On 10 May 2011 21:42, Matthew Weier O'Phinney <weie...@php.net> wrote:

Sure - I just revisited my browser history today. I turned up at least dozen annotation driver implementations (including 3 separate PHP C implementations!) just with a casual google search.

Google Searches

----------------------- Lots of pages specifically about "PHP annotations" (9,470): http://www.google.com/search?hl=en&q="php+annotations" "PHP Annotations" + patch (5,000) http://www.google.com/search?hl=en&q="php+annotations"+patch PHP annotations (1.8 million) http://www.google.com/search?q=php+annotations - you can see lots and lots of chatter about PHP annotations.

C patches for PHP

------------------------- http://labs.adoy.net/php-annotations.php (source: https://github.com/adoy/PHP-Annotations) http://code.google.com/p/addendum/ https://github.com/marcelog/AnoForPHP Guilherme Blanco's of course

PHP based annotation drivers

--------------------------------------- https://github.com/schmittjoh/annotations https://github.com/jubianchi/PHPAnnotations https://github.com/oes/pia http://sourceforge.net/projects/phpannotations/ https://github.com/veritech/Annotations (for CakePHP) https://github.com/domain51/Domain51_Tool_Annotation https://github.com/doomspork/sleepy PHPUnit :-P Doctrine and so on... ... and these are just projects which have written annotation drivers found during a quick search.

There is clearly a lot of interest from PHP userland in annotations, evidenced that annotations have been being used for years already, most notably by PHPUnit from so long ago. As frameworks adopt annotations the use of annotations will escalate more and more and it's already happening since many people use PHP frameworks: people using the frameworks inevitably use the technology in the framework. Also logically, people would not be writing drivers if there was not a demand for it. I feel like I am stating the obvious :)

If PHP accept the patch (in it's current or a modified form) it should be feature complete. There is no point neutering a perfectly good feature at it's inception. Whether it's in the docblock, or as Guilherme's designed in the code, annotations should accept more than key=>value.

PS - sorry to say this but from the other thread, all this talk of ecosystems is quite strange and full of FUD. The PHP eco-system depends on PHP and exists only because of PHP, not the other way round. If PHP adds a new syntax or new functions, the IDEs have to support it - it's only logical or they will lose users to another product which does support PHP properly and in a timely manner. IDE vendors provide a solution for people who need a PHP IDE so it only goes to reason surely? Given their scale and userbase I cant see Beans and Eclipse PDT ignoring new syntax any time soon nor any of the commercial IDEs like Zend Studio or phpStorm. If they managed with namespace support I don't see why this would be any different. I feel like I'm stating the obvious.