Devonthink Pro and Papers

Are anyone here using these to software in combination in order to classify research papers? Whay may be the advantages of using Devonthink in addition to Papers?

Hi, Auto-

I have both software, and have been trying to figure out how to use them most efficiently together. I woke up at 5 this morning with an excellent question in my head for this forum and its experts, and then found your post! Forgive me if I piggy-back and expand the details of the question…

Papers saves all my pdfs in folders by year. I would like to index these in DTPro Office, as I regularly add papers to my Papers library (and folders), and as Bill has pointed out, indexing is advantageous in that setting.

Specific questions:

  1. How do I create Groups or folders in DTPro Office to help build context upon which the AI can work?

  2. Should I get rid of the folders in Papers by year, and create folders that would match what I want to do in DTP?

  3. Are there other ways to set up a basic classification in DTP that tells it what I am thinking about these heterogeneous papers?

I have been struggling with these concepts for a bit, and have found several threads on this forum that deal with similar issues, but would love to have the experts’ opinions and input.


This might not be exactly the solution you’re after, but this is how I’ve solved the same problem. I have moved everything from Papers to Bookends. Bookends does everything Papers does, plus is a top notch reference manager — I only miss Papers’ Mac-like interface. As a result, I have Bookends move & rename all the pdfs to a specific ‘attachments’ folder (flat hierarchy, which reduces the clutter of date-based classification). I have DTPO index that folder, and the “Synchronize” script launch every time I select the indexed folder in DTPO.

The AI handles “See also” suggestions of my pdfs perfectly well, and since the attachments folder is only indexed and has no subfolders, I never use the automatic classification for this.

Thanks, Simon-

In fact, one can change the preferences in Papers to eliminate any subfolder in the main folder that contains all the pdfs. I found this since my last post. Therefore, I believe one can achieve the same thing you have with Papers.

I also have Bookends, but haven’t used it yet. I am curious if there is any function that Papers has which is not available in Bookends (sorry for the sidestep of the original question).

Please note that I have written longer comments on my workflow refinement efforts on this thread, and will probably try to focus there so as not to split the discussion among several threads:


I love Papers and am also trying to get DTP and Papers to work
together. My initial tests are working pretty well. I have a
directory structure like this:
|- Papers (automatically organized by Papers)
|- Journals (Files named by (Name)(Journal)(Year).pdf
|- Conferences
|- (Bunch of subdirectories containing posters/papers
from conferences since its now common to post all)
|- Books
|- Various electron books I have
|- Theses
|- These I have
|- Subjects
|- With more subdirectories
|- Research.dtbase2 (DTP data which has stuff I can get to)

The directories other than Papers are mostly things I’ve accumulated over the years and already had organized. What Papers does a terrible job at is helping you organize the notes, webpages, presentations, etc. that you want while you are thinking about the subject. Enter DTP. I use indexing (and thanks for the other post regarding the automated syncing – that is one of those things that should be much easier!!!) for the whole structure above.

So far, it is working pretty well just by using targeted Smart Searches. I have found that I have to be more judicious in my searching than in Papers where I can usually just search on title. Most of the time though, I can get everything I want without too much extraneous information. Once I get the group set up in the database, then I add the smart search, my notes, my weblinks, and my ToDo list). I’ll know more about how I like it in about six months, but this has been by far the most promising system I’ve found yet.

Regarding why I like Papers:
I think Papers pdf matching (using Google Scholar or Harvard ADS) is the best, and while it doesn’t have tags (my biggest gripe), the collections feature is pretty decent. When it comes time to write my research paper, I create a collection of all the refs I want to cite and then having Papers export the bibtex. My coauthors are now using Papers so I can also export my collection in an archive and give it to them to make sure we agree on references. In addition, I usually have Papers open the pdf’s in Skim so I can add notes. Overall, I think this is a fantastic tool.

I have owned Papers for about a year but was initially disappointed at what it lacked - e.g. very hard to indicate a book chapter or a book excerpt; one essentially must do a kludge.However now I have committed to it largely because it makes creating cites much, much easier.

By indexing rather than importing the article files that Papers manages, I am able to make use of them inside DT. Very nice.

Outside of Papers, I use Leap and other OpenMeta taggng tools to tag some but not all of other files - for example, not articles, but lengthy notes on articles. After tagging these notes I index them in DevonThink - it will be nice to make smart folders in DT that collect topics of interest as indicated by the tags.