The Dignity of Wear

1. What would a trace let us know?

Imagine opening a research notebook with two marks on its cover: a crease in the corner and a note identifying the last revised chapter. Both suggest use, but only one tells you something definite about the work inside. The crease might be accidental. The note can be inspected against the chapter. That difference is where I would begin a digital account of wear.

The earlier version of this essay treated permanently fresh software as permanently disposable. That was too sweeping. A plain document can preserve years of annotations; a pristine interface can open an extensive revision history. Software already has ways to remember. The useful question is which memories a reader can understand and control, and where a visual trace helps them find those memories.

I use computational patina to name a design proposal rather than a physical law: a visible trace associated with an identifiable action or retained state. A texture applied to every surface at first load is a different thing. It may be a good illustration or an enjoyable material treatment, but it does not record anyone’s use. Confusing those two meanings lets a familiar-looking surface stand in for a history it does not possess.

A trace earns its place when it helps a person answer a question. Where did I stop? What changed since the edition I read? Which note did I decide to keep? These questions can be answered with an annotation, an edition comparison or a bookmark. Simulated scratches might accompany the answer, but they cannot supply it. My preference is to design the record first and the material metaphor second.

☞

My preference is to design the record first and the material metaphor second.

2. A small surface, with a declared limit

The reading study beside this essay offers a modest example. Activating its plate adds an illustrated contact mark. A keyboard activation puts that mark at the centre. The counter describes traces local to this visit, and Restore plate removes them. Reloading also starts a new plate. These behaviours are implemented in the study; they do not establish a measured relationship between contact, physical wear and affection.

The study makes a useful distinction visible because the event and its result are close together. You activate the surface, see a mark and can remove it. The mark is a drawing. It is not evidence that ink has diffused through fibres, that brass has oxidised, or that a particular number of visits causes a particular amount of ageing. The previous manuscript’s numerical grain, spread and sheen settings were illustrative parameters, not measured material constants.

The demonstration also has a boundary that a finished product would need to address. Its trace does not tell you which passage you read or what you learned. It records a plate activation, and it is intentionally temporary. An annotation would carry more meaning; a persistent annotation would require a different retention decision. Calling both things memory does not make their usefulness or their privacy costs equivalent.

You can follow the laboratory link to compare reading colours, but that workbench does not transport an exact history of these contact marks. It supplies an adjacent contrast exercise. The essay should say what the instrument actually carries, so the reader can distinguish a useful continuation from a claimed experiment that never took place.

3. Keep the history; keep the controls readable

Consider a proposed bookmark with three states. Before saving, the page has no bookmark. After an explicit save, an edge marker identifies the saved section and a text label names it. Hiding the marker changes the appearance while preserving the saved position. Removing the bookmark deletes the position after a clearly described action. This is a constructed design example, not an observed result from a deployed bookmark feature.

The separation matters. A fresh appearance need not erase useful work, and retaining useful work need not force a material treatment on the reader. I would keep the text, action labels and focus indicator stable across all three states. The edge mark can vary; the control for removing it should not become harder to see as the bookmark gets older.

WCAG 2.2 gives the text-contrast decision a concrete test: ordinarily 4.5:1 for text, with 3:1 for qualifying large text and specified exceptions. A change in an OKLCH lightness coordinate is not that test. The rendered foreground and background pair must be checked. An appealing aged surface cannot excuse a label that becomes unreadable, and this essay does not claim that every state of the existing study has passed a full accessibility audit.

Retention deserves a similarly explicit decision. W3C’s Privacy Principles recommend restricting transferred data to what is needed for people’s goals or aligned with their wishes and interests. Applied to the bookmark proposal, my default would be an intentional saved position and a clear retention choice, rather than a hidden counter of every passage opened. The existing contact marks are temporary component state; that observation is not a privacy audit of the site’s separate analytics.

There is a real complication: forgetting can be useful. A shared device may need a fresh reading state. A person revisiting an upsetting document may prefer not to see an earlier annotation. A revision history may need an export or deletion policy rather than an attractive visual treatment. Memory is a feature with consequences, and those consequences remain even when its marker is beautiful.

I would therefore judge computational patina by a practical standard. Can the reader identify the event behind the trace, reach the record it represents, keep the surface legible and choose whether the trace remains? A design that answers those questions may make software feel inhabited. It does not need to pretend that damage is devotion, or that a clean page has no history.