Boszormenyi Zoltan <zb@cybertec.at> writes:
>> Did you think about something like the attached code?
> Or rather this one, which fixes a bug so fillPGconn() and PQconninfo() are
> symmetric and work for "requiressl".
That's incredibly ugly. I'm not sure where we should keep the "R"
information, but shoehorning it into the existing PQconninfoOption
struct like that seems totally unacceptable. Either we're willing to
break backwards compatibility on the Disp-Char strings, or we need
to put that info somewhere else.
regards, tom lane
