Open stefanceriu opened 8 months ago
Btw another easier fix for the meantime to help with https://github.com/vector-im/element-x-ios/issues/2030 and https://github.com/vector-im/element-ios/issues/7489 would be to try the following solution found by @stefanceriu to intercept that a text to speech is happening: https://stackoverflow.com/questions/32652775/ios-how-to-detect-if-voice-dictation-was-used-for-uitextfield-or-microphone-bu/53440078#53440078
Maybe we could use this to avoid using the diffing system until we are sure that the user has ended the dictation?
To improve performance by reducing the amount of data mapping. To improve maintainability by removing the need to map selection indices between the different model formats (AST / HTML / AttributedString)
Would fix https://github.com/vector-im/element-x-ios/issues/2030 and https://github.com/vector-im/element-ios/issues/7489