I don't get the advantage of using DevonThink

(Hook is not a reference app in the true sense of the meaning; it hooks documents to one another). In German there is a term “eierlegende Wollmilchsau” - the egg-laying wool-milk pig, a marvellous beast capable of delivering a wide variety of goods. Software - like pigs - is not like that. If various pieces of software play nice with one another, you can seamlessly leverage the advantages of each. Of course, if you can do all you want with one single piece of software, you’re home and dry - that’s very much up to you and your work flow.

(an example - and I realise you are doing something quite different: I use hook to interconnect e-mails (in Mail) to documents (in DT); I can import E-Mails to DT, but usually what I do is only import the very important ones, or the attachments which are actually what the mail is about; when I import the attachments, I hook back to the original e-mail, just in case I ever wonder where that came from. I do the same when I am working on something - I often take notes in Drafts, and hook them to a document in DT. The notes generally don’t need to be preserved for prosperity, so I don’t keep them in DT (which I try to keep tidy.))

Another advantage of DT is the ease with which you can process PDFs and images. My main use, before I even begin to do any searching, is to convert images, in large quantities, to OCR’d PDFs and them merge these into multipage searchable documents (and splitting them when I have made mistakes!). Finder/Spotlight won’t do that at all. For me, that is enough to justify having DT and the rest is a bonus (big bonus).

After that, it gets interesting, in the sense of how much of the organisation/searching do I do in DT, and how much in Bookends (or whatever reference manager you are using).

I have also taken to archiving emails in DT, because it seems much better at searching than any mail client I have tried.

1 Like

Use DT3 to hold articles, spreadsheets, notes and most importantly find them and find connections you forgot about. Open your favorite composing software in split screen, another tab or desktop. For the non-scripty, cut and paste works great.
I could send lots of links to how academics use DT3, but you don’t have time to read them. I am a retired life scientist, unfortunately I didn’t find DT3 until recently. It would have hugely boosted my productivity (once you’ve scaled the learning curve). Now I use DT3 for managing materials related to my diverse interests.


I have been working with DT Pro a couple of years now, published a thesis, couple of books, went to conferences, led project teams and kept track of my personal documents. Over the years, I got rid of any other file structures and just safe everything in my databases, have it handy everywhere, fully and intelligently searchable and, as well important: one-click-backupable.
All that is enabled by DT Pros powerful OCR and Search engine. If you compare prices, that alone justifies the investment.
However, there is so so so much more. “See also and sort”/related documents, hotkeys for changing groups, smart groups, concordances, devonagent integration, full scripting support, email fetch, nested tabs, indexed latex projects, document linking… for years now, I stumble upon new useful features every month and haven’t once experienced, that I wanted to do something, DT wasn’t capable of doing. I love it. Did not at first sight and had the same questions you have for months… but at some point it clicked: There are virtually now limits to what you can do with it.


Thanks again too all for your answers!
If I understand well

  • I still need to have a read-it-later app or can evenI replace Pocket with DTP?
  • I still need to have a reference manager to insert citations in a publications (so basically I need to have all references twice, once in DTP, once in the reference manager)
  • I can use DTP as file storage instead of storing them in a folder structure on the Mac. Once imported into DTP I can delete the an the Mac file system. DTP then does all what you have mentioned in your previous posts.

Is that right?

I think that’s a nice summary. Whether or not you can replace Pocket with DT is a question of personal preference; it would be easy enough to clip/save things to DT and mark them with a Tag “Later”, then collecting them in a Smart Group (or even several, based on additional criteria). That said, I like to use Drafts for things that are either active or haven’t got past the “important enough to go in DT”-barrier. It’s very much a question of your workflow and preferences.

Yes, I think you would do well to keep a reference manager - which one and to what extent you can get them to play together (saving you duplication), I cannot answer - various posts in this forum suggest that some level of intertwining will be possible.

Yes, you can (and I would encourage) use DTP instead of the folder structure on the Mac. I would caution against having a parallel universe here - neither will ever be up to date, with some things in one and some in the other universe. I urge you, however, to review any backup strategy you may have, or to develop one if you haven’t. I posted about mine here and have, in the meantime, added a monthly backup to WORM media.

1 Like

In addition to @Blanc comments, DEVONthink has a “Reading List”. Right mouse click on the document, pick “Add to, Reading List”. See page 88 of the DEVONthink Manual. I prefer that instead of adding a Tag (although I can think of reasons I might be better off with tags … choices!)

