My reading list strategy:

- Send to Feedbin (https://feedbin.com/blog/2019/08/20/save-webpages-to-read-la...)

- Never look at it again

Great writeup. I too have a long reading list - currently at 133.

I use my own little side project (Savory) to track the list. When I come across a page that I don't have the time or energy to finish right now, I save it and add the "reading" tag to it. When I have free time, I can open the reading tag and pick up something new to read.

The best part is that I usually add a couple more tags (e.g. "security" or "economics" etc.) when I save a link. This way, the reading list allows me to filter by topic. It has been an unexpected hack to attack the growing list, since I am usually able to finish multiple articles in a single run, all in the same topic, because there is usually a link between them even when I might have saved them days or weeks apart.

Anyway I like how OP actually has a reading history. I really need to add something similar in Savory. Right now, when I finish reading something, I just remove the "reading" tag and I don't get a neat history.

    22-09-05

    article Tactical Decision Game (TDG)
      - https://www.shadowboxtraining.com/news/2022/04/29/film-at-eleven/
      - title: Film at Eleven
      - scenario of the month
        - make decisions
          - make notes about your thinking
        - compare with experts

I'm still waiting for a web extension that sends a copy of the webpage I'm looking at (for more than 1 minute) to an endpoint I specify along with some metadata like the URL and user-agent. Obviously, block certain domains like banks or email.

I'd like something to build up a searchable index of everything I've read recently so I can easily find content again yet this is NOT something I want a 3rd party to do. I want to self-host something like a tiny Go or Rust server that only uses 10mb of ram to index all the pages into an embed rocks/level/badger/etc. database.

This is a neat writeup. It's fun to think about how to potentially automate this kind of tracking.

> I wish there was an easy way to filter by "independent websites"

This side comment from the post is intriguing. Other than manual curation, I wonder if there is a way to identify commercial vs independent domains? This would make a really good entry point for a specialty search engine for indie sites only.

I Print-to-PDF everything I've ever found interesting to spend longer than 2 minutes lingering on .. and as a result, I've got 20+ years of Internet articles to go through and read offline, any time.

Its very interesting to see the change of quality of technical writing over the last two decades. There's a definite, observable increase in click-bait style writing.

I have started using Obsidian (at work). And I copy/paste into it any web content, text or image I find useful from the intranet, emails, meet. I try my best to organise things. But for the most part, I use the search engine and the [autogenerated] links.

The only requirement when adding content is to figure out whether it should be added to an existing note or a new dedicated note should be created. [btw, note nesting does exist in Obsidian]

With this simple workflow, you completely eliminate the notion of provenance of the knowledge. The knowledge is here and up to your organisational habits.

After some time doing that, you end up with VERY dense notes (in term of knowledge/line ratio), and very few useless (distracting) content.

For the moment I like that A LOT !

Tampermonkey is great for this, because it can log everything and it brings its own XMLHttpRequest:

  GM_xmlhttpRequest({
            method : 'GET',
            url : 'https://myloggingurl/?client='+client+'&url='+encodeURIComponent(window.location.href)+
                '&title='+encodeURIComponent(document.title),
            responseType : 'json',
            onerror : function(e) { console.error('URL-Logger', e); },
        });

I've been logging all my web activities since 2018, it's been a great tool. On the server side, I filter out ad spam and other extraneous URLs, and then run a cronjob that converts all new HTML documents it sees to PDFs with wkhtmltopdf. It's been a great tool for finding stuff in those moments where I go "hm, I remember seeing something about this months ago..."

The author doesn't appear to have documented the bookmarklet itself. If they are here or another person, can you suggest what it might look like to have a bookmarklet collect the url, page title, meta description and image, and then set window.location.href ?