UART functions on UDB4 and Interface

[UAVflightdirector]

UDB4 and servo interface could potentially have more than one
datalogging or telemetry UART. This is a definite requirement for the
interface board now.

Aims are:
Support any datalog or GCS format on any of the ports
Support any future data transport devices (SPI, TCP/IP, I2C, CAN
virtual serial etc)

HIgh speed UARTs can be added over an SPI interface so UDB4 can grow
more ports if requried.

[UAVflightdirector]

This project is now on hold.

More critical is making updates with the MatrxiPilot repository

On Nov 4, 4:25 am, UAVflightdirector <uavflightdirec...@gmail.com>
wrote:

[Peter Hollands]

Matt,

Sound like a very good idea.

Do you need further access rights ?

Pete

[UAVflightdirector]

Pete,

It would be very useful to add the MatrixPilot CAN files to the main
repository or a branch. Maintenance would be a whole lot easier.

My policy is that the CAN interface should make no difference to the
main code. A separate MPLAB project including the alternative
selection of files/drivers is required.

Right now, the file and functions naming needs some attention. Once I
get this fixed, merging the projects should be simple.


The interface code has just come to a decision point. It can either:
Diverge away from MatrixPilot giving it more flexibilty
Merge with MatrixPilot giving it better maintainability.

The interface code is a complete mess. It is based on a many months
old version of the re-org code plus some random updates. The
attempted untangling of MatrixPilot has made a giant mess and
confusion.

I need to put in some hard work to make it maintainable. I would like
to get it to the point where it is nicely MatrixPilot compatible with
its own MPLAB project.

Regards
Matt

On Nov 5, 4:40 pm, Peter Hollands <peter.holla...@gmail.com> wrote:
> Matt,
>
> Sound like a very good idea.
>
> Do you need further access rights ?
>
> Pete
>
> On Fri, Nov 5, 2010 at 2:37 PM, UAVflightdirector <
>

[Peter Hollands]

Matt,

As a first initial step, I recommend that we give you a branch in the main Matrix Pilot repository as soon as possible.

Ben is the primary gate keeper on style, policy, and future vision for code architecture of Matrix Pilot. So I think he will need to comment on that, and potentially some of the details of the merge.

Best wishes, Pete

[UAVflightdirector]

Both the interface and the MatrixPilot code are re-worked to fit the
latest MatrixPilot build. They are checked into the repositories so
anyone can use/view them.

This task was much easier than expected. The UDB side integration was
very smooth. The interface side required overides of a large portion
of the code.

The impact on the main code is very minimal so the risk is low.

The two projects are not quite compatible with each other yet. It
will be a few days before we have a single project that can be built
as UDBx, CAN interface or UDB2/3 with CAN.

On Nov 5, 7:06 pm, Peter Hollands <peter.holla...@gmail.com> wrote:
> Matt,
>

> As a first initial step, I recommend that we give you a branch in the main
> Matrix Pilot repository as soon as possible.
>
> Ben is the primary gate keeper on style, policy, and future vision for code
> architecture of Matrix Pilot. So I think he will need to comment on that,
> and potentially some of the details of the merge.
>
> Best wishes, Pete
>
> On Fri, Nov 5, 2010 at 5:24 PM, UAVflightdirector <
>

[Peter Hollands]

Matt,

Sounds like promising progress.

Best wishes, Pete