Re: [bug-gnulib] Re: base64

[Paul Eggert]

Simon Josefsson <address@hidden> writes:

> does your proof require that SIZE_MAX is a power of two?  (For the
> "we lost at least half of the correct value".)

C99 guarantees that SIZE_MAX is one less than a power of two, and
I don't know of any pre-C99 implementations where it wasn't.

One other fairly subtle point: his proof also relies on the fact that
either (A) INT_MAX < SIZE_MAX, or (B) signed integer arithmetic is
two's complement, without runtime overflow checking.  But you needn't
worry about this, as (B) is already listed as a gnulib assumption.