Re: [PATCH v2] NEWS: mention more recent developments

On Tue, 21 Jul 2026 at 11:56, Andreas K. Huettel <[email protected]> wrote:
>
> Am Dienstag, 21. Juli 2026, 19:45:54 Japanische Normalzeit schrieb Jonathan Wakely:
> > On Tue, 21 Jul 2026 at 09:59, Florian Weimer <[email protected]> wrote:
> > >
> > > * Andreas K. Hüttel:
> > >
> > > > diff --git a/NEWS b/NEWS
> > > > index e1c1db284a..ca81dedc6a 100644
> > > > --- a/NEWS
> > > > +++ b/NEWS
> > > > @@ -27,10 +27,13 @@ Major new features:
> > > >    been imported from the CORE-MATH project, in particular cosh, sinh, and
> > > >    tanh.
> > > >
> > > > -* The SVID handling for cosh and sinh were moved to compat symbols, allowing
> > > > -  improvements in performance.
> > > > +* Many additional improvements to existing functions have been synchronized
> > > > +  from the CORE-MATH project.
> > > > +
> > > > +* assert is now supported as a variadic macro for C++26.
> > >
> > > What about this instead?
> > >
> > > * For C++26, the assert macro now accepts arguments that contain commas
> > >   which are not nested in parentheses.
> >
> > This could be read as suggesting it allows assert(a,b) which it
> > doesn't. (And formally, the commas still separate the macro arguments
> > as far as the preprocessor is concerned, so the arguments don't
> > "contain" the commas. That's super-pedantic though.)
> >
> > Maybe "now accepts expressions that contain commas, e.g. within
> > template argument lists, without requiring them to be enclosed in
> > parentheses"?
>
> This is both very nice and good, but doesnt really hit the purpose
> of NEWS - a brief, (for someone who programs for glibc) easily
> understandable message about a new feature.
>
>  * For C++26, the assert macro is now formally variadic, allowing
>    more complex arguments containing commas (which however still must
>    evaluate to a single value).
>
> Does that make sense?

Works for me. You could even drop "formally". It is variadic, there's
no doubt about that. There are restrictions on what you can use as
arguments to it, but that's true for many variadic macros.

>
> Or is it overkill and we should just drop the assert item in NEWS?

I think it's useful to mention, because this is a C++ library feature
that couldn't be fixed in libstdc++ and had to be done in glibc
instead. It's useful to make it easy to find which version of Glibc
adds support for the feature.


>
> >
> >
> > >
> > > >
> > > > -* New locale added: hrx_BR (Hunsrik language spoken in Brazil).
> > > > +* The SVID handling for cosh and sinh was moved to compat symbols, allowing
> > > > +  improvements in performance.
> > >
> > > Maybe: SVID [error] handling
> > >
> > > Thanks,
> > > Florian
> > >
> >
>
>
> --
> PD Dr. Andreas K. Hüttel
> [email protected]
> Gentoo Linux developer
> (council, comrel, toolchain, base-system, perl, libreoffice)
> https://wiki.gentoo.org/wiki/User:Dilfridge