"Nelson, Clark" <clark.nelson_at_[hidden]> writes:

| > | So how do you think we should respond to Matt's feedback?
| >
| > Agree with Matt that in less than 10 years most of these macros
| > will
| > pointless. I think implementations are progressing faster that,
| > what we
| > had at comparable stage for C++98.
|
| What I would like to know is what you think we should do this week. Should
| we not recommend any macro for these constexpr functions?

My instance on having few macros isn't just specific to constexpr.

For 'constexpr' in particular, I do believe it is still too early. LWG
doesn't even know what it wants. Having a new macro everytime that LWG
changes its mind sound overkill. If the desire is to have a macro
anyway, then I will repeat the suggestion I made previously: Have just one
macro for constexpr, call it something, and have its value updated for
new released of the committee draft.

-- Gaby