[Accessibility] Significant Screen Reader Lag in Integrated Terminal Output

[aaronr7734]

Type: Bug

Summary

There is a consistent and noticeable delay of approximately half a second (~500ms) between executing a command in the Integrated Terminal and the screen reader announcing the command's output. This creates a significant drag on productivity and disrupts the interactive flow that is expected from a command-line interface.

This issue appears to be independent of the shell, the specific command, the operating system, and the screen reader being used. This suggests the delay likely originates within the terminal component of VS Code itself, rather than from an external tool.

Steps to Reproduce

1. Ensure a screen reader is active (e.g., VoiceOver on macOS, NVDA, JAWS, or Narrator on Windows).
2. Open the Integrated Terminal in VS Code (Terminal: Create New Terminal).
3. Type a simple, fast-executing command (e.g., pwd, ls, dir).
4. Press Enter to execute the command.

Expected Behavior

The command's output is spoken by the screen reader almost instantaneously upon execution, mirroring the responsive behavior of native terminal applications like macOS Terminal.app or Windows Terminal.

Actual Behavior

There is a consistent, perceptible lag (estimated at 0.5 seconds) after pressing Enter before the screen reader begins to announce the output. The command itself executes quickly, but the speech output is delayed.

Additional Context & Troubleshooting

• Cross-Platform: This issue has been personally verified on both macOS (with VoiceOver) and Windows (with NVDA), indicating it is not specific to one OS or screen reader. I've experienced the issue on several different machines as well.
• Remote & Local: The lag is present in both local terminals and when connected to a remote server via the Remote-SSH extension.
• Troubleshooting Performed: I initially suspected the terminal success/failure accessibility signals might be related. I have since disabled all terminal signal announcements, and the lag remains identical. I've also tested this with all extensions disabled and had the same issue.

VS Code version: Code 1.101.0 (dfaf441, 2025-06-11T15:00:50.123Z)
OS version: Windows_NT arm64 10.0.26100
Modes:
Remote OS version: Darwin arm64 24.5.0
Remote OS version: Darwin arm64 24.5.0

System Info

Item	Value
CPUs	Apple Silicon (8 x 3200)
GPU Status	2d_canvas: enabled canvas_oop_rasterization: enabled_on direct_rendering_display_compositor: disabled_off_ok gpu_compositing: enabled multiple_raster_threads: enabled_on opengl: enabled_on rasterization: unavailable_off raw_draw: disabled_off_ok skia_graphite: disabled_off video_decode: enabled video_encode: enabled vulkan: disabled_off webgl: enabled webgl2: enabled webgpu: enabled webnn: disabled_off
Load (avg)	undefined
Memory (System)	19.90GB (10.93GB free)
Process Argv	--crash-reporter-id 837afa97-17a4-47af-9d34-aee526b5f154
Screen Reader	yes
VM	67%

Item	Value
Remote	SSH: fire
OS	Darwin arm64 24.5.0
CPUs	Apple M4 Pro (12 x 2400)
Memory (System)	24.00GB (0.09GB free)
VM	0%

Item	Value
Remote	SSH: fire
OS	Darwin arm64 24.5.0
CPUs	Apple M4 Pro (12 x 2400)
Memory (System)	24.00GB (0.09GB free)
VM	0%

Extensions (29)

Extension	Author (truncated)	Version
jupyter-keymap	ms-	1.1.2
remote-containers	ms-	0.417.0
remote-ssh	ms-	0.120.0
remote-ssh-edit	ms-	0.87.0
remote-explorer	ms-	0.5.0
vscode-speech	ms-	0.16.0
indent-nav	Ton	1.0.3
ruff	cha	2025.22.0
npm-intellisense	chr	1.4.5
github-copilot-code-review	Cod	0.0.4
vscode-eslint	dba	3.0.10
prettier-vscode	esb	11.0.0
copilot	Git	1.335.0
copilot-chat	Git	0.28.0
vscode-containers	ms-	2.0.3
vscode-docker	ms-	2.0.0
debugpy	ms-	2025.8.0
python	ms-	2025.8.0
vscode-pylance	ms-	2025.6.1
jupyter	ms-	2025.5.0
jupyter-keymap	ms-	1.1.2
jupyter-renderers	ms-	1.1.0
vscode-jupyter-cell-tags	ms-	0.1.9
vscode-jupyter-slideshow	ms-	0.1.6
copilot-mermaid-diagram	ms-	0.0.3
vscode-commander	ms-	0.2.0
vscode-copilot-data-analysis	ms-	0.2.2
vscode-copilot-vision	ms-	0.1.1
parallels-desktop	Par	1.5.2

A/B Experiments

vsliv368:30146709
vspor879:30202332
vspor708:30202333
vspor363:30204092
vscod805:30301674
binariesv615:30325510
c4g48928:30535728
azure-dev_surveyone:30548225
962ge761:30959799
2e7ec940:31000449
pythontbext0:30879054
cppperfnew:31000557
dwnewjupyter:31046869
pythonrstrctxt:31112756
nativeloc1:31192215
5fd0e150:31155592
dwcopilot:31170013
6074i472:31201624
dwoutputs:31242946
customenabled:31248079
9064b325:31222308
copilot_t_ci:31222730
e5gg6876:31282496
pythoneinst12:31285622
bgtreat:31268568
4gafe986:31271826
c7cif404:31314491
5b33h341:31326279
pythonpulldiagcf:31325931
996jf627:31283433
pythonrdcb7:31303018
usemplatestapi:31297334
0aa6g176:31307128
7bj51361:31289155
747dc170:31275177
9hi5e907:31325879
aj953862:31281341
generatesymbolt:31295002
convertfstringf:31295003
pylancequickfixf:31329273
jhi8h917:31327034
d7aab740:31328298
usemarketplace:31328952
nesewctrl:31323299
1292j425:31329156
cja58720:31329340
4f60g487:31327383

[meganrogge]

Thanks for the issue. This is as designed because we don't have access to the output instantaneously - we wait for the command detection and command finished to occur.