Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:82271
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: pass (pb1.pair.com: domain gmail.com designates 74.125.82.176 as permitted sender)
Message-ID: <54D89959.6060904@gmail.com>
Date: Mon, 09 Feb 2015 11:26:17 +0000
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:31.0) Gecko/20100101 Thunderbird/31.4.0
MIME-Version: 1.0
To: PHP internals <internals@lists.php.net>
References: <8703B53E-2C4A-4AC6-95C4-D4F19C6D5221@ajf.me>	<CE494877-8C5A-4315-BBF6-089FBD83F148@ajf.me>	<048c002a67c3d75afd802025b6946362@mail.gmail.com>	<54D7F1AB.1070302@gmail.com> <CAEZPtU4CR9hGGRHXCK6v2E3jx5AABAqbFnvTj84UnwgRq4Zjrg@mail.gmail.com>
In-Reply-To: <CAEZPtU4CR9hGGRHXCK6v2E3jx5AABAqbFnvTj84UnwgRq4Zjrg@mail.gmail.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 7bit
Subject: Re: [PHP-DEV] Syntactical change to Scalar Type Hints RFC
From: rowan.collins@gmail.com (Rowan Collins)

Pierre Joye wrote on 09/02/2015 00:05:
> On Mon, Feb 9, 2015 at 6:30 AM, Rowan Collins <rowan.collins@gmail.com> wrote:
>> On 08/02/2015 20:33, Zeev Suraski wrote:
>>> FWIW, while I think strict types - stricter than even strict languages -
>>> don't belong in PHP, this syntax is clearly a step up from declare(),
>>> which
>>> was definitely not intended for this purpose.
>>
>> I'm kind of intrigued what purpose it *was* intended for. I've always found
>> it a rather odd part of the language, but this seems as logical a use for it
>> as any.
>>
>> For years, it had exactly one "execution directive", which worked in concert
>> with a runtime event handler (ticks=N + register_tick_function). In PHP 5.3,
>> a completely unrelated directive was added, to specify the encoding of the
>> file (something that more obviously has to be detected by the compiler).
>>
>> Was there an original idea of what it would be used for that never came to
>> fruition? Or is there something I'm missing that connects "ticks" and
>> "encoding", but excludes "strict_types"?
> Both are being done at compile time and modify how a file is compiled.
> So no, I do not agree with "runtime" here, while one directive is
> being used at runtime.

I didn't say it was runtime, I said it "worked in concert with" runtime 
code. My point was that declare(encoding) is something purely internal 
to the compiler, affecting how it parses the file, whereas 
declare(ticks) is using the compiler to change the actual execution of 
the code, which makes them feel very different in effect.

It's quite hard to generalise from those two examples to any real 
"vision" of declare()'s purpose, so I was genuinely curious what that 
vision was.

Regards,
-- 
Rowan Collins
[IMSoP]