What do I want out of a bookmark manager?

Epistemic status: Starts one place and ends somewhere entirely different. Probably only applies to my particular set of tools and requirements.

Organization, Retrieval

Generally speaking, bookmarks tend to fall into a few distinct categories:

  1. Things I want to read/watch/listen to later
  2. Things I want to check out regularly (i.e. blogs, email, task list)
  3. Things I want to check out as part of a project (e.g. Quartz for Working with the garage door up)
  4. Things I want to check out as part of a process (e.g. A process for putting evergreen notes to work)
  5. Things I want to buy for myself
  6. Things I want to buy for others
  7. Places I want to visit

But how?

Is there one tool that fits all these use cases? Should there be one?

With Capacities, I’d been using the Weblinks object, then either creating queries (e.g. Things I Want) or referencing them in other files.

Steph Ango put together a nice clipper for Obsidian which allows you to capture full pages or just the selected text. Should I Reduce friction and just use this? In conjunction with Dataview, this could be enough to surface links based on tags or other attributes. Easy and convenient to link across notes, too. And, no reliance on another tool, another service.

Every link becomes a page where thoughts can be jotted down, links can be formed.

Read later

The one exception to this approach might be the first category: things I want to read/watch/listen to later.

These should be ephemeral. I don’t know that something is worth keeping in Obsidian until I actually read it. Of course, I should feel free to delete files whenever it makes sense. Even so, I’m leaning towards having a separate place to “triage” those items, then bringing them in to Obsidian.

Omnivore

Enter Omnivore. It’s privacy focused, open source and syncs with Obsidian. I’ve jumped between more read-it-later services than I care to admit, so who knows how long this one will stick? But it seems well worth a try.

Even better, this thing can be self-hosted. See Deploying a minimal self-hosted Omnivore using fly.io and bonsai.io. I can see it deployed to something like https://later.pcalv.es. It would be pretty interesting if I could somehow make the list of items public but have its management behind user credentials, but it feels like a rabbit hole.

Feedbin

Turns out this RSS aggregator that I’ve been using for years has a read later feature. So that’s that taken care of. Reduce & Simplify.