By way of answer, Oddey took me through the latest version of 3Com's grand plan. The company is trying to deal with three areas: Connectivity: the NIC card and firewall business. The company plans to continue making these easy to use and competitive Network switches: 3Com bundles wireless in with this part of the market, as well as voice/data integration, including the NBX digital PBX that is currently successful. Carrier infrastructure: here 3Com is not a switching vendor, but an integrator and services organisation. In the enterprise, the aim is to make stuff that is easy to use. "We're not the sexiest, or the first to deliver something, but when we do, we have it at a good price, and it is easy to use." With that aim, it is easy to see that "leading edge" products that it has talked about in the past might easily get sidelined. Embedding NT was seen as a way to deliver "policy-based networking" -- infrastructure which responded to the business priority of network traffic. This was a popular buzzword three years ago, but turned out to be difficult to implement with the technology of the time. Instead, 3Com has embedded more generic stuff on its NICs, starting with a firewall. Similarly, a leading edge user might get excited about a Bluetooth hub that sits on his desk and handles all of his communications with PDAs, printers, phones and such, without cables. 3Com has a prototype, and a year ago spoke to analysts about the product it might launch. But so far, it is too complex a product, and clashes with the company's 80-2.11b strategy for wireless, said Oddey: "We have it in our labs. There are a lot of things we could wipe the dust off and turn into products at the right time." Why should we believe in XRN? So, with 3Com's history of launching things that turned out not to be such a good idea, why should we take it seriously over XRN? Absolutely, says Oddey, because this is the next logical step in reducing network complexity. More cynically, this is a marketing-led package rather than a technology pitch. One of the main positive things about XRN, as she describes it, is that it gives resellers that rarity -- something "new" to say about core switches. "Chassis switches are flexible and redundant, but complex to expand, while stackables are affordable, but not highly reliable," she said. Making a more reliable stackable switch and letting it share a routing table with others across a "distributed fabric" (extra Gigabit connections and a shared routing table). Right now, there is just one "reliable stackable" from 3Com, and a promise that sometime this year, 3Com will issue a kit to upgrade to XRN. Next year, is when things will start to get interesting, as 10 Gigabit Ethernet enters the picture. Oddey expects the technology to move from service providers' metropolitan area networks (MANs) to enterprise backbones during 2003.