Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:128855
Feedback-ID: ifab94697:Fastmail
Precedence: list
MIME-Version: 1.0
Date: Thu, 16 Oct 2025 16:42:39 +0200
To: "Edmond Dantes" <edmond.ht@gmail.com>
Cc: internals@lists.php.net
Message-ID: <e3f5c8bc-48de-4294-9a86-291661e12b39@app.fastmail.com>
In-Reply-To: 
 <CAMW7n8CdXxYZ9=zwZdt0rohMALkNhQwCcwnJPnnmxVjVc42iWg@mail.gmail.com>
References: 
 <CAMW7n8DGbFnQzjt3s9CcOBwhp+MUOV+K3U=yY=ULOTdUXaOLHQ@mail.gmail.com>
 <bcd8253c-04f3-4046-b51f-2e387e215f6d@app.fastmail.com>
 <CAMW7n8BfP7h5rpDrjigu0oWK4Z+aRaqGjgzEFS_mDXRhq19vQQ@mail.gmail.com>
 <2b9fd3ec-50ca-41e4-985a-274f886df8b3@app.fastmail.com>
 <CAMW7n8C5GD+F2KdCZTtZwFbc2rk_4yGgdiak8fs9y5Z3UU_zxg@mail.gmail.com>
 <2e39e211-c816-41db-a079-f2c6b3934e0a@app.fastmail.com>
 <CAKXKfQtO+yoFB1=6zonwgGRpx1YExBNSkcjpso51aJ59YzaDLQ@mail.gmail.com>
 <436d4ede-c08e-4e7d-9ae2-8342f7f827ed@app.fastmail.com>
 <CAMW7n8CdXxYZ9=zwZdt0rohMALkNhQwCcwnJPnnmxVjVc42iWg@mail.gmail.com>
Subject: Re: [PHP-DEV] PHP True Async RFC Stage 4
Content-Type: multipart/alternative;
 boundary=03740351e4a24d9e97c9363613943817
From: rob@bottled.codes ("Rob Landers")

--03740351e4a24d9e97c9363613943817
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable



On Thu, Oct 16, 2025, at 15:19, Edmond Dantes wrote:
> > Would it be better to instead of having ->awaitCompletion (which fee=
ls like an implicit implementation of Awaitable without an explicit impl=
mentation -- which is the part that was bothering me), maybe having some=
thing like ->joinAll() or ->joinOnCancellation()?
> > That way someone like me won't come along and wrap them in an Awaita=
ble because it looks Awaitable.
>=20
> For example, you have an interface `DataBaseAdmin` with a `removeDB()`
> method and a class that implements it.
> You need to be able to delete the database, but with an additional
> condition. So you create a decorator class with a method
> `ContextualDBAdmin::removeDBWhen($extraRules)`.
>=20
> As a result, the decorator class is logically associated with the
> `DataBaseAdmin` interface.
>=20
> The question is: **so what?**

I think we might be talking past each other a little.

You said earlier that one of the goals here is to prevent misuse (e.g. u=
nbounded or foreign awaits). I completely agree and that=E2=80=99s exact=
ly why I=E2=80=99m suggesting a rename. From the outside, a method calle=
d awaitCompletion() looks and behaves like an Awaitable, even though the=
 type explicitly isn=E2=80=99t one. That contradiction encourages people=
 to wrap it to =E2=80=9Cmake it awaitable,=E2=80=9D which re-introduces =
the very problem you=E2=80=99re trying to avoid.

Renaming it to something like joinAll() or joinAfterCancellation() keeps=
 the semantics intact while communicating the intent: this isn=E2=80=99t=
 a future; it=E2=80=99s a container join. That small change would make t=
he interface self-documenting and harder to misuse.

>=20
> > It also might be a good idea to make specifying a timeout in ms mand=
atory, instead of a taking an Awaitable/Cancellation.
>=20
> The idea is correct, but there will definitely be someone who says
> they need more flexibility.
> And it's true you can create a `DeferredCancellation` and forget to
> finish it. :)
>=20
> There are a lot of such subtle points, and they can be discussed endle=
ssly.
> But I wouldn=E2=80=99t spend time on them.

That's what we are here to do, no? Discussion is useful if it makes thin=
gs better than the sum of their parts.

> > It also might be good to provide a realistic looking example showing=
 a "bad case" of how this is dangerous instead of simply saying that it =
is, showing how a scope is not a 'future',
> > but a container, and preemptively mention TaskGroups, linking to the=
 future scope (which it should also probably be listed there as well).
>=20
> 1. It=E2=80=99s very difficult to write a realistic example that=E2=80=
=99s still small.

I'll give it a go:

$scope =3D new Scope();

// Library code spawns in my scope (transitively)
$scope->spawn(fn() =3D> thirdPartyOperation()); // may spawn more

// Looks innocent, but this can wait on foreign work:
$scope->awaitCompletion(Async\timeout(60000)); // rename -> joinAll(...)?

> 2. The `TaskGroup` or `CoroutineGroup` class is left for future
> discussion. In the final documentation, it will be exactly as you
> suggested.

My point is that it wasn't listed in "future scope" of the RFC, though t=
hey're mentioned throughout the document, in passing.

