On Wed, Apr 20, 2016 at 02:03:16PM +0900, Michael Paquier wrote:
> On Wed, Apr 20, 2016 at 1:39 PM, Noah Misch <noah@leadboat.com> wrote:
> > On Tue, Apr 19, 2016 at 02:42:24AM -0300, Alvaro Herrera wrote:
> >> This thread seems to have stalled. I thought we were going to consider
> >> these patches for 9.6.
> >
> > Committers have given this thread's patches a generous level of consideration.
> > At this point, if $you wouldn't back-patch them to at least 9.5, they don't
> > belong in 9.6. However, a back-patch to 9.3 does seem fair, assuming the
> > final patch looks anything like the current proposals.
>
> Er, the change is rather located and fully controlled by _MSC_VER >=
> 1900, so this represents no risk for existing compilations on Windows,
> don't you agree?
Yes. That is why I said a back-patch to 9.3 seems fair.
> >> Should we simply push them to see what the
> >> buildfarm thinks?
> >
> > No. The thread has been getting suitable test reports for a few weeks now.
> > If it were not, better to make the enhancement wait as long as necessary than
> > to use the buildfarm that way. Buildfarm results wouldn't even be pertinent;
> > they would merely tell us whether the patch broke non-VS 2015 compilers.
>
> Well, they could push them, the results won't really matter and
> existing machines won't be impacted, as no buildfarm machine is using
> _MSC_VER >= 1900 as of now. Petr has one ready though as mentioned
> upthread.
Here you've presented two additional good reasons to not "simply push them to
see what the buildfarm thinks."