<<< Previous speaker next speaker >>>

Gareth Rushgrove, Web geek, GOV.UK

 Gareth  Rushgrove

Gareth Rushgrove is now a developer at Government Digital Service, part of the UK Government.

In the past Gareth worked on everything from successful marketing campaigns to enterprise content management and financial service applications. These days he's more likely to be found persuading people of the benefits of small web service based systems, automation and message queues.

Gareth has written articles on topics from mobile web design to facebook and website performance to javascript for the likes of Vitamin, Digital Web and Opera. He also featured in the 2007 edition of 24ways, the annual web design advent calendar.

When not working, Gareth can be found blogging over on morethanseven.net or uploading code to GitHub. He's previously organised local BarCamps, organises the Django user group in London and helped out on the board of the hugely interesting Thinking Digital conference.

Blog: www.garethrushgrove.com

Twittter: @garethr

Presentation: "Web Framework Application Performance Stories from Django and Rails"

Time: Friday 11:40 - 12:30

Location: Seacliff AB


Speed is good. But making web applications fast is nuanced, and never as simple as just picking the fastest language. Modern web frameworks can be large, but that doesn't mean you have to sacrifice performance - in fact many of them provide tools or hooks to make measurement and optimisations easy.

This talk will delve into the world of Ruby on Rails and Django, two popular web frameworks written in Ruby and Python respectively. We'll look at what tools they provide out of the box and at some simple configuration options that can make a big difference. We'll also delve into a world of dynamic plugins, libraries and third party tools written to make finding and resolving performance problems easier.

The entire stack from front to back should get a look-in, from object caches, fragment caching approaches and HTTP caching up to, asset compilation, profiling, log file measurement tricks and framework hooks for instrumentation. We'll end with an attempt to identify a toolkit that any upcoming framework ecosystem should aim to provide if they want their development community to build performant applications.