cmurray.org

This is a very interesting (and for me, thought-provoking) post on David Churbuck’s site titled, Content Management — the next wave.

Documentum — long a staple in the biomedical and pharmaceutical industries, but equally at home in the engineering environment I supported at Lucent — is an expensive application developed on a complex architecture. Our requirements for it at Lucent were simple: check documents in, check them out, provide a level of revision control, manage group and user priveledges, support a document repository across several divisions spanning the globe. Most of the content was in the format of Word, Excel, Visio, and some AutoCad. The web interface (customized by us because the Documentum implementation was so poor) provided simple navigation down the organizational tree and search to find documents. Once the requested documents were found, the user had the option of clicking to download the source file or poorly auto-generated HTML or PDF versions. When checking in documents, the files themselves are stored on the Unix file system, while the paths of the documents are inserted to the database.

I have on my LAMP server here at home a simple, open-source, web-based CMS called MyDMS. I downloaded and installed this a couple years ago, and with few customizations, use it for all my legal and financial documents, schoolwork, and storing billing and patient information for my wife (yes, it is an extremely hardened server). As with Documentum, the files are uploaded to the file system while their paths are stored in the MySQL database. This system provides me much of the same functionality found in Documentum, but at a cost only of my time to configure and cutomize it. (I also have installed on this server both Drupal and Joomla, just so I can get under the hood and try them out. I find neither without merit, but also difficult to customize and not particularly intuitive — thank god for the Mark Cahills of the world.)

The difficulty to which David alludes is at the presentation layer. Again, for Documentum in the engineering house, an organizational tree and a decent search usually got them what they needed. But supporting large, stylized publishing sites, like those spawned of IDG, is far more complicated. True enough, you can use something as simple as WordPress to gather and publish textual information. But to take that content and templatize it along with other dynamic and static content as well as revenue producing modules is a huge challenge. The value in these larger systems comes from their ability to use templates for gathering and presenting a wide array of content across a site in a consistent and functional manner, while also allowing producers to override them and customize pages. Search engines such as iPhrase then can provide a contextual component, delivering additonal, related content based on what the user has chosen to view. This level of complexity and functionality is not available, to my knowledge, in current open-source offerings.

But I agree that there appears no true middle ground. If my wife were to start a new practice with several other physicians, the open-source model would not provide the level of security and compliance with regulations, such as HIPPA. And equally, a Documentum or Vignette installation would fall well out of the financial scope. And I believe the open-source community will one day bridge this gap and provide truly scalable solutions across many business and financial models. And that’s what I love about open-source; their passion and their inability to resist a good challenge.