Sam Doidge

Continual improvement

PHP London 2012 Conference

Roundup: Was it worth going? Yes. There is a lot to take in and it got my brain working. As programming conferences goes it was a riot ;)

I was impressed with Rasmus Lerdof’s answers on the panel at the end of the Friday (the day I attended). Is PHP still viable for a startup to use? Yes. From the Facebook Blog:

“As a programming language, PHP is simple. Simple to learn, simple to write, simple to read, and simple to debug. We are able to get new engineers ramped up at Facebook a lot faster with PHP than with other languages, which allows us to innovate faster.”

Lerdof mentions how Facebook was getting enough performance from it up until they developed HipHop, in 2010 – which transforms PHP into highly optimized C++. So until you are the size of Facebook was in 2010, PHP is fine, and by the time you are the size of Facebook in 2010 – you should have enough money to introduce whatever efficiency enhancements needed.

Another topic: PHP doesn’t offer threading / allow asynchronous operations. Rasmus says PHP isn’t designed to do that, PHP is supposed to be simple, you don’t want it doing that – pass it off to some other technology to handle.

I enjoyed the talk by June Henriksen on Creative Coding – talking about what we (science) know about creativity and how this affects us in programming. I got me from it than just this, and it was a nice change from looking at syntax.

Derick Rethans shows us some profiling in PHP – and he was honest and frankly a little brave – he was showing how-to via the terminal for us live to demonstrate, and things weren’t going exactly as he predicted – but as a group we quickly debugged and could see profiling in action. I like that he wasn’t afraid to do try this infront of the group – others had screenshots, and that worked will too, but this was a nice change – he was enthusiastic and really came across as knowing his stuff.

The final panel also included Hugh Williams from Ebay – and he said how almost any user on the site is going to be in some sort of experiment – the site they have presented to them will often differ to another user in a slight way. I found this really interesting – and efficient – more large sites will have to take this approach with users to stay ahead. Not testing or analysing a users visit is a wasted opportunity to improve the website as a whole.


Insightful, thought provoking stuff: just make sure you choose what talks you goto well.