Login
You're viewing the assemblag.es public feed.
  • chris martenschrisamaphone@hci.social
    Oct 15, 2026, 5:25 PM

    every so often i nostalgically visit old websites i used to maintain or frequent, and i click on all the external links to find out which still work.

    has anyone done research about this? i don't know how to systemize the method exactly, but i want to know things like, "what percentage of links [over what dataset?] still work", "what are the different ways things are broken/how many in each category", "what technical/organizational decisions/circumstances impact url longevity in what ways"

    💬 4🔄 1⭐ 0

Replies

  • chris martenschrisamaphone@hci.social
    Oct 15, 2026, 5:32 PM

    also i'm curious to hear from those of you who maintain websites how you approach (a) your own use of external links and (b) others' ability to link to specific content you host.

    at some point i realized that frank pfenning (cs.cmu.edu/~fp/), who's known for the decades-old running archive of class lecture notes he makes publicly available, uses external links *extremely* sparingly, mostly to either time-limited (like gradescope for a class) or archival (DOI-associated) resources

    💬 2🔄 1⭐ 0
  • Daniel Schildtautiomaa@mementomori.social
    Oct 15, 2026, 5:30 PM

    @chrisamaphone Friend looked at their list of lists from 15 years ago, and realised that a lot of very good websites had disappeared. Not sure if it is right word for it, but "longing for the past feelings about what web was" feels like a theme.

    💬 0🔄 0⭐ 0
  • 💬 1🔄 1⭐ 0
  • chris martenschrisamaphone@hci.social
    Oct 15, 2026, 5:33 PM

    @Edent wow! this is a really cool project and exactly the kind of thing i was looking for!

    💬 0🔄 0⭐ 0
  • Dr Ro SmithRhube@wandering.shop
    Oct 15, 2026, 5:37 PM

    @chrisamaphone It's an interesting question. From time to time I still check in on an old website I adored in the 00s - it had already stopped updating when I found it, yet the maintainer still continues to save it from the ether thoughtviper.com/inexob/archiv

    As for maintaining my own websites... it's not been something I've actively done for personal websites for a while (too sick), but when I did it as part of my job, part of the review of any page was checking all the links worked. Most of the>

    💬 0🔄 0⭐ 0
  • Andrew Plotkinzarfeblong@mastodon.gamedev.place
    Oct 15, 2026, 6:12 PM

    @chrisamaphone Speaking as someone with a 30-year-old web site…

    I absolutely do not worry about it. People who browse a 2010 blog post can figure out how to use Wayback Machine.

    I could try to put my entire site under some kind of system that tests links and waybackifies them. But such a system will itself go wrong sometimes, and it's not obvious that it would be a net improvement.

    💬 1🔄 0⭐ 0
  • Andrew Plotkinzarfeblong@mastodon.gamedev.place
    Oct 15, 2026, 6:14 PM

    @chrisamaphone Also, it's essentially impossible now for an automated tool to detect broken links, because huge swathes of the web have gone behind automated-tool blockers to try to mitigate the AI scraper apocalypse.

    💬 0🔄 0⭐ 0
  • pfriedmapfriedma@pfedi.pfriedma.org
    Oct 15, 2026, 5:32 PM
    @chrisamaphone
    There are some papers on arXiv on link rot and pew did some research suggesting up to half of the Internet (pre-2013) may no longer exist.
    Wikipedia article citations are a good starting set, apparently.

    The availability question is in the back of anyone's mind that deals with the minting of DOIs for papers and datasets. And NARAs guidance is....best effort these days.
    💬 0🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0