Re: GCC_REQD/USE_LANGUAGES

[Greg Troxel <gdt%lexort.com@localhost>]

Alistair Crooks <agc%pkgsrc.org@localhost> writes:

> http://cvsweb.netbsd.org/bsdweb.cgi/pkgsrc/textproc/sentencepiece/Makefile.common.diff?r1=1.1&r2=1.2
>
> I don't see how losing information about the c++ standard used, and
> hardcoding a minimum GCC level could be "the proper way" - just seems
> wrong to me
>
> Can you revert and discuss, please?

As Alistair said, please revert, specifically including the change to
the guide.  That is two of us from PMC asking you to do that.  I am not
saying that you knew this was controversial when you made the edits, but
it has become obvious.

As a basic process matter, controversial undiscussed changes must be
reverted immediately when they become known to be controversial, and we
can can then discuss making from the position of them not being in the
tree.  This is crucial to avoid giving an advantage to those making
undiscussed changes, and not enforcing this encourages deviations from
our norms.

I will not be participating in the technical discussion further until
the revert has happened, and I ask everyone else to do the same.
Community norms are more important than speed.

Post revert, we can discuss, but with enough time for those who have
less day-to-day time to deal with pkgsrc to speak up.  That means at
least a full week (after the revert) and probably two.  We have been
living with compiler selection woes forever, so there is certainly no
emergency.