Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:129417
Feedback-ID: ifab94697:Fastmail
Precedence: list
MIME-Version: 1.0
Date: Sun, 23 Nov 2025 20:25:29 +0100
To: =?UTF-8?Q?Tim_D=C3=BCsterhus?= <tim@bastelstu.be>, internals@lists.php.net
Message-ID: <7a9286a9-17c8-44fb-8d10-3483aaf8bc6d@app.fastmail.com>
In-Reply-To: <bd67e375-ecc6-43f9-ae23-196a55d94867@bastelstu.be>
References: <e21b1992-d426-46cd-b86a-6d6a5844c934@app.fastmail.com>
 <f1f694d7-d494-45c5-a4e4-fb39809d3b67@rwec.co.uk>
 <c0bfa15e-1df2-4d87-8092-2531b2bafcb3@app.fastmail.com>
 <72f90052-fa19-415c-9f5a-ae75275fd030@rwec.co.uk>
 <be5c97b7-430d-4d4e-b413-b331c85816e1@app.fastmail.com>
 <2d9f8f1d-e568-4bb5-b30c-0a9e54a8f5fe@rwec.co.uk>
 <8978174c-04fc-451a-8bf5-cb7c54346318@app.fastmail.com>
 <1606b9e5-62b8-446b-adac-519ac19d01c3@app.fastmail.com>
 <c6ac9820-e35e-4fb2-80c5-ee0050cc5e91@bastelstu.be>
 <de42ddbf-1650-4039-b172-c1f26e0c01e1@app.fastmail.com>
 <a7d63636-f079-429f-a57d-111631187e1e@app.fastmail.com>
 <bd67e375-ecc6-43f9-ae23-196a55d94867@bastelstu.be>
Subject: Re: [PHP-DEV] RFC: Namespace-Scoped Visibility for Methods and Properties
Content-Type: multipart/alternative;
 boundary=46e9984d1ba84ca0b4eaf45498cfabb4
From: rob@bottled.codes ("Rob Landers")

--46e9984d1ba84ca0b4eaf45498cfabb4
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable



On Sun, Nov 23, 2025, at 17:00, Tim D=C3=BCsterhus wrote:
> Hi
>=20
> On 11/10/25 20:19, Rob Landers wrote:
> > Updated RFC:
> > https://wiki.php.net/rfc/namespace_visibility <https://wiki.php.net/=
rfc/namespace_visibility?utm_source=3Dchatgpt.com>
>=20
> From the RFC I'm seeing that it's legal to redefine methods in=20
> different namespaces? I've only figured that out with the last example=20
> in the =E2=80=9CPreventing Visibility Reduction and Incompatible Redec=
laration=E2=80=9D=20
> section, it might have been hinted at with =E2=80=9CHowever, visibilit=
y is=20
> enforced based on the namespace where they were declared, not on the=20
> inheritance hierarchy:=E2=80=9D but if it did, I didn't understand it.

No, it should be illegal, but it looks like the current RFC text doesn=E2=
=80=99t say that. Basically, private(namespace) is incompatible with pri=
vate(namespace) if the original namespace doesn=E2=80=99t match the curr=
ent namespace. I=E2=80=99ll update the RFC later this week.

> I feel that namespace-scoping needs to be restricted to top-level=20
> symbols to be easy to reason about. This also brings me to the next=20
> point: What about free-standing functions and constants?=20
> public/protected/private doesn't make sense for them, since they are n=
ot=20
> part of a inheritance hierarchy. But namespace-scope would be meaningf=
ul.

That would need https://github.com/php/php-src/pull/20490 (subtle wink e=
moji)

>=20
> All in all, I feel that file-private symbols would solve many of the=20
> same problems, but be much easier to reason about. The SessionManager=20
> from your example could just be file-private.

You can use private(namespace) as effectively file-private, just use a n=
amespace in a single file. And, if you need to refactor, you don=E2=80=99=
t need to be constrained to a single file.

=E2=80=94 Rob
--46e9984d1ba84ca0b4eaf45498cfabb4
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE html><html><head><title></title></head><body><div><br></div><d=
iv><br></div><div>On Sun, Nov 23, 2025, at 17:00, Tim D=C3=BCsterhus wro=
te:</div><blockquote type=3D"cite" id=3D"qt" style=3D""><div>Hi</div><di=
v><br></div><div>On 11/10/25 20:19, Rob Landers wrote:</div><div>&gt; Up=
dated RFC:</div><div>&gt;&nbsp;<a href=3D"https://wiki.php.net/rfc/names=
pace_visibility">https://wiki.php.net/rfc/namespace_visibility</a> &lt;<=
a href=3D"https://wiki.php.net/rfc/namespace_visibility?utm_source=3Dcha=
tgpt.com">https://wiki.php.net/rfc/namespace_visibility?utm_source=3Dcha=
tgpt.com</a>&gt;</div><div><br></div><div>From the RFC I'm seeing that i=
t's legal to redefine methods in&nbsp;</div><div>different namespaces? I=
've only figured that out with the last example&nbsp;</div><div>in the =E2=
=80=9CPreventing Visibility Reduction and Incompatible Redeclaration=E2=80=
=9D&nbsp;</div><div>section, it might have been hinted at with =E2=80=9C=
However, visibility is&nbsp;</div><div>enforced based on the namespace w=
here they were declared, not on the&nbsp;</div><div>inheritance hierarch=
y:=E2=80=9D but if it did, I didn't understand it.</div></blockquote><di=
v><br></div><div>No, it should be illegal, but it looks like the current=
 RFC text doesn=E2=80=99t say that. Basically, private(namespace) is inc=
ompatible with private(namespace) if the original namespace doesn=E2=80=99=
t match the current namespace. I=E2=80=99ll update the RFC later this we=
ek.<br></div><div><br></div><blockquote type=3D"cite" id=3D"qt" style=3D=
""><div>I feel that namespace-scoping needs to be restricted to top-leve=
l&nbsp;</div><div>symbols to be easy to reason about. This also brings m=
e to the next&nbsp;</div><div>point: What about free-standing functions =
and constants?&nbsp;</div><div>public/protected/private doesn't make sen=
se for them, since they are not&nbsp;</div><div>part of a inheritance hi=
erarchy. But namespace-scope would be meaningful.</div></blockquote><div=
><br></div><div>That would need&nbsp;<a href=3D"https://github.com/php/p=
hp-src/pull/20490">https://github.com/php/php-src/pull/20490</a>&nbsp;(s=
ubtle wink emoji)</div><div><br></div><blockquote type=3D"cite" id=3D"qt=
" style=3D""><div><br></div><div>All in all, I feel that file-private sy=
mbols would solve many of the&nbsp;</div><div>same problems, but be much=
 easier to reason about. The SessionManager&nbsp;</div><div>from your ex=
ample could just be file-private.</div></blockquote><div><br></div><div>=
You can use private(namespace) as effectively file-private, just use a n=
amespace in a single file. And, if you need to refactor, you don=E2=80=99=
t need to be constrained to a single file.</div><div><br></div><div id=3D=
"sig121229152">=E2=80=94 Rob</div></body></html>
--46e9984d1ba84ca0b4eaf45498cfabb4--