Subject: Re: [PATCH] Opclass parameters
To: Nikolay Shaplov <dhyan@nataraj.su>, pgsql-hackers@lists.postgresql.org,
 obartunov@gmail.com
Cc: PostgreSQL mailing lists <pgsql-hackers@postgresql.org>
References: <d22c3a18-31c7-1879-fc11-4c1ce2f5e5af@postgrespro.ru>
 <15077864.qMPJiTemUy@x200m>
 <CAF4Au4xedPHJQyN60x+6giMpEyLdKn7jFR85_kzaApy9BptRUw@mail.gmail.com>
 <2803786.3sFNWAn8Pn@x200m>
From: Nikita Glukhov <n.gluhov@postgrespro.ru>
Organization: Postgres Professional
Message-ID: <6382cb29-83d8-90f2-1dbb-b4ac606cbf1f@postgrespro.ru>
Date: Sat, 3 Mar 2018 01:32:58 +0300
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101
 Thunderbird/52.6.0
MIME-Version: 1.0
In-Reply-To: <2803786.3sFNWAn8Pn@x200m>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 8bit
Content-Language: en-US
Precedence: bulk

On 02.03.2018 19:12, Nikolay Shaplov wrote:

> В письме от 1 марта 2018 23:02:20 пользователь Oleg Bartunov написал:
>>> 2. Your patch does not provide any example of your new tool usage. In my
>>> prototype patch I've shown the implementation of opclass options for
>>> intarray. May be you should do the same. (Use my example it will be more
>>> easy then do it from scratch). It will give more understanding of how
>>> this improvement can be used.
>> Hey, look on patches, there are many examples  !
> Oups...Sorry. I've looked at the patch from commitfest
> https://commitfest.postgresql.org/17/1559/ and it have shown only the first
> file. And When I read the letter I did not pay attention to attachments at
> all. So I was sure there is only one there.
>
> Yes. Now I see seven examples. But I think seven it is too many.
> For each case a reviewer should make consideration if this parameter worth
> moving to opclass options, or fixed definition in the C code is quite ok.
> Doing it for whole bunch, may make it messy. I think, it would be good to
> commit an implementation of opclass options, with a good example of usage. And
> then commit patches for all cases where these options can be used.

There are 5 examples for GiST opclasses, not 7, and they are almost
identical -- in all of them added 'siglen' parameter for signature length
specification.

> But since it is now "Rejected with feedback", let's wait till autumn.

We don't want to wait that long.  But now we only need to сome to an agreement
about CREATE INDEX syntax and where to store the opclass parameters.

-- 
Nikita Glukhov
Postgres Professional: http://www.postgrespro.com
The Russian Postgres Company