BugVM – free open source alternative to proprietary and discontinued RoboVM
bugvm.com
bugvm.com
The problem is that LibGDX chose not to pick a "winner" and is fragmenting between all of those different forks/options guaranteeing that they will all have issues reaching critical mass.
If you are a Java developer there is one option that pre-dated RoboVM, has always enjoyed better traction and installs millions of apps every month and that's Codename One. It's taken a far more conservative approach to the VM so it is both future proof and much smaller: https://github.com/codenameone/CodenameOne/tree/master/vm
Add to that a true write once run anywhere approach there is really no one else.
I am sceptical of bitcode, and i would not want to buy into that scheme.
E.g. they announced 64 bit and gave relatively little time to make it a requirement for all new projects. Apple officially supports: C/C++, Objective-C, Swift & JavaScript. If you skip a stage (like going directly to LLVM) they will not make your lives easier.
Since we are talking about portability I guess most client developers abandoning Java go to JavaScript not C++.
On iOS, WP, Tizen and anything else Java support is quite worse than having to deal with C++, specially since it is a first class language on those platforms.
It is also the approach taken by Microsoft, Dropbox.
My work experience is that relying on second class languages always leeds to support problems down the line.
I had that with Turbo Pascal, Oberon and quite a few other languages.
Umm... I don't have hard numbers handy, but isn't the mobile OS market something like:
60% - Android
35% - iOS
5% - Everything else combined (if it's even 5%)
Quite frankly, if I were going to do a professional mobile app today (a big "if", as the bubble seems to be bursting)... then I'd just bite the bullet. Use Java for the Android version, ObjC/Swift for the iOS version, forget anything else, and keep the actual business logic on the server side.
No need to increase development costs.
My point was about those that care to write portable native code with minimal costs.
Or one just decides to stick with the biggest platform, like the PC on the desktop, and focus soleny on Android.
In terms of "viability", young Android devs and iOS devs grow on trees... including those who are comfortable writing both Java and Swift. In contrast, finding a competent C++ Android or iOS developer is much more challenging. And he or she would still have to manage a codebase with completely different GUI frameworks crammed into it.
I'm not arguing that "X" is a better cross-platform panacea than "Y". I'm arguing that the underlying premise is misguided.