Message-ID: <d9f99d9254ff638f82e1e7c91eaca039e9f9df68.camel@j-davis.com>
Subject: Re: range_agg
From: Jeff Davis <pgsql@j-davis.com>
To: Paul A Jungwirth <pj@illuminatedcomputing.com>
Cc: Pgsql Hackers <pgsql-hackers@postgresql.org>
Date: Sat, 06 Jul 2019 12:13:08 -0700
In-Reply-To: 
 <CA+renyX7jsWnyry6SOn-WSgdgQLtyFJ7SLWBVs+oatwQsMzvog@mail.gmail.com>
References: <16d71dc8-34cf-5ebd-1ce5-ccd93c0a14f9@illuminatedcomputing.com>
	 <ba323e489b98029be5ba2d5c5cbc99795f333f48.camel@j-davis.com>
	 <CA+renyX7jsWnyry6SOn-WSgdgQLtyFJ7SLWBVs+oatwQsMzvog@mail.gmail.com>
Content-Type: text/plain; charset="UTF-8"
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Precedence: bulk

On Fri, 2019-07-05 at 09:58 -0700, Paul A Jungwirth wrote:
> user-defined range types. So how about I start on it and see how it
> goes? I expect I can follow the existing code for range types pretty
> closely, so maybe it won't be too hard.

That would be great to at least take a look. If it starts to look like
a bad idea, then we can re-evaluate and see if it's better to just
return arrays.

The "weighted" version of the aggregate might be interesting... what
would it return exactly? An array of (range, weight) pairs, or an array
of ranges and an array of weights, or a multirange and an array of
weights?

> Another option would be to rename my function range_array_agg (or
> something) so that we are leaving space for a multirange function in
> the future. I don't love this idea myself but it would could a Plan
> B.
> What do you think of that?

Not excited about that either.

Regards,
	Jeff Davis