Re: [HACKERS] DOC: catalog.sgml - Mailing list pgsql-patches

From Zdenek Kotala
Subject Re: [HACKERS] DOC: catalog.sgml
Date
Msg-id 4523F5FC.6010407@sun.com
Whole thread Raw
In response to Re: [HACKERS] DOC: catalog.sgml  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-patches
Tom Lane wrote:
> Zdenek Kotala <Zdenek.Kotala@Sun.COM> writes:
>> There is new version of catalogs overview patch. This version add only
>> one column into overview table which contains Oid/Filename for each
>> catalog table. Oid information is important if someone need make
>> relation with filename on disk and related catalog table.
>
> I still say this is just confusing clutter.  The proposed patch even
> goes so far as to give the OID pride of place as the most important
> item you could possibly want to know about a catalog, which is surely
> silly.

You have right that OID is not important information in many cases, but
how I said It is useful when you want know relation between filename and
catalog table.

> People who actually want to know this information can look into the
> pg_class catalog, which has the advantages of being complete (eg, it
> covers indexes too), guaranteed up-to-date, and easily program-readable.
> I really do not see the value of putting it in the sgml docs.

You can look into pg_class catalog only if database is running. If you
have some data corruption problem, OID should help during recovery. But
you have right, that pg_class have complex information and who want to
"play" with datafiles, he must know more than OID.

    OK, thanks for response, forget to this patch

        Zdenek



pgsql-patches by date:

Previous
From: Tom Lane
Date:
Subject: Re: qsort_arg needed for msvc build
Next
From: Zdenek Kotala
Date:
Subject: Updated version of FAQ_Solaris