Improvements to MSG_VEL_COG [TES-129]

[lloydmaza]

Context: the MSG_VEL_COG spec was defined at the end of 2021 to support a customer need for non-fusion heading output. The message was defined quickly and without rigorous review, and a handful of issues were identified only after it was released. Upon further review, it was determined that we should update the message to fix these issues rather than define a new message altogether.
Its use has been limited to a single release branch (Starling v1.2.20), meaning the message has only been used by one external stakeholder. Deprecation is not imperative in this case to maintain usability, so the risk of this update appears relatively small and is favored as a quicker and less cumbersome solution.

Changes made to the message definition:

• flags field increased from 8 bits to 16
• "Velocity mode" and "INS mode" flags increased from 1 bit wide to 2 (as is standard with other velocity messages)
• "Type of reported TOW" flag added (as is standard with other velocity messages)
• "COG frozen" flag added to indicate when COG is frozen
• Wording changed on vel_d field to v_up

All other files changed as a result of updated bindings or updated test definitions.

Note: I had some trouble updating the test definitions, but eventually worked something out. My ad hoc procedure is written out here.

[reimerix]

LGTM - Thanks for writing up your process for updating the test definitions!

[silverjam]

Can we write up the rational for mutation of the existing message rather than depreciation of the existing message and the creation of a new one? We've technically released the existing message, so we need to be careful here.

[lloydmaza]

@silverjam great point, I've updated the description to include the rationalization for this approach. Let me know if you have any disagreements with the reasoning laid out - we've already encountered issues due to rushed implementation, so I'd like to avoid that going forward if we're introducing more risk than I've anticipated.

[silverjam]

Thanks for writing up the reasoning on why we are skipping deprecation