You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Switching to the Expo SQLite library seems more desirable to me:
it reduces the number of organizations we rely on for our dependencies
we know that the Expo people can be reached via office hours and will take care of critical bugs promptly
it has some bells and whistles (on-device database browser seems nice)
it makes a SQLite database available in our app in case we ever want to use one
it's based on SQLite, the most widely used database in the world
Switching to the other MMKV library would mainly be desirable because:
it'd be easier
it has better performance (switching to sqlite might make our KISS strategy of synchronously writing the whole redux store to disk on each redux update less do-able)
But we're less likely to encounter any incompatibilities switching to the other MMKV library. It's possible the only reason we didn't take this approach initially is because the expo integration didn't exist previously.
Migration implications
This changeover could be done in two ways:
phased transition, where we try to preserve the contents of users' local databases by having two libraries installed at once and writing migration code. this is more effort (probably makes this a 7+ instead of a 3)
hard transition. this is easier, but has the following user-visible implications:
users will be logged out of the app
unsynced soil data stored on users' devices could be lost
users will be shown the interstitial again
The text was updated successfully, but these errors were encountered:
We are currently using https://github.com/ammarahm-ed/react-native-mmkv-storage. We've experienced a lack of responsiveness to bug reports with the maintainer:
and are now blocked from upgrading to expo 52 and react native 0.76 because of this.
I see two potential options for other libraries to use:
Switching to the Expo SQLite library seems more desirable to me:
Switching to the other MMKV library would mainly be desirable because:
But we're less likely to encounter any incompatibilities switching to the other MMKV library. It's possible the only reason we didn't take this approach initially is because the expo integration didn't exist previously.
Migration implications
This changeover could be done in two ways:
The text was updated successfully, but these errors were encountered: