Because of my ongoing dissertation (humanities), my workflow is similar in many aspects: Wide-Screen monitor, scanning from books, Bookends. I use two scanners, a CanoScan 3200F (fast, usb 2.0), plus a ScanSnap on my desk, always waiting for input. The Scansnap eats a lot of copied articles from the library - it is not possible to borrow the majority of books here. My workflow:

1) scanning while clearing the Inbox: (with ScanSnap app. and VueScan for my flatbed scanner) All the files go into the folder “InScan”

2) import Scans into Devon, rename. I use two databases (“Dissertation” and Archive” for everything else.)

Why Devon? The “fuzzy search” is incredibly helpful to locate any file. OCR is never 100% accurate - not a good basis for a spotlight search. When writing a Paper (in Scrivener), I often paste portions of my text into devon and look for similar entries in the database. Results are impressive so far.

3 a) (in the evening): Batch OCR all the scanned PDFs in Devon (OCR needs a lot of memory) If you have many PDF files, search for “.pdf” in Devon and change the view of the results to display the file type, so that you can select only the files that are not yet “PDF+text”, convert them and delete the unconverted files afterwards (next morning).

b) sorting the documents in DT

4) (if a Bookends entry exists:) export PDF and attach file to bookends entry. For annotating PDFs, I use pdfPen - it is cheaper than acrobat (and a bit faster?). For notetaking, I use Bookends - the annotations from PDFPen are visible in BE (annotations from Skim are not).

5) BACKUP

Since I use the ScanSnap a lot, regular backups have become a top priority. I backup with an external drive and with mozy.com. I also save the DevonThink databases to a DVD-RW every weekend: Due to their size, off-site backup is too slow for these files.

With stacks and faster preview in Leopard, organizing my semi-paperless office should become easier.

Markus