I heard that my first Documentum project still lives despite three post-merger attempts to decommission it. It’s still my benchmark project because it was a real “Hoover Dam” from day one. That’s the kind of thing I don’t see anymore, real contenders for the 00000111 Wonders of the IT World involving Documentum. Sure, capturing the entire email stream of a 100,000-employee company will be impressive in a Rain Man sort of way, but we were eating up terabytes of disk and database back in 1998–back when millions of objects really meant big.
I completely lucked out after the Great Commodore Purge of 1993 (a 75% layoff with exit interviews held en masse in the cafeteria) by landing a job on a big project in a new technology for a great company with some of the brightest non-Commodore people I’ve met. So began my four year apprenticeship as software systems architect. Our team motto was “not for the faint of heart” which translates into the mother of all learning experiences.
Maybe this was just the state of Documentum in the 1990s; I knew of similar projects all over the Northeast and heard rumors of other interesting things along the Pacific Northwest. Some were about huge repositories. Others were about authoring complex documents. The best ones were enterprise-wide efforts where Documentum and (SG|X)ML intersected. My first Documentum job was all three.
Now I’m looking at my prospects as my back-to-work deadline approaches, wondering where those great jobs are now. There don’t seem to be as many, and I have a few ideas about why that is.
Money is tight and the government is saying “Jump!”
There’s plenty of Documentum work out there, but most of it is coming out of tightening regulations and compliance crack-downs. That’s not inherently interesting because it’s not about creating new intellectual property and doing interesting things with it. It is, however, mission critical.
Money’s been tight for years. The economy has been in an overall slump since 9/11, falsely propped up by creative monetary policy and a stampede away from real industry to finance. That industry’s had its share of problems lately, but even Big Pharma–companies in the business of selling life–have suffered financial and regulatory setbacks. Don’t forget extra-tight IT budgets because of the unrealistic expectations offshore software development promised–and promptly failed to deliver. Businesses (other than Apple, Google, and Microsoft) don’t have buckets of cash to throw at every big idea that wanders by.
These two facts mean that projects must really compete for dollars from the capital budget. Unfortunately, those less-interesting records management projects are scarfing up the bucks because they have one adaptation that trumps all others–a government mandate with well-defined, severe penalties. The bright side here is economies turn around and records management projects get done. Things will eventually get back to normal.
Business thinks it knows what Documentum is.
Documentum has enough mind share that people think they know what it is. They don’t. Perhaps they know what a particular piece of it does. Maybe they’ve bought into some of the counterproductive marketing from Documentum (the company) and EMC that recast the product as web content management or a way to sell more SAN. Worst of all, they may think they know what Documentum can’t do because of failed projects at the hands of integrators of questionable talent or intent, off shore and on.
These experiences, even just the fact of having so many experiences, aren’t going to attract the visionary, the radical, and the mildly megalomaniacal. You have to be a little crazy and totally overconfident to build a Hoover Dam or dig a Panama Canal. These people may already be a few software products beyond Documentum. They don’t notice business as usual, and that’s optimistically what Documentum has become.
It’s really up to EMC to get out there and evangelize Documentum as state of the art for more than their comfort zone. I would evangelize more, but that lack of a developer license to explore new features hamstrings me. IQs probably aren’t going to rise suddenly at EMC, so this problem’s here to stay.
There’s very little frontier left in document management.
In 1994, electronic document management was a frontier. The problem space was huge and mostly filled with dark matter problems, the kind you don’t know you have until you’ve solved simpler problems obscuring them. Only a few people had even considered what routes lie from problem to solution; a few of those people cashed out of Documentum years ago. Here were the raw materials for all kinds of frontier dramas, from Little House on the Prairie to The Donner Party. My kind of stuff.
Is there still room for those glorious victories–and equally glorious failures–that made the first decade of Documentum so much fun? I’m not ready to say “no” just yet, but something needs to give the problem space a really good shaking before I can answer “yes”. While the product’s certainly changed since the early 1990s, neither it nor any of its up-and-coming competitors have hit upon anything truly revolutionary. I’m not even sure what would count as revolutionary anymore, but I’d love to see some suggestions in the comments.
Are the interesting projects just lost in the crowd?
My final hypothesis is a little less pessimistic: Maybe those great projects are still out there, but they’re hard to pick out in a sea of so many other less-interesting projects resulting from a maturing market.
I’ve had words with a certain somebody about when document management is “enterprise”. She thinks it means everybody in the whole organization uses it; I’m fine with the term when it cross-cuts business aspects like department or time zone even if only a fraction of employees use it. We may be moving towards a world where she’ll be right, and the proof might be MOSS 2007.
I still haven’t touched it personally, but I did sit through a few hours of demos on Microsoft’s website. Aside from inheriting architectural problems from Sharepoint 2003, there is the potential for real document management here. It’s hamstrung by some of the usual Softie assumptions like “one document = one file”, but it’s a start. Brace yourself, because I’m the most optimistic about this case despite the Evil Empire influence.
Sharepoint 2007 could lay the foundation by really helping users “get it.” MOSS’s pretty face and stripped down functionality might entice where Documentum has traditionally overwhelmed. MOSS becomes a stepping stone to Documentum instead of a replacement. Those smarter users could become the fertile sea for a new cambrian explosion in the field that will make it fun and interesting again.