Persistent Connections have been a mixed bag. They can give you a significant performance boost by caching (pooling) connections although MySQL is already comparatively fast at establishing connections. However,connections are stored "as-is" in the cache. They are not "cleaned up".

The ext/mysqli driver takes care of this and a few other problems surrounding the persistent connections by cleaning up things like rolling back active transactions, unlocking tables, closing prepared statements and closing handlers. The trick is in a call to the C-API function mysql_change_user() (= COM_CHANGE_USER).

I like PHP, I really do. PHP5 is a great language and was a huge step in the right direction, unfortunately the community doesn't agree. It is not that the PHP community thinks PHP5 is bad, it is just that they don't agree on anything really. I am starting to wonder if the lack of cohesion is going to bring real trouble to success of PHP as a language.

Though he admits to no longer writing PHP, he's still interested to see it succeed. He just doesn't see the structure the language/community needs to make this happen and make the languagde advance. Jackson also comments on the fact that there doesn't seem to be one "solid" framework for the language, and that even the Zend offering seems on shaky ground.

He also looks at the "downward spiral" things started taking around the time PHP5 was just coming into view - arguing over petty issues, delays of releases, features being added and removed.

The luster was tarnished and the community hasn't recovered. I hope I am wrong, but it looks like it is not going to improve anytime soon.

PHP has become the BASIC of web application design, used primarily by designers without too much computer science background. The platform hasn't taken well to multiprocessing, and is doing poorly enough from a security standpoint that "PHP security" conferences have been hastily organized. PHP's woes aren't so much a problem with the PHP core as with existing libraries and applications on top of PHP and their lack of future-proofness. If PHP is to be faulted, it is because the platform designers never placed an emphasis on good software design by their users - for example, most PHP developers don't write using a model-view-controller paradigm.

While organizations invested in PHP will band-aid their existing code, new projects will move to other platforms, with designers more cognizant of both software engineering and application security.

Personally, I think that PHP, while not to the point he mentions here yet, is growing to that point rather quickly. More and more I see highly skilled PHP developers out there coming out with quality projects/products to help "the little guy" get up and running - *and* work within some standards...]]> Tue, 20 Dec 2005 06:49:57 -0600