Why did you change the test for HAVE_UNIX_SOCKETS?
Kurt Roeckx writes:
> Here is an updated patch for my ipv6 related changes.
>
> Most important change is to get my cidr support in line with the
> patch Andrew Dunstan made.
Could you explain what your patch changes to the rest of us?
> - configure now checks inet_ntop() to check if there is ipv6
> support, where we don't even use that function anymore now.
> Maybe just checking for sockaddr_in6 will be enough?
Seems reasonable.
> - pg_hba.conf.sample has ipv6 addresses in them. If you don't
> have support for ipv6 it will give an error on that. Not sure
> what to do with that.
That depends on what we decide(d?) to do with the IPv6 support in the
first place. Some people have suggested to silently disable IPv6 if the
kernel isn't set up for it, others wanted an explicit switch.
pg_hba.conf should then behave similarly.
--
Peter Eisentraut peter_e@gmx.net