Jump to main navigation, main content

Archived entry | Matt Wilcox .net

Mozilla Labs: Aurora browser concept

Adaptive Path are working with Mozilla to produce concept video’s for far-future internet browsers. While this is interesting, I can’t help but feel that the first showcase of Aurora misses the mark by a country mile.

First, be sure to check out the video. You’ll get the most out of it if you hop over to Vimeo and watch the HD version, but you’ll get the gist if you watch it embedded here:

Taken a look? OK, here’s my reaction:

Pure horror. This is really what someone feels the future should be like? The Aurora concept shown here seems to be suffering from the most severe bout of designer syndrome it’s possible to suffer from: absolutely everything is visual. Absolutely nothing conveys any inkling of meaning or use-scenario until after Jesse explains it to us. This design fails utterly for anyone other than a spatial thinker, and it’s not great for one of those either - as I’ll show later (unimportant side note: take a BBC quiz and find out what sort of thinker you are).

I felt lost from the moment it went into ’spatial’ view. While the concept is clever, I’m never going to use something even vaguely similar to this, and there are two huge reasons for that:

Firstly, why embrace this ‘digital data as physical object’ viewpoint? It smacks so hard of 1990’s thinking, when designers were scrabbling for metaphors to allow users to understand what the computer was doing. It smells of Microsoft BOB’s ‘house’ metaphor, with programs sitting on shelves and documents on virtual mantle-pieces. That metaphor was a dog then, and it’s a dog now. Haven’t we grown up beyond the need to cripple digital experience by representing it as though it were a physical experience? The one obvious and huge advantage of the digital medium is precisely that it’s not limited to physicality. An item can be in multiple places/categories at one time, and have multiple states at once, etc.

I don’t want objects shrinking and fading because of how old they are. I don’t want to have to have a bookmark live in only one ‘cluster’. I don’t want to have to hunt around to locate, and then squint to identify, a page/url by it’s tiny thumbnail. Who would?

The interface should adapt to my current method of data-hunting and sort accordingly - it should not try to represent all of the states of one item at the same time. e.g., if I’m looking for items in my history then show and sort the age states of all items whilst ignoring other factors - but if I’m not looking for their history state then ignore that property of the object, because it’s unimportant at this moment.

Secondly, 3D space for navigation of non-spatial concepts is never going to work, most especially when it’s interfaced through a 2D screen. It’s a simulated dimension, with non of the properties that make a true dimension useful. 3D space in the real world works because I can just flick my eyes around it and see everything clearly, all at once. It works because it is effortless for me as a 3D person to interact with the 3D world and because everything in a 3D world is a 3D object. You can’t do that on a screen without resorting to clumsy interface solutions, and the ’stuff’ that Aurora is throwing into 3D space are not objects but concepts. Further, those concepts lack any context in their space. You can’t ‘remember’ that you filed a url on the shelf because there’s no shelf. Just a disembodied collection of continually moving ‘pages’. For 3D space to work it needs the framework of an unchanging 3D environment into which objects can be spatialy contextualised and placed.

Interface wise, a mouse works so well because the exact motion you are making with your arm is the exact motion that’s happening on screen. But until you can reach into a screen, there’s no 1:1 perfectly intuitive input method for navigating 3D space on-screen.

Another problem with the 3D view is that the 3D environment changes over time. What makes 3D space work is that it doesn’t change until we make it change. If all of my books on my bookshelf started drifting into the distance and fading out based on the last time I read them, I’d have no idea where to find any particular book I wanted to read because it’d no longer be where I left it. In real space we don’t have a 3D anchor point, because 3D is real. But in this ‘virtual space’ the 3D anchor point is my screen. And that anchor point ‘moves’.

Things that are nice though: the zero chrome browsing concept and the ability to grab anything on the page and use it in other contexts.

I’m glad Mozilla are experimenting and being radical - but I do wish they’d not throw the baby out with the bathwater.


skip to comment form
  1. Jon posted 2hrs, 46min, 20sec after the entry and said:

    All interesting points. BTW, you need to add an RSS feed.

    Matt says:Thanks. Yeah, I need to re-do the whole site, this version never got finished. The new one is on the drawing board, along with some cool RSS features.

From the archives

Other enteries filed under:

My Two Cents

Web Development

Site information

Built with valid XHTML and CSS, designed with web standards and accessibility in mind. Best viewed in a modern browser [Firefox, Safari, Opera]

This domain and all content is a copy of my old website, for historical purposes only.