Skip to content
All fingerprint checksRender & GPU

a box placed at the viewport origin measures its origin as zero

Check id client-rect-origin-vs-position

What an ordinary browser yields

A fixed box at left:0/top:0 with no margin reports getBoundingClientRect().left and .top as exactly 0, and its size as exactly the 10×10 it was given.

What a detector infers

Zero is scale-invariant, and that single property is the whole check. A fixed-position box at left:0 and top:0 with no margin has its border-box origin at exactly (0, 0) in viewport coordinates. Every transformation the pipeline can legitimately apply on the way to a client rect — devicePixelRatio, browser page zoom, sub-pixel layout, the whole LayoutUnit fixed-point pipeline — is a MULTIPLICATION, and any factor times zero is still zero. So unlike a width, an origin at zero has no legitimate spread to hide in: a non-zero reading cannot have been scaled into existence, it can only have been added. This is why the check reads a position rather than a size. Client-rect farbling is a well-established privacy technique — it defeats element-size and font-list fingerprinting by perturbing the numbers slightly — and it is very effective on boxes that sit at arbitrary coordinates, where a fraction of a pixel is indistinguishable from ordinary layout. At the origin there is nothing to blend into. Measured on a shipping stealth browser: a constant offset applied to the entire coordinate space, identical on all four reads of the same box, identical for absolutely and fixed positioned boxes, and applied to document.body itself, while every box still measured its correct size. The offset was drawn fresh on each launch. This also closes an escape hatch. The layout-box-arithmetic check refuses to judge unless its probe box sits at an integral origin, because Blink snaps clientWidth and offsetWidth against the box's position and a shifted origin can legitimately move a snapped width by a pixel. That guard is right, but on a browser that offsets every rect it means a critical check quietly returns N/A instead of a verdict — the safety gate becomes a way through. The moved origin now gets adjudicated on its own row instead of being lost in that one.

How to resolve it

getBoundingClientRect is the layout engine's answer about a box whose position that same engine has just reported as left:0/top:0, so the two cannot be set independently. This row is deliberately not scored as a contradiction, because offsetting rects to blunt element-size fingerprinting is a legitimate thing for a privacy build to do. What is worth understanding is where the technique leaks: an offset chosen once per session disappears into the noise of any box sitting at an arbitrary coordinate, but at the origin the true answer is zero and no scaling in the pipeline can reach it, so the added constant is left standing on its own. Controls measured for this row: stock Brave 150 reports the origin at exactly (0, 0) on a real HTTPS origin and on localhost — it farbles canvas but not rect geometry — as do Chrome 150, Edge 150 and Firefox 151, the last confirming the invariant is not a Blink quirk. The unscored treatment is therefore precautionary rather than a fix for a false positive anyone has seen.

Anatomy of a browser fingerprint: every signal, and why they must agree

Read in the wild by

CloakBrowser (measured)

Every client rect translated by a constant offset drawn once per launch, on all three launches measured. The same offset appeared on a position:absolute box, a position:fixed box and document.body itself, and was identical across four consecutive reads of one box — so it is a translation of the coordinate space, not per-call noise. Box sizes were untouched at exactly 10×10, which is what isolates the origin as the thing that moved. On this browser the offset also pushed layout-box-arithmetic into N/A, so it was suppressing a critical check as a side effect.

origins (left, top) of (-0.015625, +0.046875), (-0.015625, +0.03125) and (+0.046875, -0.015625) on three launches — every value a multiple of 1/64 px, stable across 8 repeated reads each. Stock Chrome and Clearcote on the same machine both report exactly (0, 0).

CloakBrowser (measured)

Every attribution traces to a published artifact. See the sources and their limits.

Nearby checks in Render & GPU

See all 32 checks in Render & GPU
Who builds this test

Clearcote is a browser built for fingerprint coherence

It is a Chromium fork, maintained by the same people who wrote this reference. It ships as a compiled browser rather than as a stealth script injected into someone else's — which is a description of how it is built, and is not an argument about how it behaves on this check.

This audit takes no position on how Clearcote scores on Render & GPU checks, on this one, or anywhere else. It has no baseline corpus of other people's fingerprints to rank you against and no vendor scoreboard — nearly every check is self-referential, asking one browser the same question through two independent APIs and reporting whether both answers can be true at once. It runs identically on any browser, including ours. Run it on yours and read the result yourself.

See the other checks in Render & GPU — the family client-rect-origin-vs-position belongs to.