This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: C++ PATCH: PR 20599 (1/3)
- From: Benjamin Kosnik <bkoz at redhat dot com>
- To: Paolo Carlini <pcarlini at suse dot de>
- Cc: Mark Mitchell <mark at codesourcery dot com>, Douglas Gregor <doug dot gregor at gmail dot com>, gcc-patches at gcc dot gnu dot org, libstdc++ <libstdc++ at gcc dot gnu dot org>, jason at redhat dot com
- Date: Mon, 2 Oct 2006 14:01:38 +0200
- Subject: Re: C++ PATCH: PR 20599 (1/3)
- References: <450A658C.7050109@suse.de> <450F024C.9030904@codesourcery.com> <24b520d20609181627o651f77d9l7d8832ec5c7038a7@mail.gmail.com> <4rdvg2horcsuasdvh7ipmjjkndh7eu0f1a@4ax.com> <24b520d20609190734r7870eb67ob2cb3a0b28e08fe4@mail.gmail.com> <24b520d20609190743h6a7197ecrcf81eac973ad6531@mail.gmail.com> <5t50h2lpfc1ml861i9lvn6kb67apk8term@4ax.com> <m3odtbdbk6.fsf@zeus.integrable-solutions.net> <20060919173834.GR31210@synopsys.com> <4rb0h2p88dgivc267gbqccketoqjaf8o64@4ax.com> <m3u033br7h.fsf@zeus.integrable-solutions.net> <4511B690.3010806@codesourcery.com> <4516CCFB.8070903@codesourcery.com> <0535BDD0-FEEB-4C4D-823D-2EC7FC9BC2B6@osl.iu.edu> <20060928163525.8f9ac452.bkoz@redhat.com> <45202D43.10503@codesourcery.com> <20061002122100.e13b6bad.bkoz@redhat.com> <4520EBB0.1040903@suse.de>
> I only want to restate that the initial issue, variadic templates,
> remains unresolved: the library and GCC as a whole has a lot to gain
> from the availability of such feature (see the "competitive advantage"
> point often raised). But I understand the general policy to which we are
> converging, I cannot disagree.
I think the battle lines are pretty clearly drawn here: the library
people want this, the front end people are much more hesitant.
Traditionally, Jason has been the ally of extensions such as these: his
continued silence puzzles me.
It's pretty clear that you, me, Doug, and others are all for this. The
practicality is quite apparent.
> I was thinking that maybe we can be optimistic: even if variadic
> templates will not appear any time soon in a released GCC, the
> availability of a complete implementation + testcases +
> re-implementation of TR1 facilities should be fine to reassure the C++
> Committee that the feature is implementable and works well in practice!
You would certainly think this would be enough. I think Doug's gone
beyond the call of duty on this.
Certainly, I long for the day when every other C++0x proposal was so
diligently worked over....
> If that is the case, we can hope that the standardization process will
> be fast and also hope that in the window of time when the feature will
> be in draft, received favorably by the committee, but still subject to
> minor changes it will be allowed to go in GCC rather soon, maybe a bit
> sooner than per the general rules...
I would hope that by the time core decides to work on fine tuning it,
an implementation would be allowed in g++. I consider draft language in
core to be "in the WP."
Quite honestly, I believe actual implementations in wide use will
improve the quality of the final spec.
-benjamin