BlogRolling.com

» see blogrolling.com as it lived, on the wayback machine

~ the obit ~

Before the feed reader won, the way you found the next blog was to read somebody else's sidebar. That column of links was called a blogroll, and maintaining it by hand meant editing your template every time somebody moved to a new domain. BlogRolling.com took that job away. You made an account and added your links through a web form, then dropped a single line of JavaScript into your sidebar. The server rendered the list on every page load, so a change you made once appeared everywhere you had pasted the tag.

Jason DeFillippo built it, and it was live on blogrolling.com by the start of 2002. It won a Weblog Award at SXSW twice. The service also did something a static list could not: it knew when the blogs on your roll had last published, so it could mark the fresh ones. That turned the sidebar into a low-effort reader, years before anyone in the mainstream had heard the word RSS.

Tucows bought it on 27 February 2004. The press release put numbers on how big the thing had quietly become: more than 32,000 users, and over 1.5 million blogrolls delivered every day across every major blogging platform, tracking half a million weblog links between them. Tucows already sold Blogware to hosts and resellers, and BlogRolling looked like a natural piece of the same shelf.

~ the rap sheet ~

~ what the list was actually for ~

I was reading in this era rather than publishing in it, mostly on Turkish hardware and tech sites, and the blogroll was how you got out of whatever page you had landed on. You came in through a search result, found something worth reading, then went down the left-hand column to see who the author trusted. That is discovery without an algorithm, and it worked because it cost the author something to put a name there.

~ what killed it was a security scanner ~

The architecture that made BlogRolling easy in 2002 became the exact pattern the industry spent the late 2000s teaching people to distrust. A third-party script tag, hosted on someone else's server, deciding remotely what got injected into the middle of your page. By 2010 malware scanners flagged that shape on sight, and it did not matter that the payload was a list of your friends' blogs.

Ken Schafer at Tucows said it plainly when the wind-down was announced: the way BlogRolling had always worked was no longer acceptable, they would need substantial work to rebuild it to be scanner friendly, and they could not afford to do that. On 6 October 2010 the status blog carried the headline "BlogRolling Will Cease Operation November 1st" and two sentences confirming the date.

~ no export button ~

The shutdown post is worth reading for one detail. Under the instructions for what to do next, it says: sorry, there is no export option in BlogRolling. The advice was to go to a page on your site that contained your blogroll, click every link in turn, and copy down or bookmark wherever you ended up. Nine years of curation, retrieved by hand, one click at a time.

blogrolling.com kept answering with a small page through 10 November 2010, then started redirecting. The domain has not resolved for years.

What died on 1 November 2010 was not the idea. Plenty of blogs still keep a links page. What died was the shared registry underneath thousands of them, which is why so many blogrolls from that period now render as an empty div.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »