Add advanced setting to allow UIA in chromium based browsers #11079

LeonarddeR · 2020-05-01T14:03:50Z

Link to issue number:

Related to #10555

Summary of the issue:

UIA is now forcefully disabled in chrome based browsers. While this makes perfect sense, it could be beneficial for developers and/or development purpose to circumvent this without hacking.

Description of how this pull request fixes the issue:

Adds an advanced setting that allows using UIA in Chromium based browsers, particularly in Edge. This option defaults to only when necessary, which is when UIA is exposed by the browser and we are unable to inject. If yes, UIA is used, but still only when UIA is exposed by the browser (not by default in Chrome). If no, there's just no browse mode when we can't inject.
Added NVDAObjects.UIA.web and NVDAObjects.UIA.chromium. The chromium module is meant to contain (Edge) Chromium specific code. The web module is meant to contain code shared between Edge Chromium and Edge Legacy. The Edge module contains Edge Legacy specific code. The code I moved to the web module was moved mainly by guessing what we'd need and what not, so some code in there can still be considered Edge legacy specific. It is very hard to assess this, though. I haven't considered renaming edge to edgeLegacy.

Known issues (Todo)

Update documentation
Headings announcements are duplicated, probably because Edge Chromium both exposes them as control and format field. We'd have to decide what to use and whether the EdgeHeadingQuicknavIterator should be moved to the Web module.
The makeTextInfo implementation of the Edge tree interceptor has been moved to the web tree interceptor, so is also used for Chromium. Without this, I"m getting lots of errors indicating that a range couldn't be cloned. This is either a bug in Edgium, or we need another workaround, as navigation in browse mode is still pretty buggy and bouncy.
It looks like navigation by line is also broken, as I've seen several cases (e.g. at the top of babbage.com) where lines are broken into words.

Testing performed:

Tested in Edge that the config flag works.

Change log entry:

t.b.d.

codeofdusk · 2020-05-01T22:54:17Z

On my machine, NVDA completely freezes with this PR in use:

Download the Appvayor build, and opt to continue running.
Set UIA in Chrome to "yes".
Start chrome (at the run dialog "chrome --enable-experimental-ui-automation").
Tab into the document.

LeonarddeR · 2020-05-02T06:59:28Z

What version of chrome was this? I think it is likely that Microsoft works on the UIA implementation separately from the chromium project and that it's only merged into chromium when a major milestone has been reached. Therefore, we couldn't be sure that the UIA implementation in Edge 81 is the same as in Chrome 81, etc.

codeofdusk · 2020-05-02T07:00:50Z

Running Chrome 83.0.4103.34.

codeofdusk · 2020-05-02T07:16:15Z

I've just installed Chrome Canary (version 84) and retested. I opened Chrome, NVDA hung and I closed it. I get the following in the log:

DEBUGWARNING - _UIAHandler.UIAHandler.IUIAutomationFocusChangedEventHandler_HandleFocusChangedEvent (03:12:56.765) - Dummy-3 (15196):
HandleFocusChangedEvent: Ignoring duplicate focus event
DEBUGWARNING - watchdog._watcher (03:13:07.013) - watchdog (16652):
Trying to recover from freeze, core stack:
  File "nvda.pyw", line 215, in <module>
  File "core.pyc", line 545, in main
  File "wx\core.pyc", line 2134, in MainLoop
  File "gui\__init__.pyc", line 1032, in Notify
  File "core.pyc", line 515, in run
  File "queueHandler.pyc", line 83, in pumpAll
  File "queueHandler.pyc", line 50, in flushQueue
  File "sayAllHandler.pyc", line 172, in nextLine
  File "speech\types.pyc", line 49, in _flattenNestedSequences
  File "speech\types.pyc", line 49, in <genexpr>
  File "speech\__init__.pyc", line 1116, in getTextInfoSpeech
  File "treeInterceptorHandler.pyc", line 231, in getTextWithFields
  File "NVDAObjects\UIA\__init__.pyc", line 716, in getTextWithFields
  File "NVDAObjects\UIA\__init__.pyc", line 686, in _getTextWithFieldsForUIARange
  File "NVDAObjects\UIA\__init__.pyc", line 686, in _getTextWithFieldsForUIARange
  File "NVDAObjects\UIA\__init__.pyc", line 686, in _getTextWithFieldsForUIARange
  [Previous line repeated 263 more times]
  File "NVDAObjects\UIA\__init__.pyc", line 594, in _getTextWithFieldsForUIARange
  File "NVDAObjects\UIA\__init__.pyc", line 428, in _getControlFieldForObject
  File "baseObject.pyc", line 42, in __get__
  File "baseObject.pyc", line 145, in _getPropertyViaCache
  File "NVDAObjects\UIA\__init__.pyc", line 1533, in _get_positionInfo
  File "NVDAObjects\UIA\__init__.pyc", line 811, in _getUIACacheablePropertyValue
  File "comtypesMonkeyPatches.pyc", line 26, in __call__

