Hi. First off, I merged your two leaderboard entries ;)
On 04.09.2014, at 14:07 , Zeeshan Ali (Khattak) <[email protected]> wrote: > While I'm very happy about all MLS developments, rely on it a lot in > geoclue and help with collection of data as much as I can, I'm not > exactly happy with this development in particular. IMHO this will > undermine my efforts to push for geoclue to be *the* geolocation > framework every app should be using on Linux desktop (especially > Fedora, since I work for RH). It will make it harder for me to > convince about the importance of firefox needing to use geoclue and > therefore getting any contributor to make that happen [1]. > > This is also not so good for firefox actually since it can't use > modems and thus you are limited to geoip and wifi-geolocation only. I think it makes sense for Fedora to use MLS via the existing code base. All that needs to be done to get working geolocation is a build config change and adding the api key. The code to talk to MLS / Google is already there in Firefox, so this is very simple non-coding change. In my opinion switching to geoclue still makes sense for a different set of reasons. As you mentioned, geoclue can actually talk to more sources like GPS and cell modems. It can also act as an app-agnostic caching layer, so if a native maps app already got your location, maybe Firefox doesn’t have to call out to the network at all anymore, as geoclue already has a recent and good location fix. And if we’d ever get any sort of offline database, geoclue would be a much better layer to implement it than making this specific to each app. In that light geoclue makes sense to get better quality and improvements. That’s maybe a harder sell than just going from broken to basic functionality, but there’s still value to be had here. Hanno _______________________________________________ dev-geolocation mailing list [email protected] https://lists.mozilla.org/listinfo/dev-geolocation
