TechWeb

Sounding Off About The Legacy Monster

Mar 30, 2003 (07:03 PM EST)

Read the Original Article at http://www.informationweek.com/news/showArticle.jhtml?articleID=8700015



Plug-And-Play Replacement To Legacy Systems Is Job Killer
"Since it's safe to assume that no savior is going to swoop in and slay the monster, what alternatives exist?"

I disagree with this statement, and yes, I'm a little biased. Everyone thought that Linux would be the hero, but Linux is just reinventing the monster from a whole new perspective. The model hasn't changed; it's still an operating system sitting on some hardware with additional software sprinkled in to add flavor.

And talk about legacy: Ask anyone who has had to upgrade NT 4 with Exchange 5.5 to ASD on S2K with Exchange! What businesses need is a replacement for what they have in- house with the least amount of reconfiguration while maintaining the same feel to the clients inside the network. They also need a solution that won't start becoming "the next legacy" the moment they deploy it. Wouldn't it be nice if you could deploy a solution that kept up with technology for 1-to-3 years after installation!

And let's not forget what we're really trying to accomplish--users just want to do simple things like share calendars, and coordinate meetings and events. They don't want Exchange, just the ability to use Outlook. Exchange is one of those "monsters."

OK, so here it is. A plug-and-play replacement into an existing network offering Microsoft authentication, Exchange replacements, the best and fastest disaster recovery available on the market, the ability to set up complex WANs and VPNs in minutes, not weeks, the stability of Linux, technology upgrades for at least one year (more for a very nominal fee), and many more features.

It's a replacement for what you have (getting rid of the monster for something that actually works), no certifications needed to deploy or run it, artificial intelligence to keep things running smoothly ... for $2,999. Don't laugh and don't dismiss this as one of those "toys" or appliances you may have read about. We consistently win projects through our channel to replace "monsters." Big ones, too!

So what's the downside? It isn't a Windows application server, so that's one monster we can't slay just yet (we're working on it). But it does address the basic-to-complex infrastructure of the organization.

So why hasn't anyone heard of this and why aren't we at the top of the heap yet? It's a job killer, plain and simple. I see it day in and day out. I demo the box to value-added resellers and end users, and the IT guys run for cover (the ones who "get it") because they know that if it works as advertised (and it does), they no longer have a job.

The product is the Net Integrator. I know that some of the claims I made are hard to believe, so I'd be more than happy to prove any of them.
Tom Turney
West Coast Sales, Net Integration Technologies Inc.



Trust System Is Alternative To Legacy
Innovest Systems has built the first new trust system in more than 20 years.

We are offering this as an alternative to legacy systems offered by SunGard, Software Engineering Institute, and Metavante. A large legacy system provider, AM Trust, recently announced they were exiting the business.
Bill Sullivan
VP, Innovest Systems LLC



NonStop Technology Secure And Scalable
We use Hewlett-Packard's NonStop technology to help provide maximum uptime and security. NonStop HP servers are designed to run 24 hours day with no downtime.

HP NonStop servers provide the highest and most scalable solution on the planet. Our configurations can start as small as two-processor systems and grow to more than 4,000 processors in a single cluster.
Morris Hoodye
Software Developer, Hewlett-Packard



Reengineer Programs Into Reusable Components
The only way to avoid slaying this monster over and over is to reengineer your current programs into well-behaved components.

When restructured properly, the services offered can then become reusable in at least two key ways: (1) with wrappers that provide access to the existing environment via the communication "flavor of the day" whether it is as a COM object, a Web service, Corba IDL, or something new we haven't even thought of yet; (2) translated from the current program language to a new language that enables migration to an entirely new platform.

Of course, reengineering legacy code isn't a simplistic adventure, and usually goes smoother when supported by automated, analytical tools.

First, you have to understand relationships between data and the logic that alters that data. Then find candidate duplicated logic and rationalize the common logic occurrences to become new, reusable business logic. To complete the cycle, the existing programs also need to call the newly reusable business logic--you surely don't want to cause new redundancy by leaving the old code in its monolithic state.

Once well-behaved component services replace redundant hard-coded logic, those same services also become the ones we can reuse the next time technology causes our environment to change.

At ASG we specialize in offering tools that help do exactly what is described here, and more. And we don't just sell these tools; we are applying them to our own products because we also have legacy software that has to keep up with new technologies. In fact, we're going a step further and adding an open services-oriented architecture that makes it possible to utilize tool services from outside our products--an especially valuable approach for Web service collaboration that requires easy-access metadata mapping and translation offered by our repository products. In essence, we're "eating our own dog food" and not just telling others we can help them acquire a taste.
Barbara Errickson-Connor
Director, Applications Product Management, ASG



Do you have legacy issues? Let us know here.

Back to Sounding Off About The Legacy Monster main page