By Simon Willison, Program Chair, Developer Track
Tim Bray is the Director of Web Technologies at Sun Microsystems. He has a long history of involvement with technology standards, including co-editing the XML 1.0 specification and more recently co-chairing the IETF Atompub Working Group. Tim will be presenting the morning keynote on Friday the 10th of October.
Your keynote is titled “The Fear Factor: What to be Frightened of in Building A Web Application”. Can you give us a few hints?
Based on the events of the last week, we may be called upon to get useful work done without any capital investment for infrastructure. Seems scary to me. The other front-of-mind worry is the yawning disconnect between the way enterprise software development is done and the best practices emerging in the Web2.0 and OSS spaces.
Are there any skills a web programmer should be developing now that will better prepare them for the next few years?
Most obviously, don’t be “an X developer” for any value of X, whether it’s Java or .NET or PHP or Rails. A command of more than one technology makes you more employable and also deepens your understanding of each individual one.
You’ve long been a champion of developing with multi-core systems in mind, but parallel computing is still too complicated for many developers. Are there any silver bullets on the horizon, or are we all going to have to hunker down and learn Erlang?
No silver bullets. For conventional shared-nothing web deployments we’re in pretty good shape, but compute-intensive and communication-intensive programming is getting tougher and tougher.
What’s the most interesting new idea you’ve seen in the field of web development in the past year?
New database architectures: CouchDB and friends.
PHP is the web programming language that refuses to die. What will it take for an alternative such as Ruby or Python to rival it in popularity?
PHP has sufficient deployed base that it’s never going away. Plus some truly great apps, like WordPress and MediaWiki, are written in it. On the other hand, for new app development, I think things like Django and Rails, with better maintainability stories, are already stealing market share fast.
Are there any emerging web technologies you think don’t get enough publicity? Any hidden gems?
Well, I spent the last few years of my life working on AtomPub and I think it’s a real winner, but it’s catching on pretty well so I wouldn’t say it’s “hidden”.
What’s the one piece of open source software that you really wish someone would write?
A library or framework to make concurrency easy in high-level languages like Ruby or Python.