4 ms·

As an example of how bad the searchability is nowadays, I’ve been creating and expanding my own knowledge base (something like a personal wiki with links to int

by balder1991 3y ago

As an example of how bad the searchability is nowadays, I’ve been creating and expanding my own knowledge base (something like a personal wiki with links to interesting content I find) for about a year. It seems to work very well despite the effort it takes to keep it organized.

iteratethis 3y ago
Same here, but in my notes app. Been doing it for years. And if I may expand a little, not just for crappy Google, I also create alternative local knowledge bases at work. I can't find anything at work. Everything is spread out across chat, Wikis, SharePoint, email. All having different owners, content may at any time disappear or move, there's constant authorization headaches. Whenever I come across something useful that I expect to be of some future use, I make a local copy. File, web page, wiki, anything. Because our information systems are a massive failure.
- renegade-otter 3y ago
  I think this is typical of most work places. Our ability to maintain good knowledge bases has essentially collapsed. Every coding task at work I basically start with reverse-engineering the system because no one really knows how it all works anymore.
chubot 3y ago
I've maintained my own hosted wiki since 2004, and yeah I'm glad I didn't completely outsource information management. It's definitely getting hard to find certain things I know I've seen. Now I just need some kind of open source search engine to run on it ... (a bunch of text files that render to HTML, and ideally following the links 1 or 2 levels deep) ~20 years ago Google desktop search was a fantastic piece of software ... very fast and accurate on your local files. I don't think something like that exists now, and maybe never existed for Linux Search engines are extremely modular and Unix-y. You have a bunch of indexed corpora and you intermingle them at ranking time, with respect to a query. But unfortunately there is no real incentive to provide something that has measurably good results and is also open to your own data and modifications The incentive is to make a walled garden out of it
- knight17 3y ago
  Try Docfetcher https://docfetcher.sourceforge.io/en/index.html https://docfetcher.sourceforge.io/en/index.html State of desktop search is very bad but this is what I use. It is of acceptable quality.
- fsflover 3y ago
  > Now I just need some kind of open source search engine to run on it ... Here you go: https://yacy.net https://yacy.net
phs318u 3y ago
I was about to ask you to share your list, but that got me wondering: is there any tooling for curating, sharing, and most importantly, consolidating curated lists of sites (based on tags rather than categories), such that the consolidated list is then searchable?
araes 3y ago
This is basically what I use my Wikipedia User Page for. Keeping links to all the news websites I happen to have found reasonably interesting content at. Stories I like that Wikipedia will probably not accept. Articles that I want to keep, yet might not be accepted for an article, or I have little faith they'll remain in the article. Probably just need to dump my bookmarks in occasionally.

iteratethis 3y ago

Same here, but in my notes app. Been doing it for years. And if I may expand a little, not just for crappy Google, I also create alternative local knowledge bases at work. I can't find anything at work. Everything is spread out across chat, Wikis, SharePoint, email. All having different owners, content may at any time disappear or move, there's constant authorization headaches. Whenever I come across something useful that I expect to be of some future use, I make a local copy. File, web page, wiki, anything. Because our information systems are a massive failure.

renegade-otter 3y ago
I think this is typical of most work places. Our ability to maintain good knowledge bases has essentially collapsed. Every coding task at work I basically start with reverse-engineering the system because no one really knows how it all works anymore.

chubot 3y ago

I've maintained my own hosted wiki since 2004, and yeah I'm glad I didn't completely outsource information management. It's definitely getting hard to find certain things I know I've seen. Now I just need some kind of open source search engine to run on it ... (a bunch of text files that render to HTML, and ideally following the links 1 or 2 levels deep) ~20 years ago Google desktop search was a fantastic piece of software ... very fast and accurate on your local files. I don't think something like that exists now, and maybe never existed for Linux Search engines are extremely modular and Unix-y. You have a bunch of indexed corpora and you intermingle them at ranking time, with respect to a query. But unfortunately there is no real incentive to provide something that has measurably good results and is also open to your own data and modifications The incentive is to make a walled garden out of it

knight17 3y ago
Try Docfetcher https://docfetcher.sourceforge.io/en/index.html https://docfetcher.sourceforge.io/en/index.html State of desktop search is very bad but this is what I use. It is of acceptable quality.
fsflover 3y ago
> Now I just need some kind of open source search engine to run on it ... Here you go: https://yacy.net https://yacy.net

phs318u 3y ago

I was about to ask you to share your list, but that got me wondering: is there any tooling for curating, sharing, and most importantly, consolidating curated lists of sites (based on tags rather than categories), such that the consolidated list is then searchable?

araes 3y ago

This is basically what I use my Wikipedia User Page for. Keeping links to all the news websites I happen to have found reasonably interesting content at. Stories I like that Wikipedia will probably not accept. Articles that I want to keep, yet might not be accepted for an article, or I have little faith they'll remain in the article. Probably just need to dump my bookmarks in occasionally.