On Wed, Jun 23, 2010 at 12:48 AM, Ross Burton <[email protected]> wrote: > On Tue, 2010-06-22 at 21:51 -0700, Luis R. Rodriguez wrote: >> Curious if HTML5 feeds are planning to be used to be passed to >> GeoClue? Moreover, I do have to wonder how reliable >> navigator.geolocation.getCurrentPosition() is. >> >> http://slides.html5rocks.com/#slide14 > > Firefox has a patch to implement geolocation using GeoClue, so you'd > want to make sure that you don't end up with an infinite loop. :)
How can you end up with an infinite loop? > Directly supporting the various methods that browsers would use (the > REST API to Google Latitude, etc) would be a preferable way of doing > this in my opinion -- embedding a web browser into GeoClue seems > overkill to me. I was not thinking of embedding a web browser to GeoClue, I was thinking about using feeds from any user's browser's navigator.geolocation.getCurrentPosition() to GeoClue. But how reliable is navigator.geolocation.getCurrentPosition() ? Where does it get its Geolocation data? At the very least, with what certainty can one rely on the country it provides? Mind you, I am thinking with regards to regulatory purposes to use the Linux kernel's userspace regulatory hint command, which tells the Linux kernel what country we are in to further enhance regulatory restrictions. This means if your card is in JP and you move to the US you get channel 13 and 14 disabled, therefore accelerating the scans. As another example for the case when someone with a US card moves to JP, this won't allow channel 14 due to existing regulatory considerations but I like to think of the future too and if we can guarantee with 100% certainty the location perhaps different things can be done in the future. Does GeoClue categorize location feeds based on certainty by chance? Luis _______________________________________________ GeoClue mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/geoclue