I use it but only to mark those documents I put in DEVONthink. I have a longer list of web sites/pages I peruse/scan and I use the app Reeder for that, which also has a nice read later feature. I stopped using Pocket a few years ago. So, depending on how you read stuff, you might need more than DT.

At this point, give it all a go and see where you get to. I’ve been using DEVONthink for more than a decade and I continue to learn about and from it.


My strategy for “read later” stuff is, well, I call it “baroque”.

Most of my read later stuff come from Inoreader. Several times in a day, I go to Inoreader and “read” my RSS. As I’m reading fast, I “favourite” ones I want to read later and/or in detail (fast: j for mark as read and go next unread, f for favourite it).

Once at home, I go to my Inoreader favourites and start reading them. If I find something interesting of save it, stop reading, open real website, reading view and CMD-P CMD-D (this is a custom shortcut to send to DT Inbox).

I have my favourites Inoreader RSS exported and imported in DT, so sometimes I directly read in them in DT. If it is interesting, repeat print and save into DT.

My next step is read those in DTTG, delete not wanted, annotate others and go back into DT and “archive” them: I have a database called Scrapbook with a Smart Rule attached to a folder called ZZ_Cache. Then I move from Global Inbox into zz_cache (CTRL-CMD-M, type zz, press enter) and the PDFs are “archived” by year/month/day. I have equivalent for Papers, but those are organised by Tags-into-folders.


(Of course, I don’t do that at same time. Sometimes I accumulate some days in Global Inbox read/unread, or in Inoreader Favourites, and so).

1 Like

I’ve increased efficiency on that one: mine is CMD-P CMD-P :smiley:

(Oh, and even in DT you’ve got to have a scrapbook - I’ve got one too :slight_smile: )

1 Like

Thanks again, I am enjoying this “conversation” very much and find profit in it. I bought DTP3 yesterday (my trial period was over without me really trying the app) and I am looking forward to using it, together with DevonAgent and DevonSphere. I will have to find my way into all of them. I realize that my learning curve will be steep but I am decided to go for it. Some of you have been writing about a possibly nice interaction with Bookends, I would like to see how that works and eventually migrate for Papers2 to Bookends…

Once I’m here another question: Safari has that nice “Reader” mode. Is there something like that in DTP, too?

Best regards and thanks again,

Bon voyage! Report back any time you want - to seek inspiration, to tell of success or failure, steps & solutions. Doing so helps all of us here and - I believe - DEVONtech as a company. All the best :slight_smile:

1 Like

Wow. DEVONAgent and DEVONSphere. Way out side my interest, expertise, and probably need!

Before going too much further, spend a little time with the outstanding DEVONthink Manual and the “Take Control of DEVONthink 3” ebook both available on DEVON Technology’s web site. Much better resource than continuing asking a lot of questions here that are answered better in the documentation.

Re Safari “Reader”, simply open the document with Safari and use “Reader” mode.



I thought Agent and Sphere were part of a complete package and should be used together…

They can interact with each other, but are still independent applications.

1 Like

Hey, Michaell,

My initial use for Devonthink was just as a basket with search capabilities. The Finder would have worked as well. I’ve gone far beyond that, now.

Tags, for example, can be alternate folder trees. Let’s say I’m trying to untangle all the incestuous malarky you find in those shock political books. There’s no one single way to categorize things, so I might start with a Devonthink group per cabal. Each cabal would have a group for each member.

When a person belongs to two cabals, I could make a replicant of the person’s group in the other cabal. That’s like a symbolic link, or alias, in the filesystem, but with an advantage. If I discover one of the instances needs to be deleted, it doesn’t matter if I delete the “original” or the “copy”, because the original doesn’t disappear until all its references are deleted.

For writing, I like to turn off the “exclude groups from tagging” feature. That means if I have a group called “Men in black,” any file (or group) I tag with “Men in black” will appear as a replicant in the Men in black group.

Tags will also let you create ad-hoc groups, which is kind of nice.

I’ve got a database with a couple hundred notes about plot ideas, characters, and locations. I can tag notes with “Chapter 1”. If some of those ideas also pertain to Chapter 2, they can carry both tags and appear in both places. My hierarchy of notes isn’t disturbed or cluttered up. When I look at the Chapter 1 tag, it’s like looking at a folder.

