On Thu, Mar 13, 2003 at 10:29:44AM +0100, Jean-Marc Lasgouttes wrote: > Lars> Perfectly valid warning, it is only strange in this specific > Lars> case since 'a' is used and that cannot be wrong... 'å' on the > Lars> other hand... > > So, why does the compiler generate a warning, since it knows that the > value is correct?
I asked about this on the gcc list, then realised it probably makes sense: what if you compile lyx on a system where 'a' is > 128, and char signed ? (in the general sense of course). -Wno-char-subscripts is an option john
