Yet another DevonThink topic

I bit the bullet and purchased DT a few months ago having tried to make nearly every alternative work.

Evernote: Bloated, subscription based, not a great `Mac citizen in terms of design
Keep it: Just did not work for me, seemed buggy with sync and expensive for what it is.
Finder: yes worked, just saving stuff was a pain.

DT copes with everything I throw at it. I have 4 databases, Archive, Resources, Projects and Areas based on the PARA principle which seems to work quite well. It is expensive initially, and not the prettiest, but I came to realise that substance wins over style every time.

I think though you need to commit to using it for nearly everything you want to keep long term such as articles, PDF’s invoices, notes etc.


As bikers say, “Chrome won’t get you home.”


Been using PARA in one database. Thought about using 4 but thought one would be simpler. Why did you settle on four?

Killing features of DT for me:

  • Ability to sync encrypted databases via cloud services such as Dropbox. Enables me to throw sensitive documents in DT.
  • Ability to also encrypt DT databases locally (I do not deed that, already use FileVault)
  • Easy Backup of DT databases: Just move additional copies to other disks or cloud services.
  • Getting documents in or out of DT is as easy as moving documents in Finder. Including folder structures.
  • Splitting, consolidating or archiving whole databases easy as well.
  • Powerful search engine with Boolean operators (I frequently use the NEAR operator)
1 Like

Now I need to google PARA.

And I’ve had a change in job circumstances that suddenly makes DT look more attractive to me.


@KevinR, that would be funny if it wasn’t literally why I haven’t replied to this thread yet.* (I went all-in on Things for a day and then switched back. I learned a few lessons though! It wasn’t a waste of time! It wasn’t!)

I have actually fallen more in love with DEVONthink after some extreme frustrations early on. As everyone says about it, it’s too powerful. The “just play with it” approach can make onboarding extremely difficult if you don’t ratchet up your use at the right pace.

I have to write more about my current use-cases, but here are the evolving use cases that have led me to love it in recent months. Messy post ahead:

Scaffolding use-cases for making the most of DEVONthink

1. Reliable, highly searchable files repository

iCloud’s tendency to poorly manage what documents I want to access offline made it untenable for maintaining access to files on iOS. So, the first use-case for DT was to be a dependable place to access anything I’m working on.

As a secondary benefit, I have a variety of Shortcuts that use template files and lists to drive some functionality. I use DEVONthink’s URL scheme to manage these so that I can update those templates from any device while keeping these “resource” files universally accessible.

Finally, I use DEVONthink’s indexed search all the time to find specific references I used years ago.

2. PDF annotation database and study tool.

I wrote my comprehensive exams for my PhD about a year ago. I wrote a script that took PDFs in DT and automatically generated flash cards in Anki from the highlights. Each flashcard had a direct link to the page of the PDF relevant to the question on the flash card. It was highly motivating, fun, and helped me pass.

3. RSS reader and read-it-later service

I had too much noise in my previous RSS approach. I also found it hard to keep up with all the articles coming up. Finally, I wanted to save and use many of the articles in my readings as PDFs.

So, I set up DEVONthink as my RSS reader and read-it-later service. It was quite involved to set up, but now Smart Rules and Smart Groups self-manage my RSS feeds. Each feed has granular rules to dictate how long items can stay for. Every time I check my feeds I have a very manageable amount of reading to do, and I can configure individual feeds or keywords or whatever to save anything really crucial before I even have to review it. (As a bonus, I save $70/year from Feedly subscriptions and $X/year from read-it-later services.)

Everything coming in is automatically converted to a nice, clean PDF and I can mark that up with all the conveniences of the PDF ecosystem.

4. PARA/CODE system

Although I don’t strictly adhere to Tiago Forte’s PARA (Projects, Areas, Resources, Archive) system, his thinking on knowledge workers’ management of their working materials has changed how I approach personal knowledge management and personal project management. In particular, I appreciate the sentiment that we need to be faster and more intentional about Capturing, Organizing, Distilling, and Expressing (hence, CODE) knowledge materials.

So, I have a script that takes in an OmniFocus project and creates a new folder for that project in Finder, indexes that to DEVONthink, and creates a link to both in the OmniFocus project. Likewise I am trying to stick links to files important to a project in a variety of useful places. The idea is to try to make project creation and completion a faster cycle, and to bring project management and material creation closer together by interlinking all of these things.

(@MitchWagner here’s your link on PARA, and see this post for some discussion about buying into Tiago’s stuff.)

5. Notes and annotations database

This is where I am now. Incidentally, one lesson I learned from my recent Things adventure is that I put a lot of non-actionable items in OmniFocus. As a result, probably half of the items in there were just noise. Most of these items were just thinking on potential projects or links to things I wanted to investigate. Instead of putting them in OmniFocus, I aim now to keep my task manager clean and lean. Instead, I’ll rely on PARA/CODE and put those non-actionable items in DEVONthink as a notes database. I plan on working in DT3 on my desktop a lot more to make this happen.

A few generic takeaways

  • DEVONthink To Go (DTTG) is in an awful state. It works, for sure. But the lack of feature parity with DT3 is really ruinous these days. The DEVONtechnologies team say they’re working on big things with DTTG, so I’m really looking forward to whatever that may be.
  • I use indexing, not importing**, for working files. I just find it easier to interact with files this way, and it makes the on-ramping a bit easier as you can continue doing things as you did them before DEVONthink while learning more about the tool.
  • I have recently realized that my use of DEVONthink is a little like when I first learned Photoshop. It was an awful, difficult experience and I constantly produced junk results at first. I kept at it, though, and slowly learned how to use the tool more effectively and efficiently. Then it started to change my mental model of the work I was doing for the better.
  • Some part of all of this might just be attributable to Stockholm syndrome thanks to DEVONthink’s steep purchase price and the amount of time I’ve put into it at this point. :upside_down_face:

* It is still very funny.
** Incidentally the original source for that indexing/importing post is dead, but I had a copy of it in DEVONthink from a few years ago. :wink:


Interesting post on DT!

