The afterlife of hyperlinks
Links promise that a page will still be there tomorrow. Many of them break that promise quietly.
A hyperlink is a small act of trust. Whoever wrote it assumed that the address on the other end would keep pointing at the same thing. For a while it usually does. Then a site is redesigned, a project moves, a domain lapses, and the link starts leading somewhere else, or nowhere at all.
Researchers who sample old links keep finding the same pattern: the older the link, the less likely it is to work. What is lost is rarely dramatic. It is a footnote in an essay, a reference in documentation, the source of a quote. Each loss is small; together they make the past of the web harder to read than its present.
What helps
Some habits make links last longer. Addresses that describe content instead of the software behind it survive migrations better. Redirects from old paths to new ones cost almost nothing to keep. Public web archives can store a snapshot of a page at the moment it is cited, so a reader has somewhere to go when the original is gone.
None of this is new advice, and that is perhaps the point. Keeping a link alive is less a technical problem than a question of whether anyone still feels responsible for the address.