An old acquaintance from my days at The University of Chicago recommended this product. On the strength of his review, I decided to give this product a try. DevonThink is a document management system for the Macintosh platform. In plain terms, it is an electronic filing cabinet for electronic documents. Anyone who has done any kind of research on the Internet will understand the utility of document management. Search engines generate voluminous lists of data, and the act of clicking hyperlinks creates electronic paper trails faster than most of us can take notes on what we have found. What is needed for serious research on the internet are tools that help us to understand the data that we have collected, and DevonThink is an excellent tool.
DevonThink has a terrific user interface and seems robust. It can slurp most any kind of document (PDF, RTF, HTML, DOC, XLS…) and arrange them hierarchially. Once the documents are in the system, DT will take a stab at finding documents related to one another, using its “see also” feature. So if you have 10,000 documents in DT, and want to find all the documents related to, say, baseball, DT will scan through all of them and give you a result set. DT also offers some nice text analysis features such as concordances, which is basically a list of every word in a set of documents and the number of times every word appears. If you want to really get the most out of DT, however, you will need to download their other product DevonAgent. Agent will automate the task of getting web pages into your DT database.
There is a learning curve for DT. The documentation is OK but you’ll tool around with it for a while before you get the hang of it. I suspect that the complexity of using this product will to its fullest potential will turn many people off.
Interestingly, products like DT are predicated on the assumption that the existing file systems and search facilities built into the computer’s operating system are inadquate to the task of finding, labeling and presenting data. For a long time, that was most certainly true. File systems have come a long way since we used DOS. This observation has been bandied about for a long time by computer researchers. If my memory serves me, about ten years ago a research project at Yale University called the ALICE project (I did a quick search of Yale’s site, but didn’t see a mention of it or I would have provided the link), speculated that users would be better served if they didn’t have to create folders and documents with cryptic names. At the time, file systems restricted users to devising names with eight characters and a three character extension. Directories were similarly limited. The bottom line was that users nested documentes and directories so deeply that there was no hope of finding something unless you knew exactly where to look. Search tool were too primitive. The ALICE group wanted to get rid of directories and allow unlimited file name lengths. To help find documents, a powerful search facility would create a virtual file directory, based on the search criteria.
Windows NT, XP and the Macintosh have come a long way since DOS and the Early MAC OS. Long file names are now permitted and directories need not be nested deeply to categorize information. But if finding documents is a bit easier, the volume of information stored on our computers has grown far faster than any of the improvements, and for that reason finding data, storing it and turning it into the understanding is as difficult as it has ever been.
