Re: Parsing of pg_hba.conf and authenticationinconsistencies - Mailing list pgsql-hackers

From Tom Lane
Subject Re: Parsing of pg_hba.conf and authenticationinconsistencies
Date
Msg-id 15576.1218041994@sss.pgh.pa.us
Whole thread Raw
In response to Re: Parsing of pg_hba.conf and authenticationinconsistencies  (Peter Eisentraut <peter_e@gmx.net>)
List pgsql-hackers
Peter Eisentraut <peter_e@gmx.net> writes:
> Am Tuesday, 5. August 2008 schrieb korry:
>> Perhaps the --check-config option should take an (optional) file name?
>> That would allow you to validate a config file without having to copy
>> it into place first.
>> 
>> postgres --check-config=myFilenameGoesHere -D $PGDATA

> There is already an elaborate system to tell where configuration files are 
> located.  So just doing postgres --check-config -D myFilenameGoesHere should 
> do the job.

Even more to the point: in the presence of include directives with
relative file paths, the idea of checking a file that's not in its
intended place is broken to begin with.  Both pg_hba.conf and
postgresql.conf have include-type syntax, though the former doesn't
call it that.
        regards, tom lane


pgsql-hackers by date:

Previous
From: "Hiroshi Saito"
Date:
Subject: Re: unable to build libpq on Win 2003 (32 bit)
Next
From: "Merlin Moncure"
Date:
Subject: Re: plan invalidation vs stored procedures