Meta-comment: Until we have the discussion in Kona about what the scope is
for the next period, it's probably premature to publicly encourage Core
proposals. However, we should (and have) already encourage(d) Library
proposals and this applies to them too.
> Bjarne put together a document describing how to write up a
> WG21 proposal a number of years ago. I think that should probably be a
> good starting point for whatever is posted now.
It's my paper N1364:
http://www.open-std.org/jtc1/sc22/wg...2002/n1364.htm . Still
looks like the right content.
> However, the mechanics of submitting a paper are the absolute least of the
> problems facing someone who wants to submit an idea to the Committee.
> The real point that should be emphasized is that it's nearly imperative to
find
> a Committee member who will act as a "champion" of the idea,
Exactly. And this at *multiple* face-to-face meetings because there are
always rounds of feedback.
Note that LWG stated this explicitly in their minuted resolution, which
appeared also in my trip report
(
http://herbsutter.com/2011/08/24/trip-report-2011-08/).
> help put the
> paper into an appropriate form for consideration, present the paper at
> meetings, answer questions about it, receive comments and suggestions for
> future revisions, etc. (The alternative is for the proposer him/herself
to
> attend meetings and fulfill that role.) If they've made such a contact,
then
> the contact will know the mechanics of submitting a paper.
We can point to N1364, but we really don't want people just
firing-and-forgetting proposals via air mail. Even with the template, if
they're not planning to attend in person they should be working with a
committee member who cares and is actively helping them put the paper
together.
Administrative note: Our usual procedure in all subgroups is that if the
author of a paper is not present, and there's no one present willing to
present and champion the paper at that meeting, we won't spend time on it.
We will try to not lose track of such 'older papers not yet considered' --
for example, we do this as the main agenda point of our WG21 administrative
telecons, exactly to make sure things don't fall through the cracks -- but
unless and until there's an author or champion who can participate in the
discussion and answer questions and take feedback, it's a waste of WG21
subgroup time to consider the paper. Mailing in feature suggestions isn't
how we operate.
Herb
C++ 2011 : pourquoi pas les polynômes de Chebyshev ?
) de ne pas voir les polynômes de Chebyshev (très utilisés pour les méthodes d'interpolation et les méthodes spectrales) aux côtés de toutes les nouvelles fonctions (il y a bien Hermite/Laguerre/Legendre alors pourquoi pas Chebyshev ?).
Répondre avec citation
Envoyé par Kaluza
Partager