[ros-users] Standard GPS Message

[Jack O'Quin]

On Fri, May 21, 2010 at 5:02 PM, Bill Morris <morris at ee.ccny.cuny.edu> wrote:
> I want to see if the GPS message format can be standardized, and I was
> wondering if anyone had pointers on what steps need to be taken to get
> the message format(s) considered for the D-Turtle release.

WG would probably permit a community effort to define a "gps_common"
package somewhere in ros-pkg with shared messages, UTM conversions,
etc. That supports the federated development model by enabling common
interchange of data and sharing of code.

http://ros.org/wiki/QAProcess defines the steps:

  http://www.ros.org/wiki/PackageProposalProcess
  http://www.ros.org/wiki/APIReviewProcess
  http://www.ros.org/wiki/DocReviewProcess
  etc...

> The review process for camera_info_manager seemed to work well enough so
> perhaps something like that can be used.

Agreed.

I can help once I finish my current C-turtle work.

> Right now I'm writing a driver for a GPS that does not work with gpsd,
> and our lab is working on building some graphical GPS tools so it is
> probably a good time to standardize the message format before we get
> much further.
>
> I think the gpsd message format looks like a good start, but I would
> also like to see some sort of waypoint format as well. This will help
> ensure that all of the tools developed are interoperable.

> Current Formats
> http://www.ros.org/doc/api/gpsd_client/html/msg/GPSStatus.html

This looks reasonable to me too, although a slight restructuring might
be good for some uses.

> http://www.ros.org/doc/api/applanix/html/msg/GpsInfo.html

I created this one from the Player GPSI message format, but added some
fields for my own convenience. I would not necessarily recommend it
for standardization in its current form.

For example, I added the UTM easting, northing and zone data. Since
they can be derived from latitude and longitude, they should not be
part of the GPS message. But, some standard conversion routines
between the two coordinate systems would be good to have.

I have some rather messy ones here. They could be cleaned up and used,
if there's nothing better available.:

  http://www.ros.org/doc/api/art_common/html/UTM_8h_source.html
-- 
 joq