Thank you — this makes sense. For Ossetian, I would expect Cyrillic Ӕ / ӕ under the А / а key on the Russian Cyrillic keyboard. I also have a public MIT-licensed corpus-based dataset for Cyrillic keyboard support: RF Keyboard Corpora: https://github.com/Agisight/rf-keyboard-corpora The Apple-specific summary file is: world_results/world_priorities_apple.csv It provides prioritized long-press variants by Russian base key, based on corpus/frequency data. The repository also contains or is intended to contain language-specific data such as: language name script and alphabet special characters expected base key for long-press variants keyboard layout JSON corpus-based character frequencies population/coverage statistics machine-readable metadata I will also file this through Feedback Assistant with structured CSV/JSON attachments, because I understand that this is the proper channel for tracking system keyboard requests. Thank you.

Hello! I see in the beta the Tuvan and Sakha keyboards. Will you add more languages? Also I see the improved Russian keyboard, but many letters have no long press letters, for expample, Ossetian Ӕ ӕ. Do you have all longress mapping for all Cyrillic languages?

I can provide all these validated NLP data to implement text predictions. For example, I have a big text-monocorpus for Chuvash, and a huge collection of voice-text dataset