[acao-it] ACAO Tracking Protocol

Daniele Orlandi daniele at orlandi.com
Tue Jul 7 11:44:42 CEST 2015


On 05/07/2015 13:59, Thomas Sevaldrud wrote:
> Hi Daniele,
> 
> Thanks for that, that was fast, looking forward to try the server. We
> will of course test it thoroughly and give you feedback :-)

Expect it at the beginning of the next week, I'm currently on vacation.

> The original plan as I understood it was that we would work out a new
> protocol in cooperation with FAI which I presume would involve you in
> this work

A protocol from FAI is not something I'm aware of.

> We will be there the first weekend + a few days into the competition.
> After that I think Brian will make arrangements for someone else to run it.

Okay, fine.

> We have an alpha version of the viewer out. I'll send you a separate
> email with instructions for downloading and running that later.

Thank you.

>     Our system receives and sends fixes every second, in real time with no
>     delay, it does not do any queueing, so, every fix that is not received
>     is lost. You should expect fixes missing randomly or for long time due
>     to radio reception quality.
>  
> This could be a problem... One important feature of the system is that
> it shows a leaderboard with the current standings during the race. If we
> lose fixes inside waypoint sectors this will of course mess up the scoring.

I suppose the turning points will be placed in a well covered area and
that occurrence should be rare although not impossible, however I
suppose you also have a way to manually "fix" missed turns, don't you?

> The previous systems we have worked with have buffered data on-board and
> pushed everything when they got back in an area with reception, but I
> guess that is not an option here since each FLARM just sends blindly right?

Yes, FLARM transmission is blind.

> Also, the server needs to store all fixes from all the gliders, so that
> when a viewer connects for the first time, it gets the entire flight (up
> to the current time).

This is not a problem and it is already implemented as such.

Bye,


More information about the It mailing list