COMMENTARY--A few weeks ago, I wrote about the promise of portals, declaring that all roads to increased productivity in the enterprise lead to customizable digital dashboards, ending the reign of monolithic applications.
These digital dashboards provide the user interface to all the relevant data as well as the analytical and communications tools to perform your tasks from any location or device.
The problem, I pointed out, is not in creating the portal user interface, but in integrating and synchronizing all the data sources into a virtual data warehouse that can be acted upon by a variety of analytical tools.
At least one company is making broad claims to having solved this enterprise integration conundrum. In a recent ZDNet news story by Wylie Wong, BEA Systems CEO Alfred Chuang sketched out the company's plans for a technology code-named Liquid Data.
He did not offer much in the way of specifics about Liquid Data, despite the company's big ambitions for the technology. In a letter addressed to stakeholders, the company claimed that Liquid Data would become the next major business paradigm, and that the BEA WebLogic E-Business software would become the operating system for the Internet.
I am accustomed to companies touting their virtues, but those are especially audacious claims. I talked with Chet Kapoor, vice president and general manager of BEA's Integration Group, to get a more in-depth, less hyperbolic briefing on Liquid Data.
Unfortunately, Kapoor was not very forthcoming on the product specifications, ship date or packaging either. I have to assume Liquid Data is still vaporware, and the company is tantalizing the market and customers with its latest bet-the-company strategy. Presumably, BEA doesn't want to give competitors like IBM, Sybase, Oracle, Microsoft and some smaller players too much of a peek into its technology.
Liquid Data conjures up an image of data that flows easily, adapting to the shape of the containers. Kapoor described the application of Liquid Data in terms of aggregating data from multiple vendors and applications. "For example, I might have three different applications from different vendors, but could actually have one common view for all the data that exists," he said. "The distributed query technology allows you to create a view and, when you need information, the technology queries the different databases through adapters and makes the data available."
With the data available in an aggregate view, performing decision support and other analytical functions become much easier. In addition, he said that Liquid Data would use a flavor of Xquery, an XML Schema-based query language for performing distributed queries across a variety of data sources.
I subsequently spoke with Sybase CEO John Chen about his company's plans around solving the enterprise integration problem. He came up with the term "data liquidity," most likely in response to BEA's Liquid Data concept. Whatever the origin of Chen's phrase, he offered a similar concept to BEA's, and predicted that enterprise integration would be the next major battleground.
So, what can you expect in the coming months? You will hear a great deal about enterprise integration platforms, and every vendor will claim to have the right solution.
BEA is already strong on the integration front, as it doesn't focus on any single database or application base as does Oracle. In addition, BEA will work very closely with HP, and the two will likely go to market with an integration platform that leverages their customer base.
Sybase has a strong database background, and over time has put together a strong set of products, including its leading iAnywhere wireless platform, for building an integrated solution. The challenge will be in making the sum of the parts greater than the whole. You can also expect Sybase and its strategic partner PeoplesSoft to build integration platform solutions that leverage their product lines.
IBM is probably best suited to deal with the big integration conundrum. The company reaches from the mainframe to handhelds, and has major investments in middleware as well as a huge services organization to assist companies in devising an enterprise integration platform strategy. Microsoft will eventually provide a more comprehensive integration platform, and Oracle will have to deal with customers who won't go the pure Oracle route. Some smaller companies may even come up with some unique product that surpasses what the major integration platform players can achieve.
As Sybase's Chen says, this is the next big battleground. The vendors will compete with each other for share of market, and will come at the problem from different areas of strength. If the end result is that the vendors develop better enterprise integration platforms that really solve customer problems, then it will have been a good battle.
Daniel Farber is the Editor-in-Chief of ZDNet.
