6 ms·

Yeah, and, along the same lines, directories should perhaps also make a comeback -- like the old Yahoo! Directory or DMOZ. Just a carefully curated and organiz

Yeah, and, along the same lines, directories should perhaps also make a comeback -- like the old Yahoo! Directory or DMOZ. Just a carefully curated and organized list of trusted sites that are made and managed by humans and for humans. Unlike, say, DMOZ, the directory of the future should be easier to navigate and incorporate "modern" features -- and it should also be easily searchable. (All sites on the directory are indexed; no sites off the directory are indexed.) Searching for B2B products and services has become a lot tougher lately. A good directory could conceivably make a meaningful difference.

JKCalhoun 2y ago
Could work. We have databases behind PiHole, could have something similar to filter search results.
cageface 2y ago
I’ve been saying for two years now that curated directories are poised to make a comeback on the web. It just needs a critical mass of curators. I think this is why those awesome-$tech lists are so popular.
- EasyMark 2y ago
  I mean they might have a mini revival but i think we've gone too far to change the process of the general population. Indexes appeal to more organized people who like such things, most of us just throw everything in a pile in a well known location (google) and rifle through that.
  - A_D_E_P_T 2y ago
    Ideally a "modern" directory should be easily searchable as well -- just very selective in what it indexes. The basic UI could be similar to a search engine, yet with a transparent foundation that could also be perused in a different way.
renegat0x0 2y ago
I have been collecting personal sites for some time. Format JSON files. Links have tag 'personal'. https://github.com/rumca-js/Internet-Places-Database https://github.com/rumca-js/Internet-Places-Database
6510 2y ago
You could make the topical hierarchy as deep as possible and build a search that only searches the current category. I recon LLM's, with minimal oversight, could file websites as well as articles on websites under all the appropriate categories. Google is highly confused about articles vs websites about a topic. There should be a distinction. Sometimes I want a website, sometimes just an article. Making it into a distributed index complete with cached copies seems fun. Have the hosted giant monolith mirror to fail back on but attempt to do everything p2p. Maybe it should bluntly show a dialog in the corner of the screen with closest related websites (or pages) and a button to further browse the index using the current page as the query. Basically contextual advertisement only for free.

JKCalhoun 2y ago

Could work. We have databases behind PiHole, could have something similar to filter search results.

cageface 2y ago

I’ve been saying for two years now that curated directories are poised to make a comeback on the web. It just needs a critical mass of curators. I think this is why those awesome-$tech lists are so popular.

EasyMark 2y ago
I mean they might have a mini revival but i think we've gone too far to change the process of the general population. Indexes appeal to more organized people who like such things, most of us just throw everything in a pile in a well known location (google) and rifle through that.
- A_D_E_P_T 2y ago
  Ideally a "modern" directory should be easily searchable as well -- just very selective in what it indexes. The basic UI could be similar to a search engine, yet with a transparent foundation that could also be perused in a different way.

renegat0x0 2y ago

I have been collecting personal sites for some time. Format JSON files. Links have tag 'personal'. https://github.com/rumca-js/Internet-Places-Database https://github.com/rumca-js/Internet-Places-Database

6510 2y ago

You could make the topical hierarchy as deep as possible and build a search that only searches the current category. I recon LLM's, with minimal oversight, could file websites as well as articles on websites under all the appropriate categories. Google is highly confused about articles vs websites about a topic. There should be a distinction. Sometimes I want a website, sometimes just an article. Making it into a distributed index complete with cached copies seems fun. Have the hosted giant monolith mirror to fail back on but attempt to do everything p2p. Maybe it should bluntly show a dialog in the corner of the screen with closest related websites (or pages) and a button to further browse the index using the current page as the query. Basically contextual advertisement only for free.