Re: Options OUTPUT_PLUGIN_* controlling format are confusing (Was: Misleading error message in logical decoding)

Поиск
Список
Период
Сортировка
От Andres Freund
Тема Re: Options OUTPUT_PLUGIN_* controlling format are confusing (Was: Misleading error message in logical decoding)
Дата
Msg-id 20140929130308.GB25117@alap3.anarazel.de
обсуждение исходный текст
Ответ на Re: Options OUTPUT_PLUGIN_* controlling format are confusing (Was: Misleading error message in logical decoding)  (Michael Paquier <michael.paquier@gmail.com>)
Ответы Re: Options OUTPUT_PLUGIN_* controlling format are confusing (Was: Misleading error message in logical decoding)  (Robert Haas <robertmhaas@gmail.com>)
Re: Options OUTPUT_PLUGIN_* controlling format are confusing (Was: Misleading error message in logical decoding)  (Michael Paquier <michael.paquier@gmail.com>)
Список pgsql-hackers
On 2014-09-29 21:48:08 +0900, Michael Paquier wrote:
> On Mon, Sep 29, 2014 at 7:50 PM, Andres Freund <andres@2ndquadrant.com>
> wrote:
> 
> > What about the attached patch then?
> >
> 
> Thanks for this update. This looks good. Here are a couple of small
> comments:
> 1) This sentence is correct English, but I don't recall seeing in the docs
> such a formulation:
> +        Can only be used on slots using a output plugin supporting textual
> +        output.
> I'd rather rewrite 'It can only be used' or 'this function can only be
> used'. I imagine that you could add a reference to
> logicaldecoding-output-mode as well.

Hm.

> 2) s/a output/an output/g

> 3) The formulation here seems vague as a plugin that generates textual
> output can call pg_logical_slot_peek_binary_changes as well:

I've commented on this before: an output plugin doesn't call
pg_logical_slot_peek_binary_changes - if at all it's the other way
round.

> -        except that changes are returned as <type>bytea</type>.
> +        except that changes are returned as <type>bytea</type> and that it
> can
> +        be used on slots using output plugins that only support binary
> output.

Imo that's pretty much implied because it references the !binary
version. But I guess it doesn't hurt to be explicit. How about:
" ... on output plugins using any form of output, including binary."?

Greetings,

Andres Freund

-- Andres Freund                       http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, Training &
Services



В списке pgsql-hackers по дате отправления:

Предыдущее
От: Michael Paquier
Дата:
Сообщение: Re: pg_receivexlog and replication slots
Следующее
От: Heikki Linnakangas
Дата:
Сообщение: Re: pgcrypto: PGP armor headers