Re: [PATCHES] default database creation with initdb - Mailing list pgsql-hackers

From Andreas Pflug
Subject Re: [PATCHES] default database creation with initdb
Date
Msg-id 42B746A7.5000409@pse-consulting.de
Whole thread Raw
In response to Re: [PATCHES] default database creation with initdb  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
Tom Lane wrote:
> Andreas Pflug <pgadmin@pse-consulting.de> writes:
> 
>>Tom Lane wrote:
>>
>>>If it's a server-side failure it should have a SQLSTATE code.
> 
> 
>>Specifically, I'm talking about
> 
> 
>>"no pg_hba.conf entry for ....",
> 
> 
> ERRCODE_INVALID_AUTHORIZATION_SPECIFICATION
> 
> 
>>"Ident authentication failed.." (both server sice)
> 
> 
> Ditto.  Do you need to know the difference?  What exactly would client
> code do differently for these two cases?

Display different hints how to cope with this. The hint will arise for 
newbies, and it would be confusing to discuss ident auth when a missing 
pg_hba.conf line is the problem, and vice versa.

> 
>>and  "Is the server running on host ..." from libpq 
> 
> 
> libpq doesn't currently assign SQLSTATEs to internally detected errors
> ... someday someone should fix that.

Ok, I'm blind. Where in h**l do I find the SQLSTATE from a PGconn?!?

Regards,
Andreas


pgsql-hackers by date:

Previous
From: Alvaro Herrera
Date:
Subject: Re: pg_locks view versus prepared transactions
Next
From: "Victor Y. Yegorov"
Date:
Subject: make distclean keeps some files