Overwhelmed New User: Building a Scalable DEVONthink System for Writing + Archives

DEVONthink
1

Hello,

I’m a new DEVONthink user and I’m hoping for some guidance before I build a system I’ll later need to dismantle.

Last November, during Black Friday sales, I invested in several productivity and writing tools. DEVONthink was one of them, based on strong recommendations across forums and discussions describing it as a powerful knowledge management database. Now that I’m digging into the forum and handbook, I’m realizing it’s much more than a database. I see scripting, automation, wiki-style systems, AI integrations, complex linking workflows. It looks incredibly powerful, but I’m honestly overwhelmed. Part of me feels like ignoring DEVONthink because the learning curve seems steep. At the same time, I can tell it’s capable of becoming the backbone of everything I want to build.

Im trying to accomplish 4 main things:

  1. Novel Development
  2. Organizing 15–20 Years of Files
  3. Organizing my email (Probably 15 years +)
  4. AI and Automation

Below are my questions. I know it’s a long message so I am grateful if you make it to the end.

  1. Novel Development (Primary Goal)

For my novel workflow, I currently use:

  • Scrivener for drafting

  • Aeon Timeline for chronology

  • Zengobi’s Curio for visual boards and mind maps

  • DEVONthink (intended as research and long-term archive)

Is there a way to getting all of them to work together?

Within the novel workflow, I want to:

  • Store research (PDFs, articles, images, notes)

  • Organize character/world-building material

  • Create a personal wiki-style knowledge base

  • Manage images and video references

  • Potentially integrate AI tools into my research

  • Build durable cross-links between scenes, research, timelines, and visual references

I’ve also seen Hookmark mentioned frequently. How useful is it in practice? Do these tools realistically integrate into a coherent ecosystem, or am I overengineering this?

If you were designing a writing system from scratch using these tools, how would you structure DEVONthink so it complements Scrivener and Aeon rather than duplicating them?

  1. Organizing 15–20 Years of Files

I have files scattered across multiple external drives going back 15–20 years: Word documents, PDFs, music, images, random archives. It’s highly unstructured.

For a large historical archive like this:

  • Should I index or import?

  • If I index and edit a file outside DEVONthink, does it stay in sync?

  • If I import and then edit externally, what happens?

  • When does replicating make sense?

  • Do most of you treat DEVONthink as a working environment or as a controlled archive?

I want a system that is stable and searchable, not fragile.

  1. Email Organization

I also want to bring email into DEVONthink. Any best practices for avoiding clutter while keeping it searchable and usable long term?

  1. AI and Automation

I’m not an IT person at all. I’m increasingly using AI tools in my research and am thinking that automation will be vital when I organize those 15-20 years of files and duplicates. How do you integrate these DEVONthink in a structured way?

Similarly:

  • What’s the best way to manage large JPEG libraries?

  • How do you handle video files?

  • Do you centralize media inside DEVONthink or keep it external and indexed?

More broadly, I’m trying to create a networked knowledge system where writing, research, media, and archives connect intelligently. But I may have bought too many tools before designing the architecture. In addition to the writing apps above, I also own utilities like Hazel, BetterTouchTool, Default Folder X, ForkLift, PowerPhotos, Bike, Beyond Compare, and others (some bundled). I’m sure I overcomplicated things before building a foundation.

If you were starting fresh and wanted:

  • A serious writing workflow

  • A long-term searchable archive

  • Media organization

  • Email integration

  • AI-assisted research

  • A stable, scalable knowledge system

How would you structure DEVONthink from day one? Any guidance on how to make this as functional as possible is appreciated.

Thank you in advance,

A very overwhelmed user.

1 Like
2

Welcome @FilePile

It’s not completely clear as you mention writing novels, but please note that this is a user forum and AI-generated posts are not welcome. See Welcome to our user community That being said, this is a very long and complex list of questions and unlikely to be addressed in full by a single reply.

Also, have you checked out the Getting Started chapter of the built-in Help and manual?

2 Likes
3

So many people who’ve never used DEVONthink assume it’s just a storage system. They’re about one percent right!

3 Likes
4

Hello.

I am going through both the Manual and the Take Control of Devon Think PDF but it will take me some time to get through them. As I mentioned before, I’m overwhelmed with what to do with the program. I thought it was a database and now I see its much more. My goal is to fully understand it as it does seem like it will be very useful long term.

I agree with you that I posted too many questions in one place. I can delete this post and re-post per topic (one at a time) so that my questions hopefully make sense.

Thank you

1 Like
5

You’re welcome and no worries! You can leave it as-is or split it up.

I also recommend you read this blog post which many people said they found very helpful…

3 Likes
6

A few thoughts off the top of my head…

  • Should I index or import?

