Digital Information Infrastructure with DEVONthink

Having dumped ninety percent of my packing onto Gwen I tackled the hardest ten percent: my business records and files.
Writers are pack rats, mostly, whereas professional military learn to travel light, again mostly. This dichotomy could have made me schizoid were it not for the most wonderful invention for writers since the eraser on the end of a pencil: electronic files.
“The Cat Who Walks Through Walls” by Robert A. Heinlein

I freely admit that I am a squirrel. When I find something interesting, I want to keep it. (Humans excluded, I’m not a psychotic squirrel ;-)). But despite a tendency to keep and use what I find interesting, I end up being much more closer to a digital nomad than a hoarder — because, well, digital.

Whenever I get something that I want to keep, yet it is not digital but can be digitized, I do so. Books, movies on DVD, postcards, letters, whatever — with very very few exceptions it ends up digital. And with today’s high-capacity hard-disk drives (in the terra-byte range(*)), storing digital information is easy. So is creating full backups (because otherwise, one mistake or accident can destroy it all).

But that creates new problems — if you want to go beyond a one-way digitize and forget street: Where do you keep all that data? How can you make sure that you can access it again, or even be reminded about it occasionally?

Personally, I love DEVONthink in this regard. I have already written about this program a couple of times — just have a look at the entries tagged with DEVONthink. Recently, I reorganized my DEVONthink databases and — for the moment — I am very happy with them.

Currently I have 30(!) databases. That sounds — hell, is a lot, but there’s a reason for each of these databases. I also have five more which are now superfluous but I have to yet transfer the content into the right databases.

To get an idea of the structure let’s first look at the different categories of databases. There are four without a category and 26 with a category:

Without a category:

  • Collection Inbox
    I do not like the Global Inbox of DEVONthink, so I created a separate database to use as inbox. The main reason is that when I backup the DEVONthink folder, I want to backup all my information, including what I have in the inbox.
    Practically, I use the “Collection Inbox” to sort the information I get in DEVONthink in folders (groups) named after the other databases. It’s the main sorting hub which is sometimes useful when the respective database is not open.
  • Community RSS
    I have written about the use of RSS feeds to stay up-to-date with journal articles here. Likewise, you can get RSS feeds from most blogs and even YouTube channels (standard format is: — just replace USERNAME with the name of the channel, e.g., Thunderf00t ). You can create RSS feeds with the Data – New – Feed … menu item. Really useful and all of my RSS feeds are in this database. I use groups to, well, group them, e.g., YouTube, News Feeds, Interesting Blogs, Writing Blogs, Retractions and Frauds, Journals, Career, and Search Terms.
    I also use this database for correspondence I want to keep. At least, if it’s work-related. DEVONthink can easily import eMails (drag and drop works).
  • Mobile Syncing
    There are some information I would like to have easily available on my iOS devices, and while every database has a “Mobile Sync” group (which I also use), I use an own database for this information. It contains, for example, a couple of Maps (when I don’t want/cannot use the Maps App) or a couple of images I want to keep with me (among others, family, but also profile pictures). While you could also do this with GoodReader or any other program, I found it more useful to have a database for this purpose.
  • Work Projects
    Anything work-related is in this database, for example, Publications, Teaching, Career, Advisory, and Groups for specific projects. Once something is finished (e.g., a publication) it does not stay there however. It gets sorted into an Archive.

With a category:


The databases with Archive in their name (e.g., “Archive Documents and Works.dtBase2”) are just that — archives. They contain information about the past, finished projects and the like.

  • Archive Documents and Works
    This database is my main document storage. Contracts, important documents, whatever, it’s stored here. I also use it to store the final PDFs of any works I have created (e.g., books I have written). It’s the one I absolutely need to have when the apartment burns down.
  • Archive Private Memories
    Things I want to remember from my private life are stored in this database. Photos with friends or partners, family, private eMails I cherish or hate, private triumphs or failures. They are stored in groups according to the year, with subgroups about the specific person or event.
  • Archive Private Projects Full
    This rather large database contains the files and material of private projects, e.g., InDesign files, images, etc. and the finished PDFs as well. Groups first denote the kind of projects (e.g., “Books and Writings”, “Drawings”, “Gifts”, “Photography”, “Posters”, etc.), then the year, then the concrete project. If I look for a specific project, I use the search function and the Reveal icon of DEVONthink to go to this place in the database.
  • Archive Work Projects Full
    This is the equivalent for work projects, e.g., “Advisory”, “Applications”, “Ethics Commission”, “Peer-Reviews”, “Publications” (subfolders according to type), “Teaching” (subfolders according to type), etc.


