Get Some More Hats
What is the best way for a technical person to spend his or her intellectual capital in the years to come? Here is what I think is the best way to stay marketable in a Microsoft-dominated community.
Regardless of my feelings one way or another about Microsoft’s business practices, I’m still betting on its technology. There are a couple reasons. First, and maybe most important, the acquisition cost for Intel architecture machines is cheap compared with everything else. Competition for hardware is everywhere, and I can inexpensively buy equipment today from Internet auction sites, brokers, and clone vendors. For example, I just bought a network fax server from an Internet auction site for $249. It’s an SMP-ready IBM PC 365, P200, 32 MB memory, 2.5 GB hard drive, with Windows NT Workstation.
Support tools are also abundant in this market. With $1,500 and a technically competent friend, I can become a Microsoft Certified Solution Provider (MCSP), which gives me access to all of Microsoft’s latest technologies.
With Windows 2000 hitting the streets, there will be a one year window when this expertise will be worth a premium. By this time next year, armies of newly minted MCSEs will be flooding the market and the price for this skill will drop like a rock.
That’s the downside to this market -- if I can afford to be in it, so can nearly everyone else in the world. Microsoft says the Minneapolis/St. Paul area has about 17,000 companies that do business in some way with Microsoft products. Since Minnesota has between 4,000 and 8,000 businesses with 100 to 1,000 employees -- depending on who is counting -- that means plenty of people are available to service the Microsoft technologies these businesses use. So don’t get cocky about all the gloom and doom stories about a skills shortage in the United States because I’ll bet the situation is the same all over the country.
This leads to part two of the grand strategy, which is to diversify. Do this two ways: Find an application for the technology that is popular and worth learning, and learn Linux.
Lots of choices are available for applications. In my mind, the most popular include e-mail, e-commerce, and management resource planning (MRP) systems. I’d also put the truckload of software development tools into this category -- development expertise will always be in demand.
Linux is a good bet for the future. Many developers and big vendors are busy hardening it and the open source model. It won't die. Right now, Linux makes sense as a small Internet server, but it will soon play a more significant role on the desktop and in the data center.
Sooner or later companies will look at the packages they are paying a fortune to maintain and will ask themselves, why? They will try a low-cost, Linux-based package and find that it lacks all the features of the expensive package, but it works "good enough." A few early adopters will put Linux into production with these applications, followed by a few more. Before long, the trickle of early adopters will become a flood of mainstream organizations as everyone tries to squeeze a few more dollars out of capital costs.
Learning Linux also has other benefits. If Sun Microsystems or Hewlett-Packard move into Microsoft’s territory, the jump to learn Solaris or HP-UX based on a Linux skill base should be manageable.
What about other environments? Unfortunately, unless some vendors spend serious money or radically change their strategies, the other environments will continue to shrink until they disappear. For example, Compaq is still convulsing from the DEC and Tandem acquisitions, and since dumping NT on Alpha, it has no real prospects for any long-term volume with Alpha. It never built market share with Tru64, and I don’t see them spending money to port OpenVMS to IA-64. When Alpha dies, so will OpenVMS. What a shame. Other players -- such as CDC, Unisys, Stratus, and the others -- never acquired enough market share to be of interest, and mainframes are also stagnant.
Sometimes the truth hurts. But it’s time to face facts. People who bet their careers solely on any one proprietary environment need to broaden their skill base or they will face eventual obsolescence. For me, after spending more than 12 years with DEC and OpenVMS, and after watching hard-won NT skills become a commodity, never again will I be a hostage to a single vendor. --Greg Scott, Microsoft Certified Systems Engineer (MCSE), is Chief Technology Officer of InfraSupport Etc. Corp. (Eagan, Minn.). Contact him at gregscott@infrasupportetc.com.