Since we're not using the message-ID for dupe checking or threading and we
already have a mechanism in place for body text dupe detection (via CRC-32)
the hash mechanism is providing an unnecessary burden on the mail database.

hash.

Was always forcing "No" as it was using a negated pointer.

exit_code.

enter a URL and go.

or in domains.cfg, connect to the Synchronet SMTP server... by IP if mail is
bound to a specific IP, or "127.0.0.1" if it is not.

If ANYONE can come up with a reason that email which gets to the BBS should
mot be handled by the BBS if the BBS is configured for it, let me know!

work since then uifc/ANSI and uifcx could be used internall by JS scripts
for a BBS session.

Ths should fix the mystery termination bug when ran from xterm.

to avoid collision with one declared in sbbs.h (and defined in d<mumble>.c)

tv_sec is not a time_t... use one for localtime_r()