The databases for any audio files, although I differentiate between three major types.

  • Audio – Music
    I use iTunes to listen (and sync) my music, but I store all my music files in this database as well. Not that I don’t trust Apple, but I don’t trust them. 😉
  • Audio – Podcasts
    I listen to a couple of podcasts and use this database to store the files after I have listened to them. Yeah, digital squirrel, I could delete them, but who knows, there might come the time when I want to listen to them again. Especially timeless ones (e.g., non-technology related podcasts).
  • Audio – Soundfiles
    These are short soundfiles, which I sometimes use for system alerts or in Apps.


Databases I use for private ideas and projects.

  • Create – Ideas
    My main idea collection. They end up in text or rtfd files (if images are included). Currently there are about 16k files and I guess it’s time to check and tag these files for their uses.
  • Create – Projects
    I am a firm believer in focusing on one core project and — at the same time — collect ideas for seven or less central projects. Each project got a group in this database. Each group contains the relevant information.

Deep Archive

In contrast to the normal archives (see above), these archives are rarely updated. Avoids having to sync/copy large amounts of data each time you make a backup.

  • Deep Archive – Apps
    A few apps (mostly widgets and other small apps) are stored in this database. I once wanted to reinstall a widget and the site was no longer available, so I learned to keep backups of these handy little helpers. I would not want to store larger applications here, that’s something for a simple folder structure.
  • Deep Archive – Fonts
    All fonts I have and use. Given that some apps add additional fonts and you can some high quality fonts online, I need a place to store them. This is the place for them.
  • Deep Archive – Mails
    At the end of the year, I download all of my emails from icloud into this database, sort it according to the person or project (groups) and delete them on the server. Keeps the use of the mobile storage space down and allows me to access them much, much later if I ever need to. Rarely used, but nice to have.
  • Deep Archive – Social Media
    A couple of good tweets or other screenshots, which I have already digitized (OCR). But who knows, perhaps I might need the original screenshots sometime.


I love (good) images and I have collected hundreds of thousands of them. While I have a pretty good memory I need to handle these amounts somewhat. I tried to solve this problem by using the one aspect I can always remember: what type of image is it?

  • Images – Argumentative
    Argumentative contains images that make a point. These are visual jokes, PostSecret postcards, infographics, and the like. Images that specifically try to convince a person or convey something.
  • Images – Comic Strips
    In contrast to argumentative images, these images are part of an established comic series. For example, Dilbert, Calvin & Hobbes, 2GAG, PhD Comics, Incidental Comics and the like. Yeah, I visit a couple of them each day and download the strips.
  • Images – Computer Cuts
    Screenshots from Mac OS or iOS apps end up in this database, including screenshots of games (including from my PS4). Also covers of games I have bought.
  • Images – Design
    These are images that could have been put with argumentative images, but while they usually argue for something, it’s not exactly their main goal. Examples are magazine covers, t-shirts, DVD covers, even scans of money or flags. Often text and photos/drawings.
  • Images – Paintings Drawings
    Anything that was drawn or painted that is not argumentative ends up in this database. For example, paintings/sketches by Luis Royo or Leonardo da Vinci, but also cuts from comics and graphic novels. However, cuts only, not complete comics (these end up in Sources — Private).
  • Images – Photography
    Anything that was photographed, sorted in groups like Animals, Astronomy, BDSM, Documentary, Food and Drinks, Graffiti and Street Art, Human Body, Landscapes, Libraries and Books, Living, Military, People, Plants and Flowers, Steampunk, Street Photography, etc. Some of these groups have subgroups, e.g., People has subgroups like Women, Men, Children, etc. or Human Body has subgroups like Dresses, Jewelry, Make Up, Tattoos and Henna, etc.
  • Images – Video Cuts
    Screencuts from any video, from movies to series to YouTube videos. Very comfortable to do with vlc (cmd + shift + s).

My Works

These three databases contain things I have created myself, however, they contain the material to build something with it, not finished projects (these end up in an Archive database).

  • My Works – Photos Graphics
    My own photos and graphics — whether made per hand and scanned, or on an iPad, or photographed with an iPhone or my X-E1. However, I do not store all my photos here but only good ones I want to use in the future (about 5% of the photos I make). I do not use Aperture for it, because I want to prevent acci
  • My Works – Slides
    I export the presentations I have made as PDF and put the slide into different groups, e.g., “Science”, “Time and Task Management”, etc. My goal is to have these slides immediately available when I need them, e.g., in a course or online with Google Video Chat.
  • My Works – Text
    Intended for smaller texts, not sure whether I need it.


  • Sources – Private
    A fairly large database which contains books and other material (not webpages!) that are not work related. They are tagged by topic (e.g., BDSM, fiction, comics, instructions, photography, quotations, cooking and drinking, poems, etc.) and named after the author and year. Mostly they are stored in groups (A-Z), but there are also groups for special topics, e.g., instructions, lyrics, and scripts.
  • Sources – Web
    The database for PDF exports of websites. Whenever I see an interesting information/article/website online, I click the DEVONthink browser plugin to have the current webpage imported in DEVONthink as PDF. Given how fast things can change online it’s an invaluable feature, especially given that the website URL visible in the “Show Info” dialog and the date-time code of the file tell you everything you need to know to cite the website. Just make sure you actually get the website, not an automatic forward to a questionable website, or only part of the page. Usually works very well, but there are exceptions. In these cases I simply print the page as PDF or export only the text and add the URL in DEVONthink manually.
  • Sources – Work
    Work related articles and books. Tagged mostly according to topic and source (haven’t tagged all yet) — these two are the two categories I think you cannot do without. Topic is e.g., academia, creativity, critical thinking and reflection, ethics privacy and security, media technology, mobiles, persuasion, statistics and methods, teaching and learning. Source is not the specific journal, but the type of publication. It is mostly relevant for distinguishing between books, journal publications, conference proceedings, magazines, etc.
    Note that while I store all my work PDFs in this database, I also copy the articles/books I have read into my reference manager (Papers 2). But I need this differentiation between “all that I have found” and “this is really relevant and I am going to cite it”. Otherwise I would be swamped. After all, this database has about 8000 files in it.


So, these are my digital nuts and how I store them. I think it works, and I can work with them.

There are — of course — other ways to deal with all the digital information.

One other method is to go strictly be the source. Where did it come from. And in some cases, I do so, within specific databases — e.g., paintings/sketches by Luis Royo or photographs by Bettina Rheims. But what do you do with images you get from tumblr or twitter? It’s hard to identify the source. If I really want to cite an image, I might use Google Image search to find the source, but that’s something I’d do for a very small amount of the images. Not as default.

I also do not use a database for videos. The reason is that this database would be huge and given that the backup would not be incremental, but involve the whole database each time, it would take too long. Instead, I store the video files I want to keep in a simple folder structure (A-Z) in the same directory I store my DEVONthink databases in. Same with Applications (one folder only).

I also do not store Circus Ponies Notebook files in DEVONthink. I had some problems with these files when stored in DEVONthink. So they are in a simple OS X folder as well.

Why not simply use an OS X folder structure for everything? Among others, DEVONthink allows better (hierarchical) tagging, I have environments (databases) which limit the search automatically, I like the “Three Panes” preview view, the ability to easily create replicates, including replicates of folders and files into the mobile sync folder. This allows me to create a custom selection of data I can carry with me on my iPad and iPhone (several GB). Not only allows me to have the most important documents with me, but also provides me with a library in my pocket.

And yup, so far, it is working well. I love the ability to tag, or to change views and quickly scroll though a lot of images. And to get immediate feedback if I have imported an image or text a second time (DEVONthink goes by the content, not the file name, and is surprisingly accurate (but not perfect)).

How was the old saying? A place for everything and everything in its place.


(*) Note to future self in about 2024 (if I make it that far): Yeah, terrabyte was once seen as high capacity.


  1. Doesn’t having so many databases defeat one of DT’s main features (‘See also’)?
    I thought this only worked within a database. Mind you, I rarely use that feature, as I can’t seem to exclude folders from it on a per-use basis.

  2. Hoi Joao,

    good point. I don’t think it works across databases, but I do not see it as a problem either (don’t usually use it anyway). Even if I were to suddenly use it, it would still be useful given that the searches would probably be limited to the databases anyway. For example, if I look for a webpage, I would want to find other, similar websites. Given that they are all in the websites-database as PDFs the feature would work. Same with articles and books for work. In this case, I think it’s actually an advantage. Not only keeps it the amount of possible (but likely wrong) hits down, it also prevents private texts appearing in a work-related search. This can actually be an issue when you work in a public setting (open-plan office, cafe, library, conference) and private reading material might not only be a nuisance but raise a couple of eyebrows. I think this judgmental attitude is tedious and often hypocritical, but the fact remains that it’s seen (and sometimes is) unprofessional to bring private material into a work-setting. And if you use a computer for work and private purposes, that’s a constant issue that can be prevented with multiple databases. Only exception here would be the websites-database. Hmmm, probably have to look into that issue somewhat more.

