M. Edward (Ed) Borasky wrote:
> That's the JVM on x86 (32?) under Windows, Linux and SPARC/Solaris. How 
> about PowerPC Macs? Intel Macs? AMD64? Alpha/Tru64?

I develop on OS X on Intel and in Linux on AMD64. There are JVMs for all 
the platforms you list, most based on Sun's own with additional 
platform-specific enhancements. There are free JVMs for just about every 
platform available these days, and a whole raft of commercial VMs for 
the remainder. The JVM is everywhere, which means that JRuby could be in 
every one of those places with a solid VM under it.

> I'm glad to hear that. As I noted in another post, however, many more 
> people in the corporate world get paid to work *with* open source 
> software than get paid to work *on* open source software. For every 
> lucky Charles Oliver Nutter or John Lam, there are hundreds of people 
> like me who can only contribute in off hours and to things not related 
> to our employment.

And we will always appreciate you taking some of your precious off-hours 
time to help JRuby :) Of course at Sun we're all working on open-source 
software...and there may be more companies turning this direction in the 
future. You know how it is...if you can't change your job, change your job.

> Well ... there are some corporations who haven't taken a major stake in 
> Ruby like Sun and Microsoft have.

But those two moves have certainly made others in the industry take 
notice. I hope more corps will make an investment soon.

> I agree there's a lot of potential in sticking with jRuby for Sun. That 
> potential needs to be converted to profit, and that can only be done by 
> being competitive in the marketplace -- by satisficing rather than 
> optimizing.

This is actually Tom's point, but all projects have to satisfice or they 
never release. Ruby 2.0 so far is not satisficing, and is not going to 
be released anytime soon. Parrot is not satisficing, and it's unclear 
whether it will ever be done.

The other half of this is that JRuby isn't just going to be driven by 
the two Sun employees that work on it; it will be driven by its 
community, hopefully by Ruby converts that want another VM option, and 
by more big players in the future. If both Tom and I were hit by a bus, 
JRuby would go on.

> Well ... my choice of the Matz/Koichi line has more to do with my lack 
> of knowledge of the JVM and CLR than it does with the nature of how they 
> are funded. But judging by the howls in the Linux community that went up 
> over the Microsoft/Novell deal, the persistent whining around Sun's slow 
> pace at opening up Java technology, I think the Matz / Koichi path is 
> more likely to be more peaceful, in addition to being more interesting 
> in the computer science sense.

Perhaps; but it's also affected by the limitations of the current 
interpreter. Ruby 1.9/2.0 has so far been hindered by the desire to keep 
C extensions working, which prevents improvements to the memory model, 
GC, core API, method dispatch logic...etc. YARV mainly just swaps out 
the AST-walking interpreter--providing a big interpretation boost--but 
not much else will change.

Those other pieces will get attention in time, of course, but it's not 
clear now on what timeline we'll see them.

-- 
Charles Oliver Nutter, JRuby Core Developer
Blogging on Ruby and Java @ headius.blogspot.com
Help spec out Ruby today! @ www.headius.com/rubyspec
headius / headius.com -- charles.nutter / sun.com