Tim set a simple, but very much real-world challenge; write an app that determines the top 10 most popular blogs from his Apache access log. It should be fast and readable, with a subtext of illustrating how "language X" copes in terms of parallel processing and utilizing "wider" (many processor) systems.

Since PHP natively doesn't support multi-threading (well), Harry opted to go with an approach using curl_multi_exec instead. There's two pieces to the puzzle - the mapper to grab the information and extract the data and the reducer that makes the calls to grab the information from the log files.

Tim first introduced his comparison basis (Scaling, Dev Tools, Dev Speed and Maintainability) and explained his views on the 4 topics and which keywords he considers under each of them. This introduction was really good and well-founded. After that, he showed and explained a diagram to make the actual comparison.

There is a graph included in the post to help make his point in comparing the three languages (PHP, Ruby, and Java) on how well they can perform. PHP bested the rest in Dev Speed and Maintainability, but lost out in Dev Tools and Scaling abilities. Tobias has a few disagreements with Tim's findings, particularly on the topics of Dev Speed and Maintainability.

Tim will be at ZendCon this year participating in a panel discussion titled "How Do The Stacks Stack Up?" I talked with Tim by phone because I was curious why PHP developers should consider Solaris as a development and deployment environment. Here's what Tim had to say.

His answer was based around three main points - observability, virtualization "stuff", and the ZFS filesystem all Solaris systems come installed with. And, of course, he explains the thought process behind each (briefly).