Some Thoughts on Web Annotation

Yesterday, Mike Caulfied pondered how one might replace blog comments with something more connective by replacing them with annotated links, I think he was being purposefully provocative since he titled it a “…..Proposal for Killing Comments….”  Neverthelss, I think there’s a lot to be said for the idea that publishing platforms should encourage less dialogue and more broad conversation.  This goes back to Mike’s thought about tools that help people “geek out” virtually.

His choice of the term “annotated links” was important because it made me think of another annotation project that I played with yesterday ,, which counts Jon Udell of elmcity fame among its team members.  Collaborative annotation of the has for a long time been a feature just on the verge of changing everything, even before it was diigo’s killer feature.

At first glance hypothesis has much to recommend it.  There’s already a github repository, and the software is designed to run in a docker container, so running one’s own instance on a VPS should be straightforward, an important thing for anyone who lived through the life, death, and undeath of delicious.  Hypothesis is also working on Browsertrix, software that archives a web page when it’s annotated.  After all, your brilliant annotations aren’t much use if the page disappears. Federation is further down the roadmap, but a self hosted annotation server that would archive annotated documents and communicate with other servers that were doing the same thing looks to be not too far off.

This isn’t quite what Mike is talking about, but I wonder, based on the maxim “Don’t re-invent the wheel unless you have to,” how the two might fit together. Hypothesis,especially if you imagine a single user instance, seems very close to a recreation of Vannevar Bush’s memex, with electronic storage replacing microfiche.  It also shares characteristics with fedwiki, a project Mike has championed for a couple of years now, with the added benefit of maintaining copies of the source materials. I can imagine a research workflow where high level idea processing , outlining and drafting, happened in fedwiki with links to hypothesis annotated, browsertrix archived web pages (digital notecards) for documentation. Build federation into both ends and you could allow individuals and groups to create and document research publicly all the way through the process.


9 thoughts on “Some Thoughts on Web Annotation


  • Mike Caulfield
  • Harald Lordick
  • Jakob
  • Bill Seitz

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.