Readwise Reader and DT

Background statement - I love DT - anything I state in the next few sentences comes from a place of hope.

I think many of us know that Readwise has a beta Reader application that is well designed for reading PDFs and then exporting the results to Obsidian/Notion/…

Since I started using it a couple of months ago I’ve made greater headway through PDFs and articles than ever before. This leads me to a challenge.

I use DT as my system of record - a complete copy of PDFs, books, etc that are relevant to my work. I’ve tried to use it as reading platform and highlighter. It’s never stuck. The readwise people are succeeding here since this is all they do. (My gut feeling).

A few minutes ago I found a long article in the HBR that is worth saving. I sent it directly to Reader bypassing DT since it will be easier to read and annotate.

I don’t want to do this. Ideally I would have added to DT → Placed on my reading list → Synced my reading list to Reader. Is there a way to do this already? Am I making a feature request? (Perhaps Ryan has written another AppleScript to pull this off).

DEVONthink doesn’t support Readwise, is this app scriptable?

It looks highly unlikely.

Readwise is a great piece of software that iterates quickly. What strikes me most about this feature is that it adapts to many of today’s in-APP content annotations, such as articles, and unifies these annotations into Readwise. After all, from the efficiency and experience of reading, must be directly from these apps to read and comment on the most convenient.

For the endless stream of these excellent online information processing APP, combined with my current use of several software, there are the following recommendations:

1 For the mobile phone in a variety of web page notes
I use Devonthink to capture these articles, but to read them, I still use some mobile apps and annotate them, and aggregate those annotations into an online application like Readwise.

2 Reading books (PDF format)
For a large number of books, Marginnote is Maybe by far the best ( The advantage is

  1. whether it’s horizontal, vertical, scanned with mixed images, low-definition books, etc. , you can use the OCR function to extract the important parts, parse out the text, and make cards in whiteboard.
  2. after reading a book, you can create a mind map (each map node is made up of cards) by focusing on different chapters.
  3. you can easily see the frame of mind for reading a book in a whiteboard view, or you can combine and reorganize the reading notes (cards) , etc. .

Instead of reading a single PDF Book Document and annotating it on different pages in certain places, you’re likely to see one point as a reader, which is inconvenient for your association of ideas. But in the whiteboard view of Marginnote, what you see is a picture of a knowledge (annotation) framework that is useful for both memory and association of ideas.

That is to say, when you use Adobe Acrobat Reader to read a book and annotate it, the fragmentation and separation of annotations (knowledge notes) is too strong. Imagine opening a page with acrobat reader and seeing a comment that is a separate point of knowledge. This is not Ideas of the same era as the visual layout of whiteboards and cards, which is formed by Mind Mapping after reading a book.

Unfortunately, Marginnote has a lot of bugs that make me worry about its stability. although the Heptabase( is growing fast, it’s still a long way from the Marginnote experience.

3 All kinds of video courses, audio interviews
In the era of rapid development, multimedia materials are becoming more and more popular, which is also the reason for the rapid development of TikTok. A large number of people prefer to watch videos to gain knowledge, which has led many media or educational institutions to produce a large number of video lessons.

But if you need to watch the whole video or listen to the audio to know the information contained in it, is undoubtedly outdated practice.

I prefer to put these videos or audio on something like Lark minutes(View Lark minutes ) . These videos or audio after uploading, can be automatically transcribed into text, and support the translation of multiple languages. Lark Minutes also offers a lot of video storage space. When the number of videos or audio clips exceeds 10,000 or more, I can search for keywords to locate a particular conversation timestamp in a particular video to be relatively accurate in locating the data.

4 A structured database of small projects (requiring online collaboration)
For some small project repositories, especially those with structured information that requires online collaboration, I use Notion.

5 Common notes to organize, summarize (Second Brain)
As a daily contact to learn knowledge and other notes, insights and so on, are organized into the Obsidian. For example, a discussion of an issue can include many personal insights, links, references, and updates. The advantage of Obsidian is not only that it’s fast and easy to edit on computer and phone (which is easier than Devonthink in my opinion) , but also that Obsidian documents can be indexed and searched on Devonthink.

6 Visual analysis and Carding of the whiteboard for the problem event
For the visual analysis of a complex problem or event, using online whiteboard software is very intuitive and in line with the human brain habits of thinking. At the moment, I use Miro ( to analyze things and draw pictures, which is really convenient. Similar software is whimsical, Heptabase, and so on. I also used the Heptabase analysis to comb through the issues and accumulate some knowledge.

