Re: PQHost() undefined behavior if connecting string contains bothhost and hostaddr types

I drifted to come here..

At Wed, 14 Mar 2018 11:17:35 +0900, Michael Paquier <michael@paquier.xyz> wrote in <20180314021735.GI1150@paquier.xyz>
> On Fri, Mar 09, 2018 at 04:42:30PM -0500, Peter Eisentraut wrote:
> > It seems, however, that PGhost() has always been broken for hostaddr
> > use.  In 9.6 (before the multiple-hosts stuff was introduced), when
> > connecting to "hostaddr=127.0.0.1", PGhost() returns "/tmp".  Urgh.
> > 
> > I think we should decide what PGhost() is supposed to mean when hostaddr
> > is in use, and then make a fix for that consistently across all versions.
> 
> Hm.  The only inconsistency I can see here is when "host" is not defined
> but "hostaddr" is, so why not make PQhost return the value of "hostaddr"
> in this case?

https://www.postgresql.org/message-id/19297.1493660213%40sss.pgh.pa.us
> And I believe we already considered and rejected the idea of having it
> return the hostaddr string, back in some of the older discussions.
> (We could revisit that decision, no doubt, but let's go back and see
> what the reasoning was first.)

But I haven't found where the decision made. I seems to be
related to commits commits 11003eb55 and 40cb21f70 or in older
discussion. I'm going to search for that for a while.

regards,

-- 
Kyotaro Horiguchi
NTT Open Source Software Center