nqp broken -- missing get_pointer().

Andy Dougherty doughera at lafayette.edu
Thu Sep 22 19:17:54 UTC 2011


On Wed, 21 Sep 2011, Christoph Otto wrote:

> On Wednesday, September 21, 2011 5:16 PM, "Andy Dougherty"
> <doughera at lafayette.edu> wrote:
> > On Wed, 21 Sep 2011, Brian Gernhardt wrote:
> >
> > >
> > > On Sep 21, 2011, at 4:08 PM, Andy Dougherty wrote:
> > >
> > > > It appears that parrot/master no longer builds nqp/master.  Here
> > > > are the relevant error messages from an attempt to build nqp:
> > >
> > > parrot just merged the beginnings of a branch intended to clean up
> > > our default VTABLES.  This has the unfortunate effect of breaking
> > > NQP.  The needed fixes are contained in nqp/kill-useless-defaults.
> >
> > Oh.  I thought the plan for such incompatible changes was to make the
> > new way available, transition HLLs to use the new way, and then remove
> > the old way.  I thought the merge would wait till NQP was updated. I
> > guess not.
> >
> > --
> >     Andy Dougherty          doughera at lafayette.edu
> 
> There's an nqp pull request that'll get merged after Rakudo's release on
> Friday (afaiu), but we should have preserved backwards-compatibility for
> this, at least until the nqp change got merged.

Precisely.  The reason it mattered here is that I was trying to develop a 
fix for TT #2194, and was trying to check that my fix didn't break a HLL 
such as NQP.  Since I am patching parrot/master, that's what I wanted to 
use to test NQP.

-- 
    Andy Dougherty		doughera at lafayette.edu


More information about the parrot-dev mailing list