Global event stream reconnects on a fixed 250ms timer with no backoff

Description

The app's global event stream reconnects after every failure with the same fixed RECONNECT_DELAY_MS (250 ms) — there is no back-off. When the server is briefly unavailable (restart, proxy blip, network drop), the client retries in a tight loop: a long outage produces hundreds of rapid attempts, keeps the CPU busy and spams the server with connections, even though a slow cadence would work just as well for recovery.

Plugins

N/A

OpenCode version

dev (current upstream/dev)

Steps to reproduce

1. Start the app so the global event stream is subscribed.
2. Stop the backend or cut the connection for > 1 second.
3. Observe the client reconnect loop: it retries every 250 ms with no growth in the delay.

Screenshot and/or share link

N/A

Operating System

Platform-independent

Terminal

N/A (app / desktop)

Related / duplicate check

• #26416 (idle CPU on macOS) can be aggravated by a tight reconnect loop.
• #38458 (SSE stream closes) is a different root cause (mid-turn closure), not this retry cadence.
• No issue covers the missing backoff itself.