Issue metadata
Sign in to add a comment
|
Regression: A11y: Live regions are not announced by NVDA |
||||||||||||||||||||||
Issue descriptionNVDA 2017.4 I have bisected this issue and here are the results: You are probably looking for a change made after 511241 (known good), but no later than 511242 (first known bad). CHANGELOG URL: https://chromium.googlesource.com/chromium/src/+log/a95350cd78835c5e831a821594ad6bff19dc71cf..36999b26ddec5abb40845f26072cf08042db71d2
,
Jan 11 2018
Could you run the live region test page and see how many of those regressed, if any? If none of them regressed, we should add that example to our test page. It sounds like it might have something to do with changing the role of an existing element, since that's what Aaron's patch affected.
,
Jan 11 2018
,
Jan 11 2018
Live Region test page seems to pass. There are instances when NVDA repeats the ending chars of an added message (text node set data and text node replace data) NVDA will says 'succeeded.eeded.eeded' when text is only 'succeeded.'
,
Jan 12 2018
,
Jan 17 2018
David, did you get different results with JAWS? And if you left the options alone?
,
Jan 25 2018
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/03a74b82f2a1d041a426d2b13c0a822839ac0794 commit 03a74b82f2a1d041a426d2b13c0a822839ac0794 Author: Aaron Leventhal <aleventhal@chromium.org> Date: Thu Jan 25 05:43:53 2018 Do not change AX tree for attribute set to same value Setting an attribute to the same value it had before should never change the accessibility tree. For example, setting the role attribute to the same value, although not a recommended authoring practice, has negative consequences as the accessible object must be recreated to be safe. In general it is not necessary to update the AX tree unless the value changes. Bug: 801293 Cq-Include-Trybots: master.tryserver.chromium.linux:closure_compilation Change-Id: Iab61b3a9a4c30c5183046c9de073e71ab216d6cb Reviewed-on: https://chromium-review.googlesource.com/876309 Commit-Queue: Aaron Leventhal <aleventhal@chromium.org> Reviewed-by: Mike West <mkwst@chromium.org> Reviewed-by: Dominic Mazzoni <dmazzoni@chromium.org> Cr-Commit-Position: refs/heads/master@{#531821} [modify] https://crrev.com/03a74b82f2a1d041a426d2b13c0a822839ac0794/chrome/browser/resources/chromeos/chromevox/cvox2/background/background_test.extjs [modify] https://crrev.com/03a74b82f2a1d041a426d2b13c0a822839ac0794/third_party/WebKit/Source/core/dom/Element.cpp
,
Jan 25 2018
,
Feb 16 2018
JAWS and NVDA behave the same. The repro steps are fixed, but if I set the example to custom and atomic to false, nothing or short bits of text are announced by both JAWS and NVDA. Feel free to request an additional bug if this issue is not related to the solution.
,
Mar 8 2018
@dsexton, that's a different bug. Note that if you change the custom settings from "text" to "all" you will hear "Added content#X". Then you will hear "Modified" as that text is changed from "Added Content #X" to "Modified Content#X". Chrome thinks that only the first word has changed, even though all of it has been rewritten. I would check with @dmazzoni to see if we have filed that issue yet, because we do know about it. This issue is fixed. |
|||||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||||
Comment 1 by dsexton@chromium.org
, Jan 11 2018