Your comment on OmniFocus made me remember a casual remark from a person using OF as a Finder replacement, on the theory that documents are things you want to take action on.

1 Like

Exactly! Except for me, it’s become the other way around. Actions are things you do to/with some material (documents, notes, contacts, diagrams, etc.).

Do any of you do anything with email archiving and searching in DT? I like the idea of importing Google Takeout (or just doing a big IMAP sync) and being able to call up data associated both by thread title and by DT’s connections it makes itself.

I left a role at a university using Exchange, and dumped the email inbox I had on the Exchange server into an Archive database.

My mistake was putting it in the same Archive as all of the other files I archive. Search is quite messy unless I specify “not that group”. I have to move it at some point.

I also sometimes fail to remember it’s in DT. No other mail is in DT, so sometimes I’ll search my email app (Spark) and will be confused about missing results until I realize I have to switch apps to search for messages received/sent in that time period.

Interesting. I can’t find that it has any ability to do ongoing email import unless I script it. So unless it’s a final step like in your situation, I’d just be creating a new job for myself to keep up with it.


How are you automating highlight export? I’ve tried to figure out a way to do this with pocket, but am not sure if the support highlight access via the API.

Ifttt creates one text file per instapaper highlight in a Dropbox folder that DEVONthink indexes. Pocket doesn’t have that as a trigger in IFTTT…

1 Like

You use indexing, rather than working inside DEVONThink, for documents you’re likely to modify for works in progress, and store archives - documents you’re no longer using daily - inside DevonThink databases?


Particularly because other apps can easily look at the folder structure on Finder/in iCloud Drive, but it’s not so easy to open files in DEVONthink from other apps.

That said, I imported everything on a whim this evening out of curiosity. I’ll report back if I find it’s better.

1 Like

Would you be willing to share how you achieved this and your read it later workflow? I’m not too interested in rss in devonthink, but having a way to turn DT into a kind of Instapaper with PDFs I can highlight (and then summarize those highlights into markdown files) would be super valuable. Are all your read it later items coming from rss or also from random links you find?

For sure. It’s one of the things I’ve been meaning to write up in detail.


Have you tried capturing links as PDF using the clutter-free setting on Mac? That usually does it for me, and then I have a few little automations to support the process. A Smart Rule changes RSS News items to PDF by default (though you don’t need it). Just for the record, the stylesheet DEVONthink uses to style RSS items can be modified at Preferences→RSS.

On iOS, I have modified Viticci’s import shortcut into my own version, PDF to DT. Share a link to it and it sends it to (in my case) either your Resources inbox, an “Organizer” group (which sorts items automatically based on Smart Rules), or a Review group (which is my “Read it later” folder). You’ll note that the Shortcut has some built-in HTML/CSS <head> to make things a little prettier.

If you have trouble with that shortcut I’m happy to help fiddle with it. (I haven’t readied it for sharing, so you’re getting the uncut version.)

If your file is already in DEVONthink, Convert→PDF works, usually with Clutter-free.

1 Like

Update: full-on Importing was utterly foolish. iOS apps need a good Files citizen serving up items in order to cooperate—DTTG is not that. (Hopefully that will change.) It now takes dozens of taps to open up my markdown notes in iA Writer whereas Files locations loaded into iA Writer automatically.

Worth noting though: the downside to the Indexing route is that some round-tripping is required to sync things up if you make a change in a Files file on iOS. DTTG won’t recognize the change.

Therefore I use DTTG almost exclusively for PDFs, to show/share files when on the go, and to drive those resource files for certain Shortcuts.

@ryanjamurphy This has all been very interesting and useful. As you might guess from my questions, I am staying with DevonThink a bit longer. I will try your read-later workflow — though I’m ok with Pocket.

Why not use the DT clipper to save docs in DT?

I wonder whether there is some way to point Drafts and DevonThink at each other? Probably not an elegant way.

Parenthetically, I have always been confused by the difference between areas and projects in GTD. When I say “always,” I mean since I was first introduced to GTD in 2007 or so.

But now after speed-reading the PARA document, which I was introduced to in this thread, I get it.

Projects have goals and completion states. Areas don’t.

That much is what I’d been hearing for 10+ years but it still seemed confusing. My work for nearly my whole career is: I write articles. An individual article doesn’t really seem like a project. It’s an … article. I often turn each one around in a couple of hours.

But Tiago’s PARA document helped me to understand by using good examples.

“Home maintenance” is an area. “Replace the backyard fence” is a project.

Areas are never done; they just keep going until they stop. You keep having home maintenance until your kids are grown, you retire to Florida and move into a condo where it’s someone else’s responsibility.

An article doesn’t seem like a project to me because I’ve done so many of them that each one seems like a single action. I mean yeah each article is actually a dozen or more actions — preparing for and doing individual interviews, identifying and reviewing source documents, writing the headline, the lead, the rest of the article, finding a photo, entering it into the publishing software — but it all comes naturally to me. So my to-do list doesn’t have each of those steps; it just says “write XYZ article.”