For the same reason I tend to leave tabs open so they stick a little more. But I have too many tabs open so I know that my model utterly broken.
Please help.
If I need a specific bookmark, I press Shift + B in the browser and search for "linux SSH install" to get that particular page.
The reason I get that page is; I made a hashtag when I saved it as my bookmark. I try to come up with tags to find them back in seconds.
The reason I don't use FireFox tags itself is that if I imported the bookmarks into Chrome I lose the tags. And with my "solution" I don't lose them.
I either find a use for them at some point, or I don't. Doesn't bother me if I don't as I'm not trying to process everything I come across, just putting them in a trusted system where I could find them again if I needed to.
I do find it useful if I have a personal research project in an area of interest to start searching with what I already have vs. going to Google. Most of the links I have are vetted either by me or a source I trust so they don't have a bias to who has good on page SEO.
Many of them are recipes. Some are items on my wishlist. Lots of them are news articles, some essays. Lots of them are guides/tutorials howtos for Bash, Python, Linux, networking etc.
The recipes I use from time to time.
The IT-related ones I consult from time to time whenever I need them.
And since Pinboard allows for fulltext search it's quite easy to find stuff when I need it.
[Edit: One can also pay extra to enable archiving, so when articles are removed from the WWW one can still access them via Pinboard.]
At that scale, recall is indeed a problem, and one I've been planning a solution to for quite some time now. Think custom search engine with automated topic detection and tagging.
Almost all of them are saved as full text + PDF in my email (Gmail) inbox under separate labels like 'JS', 'Design', 'SEO' etc.
Gmail's search is pretty decent so it helps me find stuff that I am looking for quickly.
1) google.com 2) office (all internal links like jira, wiki, stash, etc) 3) daily-read (hacker news, slashdot, etc. ) 4) archive (the links which I like and refer as notes when needed) 5) to-read (some links which I found interesting to read but don't have time to read immediately)