I’ve got some ideas together to clone Readability v 2.0.

I’ll use someone else’s processor to make pages readable; there are a bunch of such projects being written in python right now. I want to reimplement the functionality to maintain a reading list that is accessed by web browser and can send stuff to my Kindle. I’ve been finding this very useful for getting reading done from school.

Reasons to clone:

  • Non-Free software-as-a-service is not acceptable for personal data like the articles one has chosen to read later. To its credit Readability v 2.0 lets you export your data. But it’s not good to have that amount of information on what people are reading centralised in this post-Snowden world.
  • Readability v 2.0 has a lot of pretentious quotations, designed to make one feel good because one is a member of the literary elite, and therefore feel good by using their service. This is not psychologically skillful.

Design goals:

  • Minimalist interface.
  • No social features: this is about reading.
  • Simple CGI script that users can run on shared web hosts so that reading list data may become decentralised. Use SDF as benchmark. No chunky web frameworks or obscure libraries.
  • No database requirement. Flat files.
  • Run on mobile devices too.
  • Modern CSS and JavaScript conveniences in order to make things run better on mobile devices are acceptable: this is an application using HTML as a frontend purely because of its ubiquity, not a website.
  • Can control by an Emacs interface and can submit articles by e-mail.