Who needs security when the apps do it by design.
Check your firewall logs, or use a root/no-root firewall, it's frightening.
Edit: click the screenshots at https://play.google.com/store/apps/details?id=eu.faircode.ne... to see an example. Fourth picture "Access attempts"
The only downsides I have seen to this method are it's slightly annoying to initially install and if github goes down (where the pac is hosted), phone network connectivity goes down until it comes back up.
It's a shame Android doesn't allow vpn chaining though, because netguard creates a local vpn connection you can't use it with an actual vpn simultaneously.
I think the author could take the same engine and create a new so that just blocks trackers and Facebook and people would gladly pay for it.
It would be cool if app stores showed which apps require network access and which don't.
https://play.google.com/store/apps/details?id=app.greyshirts...
For example it helped me find a trustworthy e-mail client (K-mail) by only allowing network access to my email server.
This kind of software can also be difficult to get right. There may be ways to easily bypass it.
If Google did this, I'd seriously consider going back to Android.
You can hardly compare coreutils to smartphone apps.
As someone who deals with Android fragmentation daily, you wouldn't believe how hard it is to support the whole ecosystem. Disliking analytics I can understand, but crash reporting adds measurable improvement in the quality of software.
> You can hardly compare coreutils to smartphone apps.
Indeed; coreutils works on an absurd variety of unix-family systems, is copyleft, respects privacy, and is high-quality. If smartphone apps were more like coreutils, we'd all be better off.
> As someone who deals with Android fragmentation daily, you wouldn't believe how hard it is to support the whole ecosystem. Disliking analytics I can understand, but crash reporting adds measurable improvement in the quality of software.
I do understand where you're coming from; while I don't have skin in that game, I'm passingly aware of the remarkable ways that vendors break their systems. And honestly, if you get user consent, I don't actually have a problem with having the app report problems to you. It's just the "send information about the user's system without asking" thing that I object to.
Gimp: https://bugzilla.redhat.com/show_bug.cgi?id=1828800
I wouldn't call it auto-reporting when the user has to agree first; consent is the sticking point. (Same for abrt) (granted, I should have been more precise and I suppose that is automated in a sense; I took "automatic" to mean "without asking")