The answer is: it depends, but read the In & Out > Importing & Indexing section before committing to indexing.

  • If I index and edit a file outside DEVONthink, does it stay in sync?

No. There’s nothing to sync as you’re editing the original file in the Finder. See my previous reply.

  • If I import and then edit externally, what happens?

The same thing that happens when you edit a file opened from the Finder. The Finder isn’t doing anything special. It’s just making the file available for you to edit. Imported files in DEVONthink you open externally do the same thing.

  • When does replicating make sense?

When it actually makes sense. :smiley: There is no practical answer to this as it depends on the purpose for which you’d use replicants… which is specific to you (even if others may do the same in their environment).

  • Do most of you treat DEVONthink as a working environment or as a controlled archive?

I don’t think there’s a most at play here, not even for yourself. You may have databases where you store things and others where you are composing, converting, etc.

I want a system that is stable and searchable, not fragile.

There is nothing inherently unstable about a DEVONthink database. Indexing done carelessly can certainly introduce problems, but done with understanding there isn’t an inherent instability. In either case, a database can serve you well for 5,10,15… or more years.

Remember: DEVONthink isn’t a solution looking for a problem. It’s a solution to a problem (often several). Start with a specific problem and explore the options. You’re not building a house where you have to commit to if there’s a rotunda in the center of it and you hope it’s a good idea because you’ll be stuck with it. Accept it’s possible you’re not going to build your final database on the first go around. (I’m on version 5 of a support database I’ve been iterating for almost 14 years.) Create and delete databases as you need to. Keep the good ones, examine and throw out the bad ones (cherry-picking from the contents, as needed first), and try a different approach. This is how to learn how YOU use the app.

4 Likes
7

Who says you have to use everything DT can do (right away)? No one :slight_smile: If you have a hair dryer with three buttons, you probably want to know immediately what those three buttons do. But if you use an app like DT, you don’t have to know or even master all 1,000 features. The urge to try anyway probably has something to do with the price. When you buy a “database” app that costs five times as much as the app you used before, you feel compelled to use it “to its full extent”. This leads into a psychological disaster.

I’m very happy with DT, but I would say I use less than 50% of all its features. Actually, I only use DT for one reason. For me, it’s worth it. Sometimes I discover something new by chance or thanks to this forum. I don’t need most of it and forget about it again.

So, relax. You don’t have to understand everything right away. You don’t have to understand a lot of things at all because you don’t need them. And if you do, you’ll discover it over time :slight_smile:

7 Likes
8

My reply is very similar to FrankT. I would probably suggest starting very small: one project and only one small portion within that project. Choose something that has a very clear practical goal that moves “things” forward. So I would not include “organizing my email”: that seems to be too broad. Perhaps one step withing Novel Development would work … but I’m just guessing.

A further consideration is whether you are using DEVONthink Pro and DEVONthink To Go (iOS app). If you are starting small is also advantageous. I started by trying to sync my database of receipts (from 20+ years) across the two. That took a long long time and realistically there was no reason to do so. So I backtracked an synced only a much smaller database that deals with current projects. That synced quickly and it became very easy to use/edit/add between DTTG and DEVONthink for the Mac. In turn I learned a lot more about DEVONthink because I wasn’t overwhelmed and I had a very practical use for it on my phone to add/edit/read items.

5 Likes
9

Devonthink is one of those delightful applications that remains true to its own standards. In general, DT makes things look like magic without any actual magic. Your files remain regular files, for example. Once familiar with how it works, you’ll always know what’s going on under the hood.

You don’t need to learn hundreds of features, just learn Devonthink’s philosophy. That will usually get through the day without needing to crack open the documentation.

Start by putting a few files in DT and navigate around. Learn about duplicating and replicating. Experiment with tags. Get a handle on DT’s search features, at which time you’ll know all you need to use smart groups which are really just stored searches.

Learn about annotation files with emphasis on using them in their own edit windows, not just the little peephole view in the inspector. See how you can open a text document (Markdown, RTF, or plain text) in an edit window that has its own inspector. That works for annotation files, too.

At some point you’ll probably start thinking less about what Devonthink does and more about what you can do with it. That’s when you start having real fun.

Spoiler: DT is very, very cool. You’ll probably keep learning new features for years, but that’s not a bad thing.

Good luck in your journeys!

6 Likes
10

That’s an awful lot of words to not actually say very much.

“Serious writing workflow” means very different things to different people. What sort of materials are you accumulating, and under what circumstances do you expect to retrieve them?

Are you the sort of writer who likes to “wing it” at the beginning, and then go back and research to fill in the gaps? Or do you prefer to do extensive research before you write a syllable? Are you contemplating a multi-volume epic fantasy, or standalone books in a contemporary setting?

