Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:120045
MIME-Version: 1.0
References: <CAFPFaMLghRYP4ioCkfa7WKhTAz+1gWgcZ5mWYAunkb9f3g=fdA@mail.gmail.com>
 <A97A9A6E-DBE5-4BE6-98FE-C5A7381460E4@gmail.com> <3b7db03a-1cad-3762-b0ea-d33618be9644@gmail.com>
 <a492bd9e-7147-ce9a-83d8-2263a4ea0d00@seld.be>
In-Reply-To: <a492bd9e-7147-ce9a-83d8-2263a4ea0d00@seld.be>
Date: Sun, 16 Apr 2023 17:22:01 +0100
Message-ID: <CA+kxMuRhXMpRQq57+1=j7JRbNSFqnU4SNPFe2xGp1aiAwnkfQg@mail.gmail.com>
To: Jordi Boggiano <j.boggiano@seld.be>
Cc: internals@lists.php.net
Content-Type: text/plain; charset="UTF-8"
Subject: Re: [PHP-DEV] [RFC] New core autoloading mechanism with support for
 function autoloading
From: Danack@basereality.com (Dan Ackroyd)

On Thu, 13 Apr 2023 at 11:39, Jordi Boggiano <j.boggiano@seld.be> wrote:
>
> On 2023-04-12 22:26, Rowan Tommins wrote:
> > I could just about live with that example changing so that the
> > fallback was cached, but I definitely don't think an explicit call
> > like \foo\strlen('x') should become an implicit alias for
> > \strlen('x'), which is apparently the current proposal.
> >
> > I really like the majority of this proposal, but right now would vote
> > against it based on that.
> >
> I agree here, I missed that fact while reviewing the RFC as I just
> understood it as an internal cache that \foo\strlen => \strlen, but
> function_exists('foo\strlen') should still return false IMO until it is
> defined, and once it is defined the cache should be busted.


Thanks for the code example. Code is a great of being clear...
particularly when I've gotten the wrong end of the stick.

Yeah, that would be a BC break.

For now, I'll update the RFC words to say that, but that means we need
to go and do some thinking, to figure out how to proceed.

cheers
Dan
Ack