I would like to express my disagreement with the decision to add voice transcription in intercom-ios v19.3.0 for two main reasons:
1. It is unnecessary. iOS has Dictation built-in, which allows users to dictate speech-to-text in any textfield system-wide, including the Intercom text box. Additionally, the Apple Dictation feature does live transcription, rather than waiting until the user presses "done", which is a nicer UX for those users who do want to use it. Finally, using the built-in Dictation feature doesn't require any additional iOS permissions, it simply works everywhere.
2. It was a breaking change in a minor update. Adding voice transcription requires the use of the microphone, which correspondingly requires a `NSMicrophoneUsageDescription` be provided or else the app will be rejected by Apple when you try to upload it. The fact that this breaking change was released as a minor update (ios v19.3.0, react-native v9.3.0) means it is easy for developers like myself to overlook the need for this change. The documentation says it is optional, but it is not. This is misleading and causes developer frustration, especially when trying to release timely updates, as the first sign we get that something is amiss is Apple rejecting the build, which happens right at the end of a deploy cycle.
For these reasons I would like to request that the SDK-level transcription feature be removed (or at least be made into an actually optional separate bundle / SDK). If it can’t be removed, I would like an explanation of why it is necessary to have rather than using the system-wide Dictation feature.