Your own browser can manufacture a difference

The mechanism, briefly

A browser keeps things it has already been given so it does not have to ask again. When you return to something it has, it may show you what it has instead of asking. This is ordinary, it is what makes browsing tolerable, and it is invisible by design.

Invisible by design is the important part. Nothing on the screen distinguishes a page that was just fetched from one that was kept, which is why this cause is so rarely suspected.

Why plurality makes it visible

With one address, a kept page and a fresh page are just two moments of the same thing, and a reader has no way to notice the difference and no reason to.

With two routes, a kept page on one and a fresh page on the other sit side by side, and the difference between them looks like a difference between the addresses. The browser has quietly created a comparison and attached a false explanation to it.

What it looks like when it happens

• A figure that does not match on the other route.
• Something present on one and absent on the other.
• A page that looks like the state of things earlier today.
• A route that appears not to have noticed something you did on the other one.

The specific trap of the left open tab

The worst version is the tab you opened an hour ago and came back to. It has not been reloaded since, so it is showing an hour old view while the other window shows now.

Two tabs open on two routes is the arrangement most likely to produce this, and it is also the arrangement readers use when they are deliberately comparing. The act of comparing manufactures the artefact, which is the subject of two tabs, two routes.

How to rule it out

1. Reload both deliberately rather than switching between windows.
2. Do it in the same minute, so time is not doing the work.
3. Reload the one that looks wrong first, because that is where the artefact usually is.
4. If the difference survives all of that, move on to the next cause with a much stronger observation than you had.

Why this is worth its own study

Because it accounts for a large share of every question anybody asks about mirrors being different, and because the fix takes two seconds and almost nobody applies it before forming a theory.

It is also the cause with the most misleading shape. It produces exactly the symptoms that the wrong model of mirrors predicts, so it confirms the wrong model every time it happens.

What it cannot explain

It cannot explain a difference that survives a genuine reload on both sides, and it cannot explain a page that is structurally different rather than merely older. Those belong to two front pages and the difference that matters.

Ruling this out is therefore not a way of dismissing a concern. It is the step that makes a real concern legible.

A habit worth having

Before comparing anything on two routes, reload both. It converts most of this section from a mystery into a non event and costs less than reading a single paragraph about it.

Why this is the first study to read in the section

It accounts for more reported divergence than every other cause combined, its remedy takes two seconds, and knowing about it changes how a reader conducts every other comparison they will ever make between routes.

If somebody reads only one page in this section, this is the one that will save them the most time, and it is the one they are least likely to look for, because nobody suspects their own browser of manufacturing evidence.

What this study is not. It is not a technical account of how any browser decides what to keep, and nothing here depends on those details.