[RFC] [Vote] Final constants

  114522
May 19, 2021 08:02 kocsismate90@gmail.com (=?UTF-8?B?TcOhdMOpIEtvY3Npcw==?=)
Hi Internals,

2 weeks of discussion period has passed, so I'm opening the vote. RFC link:
https://wiki.php.net/rfc/final_class_const, previous threads:
https://externals.io/message/114066, https://externals.io/message/114273

The vote will be closed on 2021-06-02.

Regards:
Máté
  114562
May 23, 2021 23:42 george.banyard@gmail.com ("G. P. B.")
On Wed, 19 May 2021 at 09:03, Máté Kocsis <kocsismate90@gmail.com> wrote:

> Hi Internals, > > 2 weeks of discussion period has passed, so I'm opening the vote. RFC link: > https://wiki.php.net/rfc/final_class_const, previous threads: > https://externals.io/message/114066, https://externals.io/message/114273 > > The vote will be closed on 2021-06-02. > > Regards: > Máté >
I'm voting against this because it makes interface constants overridable by default. And this means now an interface which extends another one can override the value of it, something which is currently not possible: https://3v4l.org/l4OOB The current behaviour for classes might be unfortunate but can make sense. If this would only target class constants my vote would probably be different. Best regards, George P. Banyard
  114563
May 24, 2021 08:14 kocsismate90@gmail.com (=?UTF-8?B?TcOhdMOpIEtvY3Npcw==?=)
Hi George,

Being able to override interface constants by default might seem like a
regression, but I believe making interface constant behavior consistent with
class constants is worth this, and also, I don't see any other good way to
fix the weird inheritance behavior Nikita highlighted than to do what the
RFC
currently proposes.

Regards:
Máté
  114690
June 2, 2021 08:05 kocsismate90@gmail.com (=?UTF-8?B?TcOhdMOpIEtvY3Npcw==?=)
Hi Internals,

I've just closed the vote, and the RFC is accepted with 28 votes in favor
and 4 against.

Regards:
Máté

Máté Kocsis <kocsismate90@gmail.com> ezt írta (időpont: 2021. máj. 24., H,
10:14):

> Hi George, > > Being able to override interface constants by default might seem like a > regression, but I believe making interface constant behavior consistent with > class constants is worth this, and also, I don't see any other good way to > fix the weird inheritance behavior Nikita highlighted than to do what the > RFC > currently proposes. > > Regards: > Máté > >