Hacker News Reader

Cool. How? Can you elaborate on "I crawl myself." Do you mean you are the search engine and your search engine is an index of stuff you found that fits your criteria? What criteria, precisely?

View on HN

It's basically a 1998-style Google which discriminates against javascript-heavy documents and uses a Personalized PageRank[1]-algorithm to promote content that is adjacent to human websites.

[1] http://ilpubs.stanford.edu:8090/422/1/1999-66.pdf (ch. 6)

Thanks. Tried some searches, good stuff. Recipes are a well known disaster area these days, it's nice to be able to find just an explanation and instructions, with nothing else.

It's a bit hit-and-miss. The index is pretty small right now, I'm working on that, as well as getting better results when you have many keywords.

Right now it's absolutely amazing if you have like a broad topic you want websites about, but kind of weak when you want something more specific.

I got tortilla-making tips, I’m good ;)

I assume you've considered this, but what about deprioritizing precisely what spam sites use to SEO, or concocting some other identification algorithm, to filter precisely those out?

While yeah, marginalia finds interesting stuff, I've not been able to find anything useful that I've tried searching for with it so far.

Reply on news.ycombinator.com