Just a bit of extra info for player 1.6.X, as a bonus for the unit mess,
the client libraries were standardised (they use meters, radians and
seconds) however the server protocol uses fixed point units which vary
by interface (hence the miliradians and so on in the docs).
In general if you doubt the documentation, which we attempt to keep up
to date but dont always succeed check out the source. In this case check
what the client lib does when it recieves a packet and you will need to
do the reverse in the driver.
Geoffrey Biggs wrote:
> In Player 1 it varies by driver. It's a bit of a mess, actually, because
> the units were standardised about a year and a half ago but not all
> drivers were updated to reflect the change. Milliradians was the new
> standard, as I recall.
> Player 2 is much better. All drivers adhere to the standard of metres
> and radians. Player 2 is due for release fairly soon, and you can start
> using it now from the CVS. I'd recommend you change as soon as you can
> simply because it's much nicer to use.
> Max Pfingsthorn wrote:
>> Dear developers,
>> I am debugging some drivers for the RoboCup Rescue Virtual Robots
>> competition and I was wondering what the unit for the rotation bits of
>> the position and position3d interfaces is. Is it really milliradians
>> (like the docs say) or degrees? Currently, the drivers return integer
>> degrees in the range from 0 to 360 (well, 359), and it seems to work
>> fine... For example, the pmap software works fine with degrees. I
>> looked at the khepera driver to check, and that one seems to output
>> degrees from 0 to 360 as well...
>> I just want to make sure we are doing it right. Or maybe the docs are
>> not right?
>> Thanks for your help!
>> Max Pfingsthorn