sso tradeoffs

[Paolo Carlini]

Paolo Carlini wrote:

>Why don't we renounce completely to sso for
>large sizeof(CharT) in favour of sizeof(string)? I mean, we keep only
>'\'0', of course.
>
I see now that sizeof(string) is actually lower bounded by
sizeof(long_rep). Still, probably swapping, f.i., can change in a
fruitful way, in some cases, it depends on the details of the
implementation.

Paolo.

P.S. I'm going to commit a tweak to sso_string that improves *a lot* the
performance of swap. I didn't consider that memcpy can be optimized much
better if the number of chars is known at compile time!