DEBUGWARNING - queueHandler.flushQueue (03:13:14.991) - MainThread (6988):
Error in func nextLine from eventQueue
Traceback (most recent call last):
  File "queueHandler.pyc", line 50, in flushQueue
  File "sayAllHandler.pyc", line 172, in nextLine
  File "speech\types.pyc", line 49, in _flattenNestedSequences
  File "speech\types.pyc", line 49, in <genexpr>
  File "speech\__init__.pyc", line 1116, in getTextInfoSpeech
  File "treeInterceptorHandler.pyc", line 231, in getTextWithFields
  File "NVDAObjects\UIA\__init__.pyc", line 716, in getTextWithFields
  File "NVDAObjects\UIA\__init__.pyc", line 686, in _getTextWithFieldsForUIARange
  File "NVDAObjects\UIA\__init__.pyc", line 686, in _getTextWithFieldsForUIARange
  File "NVDAObjects\UIA\__init__.pyc", line 686, in _getTextWithFieldsForUIARange
  [Previous line repeated 456 more times]
  File "NVDAObjects\UIA\__init__.pyc", line 704, in _getTextWithFieldsForUIARange
  File "NVDAObjects\UIA\__init__.pyc", line 489, in _getTextWithFields_text
  File "NVDAObjects\UIA\__init__.pyc", line 455, in _getTextFromUIARange
  File "comtypesMonkeyPatches.pyc", line 26, in __call__
_ctypes.COMError: (-2147220991, 'An event was unable to invoke any of the subscribers', (None, None, None, 0, None))
DEBUGWARNING - NVDAObjects.IAccessible.IAccessible._get_IAccessibleRole (03:13:16.347) - MainThread (6988):
accRole failed: (-2147024809, 'The parameter is incorrect.', (None, None, None, 0, None))
DEBUGWARNING - NVDAObjects.UIA.UIATextInfo.__init__ (03:13:16.361) - MainThread (6988):
Could not clone range
Traceback (most recent call last):
  File "NVDAObjects\UIA\__init__.pyc", line 296, in __init__
  File "comtypesMonkeyPatches.pyc", line 26, in __call__
_ctypes.COMError: (-2147220991, 'An event was unable to invoke any of the subscribers', (None, None, None, 0, None))
DEBUGWARNING - NVDAObjects.UIA.web.UIAWebTreeInterceptor.makeTextInfo (03:13:16.361) - MainThread (6988):
selection died. Using first position instead
ERROR - scriptHandler.executeScript (03:13:16.362) - MainThread (6988):
error executing script: <bound method CursorManager.script_moveByLine_back of <NVDAObjects.UIA.chromium.ChromiumUIATreeInterceptor object at 0x093FA950>> with gesture 'up arrow'
Traceback (most recent call last):
  File "NVDAObjects\UIA\__init__.pyc", line 296, in __init__
  File "comtypesMonkeyPatches.pyc", line 26, in __call__
_ctypes.COMError: (-2147220991, 'An event was unable to invoke any of the subscribers', (None, None, None, 0, None))

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "NVDAObjects\UIA\web.pyc", line 392, in makeTextInfo
  File "cursorManager.pyc", line 448, in makeTextInfo
  File "treeInterceptorHandler.pyc", line 179, in copy
  File "NVDAObjects\UIA\__init__.pyc", line 759, in copy
  File "NVDAObjects\UIA\__init__.pyc", line 300, in __init__
RuntimeError: Could not clone range

