Skip to content

beta to master#20209

Merged
seanbudd merged 3 commits into
masterfrom
beta
May 25, 2026
Merged

beta to master#20209
seanbudd merged 3 commits into
masterfrom
beta

Conversation

@seanbudd

Copy link
Copy Markdown
Member

No description provided.

LeonarddeR and others added 3 commits May 25, 2026 09:59
…on (#20203)

See actions/runner-images#14017
Summary of the issue:

GitHub Actions will migrate the windows-2025 (and windows-latest) image label to default to Visual Studio 2026 instead of VS2022. The rollout begins June 8, 2026 and completes June 15, 2026 (actions/runner-images#14017). Any workflow still using windows-2025 after that date will silently switch to VS2026 without prior testing.
Description of user facing changes:

None.
Description of developer facing changes:

NVDA is now built with Visual Studio 2026. All CI workflows use the windows-2025-vs2026 runner image label instead of windows-2025, opting in ahead of GitHub's forced migration (June 8–15, 2026) to catch any VS2026 incompatibilities early.

Note: this PR can be reverted once GitHub completes the migration (expected June 15, 2026), at which point windows-2025 and windows-2025-vs2026 will be equivalent and the explicit label is no longer needed. It is possible that GitHub eventually deprecates this label. ON revert, the changelog should stay as is, though.
Description of development approach:

Runner label changed in 5 workflow files: testAndPublish.yml, codeql.yml, fetch-crowdin-translations.yml, add-new-language.yml, regenerate_english_userDocs_translation_source.yml.
…026.2 as it is not stable enough yet. (#20206)

Fixes #19667
Fixes #19819
Reverts #19672
Summary of the issue:

Experimental code was recently added to NVDA (after 2026.1) that caused the 32 bit synthDriver shim to send its audio back to NvDA for playing, rather than playing it directly in its own process. This was done to test / prepare for the time when add-ons would be running in an app container, and therefore probably would not have direct access to the audio device.
However, currently this code is still quite unstable, causing NVDA to freeze when rapidly changing the speech rate, or arrowing up and down in lists, when using sapi4 or 32 bit sapi5.
This instability is most likely a similar issue to what was fixed in pr #19609, where RPYC is handoing the return of a call on the wrong thread. But in this case, it is calls to/from nvwave in NVDA's process.
Description of user facing changes:
Description of developer facing changes:
Description of development approach:

    Simply instruct the synthDriver process not to install the nvwave proxy that brokers back to NvDA.
    Reverts Re-enable audio ducking for proxied x86 speech synths #19672 so that 32 bit synthDrivers suspend audioDucking while they are running, as audioDucking cnanot be supported without brokering audio.
Copilot AI review requested due to automatic review settings May 25, 2026 02:11
@seanbudd seanbudd requested a review from a team as a code owner May 25, 2026 02:11
@seanbudd seanbudd requested a review from SaschaCowley May 25, 2026 02:11
@seanbudd seanbudd merged commit 1535c3b into master May 25, 2026
13 of 15 checks passed
@github-actions github-actions Bot added this to the 2026.3 milestone May 25, 2026
@seanbudd seanbudd review requested due to automatic review settings May 25, 2026 02:35
@github-actions github-actions Bot requested a deployment to snapshot May 25, 2026 07:23 Abandoned
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants