Drop attempts to set MIS_TOTAL parameter

[peterbarker]

Closes #11413

[peterbarker]

We discussed this at a devcall. There are more elaborate solutions which would fix the problem more nicely, but this works in the short term.

[WickedShell]

I'm not thrilled about the solution, as this is messy from a GCS side as there isn't an ACK and it just looks like packet loss from the protocol.

[peterbarker]

On Mon, 3 Jun 2019, WickedShell wrote: I'm not thrilled about the solution, as this is messy from a GCS side as there isn't an ACK and it just looks like packet loss from the protocol.

Yep, I realised this when writing it. OTOH, if we drop a parameter in ArduPilot you'd get the same effect, so GCSs must already cope with this in a not-too-bad fashion.

[peterbarker]

This has been modified to notify the GCS that we accepted their set - but instantly send the current value. This is so we don't break the expected parameter state machine.

STABILIZE> param set MIS_TOTAL 17
STABILIZE> APM: Param write denied (MIS_TOTAL)
< PARAM_VALUE {param_id : MIS_TOTAL, param_value : 17.0, param_type : 4, param_count : 1139, param_index : 65535}
< PARAM_VALUE {param_id : MIS_TOTAL, param_value : 3.0, param_type : 4, param_count : 1139, param_index : 65535}

param set RC4_TRIM  1600
STABILIZE> < PARAM_VALUE {param_id : RC4_TRIM, param_value : 1600.0, param_type : 4, param_count : 1139, param_index : 65535}

[tridge]

better to use a flag bit in param table

[peterbarker]

Reworked to use a new parameter bit indicting a parameter is for internal use only.

[peterbarker]

.... so am I going back to the string-matching thing on this one?

[OXINARF]

One improvement we should have in the future is allowing the top parameters to have flags too.

Points have been addressed - thanks!

[peterbarker]

@tridge @WickedShell As discussed/requested on the DevCall, I've fixed this up based on @OXINARF 's review - should be good to go now.

[OXINARF]

LGTM