Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:128121
Feedback-ID: iea414789:Fastmail
Precedence: bulk
MIME-Version: 1.0
Date: Fri, 18 Jul 2025 13:10:03 -0400
To: "PHP Internals" <internals@lists.php.net>
Cc: rob@bottled.codes, tim@bastelstu.be
Message-ID: <d8e4f236-dbe7-4dfc-b273-4d561f01aaa9@betaapp.fastmail.com>
In-Reply-To: <0d9a2968-26dd-43dd-9733-d2eb621ab9ec@app.fastmail.com>
References: <1e8634d7-ac1a-4025-b4e2-1948aabf5251@app.fastmail.com>
 <9D5043B2-1589-4FD5-B289-6E98FB1177BE@nicksdot.dev>
 <CAPyj-LDVi3ZYQuzzJz7d+cHdU6C-cxYJUSkuzdyvWpw+sWuCsg@mail.gmail.com>
 <0856c89f-2000-448a-bbbf-c145a8699f6a@app.fastmail.com>
 <2641b8bc-6337-4b75-a5a3-93dee1b03796@bastelstu.be>
 <0d9a2968-26dd-43dd-9733-d2eb621ab9ec@app.fastmail.com>
Subject: Re: [PHP-DEV] [RFC] Readonly property hooks
Content-Type: multipart/alternative;
 boundary=888837749f7a423c8f12abe139d5a722
From: php@mjec.net (mjec)

--888837749f7a423c8f12abe139d5a722
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable

On Fri, 2025-07-18 at 11:49 -04:00, Rob Landers <rob@bottled.codes> wrot=
e:
> On Fri, Jul 18, 2025, at 17:25, Tim D=C3=BCsterhus wrote:
>> A readonly class is not just a convenience shortcut to mark each=20
>> individual property as readonly. It has important semantics of its ow=
n,=20
>> because it forces child classes to also be readonly. And even for fin=
al=20
>> classes it communicates to the user that "I won't be adding non-reado=
nly=20
>> properties to the class".
>=20
> Wasn=E2=80=99t that the entire point of readonly classes? Because it w=
as painful to write readonly for every property. Then if a property is r=
eadonly, the inherited property is also readonly, so, by extension: a cl=
ass extending a readonly class is also readonly.
>=20
> There=E2=80=99s no =E2=80=9Ccommunication=E2=80=9D here; just logic.=20

All code is communication, not just to the computer, but to other humans=
 reading it. This is why the semantics of the terms are important. I agr=
ee with Tim's interpretation of those semantics, and hence vastly prefer=
 an init hook to get in these circumstances.

Importantly, this RFC can clarify the semantics of these terms for the l=
anguage as a whole. That specific meaning can be documented. If a propos=
al does not come from a principled understanding of what `readonly` mean=
s - what it indicates to users of the class - I think that is a step bac=
kwards.

mjec
--888837749f7a423c8f12abe139d5a722
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE html><html><head><title></title></head><body><div>On Fri, 2025=
-07-18 at 11:49 -04:00, Rob Landers &lt;<a href=3D"mailto:rob@bottled.co=
des" class=3D"">rob@bottled.codes</a>&gt; wrote:</div><blockquote type=3D=
"cite" id=3D"qt" style=3D""><div>On Fri, Jul 18, 2025, at 17:25, Tim D=C3=
=BCsterhus wrote:</div><blockquote type=3D"cite" id=3D"qt-qt" style=3D""=
><div>A readonly class is not just a convenience shortcut to mark each&n=
bsp;</div><div>individual property as readonly. It has important semanti=
cs of its own,&nbsp;</div><div>because it forces child classes to also b=
e readonly. And even for final&nbsp;</div><div>classes it communicates t=
o the user that "I won't be adding non-readonly&nbsp;</div><div>properti=
es to the class".</div></blockquote><div><br></div><div>Wasn=E2=80=99t t=
hat the entire point of readonly classes? Because it was painful to writ=
e readonly for every property. Then if a property is readonly, the inher=
ited property is also readonly, so, by extension: a class extending a re=
adonly class is also readonly.</div><div><br></div><div>There=E2=80=99s =
no =E2=80=9Ccommunication=E2=80=9D here; just logic.&nbsp;</div></blockq=
uote><div><br></div><div>All code is communication, not just to the comp=
uter, but to other humans reading it. This is why the semantics of the t=
erms are important. I agree with Tim's interpretation of those semantics=
, and hence vastly prefer an init hook to get in these circumstances.</d=
iv><div><br></div><div>Importantly, this RFC can clarify the semantics o=
f these terms for the language as a whole. That specific meaning can be =
documented. If a proposal does not come from a principled understanding =
of what `readonly` means - what it indicates to users of the class - I t=
hink that is a step backwards.</div><div><br></div><div>mjec</div></body=
></html>
--888837749f7a423c8f12abe139d5a722--