The URL is one of the small quiet miracles of the web. It is a piece of text that anyone can type into anywhere, and if the machine at the other end is still online, the exact document appears. Every attempt to replace this — QR codes, apps, deep links — has been worse, in specific ways that people knew about at the time and were talked into ignoring.

A long URL, historically, told you something. It told you which section of the site you were in, and often which article, and sometimes the date. You could look at the address bar and, roughly, know where you were. When links were shared over IRC and by email, this was extremely useful; you could preview a link by reading it, and know before you clicked whether it was interesting.

The last decade has been a slow migration away from this. Vanity redirects. URL shorteners. Path structures that look like abstract identifiers. Some of this was necessary; a URL is not, in fact, unbreakable, and some of the old ones had aged badly. But the aggregate effect has been to make a lot of the web anonymous at the address level. You are handed a link that could be about anything, and you learn what it is by clicking it.

The shortener deserves special blame, because it dressed a loss up as a convenience. A shortened link is not just smaller; it is opaque by construction. It cannot be read. It cannot be trusted on inspection. It can be repointed after you have shared it, so the thing your readers click tomorrow need not be the thing you linked today. We accepted all of this to save a few characters in a medium that was about to stop counting characters anyway.

There is a security dimension here that the aesthetics conversation tends to skip. A descriptive URL is a defence: you can often tell a phishing link from a real one by reading the address, if the address is legible. Strip the legibility away — hide everything behind a redirect or an abstract token — and you have removed the one check an ordinary person could perform before clicking. Prettiness and safety, in this case, point the same direction.

This is a real loss. Long URLs were how you inspected a link before trusting it. They were how you shared with context. They were how, in extreme cases, you rescued a piece of the web from its archive when the site went away. Every piece of that quiet infrastructure is worse now than it used to be.

The counterargument I hear is that nobody reads URLs anymore, so why preserve them — and it is true that the address bar has been shrinking, hiding the path, sometimes hiding everything but the domain. But this is circular. People stopped reading URLs partly because URLs stopped being worth reading. Make them legible again and the habit could come back; it was a useful habit, and it did not die of natural causes.

A good URL is also the most durable citation the web has. Long after a site's design is redone and its navigation is rearranged, a clean path-based address can still point at the exact document, and a stranger can still reconstruct what it meant just by reading it. That is a civic property, not a cosmetic one. We have been quietly trading it away for tidiness.

I would like the long URL back, at least on serious sites. Not the ugly ones with tracking parameters — those can go. The clean, descriptive, path-based ones that told you what page you were looking at. It is a very small piece of civic architecture, and I think we would notice it if we looked.