Re: MERGE SQL Statement for PG11

On 30 October 2017 at 19:17, Peter Geoghegan <pg(at)bowt(dot)ie> wrote:
> On Mon, Oct 30, 2017 at 11:07 AM, Simon Riggs <simon(at)2ndquadrant(dot)com> wrote:
>> Please explain in detail the MERGE SQL statements that you think will
>> be problematic and why.
>
> Your proposal is totally incomplete, so I can only surmise its
> behavior in certain cases, to make a guess at what the problems might
> be (see my remarks on EPQ, live lock, etc). This makes it impossible
> to do what you ask right now.

Impossible, huh. Henry Ford was right.

If there are challenges ahead, its reasonable to ask for test cases
for that now especially if you think you know what they already are.
Imagine we go forwards 2 months - if you dislike my patch when it
exists you will submit a test case showing the fault. Why not save us
all the trouble and describe that now? Test Driven Development.

> Besides, you haven't answered the question from my last e-mail
> ("What's wrong with that [set of MERGE semantics]?"), so why should I
> go to further trouble? You're just not constructively engaging with me
> at this point.

It's difficult to discuss anything with someone that refuses to
believe that there are acceptable ways around things. I believe there
are.

If you can calm down the rhetoric we can work together, but if you
continue to grandstand it makes it more difficult.

> We're going around in circles.

Not really. You've said some things and I'm waiting for further
details about the problems you've raised.

You've said its possible another way. Show that assertion is actually
true. We're all listening, me especially, for the technical details.

I've got a fair amount of work to do to get the rest of the patch in
shape, so take your time and make it a complete explanation.

My only goal is the MERGE feature in PG11. For me this is a
collaborative engineering challenge not a debate and definitely not an
argument. If you describe your plan of how to do this, I may be able
to follow it and include that design. If you don't, then it will be
difficult for me to include your thoughts. If you or others wish to
write something as well, I have already said that is OK too.

If anybody's goal is to block development or to wait for perfection to
exist at some unstated time in the future, than I disagree with those
thoughts.

--
Simon Riggs http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services