Creating a commonplace book with DevonThink?

Since I started using DevonThink two years ago, I have tried several times to use it as a commonplace book, clipping any interesting article I found on the web to Devonthink. Subject matter includes professional articles about developments in the cloud and telco business world, and also articles of personal interest, about politics, current events, history, science fiction, etc.

The last effort began a few months ago, and now I have a big, useless database of articles. “See Also” does nothing for me here. Thinking of deleting it all and just starting over again. Maybe on raindrop.io — if DevonThink won’t differentiate them, why have them cluttering up my hard drive?

Has anybody tried something like this and succeeded? What are some best practices for creating a useful archive without spending a whole lot of time on curation?

Thanks!

If the data is useless in DEVONthink, why wouldn’t it still be useless in the cloud? :thinking:
At least it’s always accessible in DEVONthink.

PS: There’s nothing wrong with creating, shelving, or deleting databases. At one point I had over 150, most that were one-off tests.

3 Likes

Yes, I do exactly that - using DT3 for a variety of personal and professional uses with multiple databases.

Everyone’s use case is different so you need to experiment with what works to retrieve things. My professional use databases are consistently organized via groups and custom metadata. My miscellaneous info, including professional articles to read someday but not linked to a specific case at present, are all lumped together without any particular organization strategy.

For me, the “Ahah” moment was drying Devonsphere for search as opposed to the built-in DT3 search. They each have their advantages, but I have found that with regard to speed and ease of quickly retrieve a random item, Devonsphere works best and is particularly nice since it is in the menu bar and thus always instantly available.

2 Likes

Interesting! Thank you! What custom metadata do you use?

My thinking is that raindrop.io might be a better interface for this kind of thing than DT. Plus it gets it off my hard drive.

I need to think about why I am saving all this stuff. It’s not like I’m referring to it again very often.

If it’s just a matter of notching up tallies of articles read, that’s fine, but there may be a better way of doing this.

I need to think about why I am saving all this stuff. It’s not like I’m referring to it again very often.

Amen! I think about this topic often. DEVONthink is amazing but it also can lead you to gather things that are only ephemeral.

You could always set up a smart rule for the junk drawer database and have it cull out anything you haven’t accessed in some period of time.

7 Likes

Various project due dates

Referring client name/email address

Booleans for which staff are assigned to the case

Booleans for various levels of alerts if a client requests an update on a case - which then trigger other smart rules

Rich Text field for internal office QA/commenting on a case

Quoted fee for the case

Paid or not paid for the case

Booleans for various processes in handling the case, such as receipt of needed documents --> with smart rules or smart groups to alert if something is missing

Overall I essentially use it as a project management system for the consulting part of my practice. But the document handling and search capabilities of DT3 are infinitely superior to actual project management software, so it works well. Plus part of my work involves identifying applicable academic research applicable for a case - and thus I can create links in a case to relevant articles.

Very interesting ideas there. I haven’t looked into custom metadata at all, although it sometimes occurs to me it might be useful. For example, when clipping articles from the web, often the publication date isn’t included in the clipping.

I use a group for each project, and include a document in that group containing the assignment, fee, deadline, etc. Later I drop the invoice in the same group, and replicate the invoice document to another group called “unpaid invoices.” Later still I move that replica to another group called “paid invoices.” (Hopefully! Almost all the time!)

Recently I remember reading a comment, here or elsewhere, from someone who says they use groups and replicas like tags, and I said to myself, “Yeah.” In my use case above, another person might tag that document “invoice” and then add another tag “unpaid” (later changing “unpaid” to “paid.”) To me, it seems to make more sense to use groups and replicas. As Bluefrog has noted, DevonThink is flexible and fluid – easy to start with one organization system and switch to another.

Query related to this thread, which I started as a new topic, for convenience of future searchers:

The relation between these queries is that a primary reason I’m creating this commonplace group is to find connections, and I want to maximize the opportunities to find those connections automatically, while minimizing the work involved in creating the commonplace book.

I use https://pinboard.in/ as my quick place to stash things. It’s a reliable old-timer that likely inspired raindrop. It also creates a cached copy in case the original source disappears.

2 Likes

“Booleans” = lists of some sort? Dropdowns, checkbox, etc.?

A boolean custom metadata field in DT3 is depicted as a checkbox

1 Like

Help > Documentation > Preferences > Data and Appendix > Metadata > Custom Metadata.

I use DT as a “maybe-read-it-later” service.

I used to employ a variety of other services for this, though they all suffered the same problem: me! However, I’ve since recognized that I’m not going to change. My appetite for new headlines will always exceed my capacity to read the ones I’ve already found. So, I’ve come to rely on DT to address that issue directly. I save everything interesting, use a deliberate and systematic tagging system to give it some context, and then I let these finds inform later work through search or browsing those tags.

It may be that I’ll never directly use many of my finds. Still, I’m neither a librarian: I don’t want to make tough, ambiguous decisions about whether something meets an arbitrary threshold for keeping every time I find something to save. Nor am I a janitor: I don’t want to make tough, ambiguous decisions about whether something meets an arbitrary threshold for not getting trashed in some sort of regular clean-up.

This is the most stable and frictionless setup I’ve had. A couple of key mechanisms make it possible:

  1. I don’t sort things myself. I have an automation that automatically places new items into a year → semester group hierarchy, based on date added.

  2. I use a “closed” tagging system. I tag every item using an automation that allows me to choose tags from a few lists. This makes tagging quick and effortless, and prevents tag overgrowth.

3 Likes

It sounds like you are exactly the sort of read-it-later packrat that I am. I think I will steal your system.

You have a group called 2020, with subgroups 1H and 2H?

Closed tagging is a great idea. How do you automate that though? Is the automation on both iOS and MacOS?

I think I’m going to need to make a habit to go through my clippings every few days at least, just to see what’s worth keeping.

My semester group hierarchy is 2020 → Fall (or Spring or Winter). It would be trivial to make it more granular but anything smaller than terms seemed less useful to me.

Indeed, I’ve set up the closed tag automation on both macOS and iOS. I’ll try to write it up soon, but roughly I:

  • create a text list of the tags I want to use, one per line
  • when the automation is triggered, pull the text from that text using the DT uuid
  • split the text by new lines
  • present the resulting options as a multi-choice prompt
  • save the selection as tags

The items stay in an “organizer” group when I first save them. Once they have tags, a smart rule automatically places them in a semester folder.

Very interesting. Thank you.

Do you do the automation with Keyboard Maestro on MacOS and Shortcuts on iOS?

You got it.

Here’s the raw Shortcut: https://www.icloud.com/shortcuts/dc633f6f7f2a45f1b9c60712f973bc18

Select a record → Share → If the record is PDF, select the “text” part of the record → Tag with Presets.

You have to replace the UUIDs in the dictionary about halfway down.

I’ll share the raw KM macro when I’m at my desk later!

1 Like

I have a commonplace database in DT3. I’ve set it out as a series of folders A - Z, another labeled ‘Quotes’ (because I like quotes!), and another labeled ‘Ideas and Words’ which are basically whimsical thoughts of my own.
I try to use the alphabetical entries fairly obviously as a main subject heading but use tags for additional subjects within the entry. It works well enough for what I want, but I’m not really pushing the envelope here, just collecting stuff as and when. Power user I am not.

1 Like