Re: ALTER TABLE ... REPLACE WITH

On Wed, 2011-01-19 at 22:16 -0500, Robert Haas wrote:

> That's another way of saying "the patch is not anywhere close to being done".

My patch is materially incomplete. Certainly we may see that as grounds
for rejection, which I would not and could not argue with. It is a
popular feature, so I submitted anyway.

When I said Noah's patch was trivial, I was referring to the amount of
work expended on it so far; no insult intended. I think the amount of
code to finish either is fairly low as well.

If we wish to continue in this release then we must decide how. What I
was trying to indicate in my earlier comments was that my focus is on
achieving the required functionality in this release, or put another
way, I would accept Noah's patch rather than end with nothing.

The main requirement, as I see it, is error checking. We need to do the
same checking however we do it; neither patch currently does it.

If Noah's patch had error checking, then it would at least be safe to
recommend people do that. Then it is a simple matter of whether we think
implicit is OK, or whether it needs an explicit command. My patch does
it explicitly because that was the consensus from the earlier
discussion; I am in favour of the explicit route which is why I wrote
the patch that way, not because I wrote it that way.

-- Simon Riggs           http://www.2ndQuadrant.com/books/PostgreSQL Development, 24x7 Support, Training and Services