I'd like to add my voice. Our softphone app uses a two-step incoming call process like many others. We send a push to the client, it posts a "registration" to our core which releases the actual call. On IOS this new developer-hostile requirements means that we are forced to post a "call" to callkit immediately we receive the push. Now we face two difficult problems: User answers too fast, there is no actual call Caller hangs up quickly - we have this stale call ringing and no communication path to our phone - we have to send another push but it can't be a voip push... The flow is terrible. Previously the push just woke up the app and it registers. It signals callkit as soon as it gets the actual call. This is the flow we use on another mobile phone platform and it works smoothly for us and the user. It is highly annoying and frustrating to discover that a competitor has been given a free pass around this hostile change.