Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:114196
MIME-Version: 1.0
References: <V7C5-r6tzKmxjhPnvy9gtfdMtxkj0z_PJphaehaqcPsOLX0pEKtGU4A8EuSKyS-cbZl5HPVA2ejXCuXXOfOksH9VACGOKT3awRsr9aSNIg8=@protonmail.com>
 <5b9f1500-615a-48f1-815f-1d48b327ef90@processus.org> <CAEPJdun9jhST-L7CPHSJZtg3Q9rFEy7aFXU0Y7NCv_cPMHrVEg@mail.gmail.com>
In-Reply-To: <CAEPJdun9jhST-L7CPHSJZtg3Q9rFEy7aFXU0Y7NCv_cPMHrVEg@mail.gmail.com>
Date: Tue, 27 Apr 2021 14:21:34 +0100
Message-ID: <CA+kxMuT-yup+ghAAw5EmbX9Lqmj+B_WEVm029jwen16MkY=pAQ@mail.gmail.com>
To: Benjamin Eberlei <kontakt@beberlei.de>
Cc: Pierre <pierre-php@processus.org>, Saif Eddin Gmati <azjezz@protonmail.com>, 
	PHP Internals <internals@lists.php.net>
Content-Type: text/plain; charset="UTF-8"
Subject: Re: [PHP-DEV] [RFC][Draft] Sealed Classes
From: Danack@basereality.com (Dan Ackroyd)

On Sat, 24 Apr 2021 at 15:03, Benjamin Eberlei <kontakt@beberlei.de> wrote:
>
> There is a much less invasive way to add new keywords/flags
> to functions by using attributes.
>
> Imho this decouples new features from the language and reduces the "risk"
> of adding them to the language.

I think I disagree with this very strongly, and plan* to vote against
any RFC that embeds another language in annotations.**

It might be quicker, easier, and more seductive to implement language
level features in them, but it is a massive trade-off in making code
hard to reason about.

> That should increase the likeliness of it
> getting accepted in my opinion.

This can also be pronounced as "makes it more likely to slip bad ideas
in core without thinking them through fully".

My experience of annotations being used as an embedded language is
Java, and it's one of the reasons why I no longer use that language.

Not only is code with many lines of annotations hard to read, it
results in behaviour that is very hard to reason about. I literally
spent 40 hours (spread over the course of 5 weeks), trying to figure
out this bug https://stackoverflow.com/q/9072749/778719 . If my
colleague hadn't been able to tell me the answer, the only way I could
have debugged the problem myself is using a bytecode level debugger,
to step through the internal details of what was happening.

I'd prefer it if we didn't repeat (what I consider to be) the mistakes of Java.

cheers
Dan
Ack

* possibly with the exception of optimization annotations e.g. 'memoize'.

** "From a language design perspective, annotations form a
mini-language embedded in Java" -
https://blog.softwaremill.com/the-case-against-annotations-4b2fb170ed67
Though probably most of the other links from "java annotations are
crap" are also appropriate.