Despite joining TV's pundits in gaffe-land by noting the certainty of an Alan Shearer goal before the England-Romania debacle, I'm still not through with tipping winners. This time, it's a long-range shot: mobile devices will very shortly be more important than what you currently have hogging desk space in your offices. OK, so there's been enough hyperbole surrounding mobile computing. Eight years ago, I heard a Toshiba spokesman predict that mobile PCs would be outselling desktops by the year 2000. Here in the 1998 real world, notebooks account for about 21 percent of all UK PC sales and Toshiba has begun offering desktop products. That says it all. There are still plenty of flaws in the argument for why mobile PCs will replace desktops. One, they're so damn expensive thanks to the incredible integration needed to manufacture these marvels. Two, they still lack the top-end features that the desktops have: giant screens, fastest processors, huge hard drives. Three, they're still too fragile. Four, the smaller formats that are truly totable still demand trade-offs: try typing on Mitsubishi's incredibly thin Pedion or on Toshiba's book-sized Libretto. I'm not saying that there won't always be trade-offs if you want mobile hardware. However, falling prices of TFTs will soon make mobiles cheaper and more attractive. The price of this single most expensive component in a notebook is collapsing as Taiwanese and Korean manufacturers reap the rewards of years of diligent R&D. Look out for usable sub-£1000 notebooks next year. This trend alone will, I believe, push waverers to buy mobile PCs instead of desktops where mobility in computing is desirable. But the PC isn't where the real mobile action is. That spot is for much smaller devices, including PDAs, smart phones and pagers. That handheld computers are an important and growing part of the enterprise is beyond dispute. Oracle's development plans for Oracle Lite for 3Com's PalmPilot, a database that can suck data from Oracle 8, is just the latest demonstration that these devices have a place in data collection and other key tasks. See also Sybase's Adaptive Anywhere strategy for connecting a wide variety of mobile users, including those using Windows CE-based devices, to its back-end servers. However, it's the decision of major handset players Motorola, Ericsson and Nokia to license Psion's EPOC 32 operating system that is the historic stage in the development of smart phones, which should create a strong platform for future development. That these companies are publicly backing an operating system that is fast, scalable and here right now is great news for anybody who believes the phone is the obvious place for all the data that is currently sitting on PDAs. If you need to carry a cellphone then you're already carrying around a decent slab of processing power with a screen. Further, there is a clear development path for GSM and exciting technologies such as Bluetooth for better local connectivity. All the forces are coming together to suggest that the smart phone - not the NC, NetPC or Java terminal - will be the next really important enterprise client. The ability to dial into the Internet, remotely access office work, make voice - or even video - calls is a compelling vision that will eventually make the PDA redundant and make the desktop workstation an increasingly rare sight.