What we collect, which is nothing

What your browser does load

Pictures, and nothing else. Every host this site loads an image from, in full:

• arstechnica.net
• co.uk
• futurecdn.net
• nvidia.com
• siliconangle.com
• wired.com
• ytimg.com

That list is read off the site itself every time this page is built, not typed here, so it cannot fall behind the sources. Video thumbnails come from YouTube's image host; the rest are publishers' own image hosts, for the small number of articles whose feed offers a picture.

Loading any image from any server tells that server your IP address and browser details. That is true of every image on the web and is not something we can do on your behalf, which is why the hosts are named here rather than glossed.

That is the complete list of what we load. No fonts, no tag managers, no pixels, no embedded video players, no social widgets, and exactly one script of ours. One more arrives without us, and it is named in the next section but one.

The one script, named — and it now does two things. _shared/constellation.js draws the drifting field of stars behind this page, and once a minute it rewrites the “5m ago” on each card so that it stays true. Those labels are written when the page is built and the site rebuilds every three hours, so without that a card could sit there saying “5m ago” about a story three hours old. The second job was added on 21 September 2026 and put in the same file on purpose, so the promise above still reads “one script” and means it.

It is served from this site, it shares its renderer with the sister sites, and neither job reaches outward: no network request, no storage, no cookie, no identifier, nothing reported anywhere. It reads the clock in your browser and writes text into the page. Until 20 September 2026 this site ran no script at all and said so here; that was a stronger claim and it was given up deliberately, for a background. Saying which is more honest than quietly softening the sentence.

Block those hosts and the site still works: you will see cards without pictures. Disable scripting and it still works: no starfield, and each card’s “5m ago” stays frozen at whatever it was when the page was built — which is why the exact date and time sits printed beside every one of them. That one never moves and never needs to.

One script we did not put here

Our web host appends its own tracking script to every page it serves, including this one. It is not in the files we upload and it is not something we can remove from our side. It arrives after the page we wrote ends, and it is two pieces: a small block that records the server and data-centre this site is served from, and tccl.min.js, fetched from img1.wsimg.com — a host belonging to the company we rent the server from.

Their own note inside it reads: “Monitoring performance to make your website faster. If you want to opt-out, please contact web hosting support.” So it is opt-out by support request, not a setting, and until that request is made and granted your browser fetches that file on every page of this site.

What this means for the promise above. Everything this page says about what we do remains exactly true: we set no cookies, run no analytics, and ship one script of our own. But the sentence a reader actually cares about is how many scripts their browser runs, and the honest answer on this host is two — ours, and theirs. We would rather write that down than let the stricter-sounding sentence stand and be wrong about the only thing it was measuring.

When you leave

Every story and every video on this site is a link to somewhere else. When you follow one, you are on that site and their privacy policy governs what happens next. Following a link to YouTube puts you under Google's privacy policy. We have no involvement in, and no visibility of, anything that happens after you click.

Every outbound link carries rel="noopener", which stops the destination from gaining control of this tab. Links from stories we have written about carry nothing more, because we vouch for them; links in the automatically selected index also carry nofollow, because we have not. The publishers page says why.

What we hold

Thing	Where it lives	How long
Headlines, links and dates from publishers' public feeds	A file on our own machine	Refreshed constantly, pruned at 30 days
Video titles, dates and thumbnail URLs from public YouTube channel feeds	The same file	The same
Anything about you	Nowhere. We do not have any.	Not applicable

Our web host will keep ordinary server logs, as every web host does. We do not read them and we build nothing from them.

Children

This site is not aimed at children and collects nothing from anybody, so it collects nothing from children either.

Your rights, and why there is little to exercise

Under the UK GDPR, the EU GDPR and laws like the CCPA you have rights to access, correct, delete and port your personal data. We hold no personal data about visitors, so in practice there is nothing for us to hand over or erase. You are welcome to ask and we will confirm the same thing in writing.

We do not sell or share personal information, because we do not have any.

If this changes

If this site ever adds advertising, analytics, accounts or an embedded player, this page changes first and says so, with the date. A policy that describes a site other than the one you are reading is worse than no policy at all.

Contact

Questions about this policy: theoriesofanything@gmail.com.