Tom Lane wrote:
> It seems rather odd and confusing that the sort order takes the second
> column as the major key.
Maybe I'm a rather odd and confused kind of guy ;-)
> I think that the sort ordering should be
> column 1 then 2. That means either putting the schema column first,
> or keeping this column ordering and sorting on relname before schema.
> I lean towards the first choice but can see a case for the second.
Here's another small patch, to implement Tom's first choice above. I
looked at re-adding your own "pg_temp_n" schema back into the results.
But I don't see a way to determine your own BackendId unless we add a
builtin function to expose MyBackendId as a user callable function.
Should we do this, or did I just miss something?
If there are no objections, please apply.
In response to
pgsql-patches by date
|Next:||From: Tom Lane||Date: 2002-07-20 21:57:33|
|Subject: Re: show() function - updated patch |
|Previous:||From: Tom Lane||Date: 2002-07-20 21:51:13|
|Subject: Re: Demo patch for DROP COLUMN |