Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:66117
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: error (pb1.pair.com: domain lerdorf.com from 209.85.216.175 cause and error)
Message-ID: <51263D89.5040308@lerdorf.com>
Date: Thu, 21 Feb 2013 07:30:17 -0800
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:17.0) Gecko/20130106 Thunderbird/17.0.2
MIME-Version: 1.0
To: Pierre Joye <pierre.php@gmail.com>
CC: Zeev Suraski <zeev@zend.com>, Lars Strojny <lars@strojny.net>, 
 Derick Rethans <derick@php.net>,
 PHP Developers Mailing List <internals@lists.php.net>
References: <alpine.DEB.2.02.1302201817350.5312@whisky.home.derickrethans.nl> <678597E6-E3A8-42E0-8DFC-F8382C9DFB41@strojny.net> <8d4e991084a1313844910ec0168eacdf@mail.gmail.com> <CAEZPtU5RZbVJS0txYRxxpTXzvk0ZvQdZBfx7qHTYwAd4cyn6Rw@mail.gmail.com>
In-Reply-To: <CAEZPtU5RZbVJS0txYRxxpTXzvk0ZvQdZBfx7qHTYwAd4cyn6Rw@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Subject: Re: [PHP-DEV] Give the Language a Rest motion (fwd)
From: rasmus@lerdorf.com (Rasmus Lerdorf)

In the slice of the "community" where I spend most of my time,
medium-to-large companies using PHP with their own custom code on
hundreds to thousands or even 10's of thousands of servers, neither
annotations nor getter/setter are anywhere on their wishlist radar. What
they most desire is performance, robustness and security. They would
love to see a PHP release that had no syntax changes, no BC changes, but
was twice as fast and crashed half as much. I realize this is just one
(small?) slice of the community but so is the part of the community
wanting annotations. This is the balancing act we have to perform. It is
not stubbornness, nor living in the past, it is recognizing that each
and every major feature addition has a destabilizing effect on the
codebase and if the addition only serves a small slice of the userbase
we have to think long and hard about the merits of it.

Personally I would love to see more RFCs focusing on performance and
less on syntax changes. Of course, a syntax change RFC, and even the
initial (often shaky) implementation of a syntax-related change is much
much easier to whip up than the deep analysis, profiling and creativity
required to find and come up with ways to make a complex piece of code
faster or use less memory.

-Rasmus