GPS as a source of time like NTP
daniel at totalueberwachung.de
Sun Jul 27 03:12:45 CEST 2008
On Sat, 26 Jul 2008 19:42:43 -0500
"Dale Schumacher" <dale.schumacher at gmail.com> wrote:
> On Sat, Jul 26, 2008 at 6:28 PM, Daniel Willmann
> <daniel at totalueberwachung.de> wrote:
> > Yeah, currently a signal is emitted if the current
> > location/accuracy/whatever packet contains a different value than
> > before. So If nothing changes you wont get a Signal either. The one
> > exception is time because I don't want to send that every second for
> > nothing. The idea at the moment is to just send a TimeChanged signal
> > once as soon as the time is valid. After that you can request the
> > time if you so desire.
> Is the GPS time used as a source for the system, like an NTP server?
There's no code for it yet, but yeah, that is being taken into
consideration. Basically I want to send a TimeChanged signal once after
I got the valid time from GPS (and not every second as gypsy specifies
as that is just wasting resources - although that probably matters if
we really route accelerometer through dbus...) and whenever a program
asks for it via a method.
Then you would have a program listening on dbus for the TimeChanged
signal and set the system clock (and rtc) according to that.
> That seems like it would be a very useful option. It would take
> advantage of the special capabilities of the FreeRunner beyond a
> typical Linux box.
That and auto timezone selection based on your location will be some
nice features, yes.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 197 bytes
Desc: not available
Url : http://lists.openmoko.org/pipermail/devel/attachments/20080727/17623660/attachment.pgp
More information about the devel