Getting the foundations right

I ran into this chain of thought while reminiscing about the previous two years of my career. When I first attended a training session on Documentum ECM. Having prior experience on J2EE & diverse architectures (like Graphtalk) aided the understanding process & I could easily grasp the concept of docbases, the distributed architecture, the interfaces, DFC, WDK and like. It all sounded like logical pieces of a tech puzzle.

What still puzzled me at the back of my mind was the concept of an ECM. And I see the same puzzling look on faces of folks newly inducted into the ECM arena. Especially if they got say some 4 years or less experience building systems using Java & related technologies. In my case, I used to think - Why do these folks need something so grandiose as an ECM. Why can't they put all of it into a database? I know, the consultants here will rip me apart - for this question - but many I asked then couldn't give me a simple explanation themselves. They used to ramble a bit about unstructured content and the fact that its a pain area and so on. But never did they articulate properly as to why current database systems are inadequate, where the challenges in building an ECM lie and so on.

My first ECM project was interesting as well. With a fair understanding of the design, interfaces and products but literally no conceptal understanding of how ECM solved problems, I landed into the murky world. It was an experience I wouldn't forget. It taught me stuff - I'm sure books would never do justice to. I had a techie's view on a subject that enthralls corporate IT departments worldwide & keeps consultant's pockets stuffed with wads of money. The clients wanted to know about proper taxonomy design, migration roadmap, infrastructure sizing, functional design - basically about how their problem areas would be resolved. And here was I, trying to map these to folder structures, dump & load tools, product installation, caching tools - everything nuts & bolts about the product - but not how it helped solve the problem. ( Of course - I still managed to complete that project with integrity ;-).

It was there, slowly & gradually, across months, that basic nirvana dawned on me - why ECM was significant. What I'd got myself into. Why the bigwigs were spending so much on this Enterprise buzz. I then set about exploring the other facets of ECM - not as a product but as an domain. Started with the histories of EDMS'es, explored the various logical designs for ECM systems, why some systems succeeded while others failed, what differentiated the cream from the rest, what industry pundits were predicting, the various streams of ECM and their significance - basically more case-studies etc about how real-life projects implemented & integrated ECM into their folds.

Today, I'm still no expert in the field. The field seems deeper as I dig each day. But yes - I have cleared a lot of those queries nagging behind my mind. Instead of getting awed by complex architectures, today I'd focus more on identifying how that helps solve the problem at hand. It is especially important as Web2.0, Serverside & XML based stuff merge together to throw up an amazing range of technology platforms. One definitely needs to keep abreast of them but always let the simple ideas and foundations behind them guide you.