Re: Trouble with some concepts - Mailing list pgsql-general

From Tom Lane
Subject Re: Trouble with some concepts
Date
Msg-id 20306.1063985674@sss.pgh.pa.us
Whole thread Raw
In response to Trouble with some concepts  (Francois Suter <dba@paragraf.ch>)
List pgsql-general
Francois Suter <dba@paragraf.ch> writes:
> Can someone explain the following:

> msgid "the target relation %u is nailed"

It's referring to a nailed-in-cache index, which is to say something
that we don't allow to drop out of the relcache (because we'd be unable
to load it back in, because it's used during relcache entry loading).

I think that this message is actually a can't-happen item because
of upstream checks, so there's probably no need to consider it
translatable.  I'll look at turning it back to a plain elog.

> msgid "improper qualified name (too many dotted names): %s"

> What's a qualified name?

Name with dots in it.  The complaint would come out for something like
a.b.c.d, which is syntactically illegal.

> msgid "database name may not be qualified"

As above.  A database name cannot look like a.b.

> Is an expressional index an index built on a calculated field (say two
> concatenated field or whatever) or something else?

Yes.  "Expressional index" isn't a really good term --- it's something
I invented on the spur of the moment to replace "functional index".
If you have a suggestion for a better term, I'm all ears.

            regards, tom lane

pgsql-general by date:

Previous
From: Richard Huxton
Date:
Subject: Re: Column defaults fail with rules on view
Next
From:
Date:
Subject: Re: Column defaults fail with rules on view