Re: GNU Classpath 0.14 released

[Chris Pickett]

Mark Wielaard wrote:
> On Thu, 2005-03-17 at 16:18 -0700, Tom Tromey wrote:
>
> > > > > > > "Etienne" == Etienne Gagnon <address@hidden> writes:
> >
> > Etienne> As usual, Mark has forgotten to include SableVM in the list
> > Etienne> of runtimes that use GNU Classpath.  It is understandable, I
> > Etienne> guess, as he does not use SableVM for development
>
>  But you just announced a
> new sablevm that used a forked sablevm-classpath CVS snapshot on our
> list an hour before and when I talked to you on irc about using sablevm
> against official GNU Classpath releases you said it didn't support this
> release out of the box and that there was no documentation for it.

I think that's fair justification for not including SableVM this time.

Since it was observed that this should have been 0.15 and not 0.14, how 
about this: give SableVM a chance to work with GNU Classpath 0.15 out of 
the box (i.e. with no extra command-line arguments, or perhaps with a 
simple argument like "--official-gnu-classpath"), and when it's ready, 
redo the announcement for 0.15.

Is that something that everyone can live with?  Or are there problems 
with my suggestion?  From what I can tell, it seems like a reasonable 
compromise between the major sticking points here, namely that Etienne 
wants SableVM to keep up to date with Classpath CVS snapshots and that 
Mark wants to encourage VM's to support official Classpath releases 
out-of-the-box and moreover to document that support.

I think that's a fairly minor amount of work for all parties (especially 
since SableVM 1.11 already works with 0.14(== 0.15) if an extra argument 
is specified and a Classpath release is installed somewhere), and if 
it's something that will improve SableVM--Classpath relations, then it 
also seems to be worth pursuing.

Cheers,
Chris