As the ideas behind responsive web design mature, and our skills and experience with them grow, we're faced with a gap between what the tools and technologies can do and what we'd like them to do.

We often end up trying to fill the gap with javascript and using it not just for behaviour but to change appearance and content too.

Javascript might be the right tool for many things, but should we really be embedding attributes that control appearance in our scripts, or might a declarative approach that uses CSS to change HTML attributes be clearer?

I've haven't written a post in ages, partly because I'm really bad at finishing posts, and partly because I've never really like typepad and so haven't paid for it this year.

Twitter is great for short conversations but it's not up to longer posts and I've got some ideas I want to share so that others can kick the tyres.

So I've switched domains, fired up Octopress, and with the help of pandoc ported the old blog content across. Once I've got the redirects in place I will be killing off the old blog.

Octopress was really pretty easy to get going, but I'm not quite sure if I'm going to stay with it as I want to experiment with some RESS type things.

I've still got some pages to write, some tweaks to the theme I want to make but it's good enough for now.

The slides from my recent talk at DigiTalks Cheltenham:

Web Performance - A Whistlestop Tour

In November 2010, Ben Strong highlighted how Google and Microsoft were cheating on TCP Slow-Start by setting the initial congestion window higher than the RFC value.

Since then there’s been quite a few articles discussing the performance benefits of increasing the TCP initial congestion window (initcwnd). Most of the articles showed how to make the changes on Linux but there was no coverage on whether and how it could be increased on a Windows Server.

Both Steve Souders and Joshua Bixby have highlighted the effect slow third party components can have on overall page load time.

Third party components have become common e.g. ads, links for bookmarking, sharing, widgets that pull in content from other sites etc, and the impact on our page load times depends on the quality of service they offer (both how the component is implemented and the reliability/speed of their service)

Steve and Joshua have highlighted approaches that can be used to decouple the components and reduce their impact on the performance of our pages but it’s early days and only some providers have implemented them (Sergey Chernyshev has a some thoughts too).

The challenge for web developers is understanding what the effect on page load times could be, when most of the time the providers’ networks remain reliable.