In-Depth
Web-to-Host Connections: The Post-Y2K Wish List
Many of you are probably relieved to have cleared the Y2K hurdle - the mother of all maintenance projects - and are conducting a final mop-up of any lingering, and, hopefully, minor problems. Now comes the inevitable flood of pent-up demands as people and funds are freed and companies act on their post-Y2K wish lists. By all accounts, IT managers will be kept just as busy as they were for Y2K remediation, this time renovating legacy applications to handle the Web.
Leading the post-Y2K pent-up demand will be monster projects that involve linking back-end legacy applications with the Web. Carl Griener, Vice President of Meta Group, sees mainframe shops evolving into "e-legacy" architectures, in which mainframe applications are renovated and updated to extend back-end parts of a system to new distribution models. Why, he asks, rip up systems that have been years in the making and work well for the company? "We just have to make sure it has a Web capability, and determine the best way to do that - by replacing the whole thing, or putting more of an e-com addition on one front-end of it."
There were a number of valuable lessons learned from Y2K, which will be instrumental in moving on to Web-enabling and e-commerce projects. First, it's notable that implementing a Web-to-host project, like Y2K, is basically not a technical challenge, but a management challenge. The technology is available, and readily deployable. But, management has to be sold on the concept, and end users have to be educated on its benefits. Y2K taught us to reach out to end users in every corner of the organization, to determine what applications are useful and need continuing support, and which can be dropped. In addition, Y2K efforts had to reach beyond the walls of the organization to business partners and vendors - as will be required in Web-to-host deployments.
It's fair to say that Y2K has drawn corporate management and IT executives closer together, resulting in a mutual understanding of each other's needs. This will be critical in moving from a traditional business to an e-business. "Y2K has succeeded in capturing the attention of top business management in a big way," says Dr. Howard Rubin, who has been tracking Y2K management projects for a number of years. "There's a real collaboration taking place. More and more business management is assuming operational and budget responsibility for IT-based projects. More than ever before, these business managers are charting a course for technology that's linked to the attainment of strategic business goals." A recent survey Rubin conducted finds that the two major goals in the post-Y2K era are e-business and customer relationship management systems.
Many organizations also have established Y2K project offices, charged with bringing the necessary resources and communications to bear on Y2K projects. The Y2K project office is an organizational entity that will have a life beyond Y2K, and evolve to a group that has the ability to do enterprisewide mass change. These offices are in a prime position to marshal necessary support for Web-to-host and e-commerce projects.
Another lesson driven home by Y2K is the continuing pre-eminence of COBOL-based applications. As companies were forced to go through and count up every line of code and application they have, it became clear that despite all the vendor hype about new-age systems, the world still runs on COBOL. In fact, at least 16,000 of the largest businesses run on COBOL, estimates Stephanie Moore, an analyst with Giga Group. In addition, there are still about two million COBOL programmers in the world - more than twice the number of Java programmers.
It's significant that so many companies have made a substantial investment remediating their mainframe-based code. Through the latter half of the 1990s, companies were forced to vote with their budgets as to what platforms they would stake their future on, and which they would kill off. MVS-based systems came out as one of the few platforms that made it through the Y2K portal, along with OS/400, Windows NT, and three variations of UNIX - AIX, Sun Solaris and HP-UX. As companies concentrate on e-business initiatives, applications on back-end systems need to be linked to the Web.
However, this is a message that hasn't reached the new Internet vendors, many of whom simply don't understand mainframe and AS/400 systems. In their world view, legacy systems don't exist or are on the verge of replacement, says Giga's Moore. Often, they end up selling CEOs and CFOs with dazzling new technologies, while virtually ignoring older systems that offer just as much, if not more, functionality and security. Even the COBOL vendors - IBM among them - push VisualAge for Java or VisualAge Generator for e-business applications over solutions that Web-enable older systems.
In addition, it's far cheaper to retrain COBOL programmers than hire new Java programmers - Moore estimates up to $30,000 just to recruit a Web developer. In addition, she points out that there's a wide variety of tools on the market - designed to re-engineer systems for Year 2000 - that can be just as effective in rebuilding systems for Web deployments. These include tools that separate the business logic from presentation logic, software configuration management tools and testing tools.
The bottom line is there will now be a multitude of demands that will be placed on IT managers to move to e-business, unfettered by Y2K. Web to host is one rapidly deployable tool available for this new era.
About the Author: Joseph McKendrick is a reserach consultant and author, specializing in technology surveys, research and white papers. He can be reached at joemck@aol.com.