4 ms·
I would use a search engine that only indexed Reddit, Stack Exchange, Wikipedia, and a small number of other sites. And that specifically blocked Pinterest, Qu
by pkamb 5y ago
I would use a search engine that only indexed Reddit, Stack Exchange, Wikipedia, and a small number of other sites.
And that specifically blocked Pinterest, Quora, most non-personal “blogs”, etc.
People suggest DDG ! operators, but I don’t want to use a site’s (bad, single-site) search box. I want a multi-site SERP that only displays results from known good sites, which are customizable.
- all2 5y agoIf I could add sites I liked to the index that'd be great. Find a blogger/hacker I like? Add to the index. Can I share my index with others? Can I include their indices in my searches? Search engine as a social media platform? If I follow you, now I can search in your indices?
- betwixthewires 5y agoYacy might serve your needs well. It is a sort of distributed engine where users run their own index and "neighbors" share their indices with one another.
- pkamb 5y agoEven rules such as “if there is a Wikipedia result in the top 10, display it first”.
- monkeybutton 5y agoToo bad they whitelist which bots can access their sitemaps!
- guynamedloren 5y agoI've been thinking about this as well. As Google search results get increasingly worse, I find myself subconsciously filtering out all the garbage and gravitating towards a small number of known sites; and, as many other HNers do, I frequently mitigate this filtering step altogether by adding add "reddit" to any search in which I'm seeking out real human sentiment. I've done similar optimizations elsewhere to counter Google's trash results, e.g., I've been beefing up my personal recipe database, with the goal being that I can avoid a google search altogether whenever possible, only hitting google as a last resort. More and more I wonder, with the modern internet, is it even a feature that the whole web is indexed? Might be a bug.