I have been kicking around this idea for a while, now, and decided to see if anybody else thought it was a good one. It seems to me that having the core (and maybe libraries) of a framework pre-compiled as a PHP extension would offer some great benefits.
Think of how many files have to be read and parsed just to make the necessary core classes available for instantiation on each and every request when you use any framework. On a modern web server, this happens very fast - it's not as though we're drumming our fingers on the desk waiting. However, as a great lover of efficiency, I'm really enthralled with the idea of compiling the core classes so they're always ready to go, such that the app can get down to loading and running the stuff it needs for each particular request. The place where the savings would really add up, of course, is something like a hosting environment, where there could potentially be hundreds or thousands of apps all running the same framework files for every request, all day long.
So, my question to you is this: would people use a framework that comes from PECL as a compiled extension? If I build it, will they come?
And if they will, would the Fuel team like to offer a compiled extension as an option for Fuel? I figure there will always be a market for the script-based solution, and it's easier to maintain than an extension, so that product isn't going anywhere. I was thinking of something like the Ubuntu LTS model for an extension, though - a periodically released version with all of the latest minor revisions rolled up, compiled, and ready to serve until it's time for the next big update.
From what I've seen so far, I really like the direction Fuel is taking as opposed to CI or any of the other variants out there. I'm impressed and intrigued. So, if the community and the development team are interested, I'm potentially offering to develop a parallel branch of Fuel in C to be compiled and made available as a PECL extension for PHP.
Donate