To sum up, my own experience with software is that different software has its own distinct advantages in different scenarios, and the software that solves problems to the top in some of these areas also gets amazing valuations. Even Notion, Miro, and Airtable, which are highly valued in the industry, still don’t solve all the problems of scraping, reading, annotating, taking notes, searching, collaborating, and so on. For example, the Notion’s search function is a far cry from Devonthink’s.

In view of the fact that more of these materials will be stored in the cloud in the future, such as Lark minutes, Notion, Miro, Heptabase, readwise… Devonthink’s ability to search for local files is very powerful.It is recommended that cloud file searches for other apps be enhanced in devonagent and more in conjunction with Devonthink to maximize file information searches, that DT will be an indispensable step in the user information workflow.

I’ve purchased Devonagent Pro, and I’m hoping that DT will continue to enhance devonagent’s ability to search for data in various online databases, cloud drives, files in apps, and information within files (close to Devonthink’s ability to search for local files) , and automatically integrates search results.

1 Like

Is there an elegant way to get DT to email a PDF? I’m happy to hook it up to AppleMail on MBP.


@terrence this is all very cool. I’m not really trying to increase the surface area of the apps I use. Reader from Readwise is a good enough tool for PDFs and they seem to be improving.

Like right clicking on a record and selecting “e-mail” from the context menu?


Sorry I’m trying to automate. So add something to the Reading list would automatically be emailed to Reader.

Emailing to another application?

Emailing to an email address on their server which then adds the PDF to my Reading list there. Example: I add a PDF (or epub) to my reading list in DT. A DT rule is fired and the item is emailed to - from there readwise adds to my reading list.

The Reading List isn’t directly scriptable and processing its items involves more processing than may be expected.

Why are you using a reading list to email to another reading list?

1 Like

I’m not picky how to achieve this. I want to automate sending things to readwise. I don’t need to use the DT Reading List function.

Maybe I’m trying to be too clever. - Mark

As a fellow user who’s also used Reader in conjunction to DEVONthink (but who eventually preferred using DEVONthink by itself), I may be able to help.

It is possible to create a smart rule that can be activated manually to email the PDF and add it to the reading list simultaneously. It is also possible to have this rule run automatically (instead of manually) based on certain triggers (e.g. a tag).

However, as far as I understand, you cannot trigger a smart rule based on an item being added to the Reading List.

As such, the best degree of automation I could think of was a smart rule like the one that follows:

To sum up the rule:

  • If you add the ‘reader’ tag to the PDF;
  • When DEVONthink next starts, the PDF will be a) sent to Reader and b) added to the Reading List.

– Matthew

However, as far as I understand, you cannot trigger a smart rule based on an item being added to the Reading List.

That is correct.

1 Like

Clever. I don’t mind the trigger issue. In the long run I look forward to getting the Readwise people to provide an onramp that isn’t email :wink: - Mark

I think maybe you are trying to be too clever :joy: I also use Reader, but I tweaked my workflow to accommodate this and control it manually.

Most webpages etc. just go straight to Reader (using their Safari integration). If I know in advance I want a copy of the page for DT, I will write it to PDF (how I save webpages) at the same time and file it myself.

Non-webpage PDFs go to DT first. I move them to Reader if and when I am ready to read them and if I don’t think I need an annotated copy of the PDF (see note below)*.

I have the “mark as read” setting disabled in DT so files are marked as unread until I manually change them, so I don’t lose track of anything new.

If something I read in Reader is worthy of going into DT and hasn’t already been added, I will do that after I’ve read it (usually this is webpages that turned out to be valuable - I go to the original webpage and print to PDF then).

*My only issue with my own Reader workflow at present is not being able to export highlighted PDFs - to me this limits its value for reading PDFs in Reader because I often need an annotated copy - I continue to use DT for this. I’ve spoken to the folks at Readwise and they will be developing this once the product is out of beta, at which point I anticipate that my workflow will be amended slightly to read all PDFs in Reader going forwards, with an annotated copy then being exported back to DT. I think I will still continue to send PDFs to DT first though and just move them to Reader when I am ready. This may not work for you, but for me I see far more PDFs than I can reasonably process promptly (I’m a scientist) and I continue to manage what’s a “high priority” versus “low priority” in DT (plus to be honest my priorities often change several times once a PDF has entered my database, and a document may get bumped up the reading list, abandoned temporarily and then bumped up the list again - I use tags and the mark as unread function to oversee this!).

I’ve had the same experience as you generally with Reader - I’ve finally started tackling my big backlog of digital reading, and it’s so much simpler having one main reading space (having newsletters and webpages in the same app is a game-changer!)

1 Like