January 25, 2008, 10:43 AM — There is a general malaise in this industry that centers around the word "manage". If you have word processor documents and spreadsheets and presentations just lying around on file systems, there is a tendency to think that they are not being managed as well as they could be. And what better way to manage data than with a (gasp!) - a database management system?
I regularly encounter situations where office-like information is bludgeoned and squeezed and pummeled into odd shapes and odd workflows in order to meet the need to "manage" the information effectively using database management systems.
Now please don't get me wrong. I am not an information anarchist. I believe passionately in information management. But, I'm also a believer in using the right tool for the right job. It is a fact of life unfortunately that tools featuring the word "management" in their classification are often more attuned to the management needs of relational data rather than office-like document information.
One fear that often drives users down the road of introducing a database for office-like document management is the fear of not being able to find things again. I used to have this fear with respect to my paper filing system. "How", I wondered, "should I label these paper filing cabinet folders so that I can find these things again?". "Should this invoice be filed under 'house' or 'insurance' or 'invoices' or 'personal' or something else?"
I finally escaped from this paralyzing classification fear by realizing that it doesn't really matter. For any given item, the number of reasonable choices for filing it are small. Rather than worry about finding the perfect one, just pick a reasonable one! When it comes to finding it again, you may need to rummage to 2-3 folders to find it but that is a better use of your time than waiting for the (non-existent) perfect filing taxonomy to drop out of the sky for you.
I do exactly the same thing when filing office-style documents electronically. I use folder structures and file naming conventions. I go with the first reasonable option that comes into my head for any given electronic filing task. It might take me 2-3 shots to find it in the future but again, that is a better use of my time than endlessly searching for the (non-existent) perfect filing taxonomy to drop out of the sky.
That is step one on the road to pragmatic, simple information management.Step two is to ensure that you cannot inadvertently lose important information. To do that, I alway keep my folder structures safely ensconced inside a revision control system such as Subversion. Using a revision control system kills three birds with one stone: centralized backup, protection against inadvertent modifications and support for collaboration.
There are many other bells and whistles that can be added to this basic setup but a surprising amount of high quality information management can be achieved with just this much.
One little extra I like to throw in. By using a file system in this way, I end up with a ready made system of unique identifiers for my information assets. The full path to any document is a unique identifier. It is very useful to be able to use these identifiers as programmable identifiers inside computer programs. To do that, I like to restrict file and folder names to simple alpha-numerics so that they can easily be turned into valid identifiers in common programming languages.
Good old fashioned file systems have much to offer as information management tools. It is well worth thinking through your requirements to see if you actually, really truly need a database in order for your content to be managed. It is possible that you do, but it is also very possible that you do not. Databases have no monopoly on the word "manage".
