Breaking VM integration interface

[Mark Wielaard]

Hi,

On Wed, Nov 07, 2001 at 06:37:31PM -0700, Tom Tromey wrote:
> >>>>> "Brian" == Brian Jones <address@hidden> writes:
> 
> Brian> This is okay, I will concentrate on getting VMs working that
> Brian> use Classpath and testing the library.  Anyone with the time is
> Brian> encouraged to do the same.
> 
> I'd like to see Bryce's VMClassLoader change go in before the next
> release.  This requires a small change to any VM that uses Classpath.
> Basically his patch changes VMClassLoader.getPrimitiveClass() to take
> a `char' argument instead of a String.

I agree that this change is good but it does break our VM interface.
What promises do we give on the stability of that interface? Currently
our releases are so far apart and include so much changes that our users
cannot reasonably expect the VM interface to stay the same but it would
be nice to give some guarantees in the future. How do people feel about
the current state of the VM integration hooks?

Note that getPrimitiveClass() is not currently documented in our VM
Integration Guide. So I will make sure to document and integrate the
VMClassLoader.getPrimitiveClass() change tomorrow and try to make the
changes to the KissMe VM to make sure it actually works.

Cheers,

Mark
-- 
Stuff to read:
    <http://www.toad.com/gnu/whatswrong.html>
  What's Wrong with Copy Protection, by John Gilmore