Chris Alexander

On Engineering

Link rot

17th April, 2016

Admittedly I have not been all that busy with blogging recently - far from it.

However, when I overhauled the site a year or so ago, I did put in place a number of automated checks to make sure that I did not cause any problems when I was posting. This takes the form of a Travis CI build that runs on pull requests and the gh-pages branch of my blog’s Github repo.

This year, when I went to post a blog (and update the copyright year a full quarter late, oops) this build kicked off and informed me that 26 links had died since I last built successfully last year.

That seems quite a high number, for 1600 or so links across 1000 or so posts. When I went back through the history and put in Wayback Machine replacements for links that I had removed, that number climbed even higher - and included people who really should have known better (such as Google, Microsoft, Twitter, and - yes - myself).

Given we have been rather lucky with things like URL shorteners so far (casting my mind back to my old old old job) and have only really lost one medium size shortener so far, I would imagine this kind of problem is only going to get worse as these fall off the radar and entire swathes of the internet start pointing to URLs that no longer resolve…

So I have made a donation to the Internet Archive to say thanks for their good work, I would encourage you to do the same!