[RFC] Unmanaged GPS setup / location source
dcbw at redhat.com
Mon Jun 30 06:41:51 PDT 2014
On Sat, 2014-06-28 at 18:08 +0200, Aleksander Morgado wrote:
> On Fri, Jun 27, 2014 at 5:45 PM, Dan Williams <dcbw at redhat.com> wrote:
> > On Fri, 2014-06-27 at 10:51 +0200, Aleksander Morgado wrote:
> >> Hey,
> >> I've been asked whether it would make sense to allow having
> >> ModemManager control the status of the GPS module in a device, *but*
> >> leaving the tty where the NMEA traces are dumped untouched. This would
> >> allow to integrate with other standard GPS applications which read
> >> from a tty, while leaving MM the control over when to enable/disable
> >> the GPS module.
> >> Any comments on this?
> > I haven't looked at the code yet, but the approach seems OK.
> >> I have a draft implementation (mmcli commands not yet done) already in
> >> the 'aleksander/gps-unmanaged' branch in upstream git, where I just
> >> define a new MM_MODEM_LOCATION_SOURCE_GPS_UNMANAGED source. When this
> >> source is enabled, GPS module is started but the GPS tty is not used.
> >> Of course, the implementation also forbids trying to enable unmanaged
> >> and raw/nmea GPS at the same time. Also, currently only done for
> >> Option/HSO (the only plugin which does GPS over a tty for now). I may
> >> add GPS support for Cinterion one of these days as well.
> > Gobi! :) The fun thing here is that you have to open the port and run
> > $GPSSTART or something like that, then it starts spewing NMEA, and to
> > stop it you have to re-grab the port and run $GPSEND. I guess that
> > would be fun if the port is still locked by the consumer.
> You mean you need to send $GPSSTART in the same port where you're
> going to get the NMEA traces?
> > Also, I think the Ericsson devices do raw NMEA to a TTY too.
> Yes, I meant to add support for it earlier, but I don't have a module
> to test with.
I can test with a couple different devices (F3507g, F3607gw, F5521gw,
More information about the ModemManager-devel