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "NVDAObjects\UIA\__init__.pyc", line 316, in __init__
  File "comtypes\__init__.pyc", line 279, in __getattr__
  File "comtypesMonkeyPatches.pyc", line 26, in __call__
_ctypes.COMError: (-2147220991, 'An event was unable to invoke any of the subscribers', (None, None, None, 0, None))

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "scriptHandler.pyc", line 205, in executeScript
  File "cursorManager.pyc", line 214, in script_moveByLine_back
  File "cursorManager.pyc", line 114, in _caretMovementScriptHelper
  File "NVDAObjects\UIA\web.pyc", line 399, in makeTextInfo
  File "NVDAObjects\UIA\web.pyc", line 402, in makeTextInfo
  File "NVDAObjects\UIA\web.pyc", line 392, in makeTextInfo
  File "cursorManager.pyc", line 453, in makeTextInfo
  File "documentBase.pyc", line 24, in makeTextInfo
  File "treeInterceptorHandler.pyc", line 173, in __init__
  File "NVDAObjects\UIA\__init__.pyc", line 319, in __init__
RuntimeError
DEBUGWARNING - NVDAObjects.UIA.UIATextInfo.__init__ (03:13:16.372) - MainThread (6988):
Could not clone range
Traceback (most recent call last):
  File "NVDAObjects\UIA\__init__.pyc", line 296, in __init__
  File "comtypesMonkeyPatches.pyc", line 26, in __call__
_ctypes.COMError: (-2147220991, 'An event was unable to invoke any of the subscribers', (None, None, None, 0, None))
DEBUGWARNING - NVDAObjects.UIA.web.UIAWebTreeInterceptor.makeTextInfo (03:13:16.373) - MainThread (6988):
selection died. Using first position instead
ERROR - scriptHandler.executeScript (03:13:16.373) - MainThread (6988):
error executing script: <bound method CursorManager.script_moveByLine_forward of <NVDAObjects.UIA.chromium.ChromiumUIATreeInterceptor object at 0x093FA950>> with gesture 'down arrow'
Traceback (most recent call last):
  File "NVDAObjects\UIA\__init__.pyc", line 296, in __init__
  File "comtypesMonkeyPatches.pyc", line 26, in __call__
_ctypes.COMError: (-2147220991, 'An event was unable to invoke any of the subscribers', (None, None, None, 0, None))

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "NVDAObjects\UIA\web.pyc", line 392, in makeTextInfo
  File "cursorManager.pyc", line 448, in makeTextInfo
  File "treeInterceptorHandler.pyc", line 179, in copy
  File "NVDAObjects\UIA\__init__.pyc", line 759, in copy
  File "NVDAObjects\UIA\__init__.pyc", line 300, in __init__
RuntimeError: Could not clone range

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "NVDAObjects\UIA\__init__.pyc", line 316, in __init__
  File "comtypes\__init__.pyc", line 279, in __getattr__
  File "comtypesMonkeyPatches.pyc", line 26, in __call__
_ctypes.COMError: (-2147220991, 'An event was unable to invoke any of the subscribers', (None, None, None, 0, None))

