My Digital Workflow
When I was finishing my doctoral dissertation, I became fascinated with the whole topic of “academic workflow”. Some colleagues, like Stian Håklev, have come up with powerful, well-design systems (Stian’s Researchr is a great example of what’s possible). Post-academia, I needed to adopt a system that was efficient, stream-lined, and worked pretty well off-the-shelf. I love hacking code, but these days I need to save that hacking for products that I’m working on rather than spending that effort on side projects. I also need to maintain a system that will work with up to four very different projects that are likely to be active at the same time.
So here’s what my current digital workflow looks like. DEVONThink Pro Office (DTPO) forms the backbone of the system. I’m no stranger to DEVONThink, having experimented with early versions many years ago. I think my engagement with DTPO started in earnest when I was a post-doctoral research fellow at Copenhagen Business School. I was working on a number of very different projects and I needed to collect resources in a way that I could easily access saved items later. Not all of my resources came in the form of journal articles and the like, so a bibliographic management system like Endnote, Bibtex, or Refman didn’t cut it for me. (I should note that I use a variety of reference management programs, but they don’t serve as the main database.)
The decision to use DTPO over technologies such as Evernote had a lot to do with the location of my resources. I prefer to have my resources stored locally. Yes, I will use “the cloud” as a conduit but I am increasingly reluctant to use some server “out there” as my primary respository. Part of that is just paranoia: I worry that the sudden disappearance of my data, or the service, could have devastating effects on my work. I’m also sensitive to the business model that has someone else using my data for their personal gains.
So, DTPO is my main respository. What goes in there? The most important components of my DTPO database are the scanned pages from my Moleskine notebook. (As a side note: I’m currenly on my final Moleskine notebook, having found the Leuchtturm1917 series of notesbooks after having been dismayed by the darker lines in my latest Moleskine.) Once every couple of weeks, I use my scanner to capture the contents of my notebook. I do this for several reasons, the most important of which are: (1) to create a persistent digital record of the contents, which is available on all my devices, and (2) to force myself to review the contents of my notebook by tagging it. My scans are added to a folder on my Dropbox account (so that it’s available to me anywhere), and I use DTPO’s facility to index external folders to pull the contents into DTPO.
The contents of my notebook are added to DTPO groups that correspond to the month and year they were written. This doesn’t take much cognitive effort, and ensures that I can browse both the contents of my notebooks as well as other digital resources at the same time. When I’m in the mood to review my notes, I tag each scanned pair of pages appropriately. Tags might consist of topics, project names, or anything else that comes to mind. I don’t get too hung up on getting the tags just right, as DTPO has some pretty powerful AI helpers that allow me to find related items pretty easily.
I also add to DTPO any email messages that I think might be worth reviewing at some later date to DTPO. Note that I set the bar pretty low for admission to DTPO. I exclude messages that are simply procedural (e.g. setting up meetings, etc.). If there is any content that I think I might want to refer back to, the message gets added to DTPO.
I’m also pretty quick to add intersting web pages to DTPO using the “Add to DEVONTHInk” plugin for whatever browser I happen to be using. Other folks enjoy using things like Diigo or some other collective bookmarking site. I don’t.
I also add things that I find via Twitter (which I look though a few times a week) and RSS feeds (which I only get to about once a month, at best). I typically do so by saving interesting articles to Pocket and then, when I have some time to indulge in going through interesting material, I go through the materials that I have saved to Pocket and then add those web pages to DTPO via the usual mechanisms.
Approximataely 60% of my DTPO comes from my scanned notebooks, about 15% comes from adjunct material that I collect as a result of reviewing my notebooks, 10% comes from tweets, 10% from web pages that I happen to have open (mostly PDFs of scholarly articles), and 5% from email. DTPO has a recently released synchronization feature that I use to maintain my information across my laptop and desktop environments.
When I need to retrieve information, it’s as easy as entering search criteria into the search box, optionally using tags and, best of all, making use of the “See also” functionality to find other materials that I have added to DTPO but have forgotten about.