hv@crypt.org wrote: > > I note also that the -A switch handling does a forbid_setid("-A"). > Any particular reason why it does that? I'd have thought that in > general a script would be expected to run more safely with -A, so > that it should positively be encouraged. I imagined that's because you can pass code to be executed via -A to the regexp. E.g. via a nasty PERL5OPT.