And now for some irony: Microsoft’s print ads touting its .Net development tools as the solution for DLL hell — a bit like a cobra advertising antivenin, don’t you think?

But wait! There’s more! If you watch much television you’ve surely seen one of Microsoft’s dramatic reenactments showing how it can quickly and easily turn your company into a model of real-time efficiency. So it was with some bemusement that Nate Viall, a sharp-eyed regular reader of this column, forwarded an e-mail advertisement for Microsoft Money he recently received that closed with this disclaimer: “Please note that it can take up to eight weeks to update customer information in our database; therefore, you may receive e-mail from us within that time period.”

The kids at Sun Microsystems, always trying to out-do Microsoft, are responsible for this longer, but just as ironic tale.

I’ve been experimenting with Linux desktops, trying to build a satisfactory replacement for Windows. One barrier is cosmetic: Linux’s font rendering has made hideous look good, so I was excited when Red Hat 8.0, which installs with OpenOffice 1.0, yielded a thoroughly satisfactory viewing experience.

Figuring many companies will prefer an office suite with corporate support, I next installed StarOffice 6.0. It ran without crashing, but to call its appearance disappointing is to call putrid “mildly unpleasant.” Red Hat made it clear this was Sun’s problem, and so, in preparation for contacting Sun’s crack technical support squad, I tried to register for my free 90 days of support on Sun’s website.

No go: The registration process crashed with a Java run-time error. Bloodied but unbowed I called Sun’s technical support number and explained the problem. The next day I received an e-mail with Sun’s diagnosis: “The cause of the problem can be definitely traced to the operating system. Our answer: Your problem description points to missing or non-existing files in your operating system. Furthermore, a certain setting or configuration may also come into question since the origin of this problem is not clearly indicated.”

Understand, the registration process crashed on (a) Internet Explorer/Windows XP on my laptop; (b) Internet Explorer/Windows 98 on my desktop; and (c) Konqueror/Red Hat 8.0, also on my desktop. So Sun Microsystems, inventor of Java and one of the industry’s most vocal advocates of Java server-based processing because it allegedly reliably runs anywhere says that if their website crashes it’s a desktop issue.

It’s enough to make me yearn for DLL hell.

Neologisms and Polonium 210 have something in common. Both have a half life of less than six months.

In the case of Polonium 210 it’s the radioactivity that declines. In the case of a new and useful word, it’s adherence to the original definition. Just as radioactive substances eventually decay into inert, dull elements like lead, so neologisms decay into inert, dull meanings.

This is the sad fate of the term, “disruptive technology.” Coined by Clayton Christensen in his groundbreaking The Innovator’s Dilemma to describe tools and techniques that disrupt and ultimately redefine marketplaces, it has deteriorated, now referring to any new technology that might achieve marketplace success.

It’s a shame, because the concept is endangered along with the term. Take wireless LANs — potentially disruptive in the original sense.

A disruptive technology isn’t, at first, a useful replacement for the older technology from which it emerged. Christensen’s original case study looked at disk technologies: Winchester drives couldn’t replace mainframe DASD (IBMese for disk); 5¼ inch hard drives couldn’t replace Winchesters, and so on. Instead, each needed new markets and uses to incubate and improve.

Christensen’s insight was that disruptive technologies never do overtake their predecessors. What they do instead is overtake the needs of their predecessor’s marketplace. 5¼ inch hard drives, for example, never exceeded Winchester drives in capacity or performance but did eventually overtake the needs of the minicomputer marketplace that Winchester drives dominated. When they did, Winchester drives ceased to have any marketplace at all.

Which brings us back to wireless LANs. Commodity 100 Mbps Ethernet outperforms the most optimistic performance improvement possible for wireless LANs — the 2,000 percent performance increase modem technology achieved. It’s simple: If wireless LANs reach, for example, 1 Gbps, just add 99 more wires and you’ve achieved cheap parity (actually, much more than parity since wired Ethernet bandwidth is switched while wireless bandwidth is shared).

So wireless LANs have to incubate where wired LANs can’t go: Airports, coffee shops, and homes. Then, there’s a self-serving favorite of mine — your facility, to connect consultants without jeopardizing security, by hooking cheap wireless hubs to a VLAN channel outside your firewall.

Incubated in niches like these, wireless LAN technology will improve. To be truly disruptive, though, it will have to overtake the needs of the wireless marketplace. I have my doubts.

But don’t let that stop you from the VLAN use. I’d really like something faster than a modem connection.

Please?