While you can argue over whether F. Scott Fitzgerald was correct when he wrote, "There are no second acts in American lives," there are certainly second, third and fourth acts in the lives of technology products. A few cases in point:
Stratus Technologies, in Maynard, Mass., has been around since 1980. Along the way, the company has had different owners, different strategies and different operating systems. Once the darling of the banking and financial industries with its focus on a proprietary but solid nonstop computer architecture, the company seemed to lose its way as the 1990s turned into the 2000s. Computer demonstrations often tend to be really boring ("Look, the inventory number went down by one!") or only interesting when the system crashes. Stratus demonstrations were a bit more interesting, as the company was always willing to yank out processor boards to show that the system would indeed keep running without a hiccup.
I stopped at the companys headquarters recently to get an update from President David Laurello. My questions centered on several things, including the state of high-availability computing and the role of clustering. It seems to me that the stuff Stratus sells (computers that dont quit) is in high demand these days and that attempts to build high-availability systems using techniques such as clustering are slowing, as clusters were intended to build scale, not reliability.
Laurello has been on a quest to build off the companys base of reliability and uptime on expensive boxes and transfer that to industry-standard Intel hardware and common operating systems, with an eye toward taking server costs under $20,000 and even lower.
"Stratus is back together again," said Laurello. Well see if the financials bear him out, but for a company in its third (or is it fourth?) life, the ability to yank those boards and keep the system running in a standard, competitively priced box is compelling.
A week after my visit to Stratus, IBM held a press conference in New York to talk about its latest Workplace initiative. This is at least a third act, where you design a robust middleware architecture to sit between the server and client to offer Web-based applications. The difference is, in this act, users will actually be interested in the applications. User devices—whether handhelds, desktops, phones or kiosks—are proliferating a lot faster than the business applications available to serve those devices.
While these applications can live in an occasionally connected world, they work a whole lot better in an always-connected space. That is no big deal when you work within a corporations walls, but it gets trickier as you travel. The telecom companies would love to keep you always connected and are now building out those networks. The difficulties of building applications that can recognize and reformat information for a range of devices are being surmounted.
Is all this activity around service-oriented architectures a threat to Microsofts monopoly on the desktop? I wouldnt count on it. If ever there was a company that had second, third and fourth lives, it is Microsoft. I suppose having multibillions in the bank helps, but it always seems that the company has been willing to rev its products enough times to counter threats it has missed. Even with the big misses—the Internet, the Web browser and search engines—Microsoft came back later with competitive products. It wont be any different for service-oriented architectures.
The other recent example (and way more than a second or third act) was the Sasser worm. Like so many security vulnerabilities, Sasser used a known hole to develop a worm with a malicious twist. Microsoft can point to a patch already developed and deployed for the vulnerability, law enforcement officials can point to the quick apprehension of the alleged worm writer, and anti-virus companies can point to a quick issuance of alerts and fixes. But how many lives do these security gaps have to have before the curtain is finally brought down on this obstacle to broad-based, safe computing?
Editor in Chief Eric Lundquist can be reached at firstname.lastname@example.org.