Go to content Go to navigation Go to search

Three foggy mornings and one rainy day.

It was a while ago that Chris Bertram announced he’d finally pulled down Junius, his old Bloggered blog. So it was a while ago I sighed and went and searched the pier for whatever links I’d made to Junius, way back when. —Turned out there was only the one, to a write-up on three-sided football, but a rotten link is a rotten link. I copied the old href, brought up the Wayback Machine, and thumbed through the archives for an appropriate copy of Chris’ old page, then copied that href and replaced the rotten link in my entry with the internet archived deal.

Then I checked the other links in that entry, just in case, and found that Tales of the Legion: the Origin of the Legion had also rotted away. Ditto and so forth.

It’s becoming more and more of a chore, this scraping the hull for linkrot. And though the pier’s been a mostly going concern for five years now, it’s only got (checks) about a thousand entries with, I dunno, three or four outbound links per, on average? There’s no way an actual jumping joint like Eschaton or Crooked Timber could even begin to think of keeping up. (Not that I’m keeping up myself. I just check when I’m specifically reminded of something. Like the tickle in the back of the brain that says hey, I think maybe once you linked to Junius, back in the day. Depsite the constant bloggering it suffered.)

—About the same time as Chris was pulling Junius down, John Holbo was trying to figure out how to avoid linkrot upfront, maybe by using WebCite® right off the bat? But that links to WebCite®’s archived copy from the get-go and not the cited site itself, mucking with traffic and googlejuice and whatnot, and anyway WebCite® only wants scholarly papers to use their service, and even if it’s free I hitch at people so profligate in their use of marcæ registradæ®.

(Also of idle note: the various Bad Actors, over the ages. It’ll be a cold day indeed before I ever again link to a Yahoo news article, or anyone’s AP piece, or the Washington goddamn Post, let me tell you. —Plus, yes, there’s the linkrot I’m responsible for, Bad Actor myself, having once used an old Movable Type link-numbering scheme that I can’t easily mask to the new, sane, easily replicable link-naming scheme. I still get hits on those old pages, from time to time. No clue what they pointed to, without Waybacking myself. I wince a little every time I see one in the logs.)

Anyway, here’s what I want, oh plugin developers, oh API jockeys, oh Web 3.0 entrepreneurs agleam in someone’s eye: I want something that will spider through my site on a regular basis, testing outgoing links in all my various entries. Anything that returns a 404 gets automagically plugged into the Wayback Machine, and the href of the archived version closest in time to the date of the entry in question is returned and replaced in the rotten link. Once a week a report is generated: here’s what was found and fixed, so I can go through myself and re-correct any overly zealous corrections. If needed.

Lazyweb powers activate! Thunderbleg explodes into action NOW!

  1. Mark Gisleson    Apr 12, 08:34 am    #

    It’s best to think of the ‘net as a Zen thing. I’ve lost at least three databases over the past three years, and each loss of old posts was a blessing, not a curse.

    Each new post represents a blog born anew, with the hope of greatness and the probability of stinkyhood.

    Broken links reflect on the intertubes, not those who make music from the intertubular bells and whistles. Leave your linkrot in place, it won’t infect other links or cause others to think less of you.


  2. Kip Manley    Apr 12, 06:50 pm    #

    Yes, Zen, and King Canute, and there’s a reason they call it the Art of Memory, yes. But I’m forever and always an idealist; it’s also why I get grumpy whenever people say the world is unfair. That’s no damn excuse. —The Long Game is such an important part of what we do and what we’re able to do out here in the Islets of Bloggerhans, and linkrot eats away at that, like endnotes to a back of the book that’s been ripped away. —What were we talking about back in then? Fitz? Frist? What? So I beat it back, when I think to do it, where I can. A tool to help would help.


  Textile Help