Two Reasons You Should Look at SPDY: 58.23% and 64.33%
Google comes up with lots of interesting technology: PageRank, BigTable, self driving cars. The most interesting invention to me at least recently has been the development of SPDY, Google’s faster HTTP replacement protocol. But will SPDY take off like Gmail? Or end up like Dart, (or Wave, or pubsubhubbub…) Luckily I’ve seen two very important signs in the last few weeks that SPDY has the potential to be hugely successful.
First, Google’s Chrome web browser, which natively supports SPDY, surpassed Internet Explorer in a Stat Counters browser usage survey. While browser stats published by tech news sites analyzing their own web server logs have shown this trend for a while now, Stat Counters is the first mainstream survey where IE is not the top browser. The growth of Chrome has been simply stunning.
The second sign was Mozilla’s release of Firefox 13 this week. While Firefox has support SPDY since version 11, Firefox 13 finally has SPDY supported enabled by default. Given Firefox’s move to automatic updates, adoption of Firefox 13 should be rapid.
Why is this important? Chrome’s share of the browser market is 32.76%. Firefox’s market share is 24.47%. That means that 58.23% of all web browsers support SPDY.
With the majority of web browsers supporting SPDY, website owners (and other web server developers) are much more likely to spend the time and resources implementing SPDY support now that so many people can take advantage of it. Of course, it gets even better. In April, Google released mod_spdy, which enabled SPDY support on Apache. Apache is by far the leading web server with a market share of 64.33%.
SPDY isn’t the only HTTP replacement effort out there. Microsoft has their own proposal as well. It’s called HTTP Speed+Mobility. But with this much momentum around SPDY, it’s likely that HTTP Speed+Mobility will die before it even gets implemented. Consider this: even if Microsoft implemented their proposal in both IE and IIS, the majority of web browsers and web servers still would not support it. And that’s not even considering mobile, where SPDY’s efficient use of network connections via multiplex is arguably more impactful than on the desktop, and in which Microsoft a marginal player at best. Ironically it is now Microsoft at the mercy of the (web) technology monopoly holders!
So, from a pure infrastructure point of view, SPDY has enormous potential to take off. However SPDY has some interesting design decisions (such as requiring SSL) which will limit its actual adoption, regardless of whether support for the protocol exists. I’ll discuss those aspects of SPDY, how to get your website ready for SPDY, and much more in a future blog post. So stay tuned.