Vitest vscode plugin is listening on the port which is needed by vite

[fourdim]

Describe the bug

Possibly related to #443.

Reproduction

1. Download the repo.

git clone https://github.com/toeverything/blocksuite

2. Install Vitest VSCode plugin.

3. Install dependencies.

yarn

4. Development mode.

yarn dev

Here Vite is not binding to the default 5173 but 5175.

Output

[INFO 4:53:41 PM] [v1.6.0] Vitest extension is activated because Vitest is installed or there is a Vite/Vitest config file in the workspace.
[INFO 4:53:41 PM] [API] Running Vitest v2.1.2 (blocksuite/vitest.workspace.js) with Node.js@v20.17.0: /home/fourdim/.config/nvm/versions/node/v20.17.0/bin/node 
[INFO 4:53:42 PM] [Worker] �[33mThe CJS build of Vite's Node API is deprecated. See https://vitejs.dev/guide/troubleshooting.html#vite-cjs-node-api-deprecated for more details.�[39m
[INFO 4:53:42 PM] [Worker] Forced re-optimization of dependencies
[INFO 4:53:42 PM] [Worker] Port 5173 is in use, trying another one...
[INFO 4:53:42 PM] [API] Vitest v2.1.2 (blocksuite/vitest.workspace.js) child process 78366 created
[INFO 4:53:42 PM] [VSCODE] Watching blocksuite with pattern **/*

Version

v1.6.0

Validations

• Check that you are using the latest version of the extension
• Check that there isn't already an issue that reports the same bug to avoid creating a duplicate.
• Check that this is a concrete bug. For Q&A open a GitHub Discussion or join our Discord Chat Server.
• The provided reproduction is a minimal reproducible example of the bug.

[sheremet-va]

Does this happen if you just run vitest in the root?

[fourdim]

If I uninstall the extension and run the vitest in the root:

I think it's expected behavior as we need to launch two instances:

• presets: bind 5173
• block-std: first try to bind 5173, then 5174

[hi-ogawa]

Browser runner taking 5173 wasn't intended and was fixed in #6700 Can you try the latest pkg-pr-new from https://github.com/vitest-dev/vitest/runs/31509661255?

[fourdim]

Browser runner taking 5173 wasn't intended and was fixed in vitest-dev/vitest#6700 Can you try the latest pkg-pr-new from https://github.com/vitest-dev/vitest/runs/31509661255?

Still, this issue persists.

[hi-ogawa]

Did you update @vitest/browser?

[fourdim]

Did you update @vitest/browser?

Issue persists even if I update @vitest/browser.

[sheremet-va]

Browser runner taking 5173 wasn't intended and was fixed in vitest-dev/vitest#6700 Can you try the latest pkg-pr-new from vitest-dev/vitest/runs/31509661255?

I can actually confirm that pkg.pr.new from https://github.com/vitest-dev/vitest/actions/runs/11343255694/job/31545435431 fixes the issue in the reproduction. It now uses 63315 by default, but the next one still tries to use 63315 - I wonder if we can track that on our side.

The CJS build of Vite's Node API is deprecated. See https://vitejs.dev/guide/troubleshooting.html#vite-cjs-node-api-deprecated for more details.
Forced re-optimization of dependencies
Port 63315 is in use, trying another one...

 DEV  v2.1.3 /Reproductions/blocksuite
      [@blocksuite/presets] Browser runner started by playwright at http://localhost:63315/
      [@blocksuite/block-std] Browser runner started by playwright at http://localhost:63316/