use forked node-speaker implementation with warning log removed
This PR removes the buffer underflow warning originating from the node-speaker's coreaudio.c implementation.
-
The update includes using our own forked repo with only the warning code line being commented out, the forked Repo is found here: https://github.com/UpstreetAI/node-speaker
-
Some package-lock updates are not directly associated with this update as they were integrated by running npm i before updating the speaker dep in package.json. (mostly ora package updates)
The latest updates on your projects. Learn more about Vercel for Git ↗︎
| Name | Status | Preview | Comments | Updated (UTC) |
|---|---|---|---|---|
| chat | ✅ Ready (Inspect) | Visit Preview | 💬 Add feedback | Nov 29, 2024 5:57am |
| docs | ❌ Failed (Inspect) | Nov 29, 2024 5:57am |
Why not include the library into packages and use file: install?
This is a good PR but we should probably ingest the change into the monorepo rather than depending on a git repo install (as far as I know we do not do this for any modules).
This is a good PR but we should probably ingest the change into the monorepo rather than depending on a
gitrepo install (as far as I know we do not do this for any modules).
- Added the suppressed speaker implementation to usdk/packages/speaker
- updated the package.json to use the local directory for installation
will have to fix build deps, moving to draft for now