Half your bookmarks are already dead

Muse does not archive the pages you save, and nothing about it stops a link from going dead. If a preserved copy of the page itself is what you are after, the tools further down this page do that properly. Muse's part in this problem is a smaller one, described honestly near the end, and it is not a substitute for actual archiving.

That distinction is worth being blunt about early, because a lot of what looks like an archiving feature in a bookmarking tool is really just a nicer bookmark: a title, a preview image, sometimes a tag. None of that is the page. The page lives on someone else's server, and it goes away whenever they take it down, redesign it past recognition, or let the domain lapse. Here is what actually keeps a copy, and what does not.

How much of what you have saved is already gone?

More than feels reasonable. The Pew Research Center collected roughly 90,000 webpages a year between 2013 and 2023 and checked them again in October 2023. A quarter of all the pages that had existed at any point in that span were no longer accessible. Pages from 2013 specifically fared worse: 38 percent were gone a decade on. Even pages from 2023, the newest in the sample, had already failed 8 percent of the time within months.

Those figures describe the open web in general, not personal bookmark files. A bookmarks list tends to skew older and more neglected than the average webpage, so treat a quarter as a floor for what has already died in yours, not a ceiling.

What actually saves a page, not just its address

Four methods do this properly, roughly in ascending order of effort.

SingleFile is a free browser extension for Chrome, Firefox, Edge and Safari that saves a complete page, including images, fonts and layout, into one self-contained HTML file on your own disk. It is the closest thing to a proper one-click archive a browser extension can offer, and the file it produces still opens correctly years later because nothing about it depends on the original site staying up.

The Wayback Machine does two useful jobs. Its Save Page Now tool captures a page deliberately, the moment you decide it matters. Separately, it is worth checking before you assume a dead link is unrecoverable, because a large share of the web already has a copy sitting there whether anyone asked for it or not.

Safari's own Web Archive format needs nothing installed at all. Open the page, choose File then Save As, and pick Web Archive from the format menu. Safari writes a local file that reopens the page as it looked, images included, without touching the network again.

Printing to PDF works in any browser and freezes exactly what was on screen. It loses anything that depended on scripts or interaction, which rules it out for a page you might want to reflow or search properly later, but for a receipt, a confirmation page or a single article it is quick and needs no extension.

How do you find out which bookmarks are already dead?

Two approaches, depending on how many you are checking. For a single page, paste its address into the Wayback Machine and see whether a version exists, or just try loading it again; sites sometimes move rather than disappear, and a search for the exact title often turns up the new address.

For a whole bookmarks file, a dedicated checker is faster than clicking through by hand. The Chrome Web Store lists several under names like Bookmark Cleaner and Bookmarks Checker, which scan your existing bookmarks locally and flag anything returning a broken response, so you are not opening hundreds of tabs to find the handful that no longer work.

Self-hosted archiving, for the more technical reader

If you want ongoing, automatic archiving rather than saving pages one at a time, three self-hosted tools are worth knowing about. ArchiveBox runs on your own server or NAS and saves each URL you feed it in several formats at once, including HTML, PDF, a screenshot and a full WARC file, the same format the Internet Archive itself uses. Linkwarden is a self-hosted bookmark manager that preserves a full copy of every page alongside the usual title and tags, so archiving is not a separate step. Wallabag takes a narrower, read-it-later approach, extracting the article text itself so it stays readable on a phone or an e-reader even after the original page is gone.

All three keep the pages on infrastructure you control, which matters if the goal is genuinely not depending on anyone else, Muse included, to keep a copy safe.

So what does Muse actually do about link rot?

Nothing that counts as archiving. What survives in Muse when a page dies is the preview image, title, tags and any note you added when you saved it. That is not the page, and no bookmarking tool, Muse or otherwise, brings a page back once the site behind it is gone.

What that preview and note do is usually enough to recognise what the thing was and go looking for it elsewhere, through the Wayback Machine or a search for the title, rather than staring at a blue link with no memory of what it used to say. If the dead links you find need clearing out rather than replacing, the method for that is in how to clean up years of bookmarks.