Measuring the Coastline
I’ve got big ideas for tota11y. It’s been a passion project of mine for over a year now, and has not gotten nearly the love it deserves.
These ideas have been written down a few times. Sometimes they’re brief, sometimes they’re super detailed, and sometimes they’re swapped out for an entirely new set of ideas. But each iteration has one thing in common — I haven’t actually done anything.
Because each time I sit down, I attempt to solve a few core problems (the details of which are not really important for this post). (1) The amount of accessibility errors on your page can be overwhelming (2) The error annotations are not user-friendly (3) tota11y is too limited as a bookmarklet.
These ideas balloon, and I inevitably hit obstacles. “Okay, so if I want to mark errors as known, how should I store them?” “Should I let users choose an arbitrary file location so it syncs with Dropbox?” “Should I let users specify their own storage mechanism?” “Should I build an intricate plugin system for tota11y to generalize all of this?”
(1) becomes (1a) which becomes (1aa) and (1ab) which becomes (1aaa) and so on and so forth.
—
The Ordnance Survey claims that the island of Great Britain has a coastline measuring 11,073 miles in length. The British Cartographic Society disagrees with this figure, and adds some larger islands to its calculation of 19,491 miles. The CIA Factbook has its own ideas, and claims the coastline measures just 7,723 miles.
What gives? See, the coastline of Great Britain is very bumpy. And when we zoom in on these bumps we see more and more bumps.
A 10km ruler may skip the neighborhood-sized bumps, a 10m ruler will skip the human sized bumps, but a 1mm ruler gets caught in all the tiny cracks and crevices which make up the British coastline.
Clearly, the coastline of Great Britain is made up of some parts. But how do we decide what those parts are? Should we draw a box around Great Britain and move on with our lives? Should we standardize a unit length? Should we slowly but surely measure each individual particle that crosses our path?
Since distance is a social construct, should we just ask Twitter?
—
My ideas for tota11y are made up of parts too. But I have no idea what those parts are. I have no idea what tiny little problem is going to derail my entire afternoon. I have no idea what the best approach to storing user settings is. I have no idea what library will work best for pinning indicators next to violating HTML elements.
Obviously I can figure these problems out — I’ve done so before, and I do so every day at Khan Academy — but on my side projects I overextend my creative freedom and strive for a perfection that I think I can figure out with pen and paper alone. I’m attempting to measure the coastline for tota11y, and it’s not doing anyone any favors.
I have no idea what subtle UX I’ll add to the tool. I have no idea how I want to organize my test results. I have no idea if people will even like my ideas for a refresh because I haven’t tested them.
So I sit here, typing away, trying to predict the future with perfect accuracy so that I can work efficiently. But it’s not efficient, because I’m not actually doing anything. tota11y hasn’t seen a new feature in almost a year, and at this point I’ve created a virtual mountain of work for myself, when in reality I have no idea if I actually have to do any of it.
—
I’ve started to cherish those tiny bursts of inspiration, and converted some of them to code. I’ve made a bit of progress, I’ve got a directly of experiments, and things are slowly moving along. I’ve also learned that I am not alone in feeling this way.
There’s a best way forward here, and I don’t quite know what it is. If I’m certain about one thing, however, it’s that sitting around doing nothing ain’t it.
