On Mon, Nov 3, 2014 at 10:39 AM, Larry Garfield <la...@garfieldtech.com> wrote:
> On 11/3/14, 10:37 AM, Stefan Neufeind wrote:
>>
>> On 11/03/2014 05:26 PM, Pierre Joye wrote:
>>>
>>> On Nov 4, 2014 1:24 AM, "Jonah H. Harris" <jonah.har...@gmail.com> wrote:
>>>>
>>>>
>>>> On Mon, Nov 3, 2014 at 9:11 AM, Chris Wright <c...@daverandom.com> wrote:
>>>>
>>>>> There are no current concrete plans and currently nothing being
>>>
>>> seriously
>>>>>
>>>>> discussed (at least, not publicly; I don't know if anyone has anything
>>>
>>> in
>>>>>
>>>>> pipeline that they haven't announced yet). The three RFCs you linked
>>>
>>> above
>>>>>
>>>>> are all basically dead.
>>>>>
>>>>> You are of course welcome to put together a proposal and/or start up a
>>>>> discussion on the subject if it is something you would be prepared to
>>>
>>> put
>>>>>
>>>>> work into.
>>>>>
>>>>
>>>> I, for one, severely dislike annotations. But, that's why there's an RFC
>>>> process :)
>>>
>>>
>>> I tend to think it is not a taste matter anymore. Symfony ecosystem
>>> (components, doctrine and co), Zend framework, etc use them. We see
>>> requests to work around user land implementation but we keep us away to
>>> get
>>> native support. Maybe it is time to the jump and get rid of our tastes,
>>> like years ago when we discussed which kind of OO we wanted in php. At
>>> the
>>> end of the day we do what we did not want back then.
>>
>>
>> The TYPO3-family (TYPO3 CMS, Flow, Neos) also use annotations.
>> So, yes it is used "in the wild" already and is there to stay. We can
>> imho just make it a bit easier to work with (maybe also performance-wise
>> in some cases) etc.
>>
>>
>> Kind regards,
>>   Stefan
>
>
>
> Drupal is now using annotations as well; not for the Symfony code we've
> inherited, actually, but for some home-grown systems for which we're using
> Doctrine's annotation library.
>
> Having first-class language support for metadata on definitions would be
> quite helpful, if for no other reason than native syntax checking and code
> assistance.  (And to help people get over the "it's code in comments!!!"
> problem, which is entirely because we have to put annotations in comments
> now as a hack due to the lack of native support.)

Whether the annotation is in a comment or not, the idea of changing
behavior at runtime based on the annotation is pretty magical. I
highly discourage using this type of feature whether it's in a comment
or not.

I will certainly vote no on any RFC on this subject, as I see it as
being significantly more harmful than helpful.

-- 
PHP Internals - PHP Runtime Development Mailing List
To unsubscribe, visit: http://www.php.net/unsub.php

Reply via email to