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "scriptHandler.pyc", line 205, in executeScript
  File "cursorManager.pyc", line 218, in script_moveByLine_forward
  File "cursorManager.pyc", line 114, in _caretMovementScriptHelper
  File "NVDAObjects\UIA\web.pyc", line 399, in makeTextInfo
  File "NVDAObjects\UIA\web.pyc", line 402, in makeTextInfo
  File "NVDAObjects\UIA\web.pyc", line 392, in makeTextInfo
  File "cursorManager.pyc", line 453, in makeTextInfo
  File "documentBase.pyc", line 24, in makeTextInfo
  File "treeInterceptorHandler.pyc", line 173, in __init__
  File "NVDAObjects\UIA\__init__.pyc", line 319, in __init__
RuntimeError
DEBUGWARNING - NVDAObjects.UIA.UIA._prefetchUIACacheForPropertyIDs (03:13:16.395) - MainThread (6988):
IUIAutomationElement.buildUpdatedCache failed given IDs of {30019, 30086, 30022, 30025, 30036, 30070, 30103, 30008, 30009, 30010, 30046, 30079}
DEBUGWARNING - eventHandler.executeEvent (03:13:16.396) - MainThread (6988):
error executing event: typedCharacter on <NVDAObjects.UIA.chromium.ChromiumUIA object at 0x0928BFD0> with extra args of {'ch': '\x17'}
Traceback (most recent call last):
  File "eventHandler.pyc", line 155, in executeEvent
  File "eventHandler.pyc", line 92, in __init__
  File "eventHandler.pyc", line 100, in next
  File "NVDAObjects\__init__.pyc", line 1047, in event_typedCharacter
  File "speech\__init__.pyc", line 966, in speakTypedCharacters
  File "api.pyc", line 264, in isTypingProtected
  File "baseObject.pyc", line 42, in __get__
  File "baseObject.pyc", line 145, in _getPropertyViaCache
  File "NVDAObjects\__init__.pyc", line 921, in _get_isProtected
  File "baseObject.pyc", line 42, in __get__
  File "baseObject.pyc", line 145, in _getPropertyViaCache
  File "NVDAObjects\UIA\web.pyc", line 321, in _get_states
  File "baseObject.pyc", line 42, in __get__
  File "baseObject.pyc", line 145, in _getPropertyViaCache
  File "NVDAObjects\UIA\__init__.pyc", line 1234, in _get_states
  File "NVDAObjects\UIA\__init__.pyc", line 814, in _getUIACacheablePropertyValue
  File "comtypesMonkeyPatches.pyc", line 26, in __call__
_ctypes.COMError: (-2147220991, 'An event was unable to invoke any of the subscribers', (None, None, None, 0, None))

LeonarddeR · 2020-05-02T07:20:25Z

I also installed Chrome Canary. The point is that, when using alpha and force UIA so it will at least work on the address bar, the address bar itself isn't very accessible. Therefore I don't think these UIA implementations are in sync.

michaelDCurran · 2020-05-07T07:11:49Z

NV access is interested in this pr, specifically to support Edge on Windows 10 S with the NVDA store app (which cannot inject in-process). Edge 81 does not have the UIA freezing bugs last time I tested. I think we should only enable UIA for msEdge. Not all of Chromium, if this is possible.

codeofdusk · 2020-05-07T07:14:15Z

I think we should only enable UIA for msEdge.

If the improved UIA implementation makes it to Chrome, wouldn't it significantly improve performance to enable it there as well?

michaelDCurran · 2020-05-07T07:18:52Z

We are still a long way off from UIA surpassing IAccessible2+virtualBuffers on the web in performance and quality. This may be a goal we are heading towards, but right now I feel we should only do it where it is absolutely necessary. Also, This pr should honor both the advanced setting, and also if NvDA can't inject (appModule.helperLocalBindingHandle is 0) similar to how Microsoft Word works.

josephsl · 2020-05-07T07:32:15Z

Hi, provided that Google follows UIA best practices in Chromium/Chrome combo. Thanks.

LeonarddeR · 2020-05-07T07:47:12Z

I'm happy to change this into something that only works in Edge, but it might help to know how Chrome and Edge UIA implementations relate to each other. @michaelDCurran, would this be something Microsoft and/or Google could inform us about?

Note that in Chrome, the UIA support is still considered to be experimental and still needs a command line parameter to work. If they are the same implementations though, I don't see a reason to block UIA from working in Chrome if someone explicitly passes the experimental uiautomation cmd parameter.

michaelDCurran · 2020-05-07T08:23:23Z

I don't believe that Google has any plans to turn on UIA by default any time soon. Therefore, I think I agree it is okay to allow UIA to be used for all Chromium implementations if it is available and either the advanced setting in NVDA is on or NVDA cannot inject in-process. But we would not want to publish the idea of UIA with Chromium too heavily here. UIA right now is not yet at all a suitable replacement where IAccessible2/virtualBuffers is possible. This may very well change in the near future, but even if it did, it would only be useful in the latest builds of Windows 10. But, for Edgium in Windows 10 S, it is necessary that NVDA can use UIA in that scenario.

LeonarddeR · 2020-05-07T09:04:51Z

But we would not want to publish the idea of UIA with Chromium too heavily here.

