Photo of a laptop and monitor on a desk

Usability—In 1883?

Updated

Similarity of Websites and Machines

When we think about the usability of a website, we tend to think of it as an inherently modern problem—how can it predate the web? Of course web usability can’t, but the problems of finding the information or functions we need are much older. A natural metaphor for finding information is a library, but what about finding settings and functionality? Many web sites have complex privacy settings, or provide sophisticated functionality (such as a metadata auditing utility!); these aren’t just information to consume, but things that we interact with.

Illustration of a 19th-century latheIllustration of a 19th-century lathe

When using a website, we tend to have a sense of where these controls and functions are located, based on where they’ve been on similar sites we’ve used in the past. When this works, we can quickly find what we need, but if the new web site doesn’t place the things we need in the places we expect, it can cause frustration or even cause users to give up. When the functionality is important, like account management or privacy settings, this can be a serious problem.

In 1883, a tongue-in-cheek article about oiling machinery was reprinted in the magazine Mechanics (vol. 3, pp. 474)—sadly without attribution. In general, the moving parts of machinery require lubrication, and will fail without it; in 1883, this often required significant regular application of a little oil through many small oil holes, each lubricating a specific point on the machine. Complex machines had many such points that needed attention to keep the machine in good working order. A common machine tool, made by many companies in many configurations and used in many workshops of all sizes, was the engine lathe for turning metal; by this time, these machines had complex devices for controlling the movement of the cutting tool, with many controls—all mechanical, and all relying on moving parts that needed to be lubricated. Setting aside some of the technical specifics, consider this passage from our unknown commentator:

“You have a lathe in the shop with oil holes in the apron [a complex assembly with many of the lathe’s controls] for every stud and gear except one or two. When you put up the lathe you notice these places and resolve to keep them oiled; so you make a tin eave spout to run oil into them, and hold it in one hand and pour oil into it with the other. That is all right, but after awhile [sic] you put a new hand on the lathe who never saw that make of lathe before. He knows that some machines have hidden oil holes, thinks he will oil all the places he can see, and trusts to Providence until he learns the machine better. He finds out at the end of two or three weeks that Providence is not in the oil hole business.”

Consider, then, the analogy with a website—which, after all, can be considered a sort of “tool” used to accomplish a task. What functionality might be “hidden,” and inconvenient for the user to find? What might not be where the user expects to find it, based on previous experience with other sites? If the user doesn’t find them, what could go wrong? In the case of the lathe, by the way, the result could be the machine literally grinding to a halt, loudly and expensively, when a critical gear seizes!