Repository navigation
2.21.0
What's Changed
- Opt-in first-recorded-failure ordering for all() with a per-run marker (#307) by @rmcdaniel in #348
- Keep activity claims through transient lease renewal failures; release an unrenewable claim as worker_lost (#352) by @rmcdaniel in #353
- Cut worker and result response bytes: capabilities digest, history-free completed result (#352) by @rmcdaniel in #354
- Follow continue-as-new in result() instead of returning None (#356) by @rmcdaniel in #357
- Keep the worker alive through transient runtime failures while it registers (sdk-php#448 parity) by @rmcdaniel in #355
- Resend once in full when a worker answer names an unknown capabilities digest by @rmcdaniel in #358
- Treat cluster-info read failures at worker startup as transient by @rmcdaniel in #359
- Rediscover a stale payload policy on 413 payload_too_large and resend the completion once by @rmcdaniel in #360
- Prepare SDK 2.21.0 for the coordinated worker release by @rmcdaniel in #361
Full Changelog: 2.20.0...2.21.0