * Joins: richardschwerdtfeger (richardschw@moz-4r0nc6.austtx.sbcglobal.net)
- # [20:20] * Quits: kkus (kkus@moz-oqbbml.fios.verizon.net) (Connection closed)
- # [20:46] * davidb is now known as davidb|afk
- # [20:47] * davidb|afk is now known as davidb
- # [20:48] <@tbsaunde> smaug: hey, have you looked at what's going on with bug 1172516 and why it causes marrionette-e10s to crash?
- # [20:48] <@firebot> https://bugzil.la/1172516 — NEW, tbsaunde+mozbugs@tbsaunde.org — fix firing caret move events for proxied accessibles
- # [20:49] <smaug> tbsaunde: sorry, I haven't
- # [20:50] <@tbsaunde> smaug: ok
- # [20:54] * Quits: ~davidb (davidb@moz-i5m.05u.207.66.IP) (Quit: Blah blah blah)
- # [20:55] * Joins: davidb (davidb@moz-i5m.05u.207.66.IP)
- # [20:55] * ChanServ sets mode: +qo davidb davidb
- # [20:55] * Quits: richardschwerdtfeger (richardschw@moz-4r0nc6.austtx.sbcglobal.net) (Client exited)
- # [20:56] <smaug> tbsaunde: based on code I don't understand why the change causes crash
- # [20:57] <@tbsaunde> smaug: welcome to the club
- # [20:58] <smaug> tbsaunde: does the crash happen reliably on try?
- # [20:58] <@tbsaunde> smaug: I believe so, but I haven't tried in a while, maybe it magically works now?
- # [20:59] <smaug> tbsaunde: let me push to try
- # [21:00] <@tbsaunde> smaug: yeah, thanks!
- # [21:00] <smaug> hmm, the patch doesn't apply to trunk
- # [21:02] <@tbsaunde> smaug: hm wonder why
- # [21:03] <smaug> ah, some indentation changes
- # [21:04] * Joins: richardschwerdtfeger (richardschw@moz-4r0nc6.austtx.sbcglobal.net)
- # [21:05] <@tbsaunde> smaug: yeah definitely broken in the patch, seems busted in my tree too though
- # [21:06] * Joins: fredw (Icedove@moz-m7n61v.rev.sfr.net)
- # [21:16] * Quits: richardschwerdtfeger (richardschw@moz-4r0nc6.austtx.sbcglobal.net) (Client exited)
- # [21:33] * Joins: richardschwerdtfeger (richardschw@moz-4r0nc6.austtx.sbcglobal.net)
- # [21:52] * Quits: ~davidb (davidb@moz-i5m.05u.207.66.IP) (Quit: Blah blah blah)
- # [21:59] * Quits: smaug (chatzilla@moz-7dv8tf.bb.dnainternet.fi) (Ping timeout: 121 seconds)
- # [22:20] * Quits: richardschwerdtfeger (richardschw@moz-4r0nc6.austtx.sbcglobal.net) (Client exited)
- # [22:23] * Joins: smaug (chatzilla@moz-3i8kib.elisa-mobile.fi)
- # [22:25] * Joins: richardschwerdtfeger (richardschw@moz-4r0nc6.austtx.sbcglobal.net)
- # [22:27] * Quits: smaug (chatzilla@moz-3i8kib.elisa-mobile.fi) (Ping timeout: 121 seconds)
- # [22:30] * Quits: @surkov (surkov@moz-i5m.05u.207.66.IP) (Client exited)
- # [22:34] * Quits: fredw (Icedove@moz-m7n61v.rev.sfr.net) (Client exited)
- # [22:41] * Joins: smaug (chatzilla@moz-37gn5j.elisa-mobile.fi)
- # [22:46] * Quits: smaug (chatzilla@moz-37gn5j.elisa-mobile.fi) (Ping timeout: 121 seconds)
- # [22:51] * Quits: anvk (anovak@moz-871.mvk.11.204.IP) (Quit: Leaving.)
- # [23:01] * Joins: smaug (chatzilla@moz-06jfaj.elisa-mobile.fi)
- # [23:02] * Joins: davidb (davidb@moz-jhmhp6.dsl.bell.ca)
- # [23:02] * ChanServ sets mode: +qo davidb davidb
- # [23:05] * Quits: smaug (chatzilla@moz-06jfaj.elisa-mobile.fi) (Ping timeout: 121 seconds)
- # [23:07] * Quits: ~davidb (davidb@moz-jhmhp6.dsl.bell.ca) (Quit: Blah blah blah)
- # [23:09] * yzen is now known as yzen_
- # [23:14] * Quits: @yzen_ (yzen@moz-ql1r6k.cpe.pppoe.ca) (Ping timeout: 121 seconds)
- # [23:19] * Joins: smaug (chatzilla@moz-lsdt2i.elisa-mobile.fi)
- # [23:23] * Quits: smaug (chatzilla@moz-lsdt2i.elisa-mobile.fi) (Ping timeout: 121 seconds)
- # [23:30] * Quits: richardschwerdtfeger (richardschw@moz-4r0nc6.austtx.sbcglobal.net) (Client exited)
- # [23:37] * Joins: smaug (chatzilla@moz-um2k5q.elisa-mobile.fi)
- # [23:42] * Quits: smaug (chatzilla@moz-um2k5q.elisa-mobile.fi) (Ping timeout: 121 seconds)
- # [23:49] * Joins: richardschwerdtfeger (richardschw@moz-4r0nc6.austtx.sbcglobal.net)
- # Session Close: Tue Aug 11 00:00:00 2015
Previous day, Next day
Think these logs are useful? Then please donate to show your gratitude (and keep them up, of course). Thanks! — Krijn