Re: PQinitSSL broken in some use casesf

On Fri, Feb 13, 2009 at 3:06 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Andrew Chernow <ac@esilo.com> writes:
>> Maybe the argument to PQinitSSLExtended should be a bit mask, making
>> this version more extendable ... PG_INITSSL, PG_INITCRYPTO?
>
> +1 for thinking ahead to the next time, but is a bit mask the right thing?

What would you suggest?  struct  pointer? varargs?

If you are down with the struct pointer idea, maybe we could revive
the idea of PQinit:
http://archives.postgresql.org/pgsql-hackers/2009-01/msg01349.php

We could fix up the windows issue at the same time, and leave room for
other things.  Fixing windows WSA issues or installing an event proc
hook into the library :D.

merlin