On Wed, 2011-01-12 at 13:17 +0000, Stefano Stabellini wrote:
> On Wed, 12 Jan 2011, Gianni Tedesco wrote:
> > On Tue, 2011-01-11 at 19:26 +0000, Ian Jackson wrote:
> > > Gianni Tedesco writes ("[Xen-devel] [PATCH 07 of 10] pyxl: Recursively
> > > scan type-tree to produce complete binding boilerplate"):
> > > > pyxl: Recursively scan type-tree to produce complete binding boilerplate
> > >
> > > I applied patches 1-6. They touch only pyxl and didn't break the
> > > build.
> > >
> > > However patch 7 contains this:
> > >
> > > > --- a/tools/libxl/libxl.idl Tue Jan 11 16:01:07 2011 +0000
> > > > +++ b/tools/libxl/libxl.idl Tue Jan 11 16:01:08 2011 +0000
> > > ...
> > > > - ("features", string, True),
> > > > + ("features", string),
> > >
> > > This makes a substantive change to both the interface and the
> > > generated code, and wasn't mentioned in the comment. I'm not
> > > convinced it's a change that's appropriate during feature freeze.
> > We discussed this off-list at the time I was writing it. I think with
> > one of the other guys. We decided that const char * in the interface was
> > a bad idea since it imposed an allocation policy where no reason for it
> > actually existed in the code. I could have worked around this by casting
> > everything in the python bindings but we felt that was an ugly solution
> > and not a good precedent to set.
> I might have verbally misled Gianni to think that a change like this
> would be appropriate.
> Bu thinking twice about it, even though it appears correct and it is
> just one line, it doesn't meet the criteria.
Well, this would have been back in November/December time, maybe
earlier. I had even forgot all about it. It may have been appropriate
then but not at this late stage.
Xen-devel mailing list