Agreed. So either the documentation should contain a huge disclaimer, or we should advertise the option as Edge specific, yet being chromium specific under the hood. I'd go for the former. We can reword the option in such a way that it mentions Edge and supported Chromium based browsers, for example.

josephsl · 2020-05-07T09:10:59Z

Hi, maybe the label would say something to the effect of “Use UIA in Chromium-based Microsoft Edge (experimental)”? Thanks.

LeonarddeR · 2020-05-07T13:27:24Z

@michaelDCurran I have updated the initial description to reflect the current state of the pr. While I started this work, I"m afraid I neither have the time nor the skills to work further on this. You really know UIA much better than I do. May be @josephsl might be interested to take a look as well?

Note that the new UIA.web module contains code from the edge module, all moved around, so that makes reviewing this code a bit easier.

josephsl · 2020-05-07T17:28:21Z

Hi, I think it might be best to rename UIA.web to UIA.chromium for sake of consistency with IAccessible.chromium so that EdgeHTML code can be kept for backward compatibility and to support PWA’s. Thanks.

…e them in format fields, as headings were being reported twice.

…completely before the start of the parent textRange. Stops a freeze on some checkboxes in Edgium and is generally safer.

…ctly in browseMode in Edgium.

…s from being used.

LeonarddeR · 2020-11-05T06:49:34Z

@michaelDCurran Thanks for taking this further!
Just to make sure, I assume you are aware that every change done to UIA.web has to be tested in legacy Edge as well, at least on older versions of Windows 10? Even on current versions, the legacy Edge implementation is still used when searching in the start menu, for example.

michaelDCurran · 2020-12-14T01:57:48Z

In order for existing logic around when NVDA should and should not report the accessible name for a control, Chromium / Edge needs to expose aria-label in the UIA element's AriaProperties property. Edge classic did this previously, but Chromium's UIA implementation does not.
See Chromium bug 1147700:
https://bugs.chromium.org/p/chromium/issues/detail?id=1147700

Adriani90 · 2020-12-14T14:12:48Z

cc: @aleventhal, @ObjectInSpace

…, and that the names of controls (that do not use their name as content) are always reported.

michaelDCurran · 2021-01-05T05:27:36Z

@codeofdusk Would you mind testing this branch again, and let me know if you see any freezes. Rememeber to turn on Use UIA in Edge and other Chromium-based brwosers in NvDA's advanced panel.

feerrenrut · 2021-01-05T07:03:22Z

@michaelDCurran Could you double check the PR description to make sure it is still accurate please?

codeofdusk · 2021-01-05T10:13:19Z

Running Chrome with the --enable-experimental-ui-automation flag now uses UIA as expected (confirmed in Python console) and documents are browsable.

Opening this page in IA2 Chrome causes a hang of several seconds. The same page in UIA Chrome loads almost instantly, but attempting to move by heading causes the browser to crash:

NVDA traceback

ERROR - scriptHandler.executeScript (05:10:01.927) - MainThread (15344): error executing script: <bound method CursorManager.script_moveByLine_forward of <NVDAObjects.UIA.chromium.ChromiumUIATreeInterceptor object at 0x072F9630>> with gesture 'down arrow' Traceback (most recent call last): File "NVDAObjects\UIA\__init__.pyc", line 297, in __init__ File "comtypesMonkeyPatches.pyc", line 26, in __call__ _ctypes.COMError: (-2147220991, 'An event was unable to invoke any of the subscribers', (None, None, None, 0, None)) During handling of the above exception, another exception occurred: Traceback (most recent call last): File "NVDAObjects\UIA\web.pyc", line 411, in makeTextInfo File "cursorManager.pyc", line 481, in makeTextInfo File "treeInterceptorHandler.pyc", line 180, in copy File "NVDAObjects\UIA_init_.pyc", line 780, in copy File "NVDAObjects\UIA_init_.pyc", line 301, in init RuntimeError: Could not clone range During handling of the above exception, another exception occurred: Traceback (most recent call last): File "NVDAObjects\UIA_init_.pyc", line 317, in init File "comtypes_init_.pyc", line 279, in getattr File "comtypesMonkeyPatches.pyc", line 26, in call _ctypes.COMError: (-2147220991, 'An event was unable to invoke any of the subscribers', (None, None, None, 0, None)) During handling of the above exception, another exception occurred:

