Web Graph Browser
TL;DR: Check out Web Graph Browser or read more to learn how I got to think idea.
For the past few nights, I’ve been experimenting with graphing the web. It started as an attempt to see if following links from RSS feeds could lead to more RSS feeds. I wondered whether unofficial online communities could be uncovered this way—figuring that members of a community might naturally link to each other. The idea was that, by starting with just one or two feeds, I could map out the major contributors to a community.
It didn’t quite work out that way. Instead, I mostly discovered that everyone links to Twitter. My scripts spent a lot of time filtering those out! I still think there’s potential in the idea, but it would take more effort than I’m currently motivated to put into it. What other ways might help people build a blogroll or find a curated set of feeds?
To visualize these connections, I asked Claude.AI to generate a web page using D3 for the graph display. The result was a decent starting point, but I ended up stripping out a lot of the code as I refined and extended the functionality. This seems to be a common pattern with AI tools, they help kickstart a project, but I still have to take it to the next level.
Ultimately, I built a “web browser” that displays only links. I’m calling it Web Graph Browser because it lets you explore the web as a network. Nodes are color-coded based on their distance from the starting point. One fun use case I’ve found is playing the Wikipedia Race, though navigating pages with hundreds of links can get pretty tricky. The most interesting part of the coding process was diving deeper into DOMParser.
I’d love to hear feedback or ideas on making web exploration more interesting, feel free to reach out!