Re: [PATCH 1/3] Introduce gdb::unique_ptr

[Pedro Alves <palves at redhat dot com>]

On 10/11/2016 12:16 PM, Metzger, Markus T wrote:

> Wow, that was a long reply to such a small question.  I was mainly
> wondering if it makes sense to write (and maintain) ones own version
> of a standard library feature.
> 
> The big step was not supporting C any longer.  Requiring C++11 looks
> small, by comparison.

Agreed, from language perspective.  The question is one of _compiler_ access
and convenience.  But it looks like I might have been wrong in my previous
perception that dropping support for older GCCs would be unrealistic at
this point.  I'd love to hear other's opinions.

> BTW, I noticed that maintainers seem very busy these days and patches
> are waiting unusually long for review.

Yeah.  Myself, I don't really know nowadays what it means to not be very busy, and
also getting the 7.12 release/branch ready took me significant effort.  Over the
past few releases, I've been considering whether an explicit "bugfixes/regressions
only" state, like gcc's stages would help -- because what happens is that
people send in patches for master and the pings and if they're not following
development closely, they won't realize the reason people are not
looking at their patches is the focus on the release, and everyone is
frustrated.  At least I am.

Thanks,
Pedro Alves