On Wednesday 17 March 2004 17:36, Tom Lane wrote:
> Bruce Momjian <pgman@candle.pha.pa.us> writes:
> > I was thinking of a GUC variable called PERFORMANCE_HINTS, which would
> > throw a message if a lookup from the primary to the foreign key didn't
> > have an index.
>
> I like the pg_advisor idea a lot better.
>
> In the first place, a lot of these sorts of checks don't have any clean
> place to insert as a test made in-passing in regular operation.
[snip]
> In the second place, you don't really want notices about bad schema
> design popping out during regular operation
[snip]
> In the third place, if we try to solve the problem by embedding checks
> here and there in the backend, we'll limit ourselves
[snip]
Fourthly - re-checking the entire schema when you have made changes to a
system is a good idea.
Fifthly - this is the sort of thing that goes into the "new features" list and
advocacy can talk about. People can write articles on it, all sorts.
Just my tuppence-worth.
-- Richard Huxton Archonet Ltd