Yesterday Horace Dediu tweeted:
A browser is an infinitely flexible interface, but is it the best interface for everything? Apps allow experiments in new interaction models
The browser is not the most advanced interface there is. It’s too easy to build the wrong features into something as flexible as a browser, and once a badly-designed feature gains traction it’s impossible to get rid of it. (See HTML5 drag and drop.)
We wanted access to DOM elements by CSS selector. Browsers didn’t support this, so we wrote scripts. Then browser vendors woke up and added native support in the form of
iPhone apps introduced swipe interaction for switching between pages (states?). The browsers can’t do it, and people started to write scripts to emulate the behaviour. Any day now the browser vendors will come to the remarkable conclusion that there should be a CSS declaration for swipe transitions.
By building a native app you win room for experiment, but lose the reach a web app offers. That’s nothing new, really, but it’s important to note that the browser is designed to be somewhat behind the state of the art, and makes up for it by its reach.
I’m around at the following conferences: