mvglasow
mvglasow
I'll wait for the F-Droid build, then. If that doesn't help, maybe you can add some extra debug output. It would be interesting to find out when we get notified...
It seems F-Droid didn't pick up v0.1.9 yet, thus I'm still on the old version. In the meantime I've noticed something else: I had been using Navit (a dev build...
Just ran the first of the two tests I described. The stale location is the last location supplied by the provider; subsequent wifi scans by other apps (Radiobeacon) do not...
Version 0.1.11 just appeared on F-Droid, I'll repeat the tests and then report.
The issue still seems to be there with 0.11.1. I currently don't have a functional wifi catalog – as soon as I do, I will pull a logcat and post...
It looks like the underlying issue is a different one: Apparently the wifi scan implementation in Android itself is buggy and occasionally returns stale scan results. While testing out some...
Great! Bear with me... I've implemented the logic to fuse the serving cell's location. Currently I'm working on penalizing stale wifis. Once this is working, I'll add lookup for neighbor...
Ran some tests today, and it looks like we have both lice and fleas. I fired up SatStat in an environment with known wifis and the modified backend gave me...
I've opened an issue for UnifiedNLP. What isn't quite clear to me: should we call `report(null)` when we can't determine the location, or should we just not call `report()` at...
Going through the code in detail, I see that `report()` never gets called if the new location is null.