From: Claus Spitzer <DocBoobenstein@...>
Date: 2004-06-11T13:14:19+09:00
Subject: Re: Introducing myself and my interest in ruby

It is difficult already to implement reification in compiled code, and
it is even more difficult to implement intercession (having just gone
through a myriad of papers covering all sorts of aspects related to
this I've gained an idea on just what a monumental job that is).
However, if (and that's a BIG IF) one's code can work without needing
such features then it makes it much more easier to be compiled. Now
the question is: how much are we willing to sacrifice to be able to
compile the code?

On Fri, 11 Jun 2004 13:00:57 +0900, Nicholas Van Weerdenburg
<nick@activehitconsulting.com> wrote:
> 
> A compiler could attempt to translate evals into other constructs. I
> think most are simple enought to do that.
> 
> Open classes and self-modifying code are another problem. Conceivable, a
> ruby class could be be redefined a 1000 times during a programs
> execution. Still, most cases are probably addressable by the compiler.
> 
> 5 mb is a big hit- maybe a special minimal gap-interpreter could be
> included that handles some of the weirder not-easily compilable cases?
> 
> Nick
> 
> 
> Sean Russell wrote:
> 
> >Hal Fulton wrote:
> >
> >
> >
> >>I like the way you're thinking. I believe the principal reason a Ruby
> >>compiler (or C translator) doesn't exist is simply that no one has
> >>written one.
> >>
> >>
> >
> >There are a few severe roadblocks to implementing this, the biggest one
> >being eval().  To support eval(), an executable needs access to the entire
> >Ruby interpreter.  This interpreter can be in the form of a dynamic
> >library, but then there isn't much difference between installing Ruby and
> >installing the interpreter library.  Or, the interpreter can be bundled
> >with the executable, which adds at least 5MB to the size of each executable
> >and still doesn't address the various library dependancies that can't be
> >discovered because of code like:
> >
> >        class Foo
> >                def initialize
> >                        eval( "require 'gtk'" )
> >                end
> >        end
> >
> >eval() is just one issue that confronts any attempt at creating an
> >application that converts Ruby to (C|Java|Native code|whatever), and I'm
> >sure there are others.
> >
> >Which makes me think that a great thing to have would be a document that
> >identifies all of these cross-implementation difficulties.  Maybe we could
> >come up with a "portable Ruby" specification -- a minimal description of
> >the set of Ruby features that can be implemented without having access to a
> >Ruby interpreter.
> >
> >
> >
> 
>