Review cursor does not follow properly browse mode cursor in Firefox

[burmancomp]

Steps to reproduce:

When caret moves review cursor and browse mode with screen layout is in use:

• move around page and use up/down arrow keys in Firefox
• use nvda+shift+. (or corresponding desktop layout gesture) to read current review cursor line.

Actual behavior:

At some stage nvda+shift+. and nvda+l gives different result.

Expected behavior:

Both should give same result all the time.

NVDA logs, crash dumps and other attachments:

System configuration

NVDA installed/portable/running from source:

installed

NVDA version:

alpha-34520,d7dc785b (2025.1.0.34520)

Windows version:

w11 23h2

Name and version of other software in use when reproducing the issue:

firefox 132.0.1.

Other information about your system:

Other questions

Does the issue still occur after restarting your computer?

yes

Have you tried any other versions of NVDA? If so, please report their behaviors.

If NVDA add-ons are disabled, is your problem still occurring?

yes

Does the issue still occur after you run the COM Registration Fixing Tool in NVDA's tools menu?

[Adriani90]

Review cursor can follow system focus, system caret and/or mouse cursor, but there is no concept implemented yet where review follows virtual cursor.
Or am I weong? Cc: @jcsteh

[burmancomp]

I think it does it (review cursor follows correctly virtual cursor) most of time (I cannot check just now). However it should do that because otherwise braille could not be tethered to review (due to unreliability) when using browse mode.

[jcsteh]

The review cursor should always follow the browse mode cursor. This does seem to be the case if you move the browse mode cursor and then read the current line at the review cursor, for example. So if braille isn't following even though the review cursor is moving, that is definitely a bug, albeit one I can't really fathom.

[burmancomp]

Thanks @jcsteh for your comment. Now I think I understand this issue correctly so I rewrote bug title and description because it is not braille issue. Braille works correctly but review cursor does not always follow browse mode cursor properly which cause braille and speech to announce occasionally wrong information.

[burmancomp]

I have noticed this when using screen layout (referring to browse mode setting). I do not know if it occurs when screen layout is disabled.

[seanbudd]

Can you please also update the issue description and add:

• more detailed and clear steps to reproduce
• a sample web page
• a debug NVDA log file

[burmancomp]

This page should be suitable for detecting this issue. I think that buttons and other elements which are not plain text may be helpful when trying to get this issue occur. It may take some time to browse (move up and down with arrows) to get this happen.

Here are essential log lines (I have removed such ones I suppose are not important):

Input: kb(laptop):upArrow
IO - speech.speech.speak (10:57:40.761) - MainThread (9724):
Speaking ['Labels ']
IO - braille.BrailleBuffer.update (10:57:40.791) - MainThread (9724):
Braille regions text: ['Labels ']
DEBUG - speech.manager.SpeechManager._handleIndex (10:57:40.802) - MainThread (9724):
Unknown index 258, speech probably cancelled from main thread.
IO - braille.BrailleBuffer.update (10:57:40.921) - MainThread (9724):
Braille regions text: ['btn Unsubscribe ']
IO - braille.BrailleHandler.update (10:57:40.921) - MainThread (9724):
Input: kb(laptop):NVDA+l
IO - speech.speech.speak (10:57:46.440) - MainThread (9724):
Speaking ['Labels ']
Input: kb(laptop):NVDA+shift+.
IO - speech.speech.speak (10:57:53.254) - MainThread (9724):
Speaking ['button', 'Unsubscribe']

[burmancomp]

I updated issue to refer to firefox. If you can get this with other browsers, leave a comment.