JavaFX port to iOS/Android status

[Mario Torre]

On Wed, 2013-04-24 at 10:38 -0700, Jasper Potts wrote:
> Tom, the prototype we did only had a interpreter on iOS which is doable for applications but takes very careful coding to get good performance. AOT would be a better solution if someone can get it working. I also found a interesting tale of a very fast ARM interpreter for Java written in assembler, see:
> 
> http://blogs.arm.com/software-enablement/182-how-do-you-make-java-fast-answer-go-down-the-pub/
> http://blogs.arm.com/software-enablement/187-how-do-you-make-java-fast-answer-go-down-the-pub-part-2/
> 
> This link has more up to date info:
> http://gbenson.net/?p=239
> 
> But looks like it has not been kept up to date.
> 
> Jasper

I didn't follow all the discussion, but both Zero and Shark are part of
OpenJDK already.

Also, it would be interesting to see what are the performances of JavaFX
with GCJ and GNU Classpath, since GCJ can do AOT (not sure if GNU
Classpath has enough class library code to run JavaFX, it will require
some work probably).

There are a bunch of commercial vendors that still distribute
GCJ/Classpath based implementations, and of course all the code is open
source with the same license of both OpenJFX and OpenJDK. This could be
a very good alternative for some use cases if you want a small but still
fast implementation [1], especially since OpenJFX is pretty standalone.

Then of course there are cacao and jamvm, those are very fast too, all
considered.

Cheers,
Mario

[1] GCJ has always had very good performances compared to similar
alternatives.