From: Tom Lane <tgl@sss.pgh.pa.us>
To: Heikki Linnakangas <hlinnaka@iki.fi>
cc: Alvaro Herrera <alvherre@alvh.no-ip.org>,
        pgsql-hackers <pgsql-hackers@postgresql.org>
Subject: Re: Removing support for COPY FROM STDIN in protocol version 2
In-reply-to: <f69a7e6b-66f6-cbab-a414-ecc5c2127208@iki.fi>
References: <20210203160958.GA2341@alvherre.pgsql>
 <3921698.1612369777@sss.pgh.pa.us>
 <f69a7e6b-66f6-cbab-a414-ecc5c2127208@iki.fi>
Comments: In-reply-to Heikki Linnakangas <hlinnaka@iki.fi>
	message dated "Wed, 03 Feb 2021 18:47:05 +0200"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <3955611.1612374792.1@sss.pgh.pa.us>
Date: Wed, 03 Feb 2021 12:53:12 -0500
Message-ID: <3955612.1612374792@sss.pgh.pa.us>
Precedence: bulk

Heikki Linnakangas <hlinnaka@iki.fi> writes:
> Since we're on a removal-spree, it'd also be nice to get rid of the 
> "fast-path" function call interface, PQfn(). However, libpq is using it 
> internally in the lo_*() functions, so if we remove it from the server, 
> lo_*() will stop working with old libpq versions. It would be good to 
> change those functions now to use PQexecParams() instead, so that we 
> could remove the fast-path server support in the future.

I'm disinclined to touch that.  It is considered part of protocol v3,
and there is no very good reason to suppose that nothing but libpq
is using it.  Besides, what would it really save?  fastpath.c has
not been a source of maintenance problems.

			regards, tom lane