Why would a JavaFXPorts/Gluon Mobile application be better than Codename One, JUniversal or Tabris. Google shows these three solutions which made me wonder, were these solutions not good enough? Especially when Codename One claims to be faster than native code. And why would one choose to start a new project and not for example collaborate with the existing solutions?
Asked
Active
Viewed 518 times
1 Answers
1
Codename One predated most of the others but took a rather different path. In fact Codename One's origins predate JavaFX as far as I know.
Codename One was started by mobile guys and as such the VM doesn't try to offer the full JDK. It's small nimble & AOT from the ground up. The whole toolchain was designed for mobile as explained in their VM project: https://github.com/codenameone/CodenameOne/tree/master/vm
Gluon is JavaFX based so it needs the full VM which back in the day relied on RoboVM. The guys from Codename One explained why they didn't go with that in the link above. Gluon now uses OpenJDK which is an interpreter and probably not the best approh for something that's already dog slow on iOS.
FX was designed by desktop people and Codename One was designed by mobile people, the difference in mindset is huge.
Tabris is a server solution and JUniversal is a Microsoft research project that came out much later. It's goals are unclear.
You missed J2ObjC whose goal is library portability and Avian which is an interesting VM. To my knowledge Codename One had a port for it in the past which underperformed.
David Cohen
- 429
- 2
- 4