How are you using DevonThink nowadays?

I’ve been feeling the limits of Obsidian lately. I’m realizing I’m a folders guy. I like to organize things in folders. Lots and lots of folders. Sometimes folders containing just two or three documents. Obsidian has limited folders support.

Also, I use a variety of document formats. I like Markdown fine, but I also am a heavy user of Microsoft Office documents, PDFs, and web pages. Obsidian is very Markdown-focused.

Earlier this week, Andrew Canion mentioned DevonThink. And that got me thinking … I used to be a heavy Devonthink user. I already have the license. Would DT solve my problems? Give me my folders, and also have internal links–which I do find useful–and more?

I know we have a few DevonThink users here. How are you using DT nowadays? How has your use changed? Are you using DT as complementary to Obsidian?

Tagging @ryanjamurphy, who has written and talked about DT/Obsidian in the past.


If you have any feelings for Obsidian and DEVONthink, my opinion is you would need both. One for the notes that you write yourself and one for all the other documents you collect. Even notes that you’ve composed yourself, once they have passed out of immediate use, can be archived in DT.


I wonder whether Obsidian gets me anything that I can’t get from DevonThink. I haven’t found a use for Canvas, Maps, or DataView, for example. I don’t even make much use of backlinks, with the significant exception of using backlinks as a breadcrumbs trail.

One important area where Obsidian shines is that it’s the best markdown editor around, IMO.


If you run short of folders I’ve got a bunch you can have :grinning:.

With the exception of Music most of my files are no more than 2 folders deep. If they ever outlaw Search I’m screwed.


I don’t see why you would ever quit using what you consider to be the “best Markdown editor around.” And there must have been some reason you left DEVONthink in the past, so why do you think it will meet all your needs now? Both types of apps are needed.

(NotePlan for me fills the role of “best Markdown editor around.” And EagleFiler archives and retrieves any document that I throw at it.)


I realize I am also a folders first person. Right now I’ve got something like 100+ folders in my Obsidian system. One really cool thing from my POV is that I can have each folder or a set of folders in a tree as a separate GIT repo and Obsidian doesn’t care. But thenotes in there are all being version controlled. Not automatic, I hav eot do my own commit and opush actiosn but that’s the same for most of the development stuff to0.

My question is what do you see as missing from Obsidian in terms of folder support? I mean it’s just a container, easy to move stuff into and out of. So maybe I’m missing some neat use of folders?

That’s a very good question, and I don’t know if I have a great answer. It just seems like folders are a red-headed stepchild of Obsidian. With DenvonThink, folders are the primary means of organizing data. The folders view is the first one you see. I can easily see the full document name, labels, tags, and other metadata. With Obsidian, folders are off to one side.

Also, it seems like it takes more action to create and manage folders in Obsidian. And it’s too easy to forget to move documents to the appropriate folder.


It’s possible I will end up using both. I know many people index their Obsidian folder in DevonThink and continue switching between the two.

As to why I switched: I got pulled into the zettelkasten cult. It’s a pandemic thing. Some folks made sourdough bread or bought Peletons, others of us thought that if we just broke down each thought into an atomic unit and linked all the atoms, we’d turn into Dr. Manhattan and be naked and blue and on the moon.

Zettelkasten never worked for me. It’s not how my mind works. But I kept banging my head against that wall, thinking if I just zettelkastened harder, it would all work right for me this time.


That could be it. I never use tags or labels on files and the only metadata I care about is when last modified and then only very rarely. I can get it in Obsidian by doing a show in filesystem on the note if I have to. Usually I embed the date I created or modified the note/file in the name.

This is another thing that’s different. I make extensive use of an inbox folder for most created documents but since my folder tree is all on the sidebar in Obsidian if I know where a file goes I right click to create the note there, same as I did in DT in the past.

1 Like

I just exported everything out of DT and am just using the filesystem now. Since turning on Advanced Data Protection everything is encrypted end to end in iCloud now, so I do t feel like I need it anymore.

Maybe I’ll miss it, we will see.


The simplicty of “just using the filesystem now” (and maybe HoudahSpot for search) holds a definite lure for me. But I like having two views of my files: one via the Finder and another, different, view through the anything-bucket tool I have chosen.


I don’t have a huge zettlekasten or anything, but I keep a lot of documents in DevonThink, with 3 “main” databases. I have my Obsidian vault indexed in my Personal DevonThink Database, as a folder.

DevonThink has backlinks now, and does an awful lot of what Obsidian does out of the box, (with the exceptions of the Graph View and Canvas), so if you just want linked notes DT does the job, and handles non-markdown better. In reality I only have Obsidian installed because of the hype, but it is sometimes fun to use to write Markdown.

I’m kind of hoping NvUltra is great.

I wonder if you could not think of it as making a change? I mean, you could just index your Obsidian folders and open DT whenever you feel like you’re missing something. If you never open it, great; if you end up living in it, also great.

I’m pretty happy with DT so won’t belabor that. Most complaints come from self-inflicted problems like:

  • hitting resource limits with massive corpuses
  • wanting it to handle structured/relational data in a way it (probably) never will
  • increased expectations from ML advances
  • awkwardness around making databases sync to multiple humans’ devices
  • wishing Readwise Reader weren’t so sexy

Zettelkasten never worked for me. It’s not how my mind works. But I kept banging my head against that wall, thinking if I just zettelkastened harder, it would all work right for me this time.

I think this is a far more common scenario than many may think – and not only with Zettelkasten.

we’d turn into Dr. Manhattan and be naked and blue and on the moon.

