I noted that in the FIT Protocol specification some details are not documented, and instead you have to look at the SDK source code.
These are the issues I found:
1. In "Table 1. Byte Description of File Header", the specification does not document that "Protocol Version" encodes the major version
(in bits 0-3, counting from the left) and the minor version (in bits 4-7).
2. In "Table 1. Byte Description of File Header", the specification does not document that "Profile Version" encodes the major version
and minor version using integer division and modulus.
3. In "Table 7. FIT Base Types and Invalid Values", the specification says that signed integers use 2’s complement format,
but it says nothing about float32 and float64. Currently this make supporting these two types not portable.
4. In "Table 7. FIT Base Types and Invalid Values", the specification does not document that "Base Type Field" encodes
both the endian ability (in bit 0, counting from the left) and the "Base Type #" (in bits 3-7).
Thanks
Manlio Perillo