PDS Aftermath. Which way NQP?

[Lucian Branescu]

On Tuesday, 1 February 2011 at 22:58, Will Coleda wrote: 
> On Tue, Feb 1, 2011 at 11:36 AM, chromatic <chromatic at wgz.org> wrote:
> > On Tuesday 01 February 2011 at 03:05, Vasily Chekalkin wrote:
> > 
> > > Strong -1 from me. "Parrot" developers doesn't have resources to
> > > maintain own very-similar-but-slightly-different version of NQP.
> > 
> > I did a bit of research on RPython, but it's not standardized either. Are
> > there any other small languages suitable for writing compilers? The relevant
> > criteria are:
> > 
> > * a published standard
> > * more pleasant to write than PIR
> > * a history of productive use
There is an unfortunate misconception about RPython. It's not a standalone language, it's just an implementation detail of the PyPy translation framework. PyPy is a framework for translating interpreters written in RPython to VMs with automatic addition of GC, JIT, object space, etc.

So if one were to write an interpreter for Parrot bytecode with PyPy (in RPython), it would get translated to a VM with free GC and JIT. But RPython isn't a system language.


Why not use Winxed or Close or similar? 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.parrot.org/pipermail/parrot-dev/attachments/20110202/d97426dc/attachment-0001.html>