Re: Recording foreign key relationships for the system catalogs

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: pgsql-hackers(at)lists(dot)postgresql(dot)org
Cc: Peter Eisentraut <peter(dot)eisentraut(at)enterprisedb(dot)com>, "Joel Jacobson" <joel(at)compiler(dot)org>
Subject: Re: Recording foreign key relationships for the system catalogs
Date: 2021-02-02 23:26:15
Views: Raw Message | Whole Thread | Download mbox | Resend email
Lists: pgsql-hackers

I wrote:
> * It would now be possible to remove the PGNSP and PGUID kluges
> entirely in favor of plain BKI_LOOKUP references to pg_namespace
> and pg_authid. The catalog header usage would get a little
> more verbose: BKI_DEFAULT(PGNSP) becomes BKI_DEFAULT(pg_catalog)
> inclined to do it, simply to remove one bit of mechanism that has
> to be documented; but it's material for a separate patch perhaps.

Here's a patch for that part. I think this is probably a good
idea not only because it removes magic, but because now that we
have various predefined roles it's becoming more and more likely
that some of those will need to be cross-referenced in other
catalogs' initial data. With this change, nothing special
will be needed for that. Multiple built-in schemas also become
more feasible than they were.

regards, tom lane

Attachment Content-Type Size
retire-PGNSP-and-PGUID-symbols.patch text/x-diff 14.0 KB

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Andres Freund 2021-02-03 00:32:50 Re: LogwrtResult contended spinlock
Previous Message Alvaro Herrera 2021-02-02 23:19:19 Re: LogwrtResult contended spinlock