> Generally speaking the compiler warnings you're talking about cause
> false alarms so often that they're more trouble than they're worth.
> For example, if P and Q are pointers into the same array and P <= Q,
> it's quite normal to compute Q-P and store the result into a size_t
> variable.
No question about this.
> Although I'm not sure why you selected the casts in question to
> attack first,
Because there were casts already.
> for casts to unsigned char how about the attached patch instead?
> The general idea is to avoid casts when possible, as they're too
> powerful in C. The characters in question are all ASCII so sign
> extension shouldn't matter.
Looks OK to me. I wished there were possibilities to improve compiler
warnings so that trivial cases like this one (and especially the
`size_t' issue you are mentioning above) don't cause a warning...
Werner
