Questions about the FlySky protocol

OK I come here from the land of Turnigy 9x in search of protocol information.

I have one of these little quads www.jperkinsdistribution.co.uk/detail.ph...6600036&activepage=1 similar to the V949 I hear. It uses the A7105 and more or less works with the FlySky module in that you can fly it but you can't switch the LEDs on or off, boo

So I go digging a little deeper and start sniffing the SPI bus to the A7105. I see you guys have already been doing just this. Your work has greatly speeded my learning up, a thousand thanks ! My goal is to make a PPM to A7105 SPI interface that I can control so it will switch my LEDs(, and eventually work with a few other models I have).

Having my sniffed packets in one hand, the A7105 datasheet in another, and your source code in my third hand, I now have a couple of questions hopefully you can help with.

Once the initialization is done I see packets like this:

My first question is the file deviation/doc/Flysky.txt suggests the data is sent in a 22 byte packet.
The source code file deviation/src/protocol/flysky_a7105.c and my observations suggest the data packets are 21 bytes
Which is right ? (both and I'm wrong ?)

Second question. Again in flysky_a7105.c the packet[] array elements 1,2,3,and 4 are 32 bit variable labelled id, but page 65 of the A7105 datasheet suggests the "ID code" is bytes 0,1,2, and 3 of the FIFO buffer ?

What am I missing ?

Thanks.
Cam.

OK, got the 21 bytes. I guessed the 05 might be getting included.

So how is the A7105 ID code set ? It's not part of the 21 byte data packet ?

EDIT: Blind from looking at the data sheet. It is in register 0x06 and pointer is reset every time you start an access. I wish I had found this earlier, been driving me nuts. What's worse I watched it get set in my captured data and thought that looks like an ID getting set ! And everything falls into place.

Is it just me or have I seen better written datasheets ?