how hard would it be to write a simple search engine, I wonder
the details behind google's page rank algorithm are well documented and all the patents have expired. if you're not trying to sell ads and drive engagement I wonder how hard it would be at this point. I also wonder what a federated/collective search engine might look like. I should be doing other things yet here I am.
@nasser I think a general search engine like Google doesn't even make sense at this point. With all of the noise out there (and a lot more coming with AI) I think the real need is for curated directories of links like the old Excite/Yahoo or Wikipedia. Good news is those are perfect for collaboration!
@nasser yea but if you're going to curate a set of approved sites you might as well just organize them into a searchable tree that has a browsable structure
@ivan but i want to say "how make css div centered" and get an answer
@nasser but that could be exactly how you organize it. "how make css div centered" is a perfect example of a tree search :)
@nasser @ivan Impulse buy the CRT, then atone for the gear fetishization by designing a printable (acrylic?) lens that can turn any LCD into a "CRT" with correct curvature. Speaking of which, recently read how the builders of the mall arcade props in The Last of Us saved money by not even sourcing 4:3 LCDs--they turned 16:9 ones vertical and "cropped" them with bigger cabinets
@ivan yeah i tend to agree with you. i suppose i was thinking of something that would search a webring or a curated set of trusted sites, or at least trusted "roots" to begin crawling from. pagerank is a sorting algorithm in any case that you run on a set of pages.
90% of my technical searches end up on official documentation or stack overflow anyway, sometimes arch linux forums depending, and more rarely on a blog if things are niche, but its a pretty closed set of sites in general...