Are you the kind of writer who covers their walls in maps, post-it notes, and visual references before sitting down with pen and paper? Or do you loathe clutter and prefer the orderly regimentation of electronic tools?

When you say you want tools to “work together,” what does that mean? All four open on your computer at once, with ad hoc links that you personally create as needed? Or an array of scripts that turns DEVONthink into the nerve center of the whole endeavor, with related files from all four tools only a single search away?

I’m not going to say that you’re overengineering. It’s your system, only you know what you need. I will say that if I composed a list like that I hope my writers group would roll their eyes and tell me to quit procrastinating and get back to work.

6 Likes
11

Personally, I would just get to work and let the “system” evolve as you work, rather than trying to build a system before you know what you need.

I think I’ve used DEVONthink for about fifteen years and I’ve barely used a quarter of what it can do. I use what I need. And over those years I’ve built and dismantled more “systems” than I can count. Mainly because I couldn’t know what I was going to need in the future, and I couldn’t predict what feature or piece of software was going to come along and change the game. Moreover, the work we do tends to change over time. It is more important to be flexible and adapt rather than trying to come up with the perfect “system” from the beginning.

9 Likes
12

Very helpful indeed.

Fair point. When I wrote the original post, I was very overwhelmed and tried to compress too many thoughts and questions into one message. I also mixed 4 big projects into one message. I will focus on the story project for now and start working on the other projects later.

A mix of both depending on what part am working on. If it’s the fiction part “I wing it” if it’s the history part I will do my research first.

Multi-volume epic fantasy set in a contemporary setting.

Definitively the writer that covers the walls in maps and post-it notes but would benefit from the order provided by electronic tools.

What I had in mind is much closer to the first scenario you described where the programs are all open on the computer at once and are linked to each other as needed. The scripted automated system with Devon Think sounds cool but I don’t have the technical knowledge to do so and I would rather focus on the writing part than on learning how to code.

When I say “serious writing workflow” I’m thinking of two things. The materials I have created and the tools that help me manage them.

I have been writing for over a year. My book initially started as a short story and has evolved into a multi book series. Up until recently I had been using Word to write everything (as that is the only program I knew about). I have about 1,000 files for the series which consist of:

  • Scene drafts (including multiple versions of the same scene – IDK what’s considered “normal” in the writing world, but some scenes might have 10-15 versions).
  • Character notes and backgrounds – some characters have changed names and descriptions over time
  • Timelines
  • Research
  • Sketches I’ve made (maps, character sketches, etc.)

Unfortunately, I did not do a good job with the file names. There are some files that contain scenes from different books in the series. I figured this was a bad idea after a few months of writing, but what’s done is done and I have to deal with it.

I have been working on more than one book at the same time as they are linked to each other and I want to make sure that those parts make sense (symbolism, easter eggs, hints, etc.). Once that is set in place I plan to go back and fill in what’s missing from each book.

At this point my biggest challenge is not the writing itself but consolidating many Word Documents, PDF’s, and sketches and making them easier to retrieve and organize. I need to be able to find scenes, compare versions, delete duplicates and rename files to that everything is very specific and organized.

That leads me to part 2 of the work flow I want to set up which is the software.

I got the following:

  • Scrivener – to write the manuscript
  • Aeon Timeline – to keep track of the chronology and events
  • Curio – I’m very visual so I need the idea mapping, ability to create character cards, keep visual track of what my world and characters look like, etc.
  • Beyond Compare – to help me compare versions  before I do that, I need to find the versions.
  • Hazel – not planning to use this one yet but I have it in case that I need to do heavy renaming

I’ve read here about Hookmark in terms of linking related materials between these programs, but I’m not sure if that’s something I should do yet.

In summary, what I mean by “serious writing workflow” is a system (software) that will help me locate scenes compare drafts, track timelines, and connect material so I can consolidate everything and create the manuscript in Scrivener.

Thank you for the questions. They made me stop and think through what I’m trying to do more clearly, and they helped me organize my thinking about how to approach this. IDK if I’m overengineering or not as this is my first attempt at this. My goal isn’t to avoid writing. That’s the fun part for me. What I want is to organize it so I can assemble the story from start to finish. I will stop procrastinate the getting everything together part and just jump at it.

Hopefully this time the words actually say something =)

Thanks again!

2 Likes
13

FWIW, IMO this is Scrivener’s wheelhouse. I’m a non-linear, chaotic pantser and my “process” (such as it is) would be utterly impossible without Scrivener.

(Full disclosure: I’m also on the Scrivener support team, but I was an enthusiastic user first.)

