parse_s3 had the S3 terminator logic inverted vs the real S3FrameParser
in framing.py. It was terminating on DLE+ETX and treating bare ETX as
payload, which caused every bare 0x03 to be swallowed — bundling multiple
real S3 frames into one giant body until a DLE+ETX sequence happened to
appear. Result: 583-byte POLL_RESPONSE 'frames' containing many real
frames concatenated, all showing BAD CHK.
Fix: mirror S3FrameParser exactly —
- Bare ETX (0x03) = real frame terminator
- DLE+ETX (0x10 0x03) = inner-frame literal data (A4/E5 sub-frames),
appended to body and parsing continues
https://claude.ai/code/session_014NczSHUz9uTzCAf4cVASTJ
SESSION_RESET (41 03) is sent before each POLL frame to wake monitoring
units. The ETX lookahead in parse_bw only checked for ACK+STX directly
after ETX, so when 41 03 followed a frame's ETX, the check failed and the
ETX was swallowed into the body as a payload byte — giving a 19-byte body
instead of 17 for POLL frames and failing checksum validation.
Fix: scan past any SESSION_RESET (41 03) sequences when looking for the
next frame start, so the real ACK+STX boundary is found correctly.
Also adds SUM8 checksum validation to parse_s3, which previously left
checksum_valid=None for all S3 frames.
https://claude.ai/code/session_014NczSHUz9uTzCAf4cVASTJ
Claude