I’ve been working at importing more content from old services I no longer need to keep around. There’s no guarantee these sites will be around indefinitely, and plenty of anecdotal evidence that they’ll vanish one day with all our data with nay a peep or warning.

It’s been fun seeing how much you can grok just from URLs. Here’s some stuff I learned looking at links from 2013:

  • Mobile-specific sites, usually with a “m” or “mobile” subdomain, were still common, especially with newspapers.

  • UTM tags hadn’t started their widespread URL pollution yet, but there was inklings that people were thinking that you could track people with get request parameters.

  • Compared to earlier lists from 2009 etc, there were even fewer addresses that ended in an extension, like html or asp. I endeavour one day to resurrect my CGI site again with .pl, just for fun. I blame Hales.

  • There were still a lot of http sites in the mix, instead of https. I still think the widespread adoption of TLS was a mixed blessing, but that’s for another time.

  • There were already a lot of URL shorteners. My scripts tried to follow as many of these as it could, so my archives reflect the original site. Most no longer resolve, like SgLinks that I mentioned yesterday.

  • Sites are mostly, and unsurprisingly, not built to last. Newspaper sites tend to be robust, but most 2013 IndieWeb stuff I linked to is long gone. It’s a bit humbling to think the only records of these sites existing are in the sites of schmucks like me, and the Wayback Machine. Me and the Wayback Machine sounds like a ELO cover band.