Re: Postgresql Materialized views

Sean Utt wrote:
>
> As Joshua Drake has pointed out before, most of the core people
> working on PostgreSQL don't actually use it for anything themselves. I
> will expand a little on that and say that this means that while they
> are extremely good at what they do, they really don't have a clue what
> might be useful to someone "in the wild". Sort of like automotive
> engineers who in the 1970's made the Cadillac's engine so large that
> you couldn't change the spark plugs without taking the motor mounts
> loose and lifting the engine.
>
>

This is both gratuitously offensive and based on a demonstrably false
premise. The definition of "core people working on PostgreSQL" is
somewhat vague. But if you were to take it as, say, the group of active
committers, then I would say that the majority of us earn our living in
whole or in part using PostgreSQL. Certainly I do (there's a reason I
use an elephant logo for my business).

Many of the things I have contributed to PostgreSQL have been in
response to user requests, and often things I personally have no
immediate need for. The same is true of many contributors. And almost
all of mine have been aimed at increasing usability (e.g. dollar
quoting, CSV import/export). So I rather take offense at the quite
inappropriate spark plug analogy.

As for feature requests, everyone has a right to air an opinion.
Personally, I will pay more attention to people who contribute to the
community than to those who don't. That doesn't just mean coders, though
- contribution comes in many forms. In the case of the present request,
Jean-Michel is a contributor, and any suggestion to the contrary is
right out of line.

The real question for me is whether we want to support Materialized
Views at the grammar level. If we do then it should be worked on. If not
then it can probably be dealt with via a bolt-on module. Personally I'm
inclined to say we should support it via the grammar.

cheers

andrew