Re:The future of java
by michaelneale on Wednesday April 22, 2009 @07:53PM
(#574808)
In some ways it has happened a little with smalltalk: Gemstones MagLev for ruby work demonstrates that it is possible.
Whilst for many things I like the VM "platform" idea, I still think of lower level compile-to-the-metal languages are more real to me, and always necessary for some things, but perhaps thats my age and experience, and in time VMs will be the norm and native compilation will be like asm code... who knows..
If you need this for jvm based languages you can always use excelsior jet. It's awesome.
Re:The future of java
by michaelneale on Thursday April 23, 2009 @07:13AM
(#575383)
years ago I used it (back 2002-ish?) and it worked pretty well back then. Have also used GCJ for some things - but it always seems like its just creating a little JVM hidden in the native code - still not quite the same "feeling" as macro assembler (which is how I saw C back when I did it) - still many layers above.
I know, its probably all in my head, but still, in my head the real stuff happens on the metal, not in abstractions (I am an elec engineer at heart, trained as one and will always be one).
How is it better than the CLR?
!troll, I'm genuinely curious.
I don't have any data with which to back this up, but as far as I can tell it's a little faster and it has better multithreading guarantees and performance. Oh, and it's open source (mostly. sigh.) and has some pretty cutting-edge languages, e.g. Clojure and Scala.
Mono is one to watch, though.
LLVM is interesting, but aims a bit below all of these. What I'd really like to see is some of the other language specific VMs try branching out. Lua, Io, Erlang, heck -- even Smalltalk, could all be viable candidates.