After 4 years of doing cross-platform mobile, I loathe React and React Native by now, and I dislike Expo. But… suggesting any other alternative is tricky.
True native development is, indeed, much better. But a small team will be hard-pressed for doing good work on both platforms, especially once you start fiddling with more important APIs. Configuration and getting your system to compile things to properly is almost a job in itself. Sigh.
KMM, Flutter, and MAUI are all promising, but with different trade-offs.
So we’re stuck with RN/Expo… At least Expo promises some stability between upgrades, by handling the package compatibility checks for you. Well, when it works.
“Make a progressive web app”: yes, that might be the way to go; provided you can get the users to ‘install’ it on their devices. For some companies that’s not optional due to different pressures (e.g. teachers would expect students to have our app installed on their phones).
Neither of these 2 solutions is that simple. There are trade-offs involved, and each team must compromise where they can/are weaker. All in all, if you MUST do RN, at least Expo saves you a lot of headaches with config/build/compatibility. It’s mediocre, but worlds better than pure RN (at least for me).
I came across a standard notes blog recently about how they “migrated” to a single codebase by relegating RN to render a single component of their web view, which is the same view that’s rendered in their desktop app and website, using a native “DeviceInterface” wrapper for each OS.
Sounded like a preferable approach to dealing with the complexity and limitations of RN and multi-platform. I wonder why more teams haven’t gone with this type of approach…
> “Make a progressive web app”: yes, that might be the way to go; provided you can get the users to ‘install’ it on their devices.
Assuming you're talking about the not-well-known process of adding PWAs to the Home screen, it's worth noting that you can package web apps for app store distribution as well. https://capacitorjs.com/
This seems quite good, if reviewers don’t reject the app on the basis that not enough of it is native. We’ve gotten push back from Apple for all sort of minutiae.
True native development is, indeed, much better. But a small team will be hard-pressed for doing good work on both platforms, especially once you start fiddling with more important APIs. Configuration and getting your system to compile things to properly is almost a job in itself. Sigh.
KMM, Flutter, and MAUI are all promising, but with different trade-offs.
So we’re stuck with RN/Expo… At least Expo promises some stability between upgrades, by handling the package compatibility checks for you. Well, when it works.
“Make a progressive web app”: yes, that might be the way to go; provided you can get the users to ‘install’ it on their devices. For some companies that’s not optional due to different pressures (e.g. teachers would expect students to have our app installed on their phones).
Neither of these 2 solutions is that simple. There are trade-offs involved, and each team must compromise where they can/are weaker. All in all, if you MUST do RN, at least Expo saves you a lot of headaches with config/build/compatibility. It’s mediocre, but worlds better than pure RN (at least for me).