Traceback (most recent call last): File "scriptHandler.pyc", line 208, in executeScript File "cursorManager.pyc", line 255, in script_moveByLine_forward File "cursorManager.pyc", line 129, in caretMovementScriptHelper File "NVDAObjects\UIA\web.pyc", line 418, in makeTextInfo File "NVDAObjects\UIA\web.pyc", line 421, in makeTextInfo File "NVDAObjects\UIA\web.pyc", line 411, in makeTextInfo File "cursorManager.pyc", line 486, in makeTextInfo File "documentBase.pyc", line 24, in makeTextInfo File "treeInterceptorHandler.pyc", line 174, in init File "NVDAObjects\UIA_init.pyc", line 320, in init RuntimeError

feerrenrut · 2021-01-13T09:55:27Z

I've been looking at this, but I'm not sure how to review it effectively. Without tests or further documentation about why the changes are made and the circumstances in which they are required I'd only be rubber stamping it.

I'd like to see a plan for making this area of the code easier to understand. How can we move away from large recursive functions with many pathways?

josephsl · 2021-01-13T14:36:55Z

Hi, one thing to keep in mind is what Narrator is doing these days – accessing Chromium’s IAccessible2 implementation through UIA, and a discussion around this time last year about making NVDA work with Edge 79 and later through IAccessible and a renewed effort to use UIA for Chromium support (required on Windows 10 in S mode). Thanks.

feerrenrut · 2021-01-14T00:31:15Z

One option is to rebase these commits to clean up the history. So that each commit message can explain the need for the change. Move classes / functions around first (without further modification), then further commits to fix up the things that need fixing with full explanations of why it needs to change in the commit message (eg the scenario where there is an issue). This would be with the goal to merge this rather than squash merge.

feerrenrut · 2021-02-02T14:15:13Z

Closing this PR as it is replaced by #12025

LeonarddeR added 3 commits May 1, 2020 15:55

Add config flag

ae097b6

Use the config flag

3b7bb3f

Add experimental, very broken tree interceptor

1724ead

LeonarddeR mentioned this pull request May 1, 2020

Support UI Automation in Microsoft Edge Chromium and Google Chrome #10555

Closed

LeonarddeR added 4 commits May 1, 2020 19:01

Add web module

8600982

Fix linting, for now

dcbe9be

Attempt to get a somewhat working experience

b6aa782

Please the linter

75e7144

nvaccess deleted a comment from AppVeyorBot May 1, 2020

Fix miscdeps

83e65e7

LeonarddeR added 2 commits May 7, 2020 14:12

Merge remote-tracking branch 'origin/master' into enforceIA2

57227de

Rewording and use of local helper binding handle

8501d00

Merge remote-tracking branch 'origin/master' into enforceIA2

4966f2b

michaelDCurran added 4 commits November 5, 2020 09:55

ChromiumUIA: as headings are exposed in the element tree, don't expos…

fa896fa

…e them in format fields, as headings were being reported twice.

UIATextInfo._getTextwithFieldsForUIARange: skip children that appear …

61c73b4

…completely before the start of the parent textRange. Stops a freeze on some checkboxes in Edgium and is generally safer.

UIA: make sure interactive / presentational lists are presented corre…

636f581

…ctly in browseMode in Edgium.

embedded object characters in Edgium should not stop overriding label…

b4a2863

…s from being used.

This comment has been minimized.

Sign in to view

Fix linting issues

af397b1

michaelDCurran added 2 commits November 10, 2020 09:56

Merge branch 'master' into enforceIA2

58d2d9e

Merge branch 'master' into enforceIA2

0234da2

michaelDCurran added 3 commits December 15, 2020 16:54

Chromium with UIA: ignore layout tables when configured to do so.

a91d16f

Chromium with UIA: ensure the value of comboboxes are used as content…

5da2437

…, and that the names of controls (that do not use their name as content) are always reported.

Merge branch 'master' into enforceIA2

3a51721

michaelDCurran requested review from codeofdusk and feerrenrut January 5, 2021 05:24

michaelDCurran marked this pull request as ready for review January 5, 2021 05:25

This comment has been minimized.

Sign in to view

feerrenrut mentioned this pull request Feb 2, 2021

Add support for UIA with chromium based browsers #12025

Merged

feerrenrut closed this Feb 2, 2021

nvaccessAuto added this to the 2021.1 milestone Feb 3, 2021

CyrilleB79 mentioned this pull request Feb 4, 2021

Fixes somes context help issues in advanced settings panel. #12034

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add advanced setting to allow UIA in chromium based browsers #11079

Add advanced setting to allow UIA in chromium based browsers #11079

LeonarddeR commented May 1, 2020 •

edited

Loading

codeofdusk commented May 1, 2020

LeonarddeR commented May 2, 2020

codeofdusk commented May 2, 2020

codeofdusk commented May 2, 2020 •

edited

Loading

LeonarddeR commented May 2, 2020

michaelDCurran commented May 7, 2020

codeofdusk commented May 7, 2020

michaelDCurran commented May 7, 2020 via email

josephsl commented May 7, 2020 via email •

edited by feerrenrut

Loading

LeonarddeR commented May 7, 2020 •

edited

Loading

michaelDCurran commented May 7, 2020 via email

LeonarddeR commented May 7, 2020

josephsl commented May 7, 2020 via email •

edited by feerrenrut

Loading

LeonarddeR commented May 7, 2020

josephsl commented May 7, 2020 via email •

edited by feerrenrut

Loading

This comment has been minimized.

LeonarddeR commented Nov 5, 2020

michaelDCurran commented Dec 14, 2020 •

edited

Loading

Adriani90 commented Dec 14, 2020

michaelDCurran commented Jan 5, 2021

This comment has been minimized.

feerrenrut commented Jan 5, 2021

This comment has been minimized.

codeofdusk commented Jan 5, 2021 •

edited by feerrenrut

Loading

feerrenrut commented Jan 13, 2021

josephsl commented Jan 13, 2021 via email •

edited by feerrenrut

Loading

feerrenrut commented Jan 14, 2021

feerrenrut commented Feb 2, 2021

Add advanced setting to allow UIA in chromium based browsers #11079

Add advanced setting to allow UIA in chromium based browsers #11079

Conversation

LeonarddeR commented May 1, 2020 • edited Loading

Link to issue number:

Summary of the issue:

Description of how this pull request fixes the issue:

Known issues (Todo)

Testing performed:

Change log entry:

codeofdusk commented May 1, 2020

LeonarddeR commented May 2, 2020

codeofdusk commented May 2, 2020

codeofdusk commented May 2, 2020 • edited Loading

LeonarddeR commented May 2, 2020

michaelDCurran commented May 7, 2020

codeofdusk commented May 7, 2020

michaelDCurran commented May 7, 2020 via email

josephsl commented May 7, 2020 via email • edited by feerrenrut Loading

LeonarddeR commented May 7, 2020 • edited Loading

michaelDCurran commented May 7, 2020 via email

LeonarddeR commented May 7, 2020

josephsl commented May 7, 2020 via email • edited by feerrenrut Loading

LeonarddeR commented May 7, 2020

josephsl commented May 7, 2020 via email • edited by feerrenrut Loading

This comment has been minimized.

LeonarddeR commented Nov 5, 2020

michaelDCurran commented Dec 14, 2020 • edited Loading

Adriani90 commented Dec 14, 2020

michaelDCurran commented Jan 5, 2021

This comment has been minimized.

feerrenrut commented Jan 5, 2021

This comment has been minimized.

codeofdusk commented Jan 5, 2021 • edited by feerrenrut Loading

feerrenrut commented Jan 13, 2021

josephsl commented Jan 13, 2021 via email • edited by feerrenrut Loading

feerrenrut commented Jan 14, 2021

feerrenrut commented Feb 2, 2021

LeonarddeR commented May 1, 2020 •

edited

Loading

codeofdusk commented May 2, 2020 •

edited

Loading

josephsl commented May 7, 2020 via email •

edited by feerrenrut

Loading

LeonarddeR commented May 7, 2020 •

edited

Loading

josephsl commented May 7, 2020 via email •

edited by feerrenrut

Loading

josephsl commented May 7, 2020 via email •

edited by feerrenrut

Loading

michaelDCurran commented Dec 14, 2020 •

edited

Loading

codeofdusk commented Jan 5, 2021 •

edited by feerrenrut

Loading

josephsl commented Jan 13, 2021 via email •

edited by feerrenrut

Loading