Dr. Manhattan rocks! :smiley:


Weird because Obisidian has folders. I too like folders; the same way I do on macOS. Organising material in related collections. Only been using Obsidian for a while (since November 2022) and therefore not pushed the folder scheme to the max but I have folders with sub-folders with sub-sub-folders with sub-sub-sub-folders. (I suspect that the ultimate limit is whatever macOS imposes on the depth of nested folders.)

The one thing I dislike about Obsidian, and this goes for its folders too, is the strict alphabetisation of notes. Sometimes notes and folders need to be listed in creation or some arbitrary order — you should see the bookshelves in my study because the shelving sequences make sense only to me — however this seems to be coming as an “insiders” release has some new feature that permits different ordering.

One virtue of Obsidian over DEVONthink for me is cost. Osidian free, DEVONthink $99 (or £81).

The Obsidian developers promote their Markdown-ness as an advantage by freeing users from the tyranny of propriety formats.

One program that I use for organising and storing final forms of documents (PDFs etc) is Scrivener. Plus it allows and promotes folders and their arbitrary ordering even to the point of providing simultaneous variations of ordering via its Collections feature. I do my full-length writing in Scrivener and have done for decades; I wasn’t an early adopter per se but was certainly an early user. Its storage feature is adequate for small/medium size projects. While Scrivener is not free for me it is an historic cost and now therefore irrelevant but when time comes to get the next major version I will happily pay the £50 or whatever the licence fee is then.

These days if I wanted large or mega projects then Apache’s Lucene/Solr with Tika (for dealing with different file format) would be my goto solution but then my background is in text retrieval/document management/document databases. I would throw in their Nutch for web crawling. All of them provide extensive APIs for building bespoke applications though they use Java, which is not one of my favourite programming languages. I have seen and use a corpus lingusitcs program (Lancsbox) that utilitises those APIs to create a targeted application capable of storing 100 million word corpora.

I would not want to write using Lancsbox not that it has that capability but equally would not want to store 100 million word copora in Scrivener. Obsidian might struggle. Not having spent nearly $100 for the basic version I’ve no idea how DEVONthink compares. Been attracted to DEVONthink several time but when I need to do something it claims to do I realise I already have tools installed to achieve the same objective.

Long story short I use the most appropriate tool for each project. But the write up is always in Scrivener with its choice of output formats from Word documents through Markdown to LaTeX and PDF.


This is excellent. I may not get a chance to reply in more depth, but I want to at least say that much.

I’m starting to think of ideas for using DT and Obsidian together.

Loving this discussion

1 Like

There are some Obsidian plugins which substantially improve/extend how Obsidian deals with folders.

In particular:

  • Auto Note Mover
  • Custom File Explorer Sorting
  • File Tree Alternative

I use it for everything work related, and always want all my files, meeting notes and planning in one place. It’s one of the few apps I use constantly and have come to love. I really appreciate having full confidence when searching and its ability to find information in slides, PDFs and spreadsheets. I’ve tried Eaglefiler, Obsidian, Keep It and Finder and none had a search as good and effective.

I had some sync issues in the past but since switching from iCloud to Dropbox they’ve gone.


DT has had an unofficial graph view script since late 2020, and it’s been part of the “official extras” since Christmas (for DT on Mac, not iOS). Here’s how to turn it on, here’s the original coder discussing it in the forum. (I installed it when it was “unofficial” so I can’t speak to what it’s like now as an official script, although I assume the answer is “pretty much the same”.)

If you’re bored, there’s a huge discussion about DT and Obisdian in the DT forum, and I’m sure you can probably find it’s counterpart in the Obsidian forum (I don’t loiter over there so much so haven’t noticed).

I tried Obsidian alongside DT. I love a good folder structure, I do tend to navigate by folder unless I can’t find what I need, I’m in a rush, or when I’m doing a broad search to see what files I have that might be useful to whatever I’m thinking about (I don’t see how the native file system can replicate the search powers of DT, but that’s a separate topic). I didn’t want my notes separate from the files they relate to, so Obsidian just caused me unnecessary friction. I did like it as a markdown editor, but I don’t feel that is a valid reason (for me) to put barriers in my way of working. Plus honestly, the DT Markdown editor is good enough: I’m here to write notes I need, not make things look nice (I’ve settled on a CSS I like for DT and I’m happy enough with it). You can set DT to open Markdown files in a different editor (I don’t).

Most my files are PDFs rather than Markdown, so I needed an app that could file and search these. I have a few other file types too (Pages, Excel, RTF, etc.) and I didn’t want a system that separated them. (I actually have that problem in MacOS right now - I keep a small number of folders outside DT that sync on iCloud Drive, and a couple of GIS files are too big and jamming my iCloud sync. So now I’m stuck trying to figure where I should file them instead where it won’t matter that they’re separate from all their related files. Who needs that friction :roll_eyes:)

For interest since you’re familiar with DT, I only have 5 databases, with one that is my main one that I use all the time. Nearly everything goes in that main one. I dithered for ages over the merits of different databases for different topics, but I’d rather just have all my notes about disparate topics in the same system, organised by folder. I have separate smaller databases that store specific things that I don’t want in my PKM: a database for archived emails (search is so much better!), a database for timelogs, archived task notes, etc, (I don’t want that in my PKM), one for a specific hobby that I didn’t want turning up results in my PKM, and a new one I’ve been experimenting with for some work files because of the iCloud issue I mentioned above (it’s going to be rejected, most my work is in my employer’s ecosystem, and iCloud suffices - most the time - for the few files I do need to store locally, plus I resent having to try a database for this when what I really need is the power to opt a folder or file out of iCloud sync).