Erics Predictions for the Future of the Web

So Ive been making these statements on the future of the web for quite some time, but just so I can prove I was “there first”, I thought I should put them into writing!

Prediction 1: Browsers will download complete environments from a web site to display a web site.

And no, I dont mean like what we have today, where a browser downloads markup instructions, image assets, and some JavaScript to glue it together. Or, where a website is written in Flash, and runs in the Flash plugin in the browser. I mean a full environment, where I, the web site content developer control everything, and provide it all in a simple standardized method to you, the browser who consumes the site. I may write my sites client code in Ruby, Erlang, Lua, or build a strange 3D world. And the browsers will expect to visit a site, find the download links, and automatically start up these environments, in a safe clean sandboxed environment as simply as they find favicon.ico today. Heck, I could see a browser downloading and firing up a VMWare image of Windows 9 to browse “My Extremely Enterprisy Website”.

By allowing browsers to download and execute really rich code natively, without weird hacks like Google Gears, we eliminate the requirement that all advancements in HTML standards be matched in lock step by all the browsers. Especially with the rise of mobile devices, I expect to see more and more browser types, not fewer, so keeping them all up to date and current will be harder and harder. And we open up the ability of content developers to deliver their content in whatever way makes sense, not just in whatever way the browsers support.

Prediction 2: “Connect” apps will become just part of the internet

Credit for the word “Connect” apps goes to David Recordon, however Ive been thinking for a while that all these social graphs just duplicate the graphs of pages on the internet. And I predict that eventually something like OpenID will be your entry point into these graphs. Youll browse to and depending on who you are, Facebook, LinkedIn, or a random stranger, youll get a list of nodes that describes my links to other entities, regardless of if they are people, sites, content, etc.