[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Paparazzi-devel] MAVLink / Ivy / ...
From: |
Chris Gough |
Subject: |
Re: [Paparazzi-devel] MAVLink / Ivy / ... |
Date: |
Tue, 6 Sep 2011 23:42:27 +1000 |
Hi Oswald,
> the main problem for a more comprehensive solution, but this applies to
> working with mavlink alone as well, is that there is hardly a canonical
> set of messages apart from IMU data, that suits all and everyone. in
> that regard, if both systems used an identical message description XML,
> that would make things easier.
Yes, each protocol is somewhat redundant too, so there isn't a way to
map between two message.xml files unambiguously.
If the translator was to work between more than 3 protocols, it would
probably need it's own internal 'canonical' language. The parser would
convert messages into the internal language with simple mappings, but
then all sorts of heuristics and duct tape (including state machines)
would be needed on the generator side. yuck.
This site http://www.cdlsystems.com/index.php/ucs-functional-acrhictecture
describes STANAG-4586 architecture as having "vehicle specific
modules" that adapt generic protocol (specifically, STANAG-5486) to
specific systems. That seems practical, cobble together the minimum
sufficient generator heuristic and test it in context.
Chris Gough
On Tue, Sep 6, 2011 at 6:33 PM, Oswald Berthold
<address@hidden> wrote:
>
> hi,
>
> i've been messing with this issue lately as well. we are using our
> homebrew (mavlink-based) controller framework and in effect i wanted to
> do, among other things, the reverse, that is, use the pprz plotter and
> settings tools instead of qgroundcontrol (which is a cpu hog). this
> works all quite well.
>
> http://www2.informatik.hu-berlin.de/~oberthol/QK/img/
>
> the main problem for a more comprehensive solution, but this applies to
> working with mavlink alone as well, is that there is hardly a canonical
> set of messages apart from IMU data, that suits all and everyone. in
> that regard, if both systems used an identical message description XML,
> that would make things easier.
>
> much left on the system inter-operation side .. (UAVTalk?, OSC, ...)
>
> bst, opt
>
> Chris Gough <address@hidden> writes:
>
>> Hi Jesus,
>>
>> I don't think so. Changing paparazzi to use MavLink nativley is a big
>> job. I did muck around with an adapter that bridged between an IvyBus
>> and MavProxy, translating "identity and location" messages so that I
>> could use a paparazzi-based tracking antenna to follow vehicles
>> speaking either protocol (and vica versa). I only got as far as a
>> proof of concept, enough to satisfy myself that the "translator
>> approach" is promising for that limited purpose.
>>
>> I'm not sure the translator approach could practically be taken as far
>> as using qgroundcontrol with paparazzi (for example).
>>
>> Chris Gough
>> (also known as Rufus)
>>
>> On Wed, Jul 20, 2011 at 9:34 PM, Jesus Martin Sanchez <address@hidden> wrote:
>>> Hello:
>>> Anybody are using MAVLink with paparazzi? Rufus, if I´m not wrong you are
>>> working about this, isn´t it?
>>> http://qgroundcontrol.org/mavlink/start
>>>
>>> Thanks
>>> Jesus
>>> _______________________________________________
>>> Paparazzi-devel mailing list
>>> address@hidden
>>> https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
>>>
>>>
>
> --
> Sent from my fingertip ...
>
> _______________________________________________
> Paparazzi-devel mailing list
> address@hidden
> https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
>
--
.