Of all the changes coming out of Google I/O 15, the new permissions model will have the greatest longer term impact on developers. I say ‘longer term’ because for existing apps on existing phones, users will see no difference. Existing apps on phones running ‘M’ can have their permissions taken away by the user which will, no doubt, cause them to malfunction and possibly cause confusion for the user (and people supporting the app).
The real power and possibilities come when an app is built for ‘M’ (or in the future ‘M or later’). Developers and designers now really have to think hard about their messages to users. There’s an article on Medium about the Cluster app, on iOS, that covers similar issues and the complications.
The good thing is that (‘M’ built) apps will no longer sit around, as now, waiting for the user to upgrade when they need more permissions. The down side is that the user disallowing a permission can seriously bazooka your app and you need to really try your hardest to explain to the user why a permission is needed.
Obviously, more transparent permissions is good for everyone – in theory. In practice, I find many clients and some designers already have trouble agreeing/deciding how screens and messages should look and feel. Adding permissions screens/dialogs into the mix is bound to lead to protracted mobile development, more testing of permissions-based scenarios and more support needed when apps don’t work due to revoked permissions.