[std-interval] More on interval computations as proofs

[Gabriel Dos Reis]

Hervé Brönnimann <hbr at poly.edu> writes:

[...]

| Now about the standardization aspect of this, I'm now quite familiar
| how errno is mentioned in the current revision of the C++ standard,
| so I looked it up, and it is how I thought:  only by reference to
| <cerrno> and <errno.h> from the C standard.

Precisely.  

And I don't think the lack of reference to errno is just a consequence
of lack of concurrency.

For example, you'll notice that the specification of std::complex never
ever talks about errno.  I believe you should really think about it.
There is no consistency to be consistent with here.

>From the standard process point of view, I would recommend that you
put a discussion paragraph in the proposal, discussing each
alternative (pros and cons.)  That is the usual way it is done when
we face many choices, none of which obviously stands out as clear winner.
Then pick your choice and explain why.  The discussion paragraph is
important for the committee and other experts to form opinion.

-- Gaby