I like the general theme of this post very much — we do pay a price, often too high a price, for adding features to our software (and other things) — but this paragraph is wrong:

A perfectly blank sheet of white paper is a tool of infinite possibility. For input you could use a pencil, a pen, a crayon, a marker, a stamp, a brush or more. You could use all of those at once. You can write or draw or paint in any direction. Even multiple directions on the same sheet. You can use any color you want. How you enter data onto it and how that information is structured seems almost limitless. That flexibility and power is available to you because of [its] lack of features. In fact, it is featureless — devoid of them.

No, a sheet of paper has many features — traits — and they are worth noting. Compared to most things in our world, it is remarkably expansive in the two dimensions of height and width, considering its lack of depth. It’s also flexible and foldable. These features are sometimes wonderful (e.g., when you want to write a detailed letter and you only have one sheet of paper, which you can fold into a small square and stick in your back pocket) and sometimes regrettable (e.g., when the letter gets all crimped from staying in your pocket, or when you can't find anything flat and solid to write on). Most of the paper we see every day is also designed so as to receive quite easily all sorts of marks and impressions: the same sheet of paper can go through many different kinds of printers, can be typed on with a typewriter, and, as the post notes, can be marked on by pencils, pens, markers, paint brushes, and who knows what else. This is a feature, not featurelessness. Curiously enough, a piece of paper is rarely square, and as a rectangle offers us the choice of portrait or landscape mode; however, once one of those modes has been chosen it’s not possible to change completely to the other. Unless you turn the page over. We find the spatial proportions of an ordinary piece of paper — or a stack of such pieces, in the form of a codex or notebook — so appealing that we design electronic readers to resemble it. We like its receptiveness to marking so much that we keep hoping for someone to design a really excellent tablet-and-stylus computer. A plain sheet of paper, then, has a great many features, and all of them are worth thinking about. Similarly, when I write in a text editor rather than a word processor, that’s not because my text editor has fewer features than Microsoft Word, but rather because the features it does have are better suited to the task of writing.


  1. Alas, as a Linux/Emacs addict, I no longer have the patience to wait for a committee of software developers to decide what features I need. If some functionality doesn't exist in a piece of software, I just find another program that has it.
    Or, as is too often the case, I write it myself.

    All this, I suppose, confirms the blog author's point. The ability to rest content with what the developers provide is a virtue. Too many features/possibilities = too much tinkering and wasted time.

  2. The physics of the world we live in give many of the "features" that are assumed with a piece of paper. It is hard to imagine a piece of paper that you could not reorient and always pointed north no matter how hard you pushed on it (even a compass will give easily). So the paper doesn't have features in that it is material and we can assume physics/mathematics to work. In the computer world there is a different physics that we get to assume, for instance we can be pretty assured that if you don't save your work there is at least some lag (however small) between when you hit a key on the keyboard and when that byte hits the hard disk (or whatever your "permanant" storage may be). If you loose power you loose that byte (maybe more). The trick of making a good user interface is making a world where the user feels like they are interacting with material, not features. If you could just mimic the physics of the real world it would be simple.

Comments are closed.