Skip to content

HiveMinds/browse-like-us

Repository files navigation

Browse Like Us

Python 3.10 License: AGPL v3 Code Style: Black Code Coverage

The Instagram of browsing experiences, except you get what you see. This is a social network that allows you to share and adopt your browsing experience based on how your peers experience the web.

Why

Raymond Hill made a wonderful application called uBlock Origin. It allows you to filter (web) content. There are hundreds if not thousands of lists that filter web content to make sites look nice and simple. The creator of the browser-extension even declines funding and suggest you donate to those who maintain the lists. uBlock filter maintenance can be seen as an everlasting battle;

  1. Content-providers want to push their adds/nonsense.
  2. uBlock list maintainers block the adds/nonsense.
  3. Content-providers catch up to this blocking, and try to circumvent it by updating their website.
  4. Back to step 2.

I think there is a lot of value in a single person tediously cleaning up 1 tiny corner of the internet really well, and in my experience it is not easy to find these persons and integrate their work in your browsing experience. That is why this repository exists.

What Stage I

If you visit a site, you can quickly scroll through the list of UBlock Origin filter lists created by the uBlock users. The one you like sticks, and you can add your own mods on top of that. These mods then become available to others.

Some kind of rating/usage system should be used to filter the good mods from the bad mods. A dial could be used to determine what you want to filter (content vs cosmetics), and how strict you want to minimalise your browsing experience.

What Stage II

Large adoption creates a dataset with:

  • pairs of: [HTML source code, uBlock origin element filter list] as input
  • filter adoption/usage as labels/scores of the filter (list(s))

I think sufficient adoption (1M+ curated websites) may allow one to automate filtering the adds and nonsense using machine learning techniques.

What Stage III

The network could be decentralised to automatically pay out the people cleaning up websites, based on your usage. (Like spotify except without me/this as middle-person.) (Like brave tokens without the 30% cut.)

Contribute

  1. Have a look at the roadmap and issues
  2. Pick an issue you like
  3. Build it 🚀
  4. Send me a pull request :)

Most issues can be solved in parallel. Currently I focus on graduating, and I am not actively generating solutions/work on this, I am willing to perform maintenance, quality control and CI :)