Separate microphone and recognizer permissions #55
Conversation
|
There was a problem hiding this comment.
Looks good! These granular permissions APIs are definitely going to be a great improvement for the library! The only thing I'm just a bit iffy on mostly has to do with communicating the API usage in our documentation because it only really affects a semi-niche use case.
Co-authored-by: jamsch <[email protected]>
Co-authored-by: jamsch <[email protected]>
There was a problem hiding this comment.
Thank you @msschwartz, this looks pretty much ready to go! When I have some time later today I'll clone your repo and test it out on iOS/Android/Web to make sure everything's working okay and update the README to document the new APIs.
* skip permission if not needed * separate permissions * dev * Update App.tsx * Update Podfile.lock * Update App.tsx * moving permission check * deprecated * Update example/App.tsx Co-authored-by: jamsch <[email protected]> * return granted response on web * Update src/ExpoSpeechRecognitionModule.types.ts Co-authored-by: jamsch <[email protected]> * More docs update and removing deprecated * Adding Android permission functions --------- Co-authored-by: jamsch <[email protected]>
When using on-device recognition, the device does not need speech recognizer permissions. By separating the permission checking we can avoid showing the following message to users:
This message is misleading since data is NOT sent to Apple when using on-device. Many users prefer to keep their data private for apps with sensitive data.
If this is something we should move forward with, we'll need to update web and Android code.