Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:126235
Feedback-ID: ifab94697:Fastmail
Precedence: bulk
MIME-Version: 1.0
Date: Wed, 29 Jan 2025 21:16:11 +0100
To: =?UTF-8?Q?Tim_D=C3=BCsterhus?= <tim@bastelstu.be>,
 "Volker Dusch" <volker@tideways-gmbh.com>,
 "php internals" <internals@lists.php.net>
Message-ID: <3b4d9be8-9255-44ab-95c5-ea3045212cf4@app.fastmail.com>
In-Reply-To: <3f170549-cfd6-441e-b892-cf51d726dbe8@bastelstu.be>
References: <a995ac24905030e168e7f6006b996cd5@bastelstu.be>
 <a3d14e31-53eb-4a04-8fb2-01f58809fec1@app.fastmail.com>
 <3f170549-cfd6-441e-b892-cf51d726dbe8@bastelstu.be>
Subject: Re: [PHP-DEV] RFC: Marking return values as important (#[\NoDiscard])
Content-Type: multipart/alternative;
 boundary=c666f93a47be4631aa878b752551c3c4
From: rob@bottled.codes ("Rob Landers")

--c666f93a47be4631aa878b752551c3c4
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable

On Wed, Jan 29, 2025, at 20:50, Tim D=C3=BCsterhus wrote:
> Hi
>=20
> On 1/29/25 20:31, Rob Landers wrote:
> > This looks promising!
> >=20
> > Why do we need the (void) cast? Wouldn't "@" work and avoid a BC iss=
ue?
>=20
> The `@` operator also suppresses any other warning or notice within th=
e=20
> call chain. That is likely undesirable because it might hide issues on=
e=20
> is interested in. See this example (https://3v4l.org/NDtR7):
>=20
>      <?php
>      function inner() {
>          var_dump($undefined);
>      }
>=20
>      function outer() {
>          inner();
>      }
>=20
>      outer();
>      outer();
>      @outer();
>=20
> The `(void)` cast makes the intent of not being interested in the retu=
rn=20
> value clear. In C casting to `(void)` is an established pattern of=20
> "using" a value that the compiler would otherwise consider unused.=20
> Likewise will the `void` operator suppress the TypeScript-ESLint warni=
ng=20
> of not using returned `Promise`:=20
> https://typescript-eslint.io/rules/no-floating-promises/#ignorevoid
>=20
> Best regards
> Tim D=C3=BCsterhus
>=20

Hi Tim,

I understand what you are saying, but I can also just remove the warning=
 via:

$_ =3D outer;

Which, to me, is clearer on the intent than (void)outer(); It also happe=
ns to make the diffs more natural looking when and if the return value g=
ets used during a code review. I'll also note that having (void) doesn't=
 preclude me from using $_ either.

I was mostly just curious as to the reasoning; I don't like it, but it m=
akes sense.

Best of luck!

=E2=80=94 Rob
--c666f93a47be4631aa878b752551c3c4
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE html><html><head><title></title><style type=3D"text/css">p.Mso=
Normal,p.MsoNoSpacing{margin:0}</style></head><body><div>On Wed, Jan 29,=
 2025, at 20:50, Tim D=C3=BCsterhus wrote:<br></div><blockquote type=3D"=
cite" id=3D"qt" style=3D""><div>Hi<br></div><div><br></div><div>On 1/29/=
25 20:31, Rob Landers wrote:<br></div><div>&gt; This looks promising!<br=
></div><div>&gt;&nbsp;<br></div><div>&gt; Why do we need the (void) cast=
? Wouldn't "@" work and avoid a BC issue?<br></div><div><br></div><div>T=
he `@` operator also suppresses any other warning or notice within the&n=
bsp;<br></div><div>call chain. That is likely undesirable because it mig=
ht hide issues one&nbsp;<br></div><div>is interested in. See this exampl=
e (<a href=3D"https://3v4l.org/NDtR7">https://3v4l.org/NDtR7</a>):<br></=
div><div><br></div><div>&nbsp;&nbsp;&nbsp;&nbsp; &lt;?php<br></div><div>=
&nbsp;&nbsp;&nbsp;&nbsp; function inner() {<br></div><div>&nbsp;&nbsp;&n=
bsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; var_dump($undefined);<br></div><div>&=
nbsp;&nbsp;&nbsp;&nbsp; }<br></div><div><br></div><div>&nbsp;&nbsp;&nbsp=
;&nbsp; function outer() {<br></div><div>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp; inner();<br></div><div>&nbsp;&nbsp;&nbsp;&nbsp; }<br><=
/div><div><br></div><div>&nbsp;&nbsp;&nbsp;&nbsp; outer();<br></div><div=
>&nbsp;&nbsp;&nbsp;&nbsp; outer();<br></div><div>&nbsp;&nbsp;&nbsp;&nbsp=
; @outer();<br></div><div><br></div><div>The `(void)` cast makes the int=
ent of not being interested in the return&nbsp;<br></div><div>value clea=
r. In C casting to `(void)` is an established pattern of&nbsp;<br></div>=
<div>"using" a value that the compiler would otherwise consider unused.&=
nbsp;<br></div><div>Likewise will the `void` operator suppress the TypeS=
cript-ESLint warning&nbsp;<br></div><div>of not using returned `Promise`=
:&nbsp;<br></div><div><a href=3D"https://typescript-eslint.io/rules/no-f=
loating-promises/#ignorevoid">https://typescript-eslint.io/rules/no-floa=
ting-promises/#ignorevoid</a><br></div><div><br></div><div>Best regards<=
br></div><div>Tim D=C3=BCsterhus<br></div><div><br></div></blockquote><d=
iv><br></div><div>Hi Tim,<br></div><div><br></div><div>I understand what=
 you are saying, but I can also just remove the warning via:<br></div><d=
iv><br></div><div>$_ =3D outer;<br></div><div><br></div><div>Which, to m=
e, is clearer on the intent than (void)outer(); It also happens to make =
the diffs more natural looking when and if the return value gets used du=
ring a code review. I'll also note that having (void) doesn't preclude m=
e from using $_ either.<br></div><div><br></div><div>I was mostly just c=
urious as to the reasoning; I don't like it, but it makes sense.<br></di=
v><div><br></div><div>Best of luck!</div><div><br></div><div id=3D"sig12=
1229152">=E2=80=94 Rob<br></div></body></html>
--c666f93a47be4631aa878b752551c3c4--