[std-interval] More on interval computations as proofs

Gabriel Dos Reis gdr at integrable-solutions.net
Fri Oct 6 15:50:04 PDT 2006


"R. Baker Kearfott" <rbk at louisiana.edu> writes:

| Certainly, it is desirable that sqr(x) return [0,1] for x=[-1,1]. 
| However, isn't the question of x*x returning [0,1] instead of [-1,1]
| a question of optimization, and of proper interpretation of what
| the programmer intended?

Certainly there is a syntactic difference.  The question as to whether
there is a semantics difference is not clear to me.  I'm not sure, it
should be left to the compiler/library writer to second guess what the
programmer intended.

In the current C++ standard, an implementation is free to yield 5 for 
2 + 2 and still be conforming -- the argument could be made that if
the programmer intended 4 he/she would have written so (yes, it is a
facetious argument, but you get the point).  I would be hard pressed
to say that it is just a question of optimization.

|  Also, have we distinguished what should
| be mandated in the standard and what is just desirable in an
| implementation?

>From the "standardese" point of view, you can only put requirements.

-- Gaby


More information about the Std-interval mailing list