But categorizing is fine if you can put multiple categories on things. That's the whole difference between flexible tagging and rigid folder, i.e. the one thing GMail got right vs. the real world analog of a folder.

The real world solves it too btw. in the same way a relational database did: Indexes ;) Tagging is nothing but different indexes of the same underlying thing. Be it a book or an email or whatever else.

And this top level comment is actually weird to me. Maybe I'm just a non-Chrome-user? I use Firefox. I never bookmark anything any longer. I literally just know what to type to get to what I need to from the history. Like HN. I literally just type n into a new tab and HN is the first suggestion based on how often/recently I went here. No manual bookmarking or anything non-local needed.

Similarly at work I just know how to find certain things by typing into the search bar. And it even learns from recency. Our app URLs recently changed and I had to start typing the new unique parts of the URL recently for a few tries until it went up in the recency bias calculation. Sure like I had to scroll down a few entries in the history search a few times. Now I just know what to type to get where I want to again.

I see others struggle with rigid bookmark structures that they can't find anything in like ever on the regular though (URL change or not) :( And (at least with firefox) those bookmarks would even get searched too! Of course, unlike history they would have to manually update bookmarks. I really don't get it how people even still use bookmarks any longer.

I personally dont usually use bookmarks unless its something truly obscure that I may want to refer to later (like an article or something.)

I like the concept of indices but a lot of indices i see (like in textbooks) are incomplete. I have to think "ok what might be related to this?" And it certainly is not an easy problem!

That's not the problem of the concept of an index. It's bad implementation. If your index only indexes as if it were a static folder structure, then you have to think exactly like the guy that built the index. If you instead allow multiple indexes/index entries for the same underlying thing, then you have tagging.

Bookmarks are folder structure. Theoretically you can bookmark the same thing under a different part of the structure of course. I've never really seen people do that though, except accidentally because they found something the liked, bookmarked it in on place, later found it again, didn't remember they already bookmarked it and bookmarked it in a different part of their folder structure. In a way they also "tagged" except that the actual bookmark, the thing is not the same. It's a copy. If your bookmark goes to some address that might change while pointing to the same actual thing, you'd now have to update (and first find) a multitude of bookmarks in your folder structure. A file system that has hardlinks could implement "tagging" too :)

The last startup I worked for (Edvo, since shut down) was actually trying to solve this, not in a bookmarking way but in a way that was like "bookmarks represent knowledge or ideas" and was meant to be a graph-based mind mapping tool.

The tech was cool, but I think they ran into the issue you just suggested - people dont really bookmark, and it was hard to make people see it as anything other than a bookmarking tool with categorization.

I use Chrome at work and Firefox at home, and I still bookmark on both. It never occurred to me to not.

I agree, but my NAS is the messy room and I don't want to clean it but I want it cleaned.

I need to find the LLM tool that "cleans my room" for me by rearranging my 20+ year old dir layout that just keeps growing.

There are aptitude tests for those two traits: Analytical Reasoning and Ideaphoria at Johnson O'Connor. You, like me, probably will score high on both.

That looks cool! Id be interested in taking a test but I'm less interested in paying $1k for it :)

FWIW, it's about a 6-7 hour series of tests. Worth it, imo, but yeah, it's pricey.