[nanomad]

Ends the heated seat and heated steering wheel commands with PARAMS_MAX (param 255), like the other control commands.

Changes

• control_heated_steering_wheel: sends {24: 1 or 0, 255: 0} (was {24}).
• control_heated_seat: sends {seat, 255: 0} (was {seat}).
• 255 is encoded as 4 zero bytes, the same as control_heated_seats, climate and rear window heat.
• Tests and docstrings updated. control_door_windows is unchanged.

Why

The iSmart app appends 255 to these commands. The first versions (#49, #50) left it out, matching a capture from one car and app build.

Draft: needs a test on a real car

• The old form (no 255) is the one proven on an MG S6 EV. This form is not tested on any car yet.
• The app's own value for 255 sometimes depends on a build flag (for example 1 when turning a front seat off). This PR always sends 0, the value on the app's normal path. If a car or build wants 1, it's a small follow-up.
• Please test heated seats and the steering wheel on a real car, on and off, before merging. @townsmcp, could you try this on your MG S6 EV?

https://claude.ai/code/session_013xKSpu8ev4cxQNsitcCidm

[townsmcp]

@nanomad, I tested #57 on my MG S6 EV (UK, iOS app). Both commands work with the 255 marker:

• Heated seat, front right (driver), request type 5, params [(18, 01), (255, 00000000)]: accepted. frontRightSeatHeatLevel went to 1, and level 0 turned it off again.
• Steering wheel, request type 8, params [(24, 01), (255, 00000000)]: accepted, and the wheel heated up.

They also work without it, which is how the iOS app sends them in my capture, and how mg-saic-ha has sent them since September. So the car accepts either, and you're right that it's harmless.

One thing worth knowing for the status: with the wheel on, the car reports steeringHeatLevel = 3, even though "on" is sent as 1. Anything reading it should treat any value above 0 as on.