Conversation
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 3c5881892f
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| conn, err := dialRPCWebSocket(ctx, c, wsURL) | ||
| if err != nil { | ||
| return nil, err |
There was a problem hiding this comment.
If the local cache points at a deleted remote file/revision, EnsureUploaded can return stale IDs and the first WebSocket handshake to /files/{id}/xlsx/ws will fail with HTTP 404. This path currently returns immediately, so witan xlsx rpc cannot start even though other files-backed commands recover by reuploading after not found. Add the same stale-cache fallback here (reupload and retry dial) so startup is resilient to stale cache entries.
Useful? React with 👍 / 👎.
There was a problem hiding this comment.
Nothing to action: for that specific “initial WebSocket dial gets 404 due to stale cache” issue, I don’t see anything to action in either repo right now.
Current behavior is:
witan-alfred: files-backed /xlsx/ws upgrades first, then session creation can emit FILE_NOT_FOUND and close 4404.
witan-cli: already handles that stale-cache path by reuploading, reconnecting, and replaying the first RPC line.
1 participant
No description provided.