Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:129881
Precedence: list
MIME-Version: 1.0 (Mimestream 1.9.7)
References: <bfea3ae503ea2a4d375da077ecef2097@bastelstu.be> <CAP1Jc10PPAC1cwoU3oow=K6iWFRcCUz+bm39ARN8fvVH6Mn3UQ@mail.gmail.com>
In-Reply-To: <CAP1Jc10PPAC1cwoU3oow=K6iWFRcCUz+bm39ARN8fvVH6Mn3UQ@mail.gmail.com>
Date: Fri, 23 Jan 2026 03:59:38 -0600
Message-ID: <CAEPJdumfncn+bimQzj90FhdAgkVwEj7BMau8FYc0DMnZP06LmQ@mail.gmail.com>
Subject: Re: [PHP-DEV] [VOTE] let construct (Block Scoping)
To: Arnaud Le Blanc <arnaud.lb@gmail.com>
Cc: php internals <internals@lists.php.net>, Seifeddine Gmati <azjezz@carthage.software>, 
	=?UTF-8?Q?Tim_D=C3=BCsterhus?= <tim@bastelstu.be>
Content-Type: multipart/alternative; boundary="0000000000002e677a06490b377e"
From: kontakt@beberlei.de (=?UTF-8?Q?Benjamin_Au=C3=9Fenhofer?=)

--0000000000002e677a06490b377e
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Am 22.01.2026, 18:57:04 schrieb Arnaud Le Blanc <arnaud.lb@gmail.com>:

> On Thu, Jan 22, 2026 at 10:17=E2=80=AFAM Tim D=C3=BCsterhus <tim@bastelst=
u.be> wrote:
>
>
> Hi
>
>
> we just opened the vote on the =E2=80=9Clet construct (Block Scoping)=E2=
=80=9D RFC
>
> (which was originally proposed as =E2=80=9Cuse construct=E2=80=9D, but th=
e keyword
>
> changed as a result of the discussion).
>
>
> Hi,
>
> Thank you for your work on this.
>
> I voted no because the RFC is presented as a way to dispose of
> resources immediately (quoting: "This feature provides immense value,
> particularly for modern applications built on long-lived servers,
> where disciplined and immediate resource cleanup is not just a best
> practice, but a necessity for stability and performance."), yet it
> doesn't in practice as it relies on reference counting as a proxy for
> cleaning up at block exit.
>

I agree that relying on reference counting can bite you, but I still voted
Yes, because


   - We only have reference counting to rely on in PHP as far as memory
   goes, and thats what people writing cleanup critical code rely on alread=
y.
   This gives a much nicer syntax.
   - The future scope lists improvements as next steps, like the error when
   RC > 1 or Disposable interfaces, this is a first step towards something
   that is missing dearly.



> While reference counting is deterministic, ensuring that a value is
> uniquely referenced before leaving a block is impossible, and cannot
> be enforced by the block itself. In practice, useful code will send
> the resource to other functions or 3rd party code that may retain a
> reference to it. Moreover, there are many non-obvious ways in which a
> reference count can be increased or a value=E2=80=99s lifetime extended:
> exceptions or backtraces may capture arguments, fibers, generators, or
> closures with non obvious lifetimes retain their local variables,
> closures tend to create cycles, foreach variables are not always
> unset, etc.
>
> Therefore there is a non-zero possibility that a let() statement will
> not in fact cleanup the resource, despite the stated goal. Fixing this
> later would be a BC break as people may rely on this behavior.
>
> The RFC refers to similar constructs from other languages, but none of
> them rely on reference counting for this purpose. I'm not aware of
> languages relying on reference counting for this.
>
> Best Regards,
> Arnaud
>

--0000000000002e677a06490b377e
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<html><body><div dir=3D"ltr">
    <br>
</div>
<br>
<div class=3D"gmail_quote">
    <div dir=3D"ltr" class=3D"gmail_attr">Am 22.01.2026, 18:57:04 schrieb A=
rnaud Le Blanc &lt;<a href=3D"mailto:arnaud.lb@gmail.com">arnaud.lb@gmail.c=
om</a>&gt;:<br></div>
    <blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;bor=
der-left:1px solid rgb(204,204,204);padding-left:1ex" type=3D"cite">
       =20
<div>
<div>
    On Thu, Jan 22, 2026 at 10:17=E2=80=AFAM Tim D=C3=BCsterhus &lt;<a href=
=3D"mailto:tim@bastelstu.be">tim@bastelstu.be</a>&gt; wrote:<br><blockquote=
 type=3D"cite"><br></blockquote><blockquote type=3D"cite"> Hi<br></blockquo=
te><blockquote type=3D"cite"><br></blockquote><blockquote type=3D"cite"> we=
 just opened the vote on the =E2=80=9Clet construct (Block Scoping)=E2=80=
=9D RFC<br></blockquote><blockquote type=3D"cite"> (which was originally pr=
oposed as =E2=80=9Cuse construct=E2=80=9D, but the keyword<br></blockquote>=
<blockquote type=3D"cite"> changed as a result of the discussion).<br></blo=
ckquote><br>Hi,<br><br>Thank you for your work on this.<br><br>I voted no b=
ecause the RFC is presented as a way to dispose of<br>resources immediately=
 (quoting: &quot;This feature provides immense value,<br>particularly for m=
odern applications built on long-lived servers,<br>where disciplined and im=
mediate resource cleanup is not just a best<br>practice, but a necessity fo=
r stability and performance.&quot;), yet it<br>doesn&#39;t in practice as i=
t relies on reference counting as a proxy for<br>cleaning up at block exit.=
<br></div></div></blockquote><div class=3D"gmail_quote"><br></div><div clas=
s=3D"gmail_quote" dir=3D"ltr">I agree that relying on reference counting ca=
n bite you, but I still voted Yes, because</div><div class=3D"gmail_quote">=
<br></div><div class=3D"gmail_quote" dir=3D"ltr"><ul style=3D"margin:0px" d=
ir=3D"ltr"><li>We only have reference counting to rely on in PHP as far as =
memory goes, and thats what people writing cleanup critical code rely on al=
ready. This gives a much nicer syntax.<br></li><li>The future scope lists i=
mprovements as next steps, like the error when RC &gt; 1 or Disposable inte=
rfaces, this is a first step towards something that is missing dearly.</li>=
</ul></div><br><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0p=
x 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex" type=3D"ci=
te"><div><div><br>While reference counting is deterministic, ensuring that =
a value is<br>uniquely referenced before leaving a block is impossible, and=
 cannot<br>be enforced by the block itself. In practice, useful code will s=
end<br>the resource to other functions or 3rd party code that may retain a<=
br>reference to it. Moreover, there are many non-obvious ways in which a<br=
>reference count can be increased or a value=E2=80=99s lifetime extended:<b=
r>exceptions or backtraces may capture arguments, fibers, generators, or<br=
>closures with non obvious lifetimes retain their local variables,<br>closu=
res tend to create cycles, foreach variables are not always<br>unset, etc.<=
br><br>Therefore there is a non-zero possibility that a let() statement wil=
l<br>not in fact cleanup the resource, despite the stated goal. Fixing this=
<br>later would be a BC break as people may rely on this behavior.<br><br>T=
he RFC refers to similar constructs from other languages, but none of<br>th=
em rely on reference counting for this purpose. I&#39;m not aware of<br>lan=
guages relying on reference counting for this.<br><br>Best Regards,<br>Arna=
ud<br>
</div>
</div>
    </blockquote>
</div></body></html>

--0000000000002e677a06490b377e--