Re: Using C++11

[Murray Cumming <murrayc murrayc com>]

On Wed, 2015-07-01 at 19:44 +0200, Murray Cumming wrote:
> On Wed, 2015-07-01 at 09:18 +0200, Murray Cumming wrote:
> > On Tue, 2015-06-30 at 17:30 +0200, Murray Cumming wrote:
> > > > But can we synchronize an ABI break in gtkmm and friends with the
> > > ABI
> > > > break that distros introduce when they change the default value of
> > > > _GLIBCXX_USE_CXX11 from 0 to 1? That ABI break will probably occur
> > > at
> > > > different times in different distros, perhaps with different
> > > versions
> > > > of gtk+/gtkmm.
> > >
> > > It looks like it's going to happen at almost the same time for almost
> > > all distros. So it looks like now is the time.
> >
> > Overall, I think now is the time to break ABI. We won't have another
> > chance until GTK+ 4, which might never happen. I think we should combine
> > this with using and requiring C++11, to get that out of the way too. I
> > suspect that using C++11 would cause ABI breaks too, but plenty of
> > better-informed people doubt that, so I'm just being pessimistic.
> >
> > Does anyone object to the ABI break?
>
> Then again, if Ubuntu breaks ABI now (or if they do parallel installs),
> but doesn't use gtkmm 3.18 until the next Ubuntu version, the second ABI
> break will be our fault.

Does anyone object to us requiring C++11 as long as we don't break ABI?
There's a reasonable chance that this won't break ABI.

This might not be of much benefit to application coders, but it can make
our own *mm library code slightly nicer to work with.

-- 
Murray Cumming
murrayc murrayc com
www.murrayc.com