Re: [Paparazzi-devel] 4.0 release preparation

[Hector Garcia de Marina]

I agree with Esden,

as the stm32f4 is "99%" compatible with stmf103 series, I believe that keeping this algorithm is not a bad idea.

Héctor

On Thu, Feb 2, 2012 at 6:30 PM, Piotr Esden-Tempski <address@hidden> wrote:

Hey guys,

This is a great thing to do! I love the initiative.

@ Antoine: the float version is a great thing actually. The stm32f4 has hard float support that a bunch of people is exploring and benchmarking at the moment. So it may become actually something we would want to run on a vehicle at some point.

I am not sure if we need to include the float estimator stuff in the release yet…

Cheers Esden

--
Transition Robotics Inc.
http://transition-robotics.com
the makers of the Quadshot
Co-Founder, Embedded systems Engineer
Mobile: +1 831 440 7454
Skype: esdentem

On Feb 2, 2012, at 9:18 AM, Felix Ruess wrote:

>>  fixed point version only does quaternions, as I did not want to go
>> through the hassle of writing the matrix re-normalization in fixed
>> point.
>
> yes, that is why I named the subsystem int_cmpl_quat.
> - So I would propose to rename the files for it to
> ahrs_int_cmpl_quat.* as well (same name as subsystem and more explicit
> since it emphasizes that it uses quaternions.
> - rename the float_cmpl_dcm subsystem to float_cmpl and the files to
> float_cmpl.* to make it clear you can use either rmat(dcm) or
> quaternions representation/propagation. Or rather make two subsystems
> called float_cmpl_rmat and float_cmpl_quat and define
> AHRS_PROPAGATE_RMAT or AHRS_PROPAGATE_QUAT respectively to the
> subsystem makefile?
>
> _______________________________________________
> Paparazzi-devel mailing list
> address@hidden
> https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

_______________________________________________
Paparazzi-devel mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

-- 
Héctor