Some of you guys are very confident about your predictions! - I have to say, I think time may tell a different story.

The problem with 11g isn't just compatibility with other 11g devices - I'm sure that will be sorted out. The problem is that it's 2.4 GHz, and quite long-range.

The result is that there is a real problem with overlap in offices.

In most office buildings in big cities, if you provided WLAN for every floor of a multi-storey building, you'd find that there simply aren't enough channels. No matter how you plan, you'll find someone else with the same channel covering the same square feet of floor space; and when this happens, coverage is TERRIBLE!

There are, also, an awful lot of 2.4 GHz devices. Video digital senders, Home RF, microwave ovens, Bluetooth, 802.11b and 802.11g all use the same spectrum - and there are others.

In /m6 someone suggested that 11a won't be ratified. It is, of course, already ratified! - I think the poster must mean that it won't become adopted. I think the poster is simply mistaken; I think dense WLAN requires 11a.

The reasons: there are a lot more channels, and 11a doesn't penetrate floors or walls as well as 2.4 GHz wireless. So you can isolate access points. Channel overlap ceases to be an insoluble problem.

For a nightmare description of just how bad 802.11b/g overlap can get, read one report from the German CeBIT exhibition last March! Would you use a WLAN that gave 10 kilobits per second throughput?

Finally, most suppliers of WLAN equipment are planning to move to dual 11g/a devices within the year. Some are using chipsets which do this (like Synad's) and others are building both standards into the AP. The next version of Centrino will include 11a/g clients, and Intel is going to promote 11a very strongly.

Guy