My Digital Transition
After coming back from Microsoft with another bundle of printed research papers in hand I found that, in the course of a year, I’d amassed a stack of read (and needing-to-be-read) research papers that filled a set of binders almost a foot thick. This would be fine if I had categorized them and knew exactly what went where, but I hadn’t and I didn’t. In fact, I had no idea what was in those binders. Furthermore, on several occasions I’ve found myself saying, “Damn, I know there’s a paper about this that I’ve got in these binders …” and not finding anything after flipping through them for about 10 minutes. “Self,” I said to myself, “there’s got to be a better way!”
I identified three problems that needed to be solved:
- A foot of paper a year, if it continued growing at that rate, would be as tall as me before I graduated. That just isn’t sustainable.
- Even if I kept all six feet of paper, it would be impossible to find any one of them in that stack.
- Often, I’m looking for a particular topic or a group of related papers rather than a single paper.
- I write notes in my papers, and I’d like to retain the annotating flexibility of scribbling on a paper with a pencil when transitioning to a digital system.
At this point, I think I’ve got all but the last one figured out.
I had considered just getting a Fujitsu ScanSnap document scanner, as it translates paper directly into PDFs, but $500 for a document scanner was way too pricey. One of the things I had going for me was that all of the research papers I’d printed started out as PDFs. The easiest and cheapest solution was to find the original PDFs, index them, add any annotations I’d put on the papers beforehand, and recycle the originals. In the end I chose to not transfer the annotations; frankly, I couldn’t decipher most of them and it would have taken too much time. Armed with Google Scholar I was able to find PDFs (and bibliographic information) for my entire paper stack in about 90 minutes.
Now that all the PDFs were downloaded, I inserted them all into Referencer. Two nice things about Referencer are its ability to store bibliographic data (in the form of BibTeX citations) along with papers and its ability to associate papers with tags, descriptive words or phrases (similar to how del.icio.us does bookmarks). This really helps when searching for papers that fit a given topic and is a lot more flexible than any fixed filing system. It also allows you to give each paper a text “note”; I find that, if I take notes on a paper, this forces me to be more concise and structured in writing my thoughts about it than scribbling on a page’s margins would be.
Now that the papers were digitized and tagged, they needed to be searchable. If it’s one thing that Google and Spotlight have taught me, it’s that if it’s not searchable, I won’t find it. Beagle does an admirable job of indexing the text of all my PDFs automatically.
The last step in the process was to make sure that when(!) my hard drive died I wouldn’t lose all my paper data. This was relatively easy, since UCSD just got a shiny new NetApp file server with a boatload of redundant storage. I’ve set up a cron job to synchronize my home folder to that server every night at midnight.
Now I’ve got a much more accessible paper library that’s really easy to maintain. The march toward paperlessness doesn’t stop there, however; just a couple weeks ago I recycled several pounds of software manuals that, if I ever needed them again, I could find online. As a result I’ve got one less storage box in my closet which is a big help given that I live in a pretty small space. I have a feeling my “important documents filing box” is next.
