Helium314
And without microG?
If I understand it correctly, this replaces the google services that tend to provide location data within seconds, but needs microG. I have a completely G-Free phone, are you aware of any usable alternatives?
You could try UnifiedNlp, which is only the location provider component of microG. But the standalone app is 6 years old, and didn't work for me when I tried it a few years ago.
On my phone I have microG installed only for location purposes, and all the Google service replacements disabled.
On my phone I have microG installed only for location purposes, and all the Google service replacements disabled.
Ah, I wasn't aware that was an option; I might look into it at some point.
The UnifiedNlp developer brought forward some thoughts on why it wasn't working anymore: https://github.com/microg/UnifiedNlp/issues/247#issuecomment-1468275786
UnifiedNlp uses [the Binder RPC system to bind to location modules]. Previously, such a bind relationship would keep the bound package process running and fully operational if and as long as the binding package process remains running. This is no longer the case in latest Android versions and Android would under certain conditions pause the bound process, resulting in the binder interface being stuck until the process is unpaused or terminated. UnifiedNlp itself is exempted from any such pausing while being registered as a system location provider.
It probably would work if we had UnifiedNlp and the modules as a privileged system app
I'm thinking of making location modules I have installed system-priv.
And without microG now? MicroG is dropping support for UnifiedNLP modules, for technical reasons, and incorporating location monolithically instead.
But I see that this project is still being worked on, as have other NLP backends been recently... is there a plan to keep these alive? Or is the assumption that people will just keep using the last microG version that had UnifiedNLP support?
See the pinned issue, especially https://github.com/Helium314/Local-NLP-Backend/issues/18#issuecomment-1678946066
