by Alex Woodie

Knowledge management has evolved quite a bit over the millennia. From etchings on a cave wall to the Dewey decimal system to computer output to laser disk (COLD) storage, man has dynamically adapted his information storage and retrieval techniques to cope with the continuously growing base of knowledge.

Now, in the 21st century, the cutting edge of knowledge management resides in the networked realm. The Internet poses its own peculiar problems, particularly those surrounding the exponential growth of unstructured data on the Web. But the Internet also provides a path of solution.

Gauss Interprise, a German company that sells enterprise content management software, hopes to bring its large installed base of American OS/400 customers forward into knowledge management using the Internet. Gauss inherited about 1,000 AS/400 customers in June 2000, when it acquired Magellan Software, an Irvine, California, company that developed and sold a suite of document management software called SpyVision. Before the merger, Gauss' Web-based content management products, in the Versatile Internet Platform (VIP) suite, were strictly Unix- and Windows-based offerings.

With the recent release of VIP Enterprise 8, Gauss has merged the old Magellan code with its VIP suite. As a result of the integration, which Gauss says took 16 months and cost $20 million, Gauss is now providing an upgrade path for AS/400 shops. Gauss hopes those companies, which previously focused on back-office document management, will start managing the full scope of their content, and do it using the power of the Internet.

VIP Enterprise 8 has four core modules: VIP DocManager, VIP WorkflowManager, VIP PortalManager, and VIP ContentManager. Magellan's old SpyVision is now called the VIP DocManager, and Gauss' VIP products have become the VIP PortalManager and the VIP ContentManager. All of the modules, except VIP WorkflowManager, run on AS/400 or iSeries servers. The complete set of modules runs on Unix, Linux, and Windows servers.

Underneath the modules runs a layer of Java and XML code that Gauss calls the VIP Enterprise Objects (VEO), a collection of APIs for connecting VIP Enterprise 8 to existing applications. In order for SpyVision customers to upgrade to the VIP Enterprise 8 offering, they need to have some knowledge of Java APIs, or hire somebody who does. An upgrade from SpyVision to VIP Enterprise 8 takes from three to five days, Gauss officials said.

When customers do upgrade to VIP Enterprise 8, they'll benefit by not having to cobble together a collection of programs in order to achieve sophisticated Web site content management capabilities that are integrated with existing document management and ERP systems, Gauss said. By having a single product that does everything from capturing documents from ERP systems to managing the workflow process among dozens of people to publishing content on the Web and managing the various versions thereof, users can manage their information and distribute their knowledge more efficiently, the company said.

If the promise of a Web site content management system that is seamlessly integrated with back-office applications doesn't do the trick, Gauss has other carrots on sticks to lure SpyVision customers into upgrading to the VIP Enterprise 8 document manager, such as new revision control and PDF capabilities. But make no mistake about it: Gauss wants its SpyVision customers to buy the other modules as well. And if its customers' Web development platforms aren't an AS/400 or iSeries, that's fine, too, because the VIP Enterprise 8 suite can be deployed across multiple, non-similar platforms, including Solaris, Windows, HP-UX, AIX, and Tru64 servers, as well as Oracle, SQL Server, and DB2 database management systems.

In order to work, VIP Enterprise 8 requires a JavaServer Page server and an HTTP server. Gauss recently announced that it will sell versions of VIP Enterprise 8 that have BEA Systems' Java-based WebLogic Web application server embedded into the product at no additional charge, as well as non-WebLogic versions. The company said it made the deal because IBM was loading WebSphere Standard Edition 3.5 free of charge on the iSeries, which would have stacked the deck in favor of WebSphere. WebSphere and WebLogic account for 72 percent of the Web application server market, according to an October report from Gartner Dataquest, and both products run on OS/400. Gauss officials said that when IBM stops giving away WebSphere 3.5 in September 2002, the company is considering a similar OEM deal for customers who want to run WebSphere. For an HTTP server, Gauss supports Apache and other servers.

by Dan Burger

Paul Schlieben is president of SoftLanding Systems, a Peterborough, New Hampshire, software company that has been helping application developers manage software since 1989. I asked Paul for a few comments about trends in the IT industry and the technology and planning that pertains to a change management strategy.

