In addition to silencing an unnecessary fatal error when nmap-os-db is
not found but user has requested -6, this should make start times a bit
more efficient, since only 1 or the other database is loaded. Patch by
Alexandru Geana
Fixes#97
Increased rarity of TLSSessionReq from 2 to 7, since SSLSessionReq
should catch most things.
Moved some obvious SSL ports (989-995) to sslports directives.
Added a few SSL ports to the existing SSL* probes.
Previously, we would only request IPv4 addresses by default, or only
IPv6 when scanning with -6. Now, we'll request both (by not passing an
address family in the hints to getaddrinfo) and report them all in the
"Other addresses" line. This should encourage more users to scan with
-6. Additionally, it allows us to catch and report attempts to scan
IPv6-only hosts via IPv4, where previously we would just say "Failed to
resolve"
Closes#76
Using the macports version, which is up-to-date, instead of the outdated
version included in the 10.5 SDK. As we update the minimum supported
version of OS X, that would improve, but not by much: 10.9 is still on
0.9.8y
Libpcap only needs libnl for monitor-mode sniffing. Disabling it lets us
not worry about which -lnl or -lnl-3 linker argument to include in our
own LIBS, so Nmap builds properly no matter whether libnl is present or
not.
Mostly just eliminating concatenation-reassignments by chaining
concatenations, reordering assignments to allow better use of bin.pack,
and using tables to store intermediate results before concatenating
them. Used strbuf as a quick fix in dhcp.lua. Eliminated some unused
string variables in vulns.lua.
Unnecessary, probably, but this eliminates 2 of 7 string
creations/collections due to concatenation and reassignment. Also
eliminates 2 unneeded table lookups and tests (since host always has
host.ip and port always has port.number), and eliminates 1 extra test
for prerule and postrule scripts, since we can't have a port if we don't
have a host.