If I want to re-do my Chapter 1 musings, I can delete the Chapter 1 tag, in one stroke de-tagging everything that was in Chapter 1 but leaving other tags intact and not deleting any documents.

My financial records are all scanned into a Devonthink database along with grumpy letters to vendors, personal notes, and stuff that doesn’t necessarily interest my accountant. I can select the groups that contain the receipts and reports the accountant needs and export them to web format. That gives him a browsable, categorized, copy of everything he might need. He won’t need Devonthink to see my records.

Devonthink is easy to write in. Save an empty word processing document as a template, then create documents from the template. I like setting double-click to open documents externally.

There’s a better way, too.

Nisus is a particularly good fit with Devonthink because you don’t need a template. Just create an RTF document. I’ll edit in DT’s editor if I just want to jot down something quick. Later, if I want to expand it into an essay, a double-click (or right-click and select open with) will open it in Nisus. I have a Nisus macro that loads the style set I usually use and replaces double newlines with single newlines. Bingo. From bare DT RTF to full-on styles in mere mouse clicks that only need be done once. Should I ever need it, Nisus supports bibliography software.

There are of course many ways to organize data. I’ve tried other utilities but keep coming back to DT.


DT has a “clutter free” web capture tool option that can generate RTF, MD, Formatted Note, Web Archive and pure HTML (if I don’t forget anything). It can be customised with CSS (only advanced users). But as this kind of tools, it work better or worse depending in what website are you going to “scrap”.

DT one is very good doing its job, but sometimes I have to “touch” the result.

However, as a matter of preference, I’m in love the way Safari presents Reading View and sometimes I have set to show some sites as Reading View and the two shortcuts described above do the best capture. When this does not work, I use DT way to Formatted Note and then edit once captured. And other Smart Rule converts my edited Formatted Note into the final PDF.

I’ve thought the very same thing over the years. I’m a bit of a minimalist, I like keeping my mac clean and clutter free, with as little software as possible. In my opinion, software has to earn it’s right to be on my Mac, which means that it must do things that I find useful enough to not only pay for, but to add to the complexity of the machine that serves as my external brain. DEVONthink is a beast of an app, it tries to do so much that is often better served by specialized applications (text editing comes to mind). But, there are a few things that keep me coming back to it, upgrading, and most importantly, earning a spot in my Dock.

  1. End-to-End Encryption - Apple really should build this into iCloud, but alas, if you drop a document in iCloud Drive it syncs to Apple’s servers exactly as it is. Not a big deal for most things, but if you deal with sensitive client data like SSL certificates and SSH keys like I do all the time, the risk of exposure is unacceptable. DEVONthink lets me sync documents, and not worry about them being stolen.

  2. Data Collection - The DEVONthink importer is not as good as Evernote’s, but it’s good enough for me. I collect technical documents for reference in building out our cloud architecture quite a bit, and have to keep up with new developments in the field. DEVONthink makes that very easy, and once you have a significant collection, the next point makes this experience even better:

  3. AI-Assisted Filing - Most of the time, DEVONthink’s classify function is better at putting files in the right place than my hand-crafted rules in Hazel are, especially when dealing with technical documentation.

I also really like DEVONthink’s markdown support. It lets you set a CSS file for viewing, enabling syntax highlighting with Prism. When I’m reading a markdown file in DEVONthink and want to edit it, I hit ⌘⌥O to edit the file in BBEdit, then save it back to the database for viewing. Gives me the best of both worlds, the power of BBEdit and the power of DEVONthink.

Everyone’s needs are different, and honestly if iCloud Drive offered end-to-end encryption, I’d have to give DEVONthink a very serious look at if it still deserved a place on my Mac. For now, and for the foreseeable future though, it’s a vital part of my workflow. My job really comes down to knowing things, and DEVONthink helps make sure I don’t forget anything I need to know.



If your main concern is linking your data to your references, you should know that Devonthink works very well with Bookends. There is even a course for dummies like me who can’t figure it out for themselves. See

1 Like

I particularly like Smart Groups in DEVONthink. Finder has saved searches, but I’ve never really found a good way to edit them, or even find them after I’ve created them sometimes. Smart Groups just work really well.

One of my favorite features is replicants. I use them to collect reference material from a common library for different consulting projects. It’s also great to be able to have the same reference document stored in different groups based on perspective. I also find that tagging in DEVONthink is much easier than MacOS.