There have been some dramatic changes in this industry over the past several years. How do you sort out this progression, and what has it meant to your company?

We've seen an increase this year at SoftLanding, and part of it is a reflection of a flat year last year relative to 1999. All of that is attributable to the year 2000, when people were so focused on getting it done that they weren't considering a tool to help them.

After the last half of 1999 and the first half of 2000, people just sat back and said, "Whew! Thank God that's over. Now what are we going to do?" After that, it took five or six months to get things going again.

There were some companies were working on their infrastructure, but for the most part they were casting about, not knowing whether a client/server or Web server was going to be the best thing.

After 2000, people started thinking about the future. I think the answers materialized rather quickly. It was clear because part of the market had already gone there. It made people look at things like WebSphere, Java, and so forth.

Now, in 2001, not only is that clear, but the tools and the concepts--what were formerly the missing pieces--are now in place. Now people understand that they can leverage their existing code and work within the context of WebSphere and Web server applications.

Anecdotally, I hear people espousing the view of the future, including things like WebSphere, Java, and using existing code. That vision is becoming fairly well accepted by people I talk to in the industry.

That said, I frankly don't think a lot of people have done anything yet.

I think a fair number of companies are in the training stage. They have a "sandbox" computer someplace that they are playing around with and learning on. They are poised to take that next step of implementing Web applications--things that incorporate database back ends with a Java front end.

I think people are starting to experiment in that direction. It's exciting.

The thinking is more like "now I understand where we are going and how we are going to get there and what the tools might be to get me there. Now I can get started and start getting people trained."

Outside of an increase in developing Web applications, what trends have you noticed?

We see tremendous interest in high availability as it relates to software management.

We see more and more people using the iSeries as a Web server --we are at the tip of the iceberg there (I hope there's an iceberg below that ice cube!)--and we expect there will be more and more people maintaining high availability with their software.

How does this affect change management software?

Well, for instance, those who are developing Web applications on the iSeries and are considering high availability have to ask themselves, "How do you manage database changes without bringing the system down?" That's what our PDQ product does.

People are realizing they need more discipline in the process as they get exposed to the greater world, not just their own back- office operations. They don't want to have to apologize for being down two hours. For customers who are inquiring or trying to place orders, the cost of downtime becomes a real factor. I think we will see more and more of that. And that has nothing to do with hardware or the operating system or the database; it has to do with people making changes. When are you most susceptible to problems? When something changes. And that's what you have to have well-managed. And that's what we do.

When anybody's world starts to break down, they start changing their applications. Unless they have a really good engineering process in place, that's where their risk is. That's where most experience difficulties.

From an application point of view, that's where there needs to be a focus--because if you have problems, your customers don't care what it is. If a company's Web application software is not working, the customer's attitude is, "I'll go somewhere else to put my order in." Serious downtime will cost a company real bucks.

You need to reduce risk as you move forward. Nobody wants to go one step forward and two steps back. People need to make sure their existing environments are under control. They need a structure, a solid foundation for managing software in place.

We provide the tools to provide that solid foundation. And the tools for the future. Tools to manage WebSphere and Java and content. Web management is a lot about content.

How do you explain the plan for improving software management capabilities?

Start with basic change management functions that secure their source and objects. Establish the development of test and production environments. That helps put a lid on the chaos. Then you can build on top of that.

Generally start with change management and source management fundamentals. Get it [the software] under control, as well as the deployment in place. Then implement some testing tools and debugging tools to build a better development environment. And, finally, implement some diagnostic tools, so that when there is a problem, you can get in there and fix it quickly.

The process changes, depending on where you are in the continuum from chaos to control.

Take a comprehensive, end-to-end approach to the entire software management infrastructure problem.

The value of a system to manage software is that it has to work well in each of the elements, but they [the system and each of the elements] have to work well together, because integration is absolutely necessary. It needs to alert you--when you are making a change--that you need to consider another area of your application. If one part of the application is ready to be implemented, there are things that need to be done, because it affects another area. That kind of intelligence is built into the software management tools. It's not going to be there if you are piecing together disparate tools. They have to be integrated to arrive at a level of intelligence.