MIME-Version: 1.0
In-Reply-To: <993142.1742484654@sss.pgh.pa.us>
References: <E1tvHdt-0009Bj-1j@gemulon.postgresql.org>
 <993142.1742484654@sss.pgh.pa.us>
From: "David G. Johnston" <david.g.johnston@gmail.com>
Date: Thu, 20 Mar 2025 08:39:53 -0700
Message-ID: 
 <CAKFQuwbA6-fXn6w5mS_k5_UOS67+0ogt_FyjpbXFnQaAPm98bA@mail.gmail.com>
Subject: Re: pgsql: Add vacuum_truncate configuration parameter.
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Nathan Bossart <nathan@postgresql.org>,
	"pgsql-committers@lists.postgresql.org"
 <pgsql-committers@lists.postgresql.org>
Content-Type: multipart/alternative; boundary="0000000000000699220630c7f36f"
Archived-At: 
 <https://www.postgresql.org/message-id/CAKFQuwbA6-fXn6w5mS_k5_UOS67%2B0ogt_FyjpbXFnQaAPm98bA%40mail.gmail.com>
Precedence: bulk

--0000000000000699220630c7f36f
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Thursday, March 20, 2025, Tom Lane <tgl@sss.pgh.pa.us> wrote:

> Nathan Bossart <nathan@postgresql.org> writes:
> > Since there's presently no way to determine whether a Boolean
> > storage parameter is explicitly set or has just picked up the
> > default value, this commit also introduces an isset_offset member
> > to relopt_parse_elt.
>
> Uh, what?  Why is it a good idea to distinguish those states?
> Seems like that risks some very surprising behavior, ie if the
> default is "true", why shouldn't that act exactly like an
> explicit setting of "true"?
>

In order to implement what amounts to coalesce(=E2=80=A6) for settings (use=
 global
value unless the table value overrides) one needs a sentinel value that
means unset because settings cannot take on the null value.  There is no
such possible sentinel value for boolean so another field is required.  The
hazards of choosing Boolean instead of text for settings.

David J.

--0000000000000699220630c7f36f
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Thursday, March 20, 2025, Tom Lane &lt;<a href=3D"mailto:tgl@sss.pgh.pa.=
us">tgl@sss.pgh.pa.us</a>&gt; wrote:<br><blockquote class=3D"gmail_quote" s=
tyle=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Nath=
an Bossart &lt;<a href=3D"mailto:nathan@postgresql.org">nathan@postgresql.o=
rg</a>&gt; writes:<br>
&gt; Since there&#39;s presently no way to determine whether a Boolean<br>
&gt; storage parameter is explicitly set or has just picked up the<br>
&gt; default value, this commit also introduces an isset_offset member<br>
&gt; to relopt_parse_elt.<br>
<br>
Uh, what?=C2=A0 Why is it a good idea to distinguish those states?<br>
Seems like that risks some very surprising behavior, ie if the<br>
default is &quot;true&quot;, why shouldn&#39;t that act exactly like an<br>
explicit setting of &quot;true&quot;?<br>
</blockquote><div><br></div><div>In order to implement what amounts to coal=
esce(=E2=80=A6) for settings (use global value unless the table value overr=
ides) one needs a sentinel value that means unset because settings cannot t=
ake on the null value.=C2=A0 There is no such possible sentinel value for b=
oolean so another field is required.=C2=A0 The hazards of choosing Boolean =
instead of text for settings.</div><div><br></div><div>David J.</div><div><=
br></div>

--0000000000000699220630c7f36f--