  3. Might start implementing (a couple of) multiple databases for one reason.
    My referenced bibliography pdfs are just too many, and interfere with my notes on searches. Might do a small db just for notes.
    One thing I’d like to see if you ever manage, is a review on the (no longer new) Devon app Devonshpere express. At first it looked like what I needed (Devonthink’s tools in the finder and not limited to a database), but initial impressions in the forum gave it a thumbs down due to heavy cpu/memory usage. Do you have any experience?

  4. Never tried out Devonsphere express and I’m not sure that I would want tools interfering (or “enhancing”) the finder. One of the nice things of DEVONthink is that it’s an additional program that handles itself, without interfering with other tools/apps. And if something hogs the memory or cpu … meh. I prefer DEVONthink.

  5. Let me ask for clarification. Are you saying that you have DTOP databases that are just audio? I was going to write that I didn’t think it was possible. But if you can import audio, why not.

    So all these databases are in DTOP?

  6. I’ve got three DEVONthink databases that store my audio files — Music, Podcasts, and Soundfiles. They are for storage only (in case something happens to my iTunes collection). Given that DEVONthink Professional Office allows you to create as many databases as you wish I find this a nice solution compared to storing them directly in the Finder.

  7. Thanks, Daniel. BTW, I really. really. really appreciate your posts on your uses of DT. I started with DT years ago–one of the first when it was still proprietary–went off to other solutions, but I’ve come back to it because it is so damn reliable.

    This post is causing me to completely rethink my digital life. I thank you for it.

  8. Thank you, always good to hear when something is useful. DEVONthink is one of the most powerful programs I know, but also one of those programs that don’t make this power obvious (and sometimes seems to hide it). It’s a bit like Microsoft Word in this regard, with the difference that DEVONthink actually does work for me, and not against me. Although I have found a way to use DEVONthink to my advantage, I’m still learning new stuff about it (“Devonian Times” newsletter is helpful here).

  9. Hi Daniel,
    Let me thank you again for keeping your blog updated.
    I have created one source database (work + websites partly because of the “see also” feature) for all of my sources related to my field of research. Therefore, all of the sources I used/cited that related to my field are compiled within one single database. I assume you do the same; in another words and in your case, you place all psychology related books and articles in one database. Accordingly, the database is getting quite large but I guess that’s where DevonThink app becomes even more helpful.
    1) Have you ever used DevonThink “table of contents” feature? This is what DTP people have to say about it:
    “A table of contents helps you stay ahead of the ever-growing pile. Select the documents for which you want to create a table of contents, then choose Data > Create Table of Contents. The application builds a nicely arranged rich text for you with links to all the other documents.”
    – For the following two questions, I assume, the lack of using A – Z folders and tagging feature for the DTP forum people is that they do not utilize CPN app:
    2) Do you create and store your files in A – Z folders in every single DTP database including databases not related to your field of study? DTP forum member users do not seem to use A – Z folder structure.
    3) Tagging: in addition, DTP forum member users are concerned with using tags. Here are several concerns:
    “don’t recommend copying or moving tags between databases because of some unwanted side effects related to inadvertently copied documents” or more importantly, “because tags are highly relevant to a specific project, in my experience they wouldn’t be useful to the next project I undertake and may even be impediments to that work”
    Again, to my knowledge, one of the reasons they have a different approach/concerns is that they haven’t been using CPN the way you/me do.
    Kind regards

  10. Hoi Mio,

    these are good questions, I’ll try to deal with them in a posting.

    Best regards


  11. How do you deal with the fact that Devonthink sync is basically non-existent? You say that you use the mobile syncing but I am deathly afraid of losing my materials.

  12. Hoi Jason,

    hmmm, as far as I know, I never lost files. But then I might have missed losing them. Nah, don’t think so. I religiously only replicate files into the Mobile Sync folder and delete files only on my Mac. Never had — AFAIK — any problems with sync. I think if you do sync, replication is the key.

    Have you had any bad experiences or heard about them regarding sync?

    Best regards


  13. Hoi Bruno, heard about Alfred, but never used it. But thank you for the tip. 🙂

3 Trackbacks / Pingbacks

  1. Some DEVONthink Considerations and Issues | ORGANIZING CREATIVITY
  2. Devonian Times » “I am a squirrel”

Comments are closed.