3 Likes
14

My brain..lol. Ocasionally it decides it needs to become superhuman and it sets up unrealistic expectations like learning DT in 4 weeks.

In all seriousness. Thank you so much for this message. It helped me take a breath and step back a bit. I think part of what was happening is exactly what you described. Plus, reading about more advanced setups on the forum made me feel like I had to understand everything at once in order to use DEVONthink properly. Your reply makes a lot of sense. I will focus on my main goal which is to organize all the files for my novel. Im sure that as I work through it I will discover be able to understand it better.

5 Likes
15

Thank you!!

1 Like
16

First of all, I’d say don’t overthink it. It’s pretty easy to move things around in DEVONthink later on, even between databases.

For writing and filing, I have the following databases:

  • Input — stuff I’ve collected from elsewhere. Web pages, PDF scientific papers, notes I’ve taken, that sort of thing.

  • Output — stuff I’m writing, or reference copies of stuff I’ve written.

  • Bureaucracy — taxes, medical info, investment reports. Things I don’t want to have to think about but do.

  • Work — job-related input and output, including reference material like RFCs and other standards.

  • Manuals — PDF copies of manuals for everything in the house (and garage). A separate database partly because it’s so big.

  • Lifestream — kind of an “everything else”, a stream of all the random junk I might need, like receipts, vaccination dates, clothing sizes, eyeglass prescriptions, notes of when particular events occurred. I don’t organize this as much, I have some categories like “House”, “Car” and so on, but I mostly rely on search.

I don’t put movies or images into DEVONthink. I keep my photos in Apple Photos, and everything else is organized in folders. I guess I could index those folders in DEVONthink, but I haven’t found a compelling need to do so.

I don’t bring email into DEVONthink, I just keep it on an IMAP server on my home server.

For automation I use Hazel, as I like my automation to be deterministic.

For longer writing, I use Scrivener. For short things I might use Drafts or Itsypad or DEVONthink, but once it’s done a reference copy moves into DEVONthink unless it’s on my public web site.

I’ve currently got a couple more small separate databases, one for Inform (interactive fiction) projects, and one for tabletop RPG planning. I may roll those into one of the main databases at some point.

So I’d say maybe try with one or two projects and see how it goes. Or try moving a hundred or so files in, enough to get a feel for whether your structure is working but not so many that it would take too long to re-file them elsewhere.

If you really want to consider a detailed system thought out in advance, you might be interested in Johnny Decimal. I don’t use it myself, but it’s an interesting idea.

4 Likes
17

My email archives are stored/organized using DEVONthink
imported as .eml files
It seems to be the best location

I’m still using a single database
Organized with tags: minimal folders

3 Likes
18

Another strength of DT. We all get to use it in ways that fit our needs.

I have a bunch of databases. One for finances, one each for some long term projects, and one database each for my writing projects, or at least one Devonthink database for each group of related documents.

I think of a DT database almost like a metadocument.

7 Likes
19

As BlueFrog said, your questions cover a broad range. Some of them I can advise you on. I’ve been using DEVONthink for well over 15 years and doubt if I use more than a fraction of its capabilities. Version 4 substantially increases them. To give you some context, I have no experience writing fiction, but I do a lot of professional writing. My degree is in the history of science and most of my paid career was spent managing digital and hard copy information, primarily in the US government at the NIH, the National Archives & DoD. Since retiring I’ve been involved in multidisciplinary research collaborations, all of them dealing with AI in some way.

I have 2 major parts of my DT database: one for my own work and the other for the disciplines that I build on. The include computer science, archaeology, history, linguistics, semiotics and archival theory and practice. In each of those groups I put publications (>3,500 so far)that are authoritative in the relevant field. Things that are particularly relevant to topics of themes in my own research I replicate to the appropriate groups in the 1st part. I also create a lot of smart groups (saved advanced searches) that pull together hits from whatever parts of the database I specify.

I also use replication in the other part. I do a lot of annotations of stuff I read. By replicatiing, all my comments are available now matter how I navigate to an item. DT also automatically creates a smart group for annotations. Beyond that, for any file you’ve annotated, DT can create a summary of the annotations and export it as a table. I do that and import the tables into Apple Numbers spreadsheets, one for each of the major topic of my research. I add some column using Numbers functions to collect things like special terms and themes. That allows me to sort or categorize my comments by any of my added columns. Great for writing. A great feature of the DT tables is that the last column has the link from the annotation to its location in your database. Clicking takes you to the exact spot in the document in DT.
As someone who managed information at the largest biomedical research institution in the world and led the development of the systems that the National Archives uses, still after 15 years, to manage the digital records of the federal government, I find DT to be an indispensible archival and research tool/

3 Likes
20

Very helpful. Thank you!