This morning I was curious to see how the battle for domination between
PHP frameworks was shaping up, and which one was most economically
sensible for a developer with limited time to learn. I thought I'd share
my results with the list, as this may be of interest to some of you.

To gauge the demand for knowledge of particular frameworks, I did an
informal survey of job descriptions on the SF Bay Area Craigslist's job
posting. (I did not look act gigs--computers.) I entered the query "php
framework" and got 38 results. Of these, I threw out results that were
looking for someone to work primarily in another language (such as C#,
Java, Python, etc.) or doing something not particularly PHP oriented,
such as client-side development. I also threw out duplicates (there were
several) and positions where the word framework was used to refer to
something else, such as the LAMP technology stack, or an aspect of the
application to be built. I eliminated 18 positions, leaving me with 20
descriptions.

For my purposes, I defined a framework as a piece of prebuilt software
that facilitates the development of Web applications using the MVC
design pattern. Smarty (a template engine), Doctrine (a standalone ORM
utility), and Drupal were described as frameworks twice, once, and once,
respectively. Utilities for other languages such as Spring, Hibernate,
and Django also generated hits, but were not counted. If the job was
clearly a PHP job and mentioned knowledge of a framework or MVC design
patterns as desirable, I counted it as "Unnamed". If more than one
framework was mentioned, I tallied each one named. In those cases, no
job description surveyed named which framework the company used or
favored.

The results were pretty lopsided:

Unnamed: 8
Zend: 11
CakePHP: 4
Symfony: 4
Code Igniter: 2

Overwhelmingly, if only one framework was specifically named, it was
Zend. 

I also did a tally of JavaScript libraries named in qualifying
positions. The results are:

JQuery: 4
Prototype / Scriptaculous: 3
Yahoo! UI: 1
ExtJS: 1
Dojo: 1

Given my small sample size, I don't think these results are quite as
conclusive.

- David

Email Disclaimer: http://www.co.marin.ca.us/nav/misc/EmailDisclaimer.cfm