If you accept Larry Ellison's premise that "internet computing is the final stage in the evolution of the industry", then replacing the traditional filesystem with a repository in the database makes perfect sense (and promises a whole boatload of Oracle server license sales).
The iFS paradigm allows you to use whatever method you prefer to create content and then makes that content accessible anywhere there is a browser-based connection to the database. Your content is indexed, searchable, and can be served up without regard to the software that is being used to access it. I have enough problems dealing with files that were created with a more recent version of Word or Excel than what I am running. This paradigm makes HTML or XML the consistent "least common denominator".
Storing everything in the database (data, files, video, audio, etc.) also allows you to implement a single, standardized disaster recovery plan. It gets much closer to Larry's vision of large, professionally managed servers than having to deal with NT filesystems distributed throughout the organization. Believe me, I don't see a line of people tossing their NT (or Netware) servers out the window anytime soon. I do think that as a strategic direction, the database as filesystem is an idea whose time will come (then again, how many of us have access to video-on-demand served up by Oracle Media Server!?).
Making Oracle smaller and easier to use is just not going to happen. Look at OracleLite to fill that need. Using Oracle Workgroup Server (on an NT box, perhaps) is also pretty straightforward. Does anyone really expect an enterprise-wide database to have the same footprint and management requirements of MS Access? Several of us were discussing (at the NYOUG meeting yesterday) when the need for a DBA to manage an Oracle database first materialized. The consensus seems to be that prior to Oracle V6, there were really only Oracle developers. From V6 forward, the database became a larger more complex beast that required dedicated, professional care and the Oracle DBA was born. At the low end, buying decisions are influenced by price and ease-of-use. No one wants a $500 piece of software that requires a $100K a year professional to manage it. If you are building a petabyte database, spending $1MM on hardware and 250K for a database license, the costs of acquiring the knowledge and the staff to keep it running are not the driving issue.
-Michael |
