A Tasmota patch keeps boiler values after a missed response
A Claude Code-assisted OpenTherm patch retries three times before dropping an unsupported value. The contributor reports testing it on a Siemens LMU 64 boiler and ESP32-S3.
Credits
Earliest dated source: Details reviewed:
The full story
A single timeout or invalid reply could make Tasmota drop an OpenTherm data value until the next handshake or midnight reset. On a Siemens LMU 64 in a MAN ProCon GWB 25 HS boiler, the contributor saw hot-water temperature or maximum-temperature values disappear for the rest of the day.
The proposed change uses an existing retry check so three failed attempts are needed before marking a value unsupported. The author explicitly credits Claude Code and says they manually verified the change on an ESP32-S3 running Tasmota 16.0.0 plus the patch. Supported values remained in telemetry, while six genuinely unsupported IDs were still excluded. The pull request is open; it is not a released Tasmota fix.
Before and after
- Starting problem
- One missed first response could remove a supported boiler value until a protocol reset.
- Reported result
- The contributor reports supported values surviving a restart after applying a three-attempt retry rule.
Device and compatibility
- ESP32-S3 with Siemens LMU 64 (MAN ProCon GWB 25 HS) · Tasmota 16.0.0 plus the proposed patch
The hardware test is the contributor's report; the PR remains open.
Source ↗
What Claude did
- OpenTherm retry-condition fix, explicitly credited as Claude Code generated and manually verified by the contributorClaude Code Source ↗
How it was done
Tools
Themes
Loading themes…
Sources
From the maker's own write-up or code. We summarise what they report; we haven't tested it ourselves.
- GitHub: OpenTherm: retry a command 3 times before marking it as not supported
The contributor's proposed patch, test report and explicit Claude Code credit.