NB: Since this entry was written, file management in Papers 3 has improved. Follow further developments in Papers app on Macademic.
I have started trying the beta-version of Papers 3, an update of Papers 2, my favorite citation, bibliography and pdf management app. Papers 3 has an entirely new user interface which takes some time to get used to. Nevertheless, I have gradually started to appreciate the smoothness of its many features, such as search for new references, editing authors, assigning keywords and collections, and annotating by underlining text. However, I also have a major concern, so serious that it may force me to stay with Papers 2 or switch to another reference management software.
Let me explain. One of the great features of Papers 2 (and its main competitor Sente) is its logical and accessible handling of files. The user has high degree of control over how PDF documents are named and stored in Papers 2. This is how I manage reference files using Papers 2:
- When I save a pdf document (e.g. a scientific article) attached to an email or download such a document from the Internet, I tag it (Mavericks, the next version of OS X, has built-in tagging compatible with OpenMeta);
- The document is then automatically imported into Papers using a Hazel rule triggered by a tag (“papers” in my case). (Alternatively, Papers can automatically import all files added to a particular folder).
- If the imported document has a DOI (as most scientific articles), Papers automatically matches it with an online database such as Google Scholar or Web of Science and assigns all bibliographic information (author, year, title, journal, etc.). In other cases, I enter the bibliographic data (most importantly author, year, and title) manually.
- When the document receives its proper bibliographic information, Papers 2 names and stores it in accordance with the Library Preferences which I set as follows:
- Now my files are not only available in Papers for reading, annotation and citations, they are also accessible to other applications. For example, I can index them with DevonThink (this is great for relating pdf documents to notes, emails, images, web-pages, and other content which is tagged similarly or is otherwise related), find and tag them through Tags, PathFinder or Spotlight, share them through the Dropbox, etc.
If in the future I decide to use another reference management software (or discontinue using reference managers at all) I will still have access to all my files neatly named and arranged.
All this seems to be gone in Papers 3, at least in the current beta version. All documents are stored together with meta-information in a bundle not accessible to other software. You can open this package in Finder but it’s pretty difficult to locate the file you need. Moreover, Papers 3 does not seem to name files logically when moving them into its library bundle (for the Dropbox-synced files). For example, yesterday, the file originally named “WEF 2012 New Energy Architecture” was re-named when imported to FCFB7D16-677C-4F58-805F-B0014DF73155.pdf. Imagine confusion of my colleagues whom I sent this file by email!
Update on October 19, 2013
Responding to my query, Papers 3 staff confirmed that switching on the Dropbox synchronization (a new feature in Papers 3) removes user control over how the documents are named and stored. They added:We are looking into a way how we could re-introduce the file organization with Dropbox syncing but we can’t guarantee if/when it would be available.
[addition on December 20, 2013. This entry describes the implementation of file access in the first official release of Papers 3]
If the Dropbox synchronization is not switched on there is still a possibility to control naming and storage of files, much like in Papers 2. However, the files are still kept in a bundle not accessible to Spotlight. The support staff says that Spotlight index will be added in future versions.
Moreover, I have noticed that if you switch on Dropbox synchronization once you won’t be able to restore control over organizing your files even if you later switch the synch back off and move your library to a local folder. This behavior is unexpected and I hope it can be corrected in the future. Otherwise an accidental click on a button in the Preferences may cost people their whole libraries!
I hope Mekentosj can reconsider these changes or mitigate their negative impact. The ability to work with the same data using different software is really invaluable in academic work on a Mac (consider for example plain-text note-taking where different software can interplay because the content is kept in a simple and logical format). Files referenced in Papers should be accessible to other programs and thus be stored and named in a way that is generic, logical and understandable.