Re: Unreviewed C++0x patches

[Mark Mitchell <mark at codesourcery dot com>]

Richard Guenther wrote:

> >  I think it's appropriate to fix bugs in new features in the branch.  Any
> >  release managers have an opinion?
>
> If you are 101% sure it only touches C++0x parts then I have no objection,
> but please be sure that these changes cannot cause C++ regressions
> in non-0x mode.

If we're not going to fix a long-standing bug in non-0x mode because 
it's not a regression -- even though it might affect relatively many 
people -- then fixing C++0x bugs (which presumably affect fewer, given 
how new it is), seems wrong.  We're introducing risk for relatively 
little upside.

This is one of these black-swan things; most of the time nothing bad 
will happen, making us think that nothing bad will ever happen, and then 
we'll manage to break something horribly and not notice until after we ship.

So, I think the bar ought to be very high.  In particular, I think we 
ought to consider this only for silent miscompilation, and only if all 
changes are isolated with C++0x conditionals.  And, if we're going to 
consider these patches, then I think we ought to consider fixes for 
other silent miscompilations as well, even if not regressions.

Of course, I've no objection to creating a C++0x branch, based off of 
the release branch, for interested C++0x users to use.

--
Mark Mitchell
CodeSourcery
mark@codesourcery.com
(650) 331-3385 x713