Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:130817
Feedback-ID: ifab94697:Fastmail
Precedence: list
MIME-Version: 1.0
Date: Sun, 10 May 2026 22:10:42 +0200
To: internals@lists.php.net
Message-ID: <7e84847b-f857-47e1-be21-2784f1821674@app.fastmail.com>
In-Reply-To: 
 <CALaXqqTdx9x5Ccvs4pLVyY6=poe1Xe-ub-d9jP=u_4XHM5gy8Q@mail.gmail.com>
References: 
 <CALaXqqTdx9x5Ccvs4pLVyY6=poe1Xe-ub-d9jP=u_4XHM5gy8Q@mail.gmail.com>
Subject: Re: [PHP-DEV] [RFC] [Discussion] Friends
Content-Type: multipart/alternative;
 boundary=cfa845b9ced66d9fc7a4a104b27cfaeed5999898
From: rob@bottled.codes ("Rob Landers")

--cfa845b9ced66d9fc7a4a104b27cfaeed5999898
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable

On Sun, May 10, 2026, at 18:28, Daniel Scherzer wrote:
> Hi internals,
>=20
> I'd like to start the discussion for a new RFC about adding friendship=
 in PHP. This is a follow-up to a pre-RFC discussion thread https://exte=
rnals.io/message/130710.
>=20
> * RFC: https://wiki.php.net/rfc/friends
> * Implementation: https://github.com/php/php-src/pull/21937
>=20
> Thanks,
> -Daniel

Hi Daniel,

I worked on the namespace visibility RFC before running out of time, and=
 life isn't slowing down anytime soon; thus I wish you the best of luck =
with this one.

First of all ... the RFC doesn't address several inheritance/override in=
teractions worth working through. I'd invite you to read the thread here=
: https://externals.io/message/129147 -- you're going to run into a lot =
of the same issues in this RFC (and especially the follow-up namespace o=
ne), and you have some of the same problems.

Take this example:

class P {
  friend F;
  private int $x =3D 0;
}
class C extends P {
  protected int $x =3D 0;
}
class F {
  static function set(P $p, int $v) { $p->x =3D $v; }
}
F::set(new C, 5); // fatal

The friend grant on P creates a non-local invariant that subclass author=
s of P can break without realizing. C's author, adding `x` for their own=
 internal reasons, doesn't know they've broken some F that depends on th=
e parent contract. C's tests pass. F's tests pass. Integration breaks at=
 runtime in production. private(namespace) had the identical pathology.

=E2=80=94 Rob
--cfa845b9ced66d9fc7a4a104b27cfaeed5999898
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE html><html><head><title></title></head><body><div>On Sun, May =
10, 2026, at 18:28, Daniel Scherzer wrote:</div><blockquote type=3D"cite=
" id=3D"qt" style=3D""><div dir=3D"ltr"><div>Hi internals,</div><div><br=
></div><div>I'd like to start the discussion for a new RFC about adding =
friendship in PHP. This is a follow-up to a pre-RFC discussion thread&nb=
sp;<a href=3D"https://externals.io/message/130710">https://externals.io/=
message/130710</a>.</div><div><br></div><div>* RFC:&nbsp;<a href=3D"http=
s://wiki.php.net/rfc/friends">https://wiki.php.net/rfc/friends</a></div>=
<div>* Implementation:&nbsp;<a href=3D"https://github.com/php/php-src/pu=
ll/21937">https://github.com/php/php-src/pull/21937</a></div><div><br></=
div><div>Thanks,</div><div>-Daniel</div></div></blockquote><div><br></di=
v><div>Hi Daniel,</div><div><br></div><div>I worked on the namespace vis=
ibility RFC before running out of time, and life isn't slowing down anyt=
ime soon; thus I wish you the best of luck with this one.</div><div><br>=
</div><div>First of all ... the RFC doesn't address several inheritance/=
override interactions worth working through. I'd invite you to read the =
thread here:&nbsp;<a href=3D"https://externals.io/message/129147">https:=
//externals.io/message/129147</a>&nbsp;-- you're going to run into a lot=
 of the same issues in this RFC (and especially the follow-up namespace =
one), and you have some of the same problems.</div><div><br></div><div>T=
ake this example:</div><div><br></div><pre style=3D"border-top-width:1px=
;border-right-width:1px;border-bottom-width:1px;border-left-width:1px;bo=
rder-top-style:solid;border-right-style:solid;border-bottom-style:solid;=
border-left-style:solid;border-top-color:#ccc;border-right-color:#ccc;bo=
rder-bottom-color:#ccc;border-left-color:#ccc;border-image-source:initia=
l;border-image-slice:initial;border-image-width:initial;border-image-out=
set:initial;border-image-repeat:initial;border-top-left-radius:3px;borde=
r-top-right-radius:3px;border-bottom-right-radius:3px;border-bottom-left=
-radius:3px;background-image:initial;background-position-x:initial;backg=
round-position-y:initial;background-size:initial;background-repeat:initi=
al;background-attachment:initial;background-origin:initial;background-cl=
ip:initial;background-color:#f6f6f6;font-family:menlo, consolas, monospa=
ce;font-size:90%;margin-top:7px;margin-right:0px;margin-bottom:7px;margi=
n-left:0px;padding-top:7px;padding-right:10px;padding-bottom:7px;padding=
-left:10px;white-space-collapse:preserve;text-wrap-mode:wrap;overflow-wr=
ap:break-word;">class P {=0A  friend F;=0A &nbsp;private int $x =3D 0;=0A}=
=0Aclass C extends P {=0A  protected int $x =3D 0;=0A}=0Aclass F {=0A  s=
tatic function set(P $p, int $v) { $p-&gt;x =3D $v; }=0A}=0AF::set(new C=
, 5); // fatal<br></pre><div><br></div><div>The friend grant on P create=
s a non-local invariant that subclass authors of P can break without rea=
lizing. C's author, adding <code style=3D"border-top-width:1px;border-ri=
ght-width:1px;border-bottom-width:1px;border-left-width:1px;border-top-s=
tyle:solid;border-right-style:solid;border-bottom-style:solid;border-lef=
t-style:solid;border-top-color:#ccc;border-right-color:#ccc;border-botto=
m-color:#ccc;border-left-color:#ccc;border-image-source:initial;border-i=
mage-slice:initial;border-image-width:initial;border-image-outset:initia=
l;border-image-repeat:initial;border-top-left-radius:3px;border-top-righ=
t-radius:3px;border-bottom-right-radius:3px;border-bottom-left-radius:3p=
x;background-image:initial;background-position-x:initial;background-posi=
tion-y:initial;background-size:initial;background-repeat:initial;backgro=
und-attachment:initial;background-origin:initial;background-clip:initial=
;background-color:#f6f6f6;font-family:menlo, consolas, monospace;font-si=
ze:90%;padding-top:1px;padding-right:3px;padding-bottom:1px;padding-left=
:3px;">x</code>&nbsp;for their own internal reasons, doesn't know they'v=
e broken some F that depends on the parent contract. C's tests pass. F's=
 tests pass. Integration breaks at runtime in production. private(namesp=
ace) had the identical pathology.</div><div><br></div><div>=E2=80=94 Rob=
</div></body></html>
--cfa845b9ced66d9fc7a4a104b27cfaeed5999898--