=E2=80=94 Rob
--03740351e4a24d9e97c9363613943817
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 Thu, Oct 16, 2025, at 15:19, Edmond Dantes wrote:</=
div><blockquote type=3D"cite" id=3D"qt" style=3D""><div>&gt; Would it be=
 better to instead of having -&gt;awaitCompletion (which feels like an i=
mplicit implementation of Awaitable without an explicit implmentation --=
 which is the part that was bothering me), maybe having something like -=
&gt;joinAll() or -&gt;joinOnCancellation()?</div><div>&gt; That way some=
one like me won't come along and wrap them in an Awaitable because it lo=
oks Awaitable.</div><div><br></div><div>For example, you have an interfa=
ce `DataBaseAdmin` with a `removeDB()`</div><div>method and a class that=
 implements it.</div><div>You need to be able to delete the database, bu=
t with an additional</div><div>condition. So you create a decorator clas=
s with a method</div><div>`ContextualDBAdmin::removeDBWhen($extraRules)`=
.</div><div><br></div><div>As a result, the decorator class is logically=
 associated with the</div><div>`DataBaseAdmin` interface.</div><div><br>=
</div><div>The question is: **so what?**</div></blockquote><div><br></di=
v><div>I think we might be talking past each other a little.</div><div><=
br></div><div>You said earlier that one of the goals here is to prevent =
misuse (e.g. unbounded or foreign awaits). I completely agree and that=E2=
=80=99s exactly why I=E2=80=99m suggesting a rename. From the outside, a=
 method called awaitCompletion() looks and behaves like an Awaitable, ev=
en though the type explicitly isn=E2=80=99t one. That contradiction enco=
urages people to wrap it to =E2=80=9Cmake it awaitable,=E2=80=9D which r=
e-introduces the very problem you=E2=80=99re trying to avoid.</div><div>=
<br></div><div>Renaming it to something like joinAll() or joinAfterCance=
llation() keeps the semantics intact while communicating the intent: thi=
s isn=E2=80=99t a future; it=E2=80=99s a container join. That small chan=
ge would make the interface self-documenting and harder to misuse.</div>=
<div><br></div><blockquote type=3D"cite" id=3D"qt" style=3D""><div><br><=
/div><div>&gt; It also might be a good idea to make specifying a timeout=
 in ms mandatory, instead of a taking an Awaitable/Cancellation.</div><d=
iv><br></div><div>The idea is correct, but there will definitely be some=
one who says</div><div>they need more flexibility.</div><div>And it's tr=
ue you can create a `DeferredCancellation` and forget to</div><div>finis=
h it. :)</div><div><br></div><div>There are a lot of such subtle points,=
 and they can be discussed endlessly.</div><div>But I wouldn=E2=80=99t s=
pend time on them.</div></blockquote><div><br></div><div>That's what we =
are here to do, no? Discussion is useful if it makes things better than =
the sum of their parts.</div><div><br></div><blockquote type=3D"cite" id=
=3D"qt" style=3D""><div>&gt; It also might be good to provide a realisti=
c looking example showing a "bad case" of how this is dangerous instead =
of simply saying that it is, showing how a scope is not a 'future',</div=
><div>&gt; but a container, and preemptively mention TaskGroups, linking=
 to the future scope (which it should also probably be listed there as w=
ell).</div><div><br></div><div>1. It=E2=80=99s very difficult to write a=
 realistic example that=E2=80=99s still small.</div></blockquote><div><b=
r></div><div>I'll give it a go:</div><div><br></div><div><span class=3D"=
font" style=3D"font-family:menlo, consolas, monospace, sans-serif;">$sco=
pe =3D new Scope();</span></div><div><span class=3D"font" style=3D"font-=
family:menlo, consolas, monospace, sans-serif;"><br></span></div><div><s=
pan class=3D"font" style=3D"font-family:menlo, consolas, monospace, sans=
-serif;">// Library code spawns in my scope (transitively)</span></div><=
div><span class=3D"font" style=3D"font-family:menlo, consolas, monospace=
, sans-serif;">$scope-&gt;spawn(fn() =3D&gt; thirdPartyOperation()); // =
may spawn more</span></div><div><span class=3D"font" style=3D"font-famil=
y:menlo, consolas, monospace, sans-serif;"><br></span></div><div><span c=
lass=3D"font" style=3D"font-family:menlo, consolas, monospace, sans-seri=
f;">// Looks innocent, but this can wait on foreign work:</span></div><d=
iv><span class=3D"font" style=3D"font-family:menlo, consolas, monospace,=
 sans-serif;">$scope-&gt;awaitCompletion(Async\timeout(60000)); // renam=
e -&gt; joinAll(...)?</span><br></div><div><br></div><blockquote type=3D=
"cite" id=3D"qt" style=3D""><div>2. The `TaskGroup` or `CoroutineGroup` =
class is left for future</div><div>discussion. In the final documentatio=
n, it will be exactly as you</div><div>suggested.</div></blockquote><div=
><br></div><div>My point is that it wasn't listed in "future scope" of t=
he RFC, though they're mentioned throughout the document, in passing.</d=
iv><div><br></div><div id=3D"sig121229152">=E2=80=94 Rob</div></body></h=
tml>
--03740351e4a24d9e97c9363613943817--