Firstly, to convince you that Wi-Fi is this year's network essential, just look at the efforts of people like HP and Nortel. They are desperate to make sure they are on board as the bandwagon moves off.
The pattern here is to have a set of access point products, but also to have a network-aware security-and-roaming switch to link them up in the wiring closet so that users can roam round the building, without confusing the access points, or the business applications you are talking to across the wireless network.
It's a pretty clear strategy. The enterprise system is bigger and more complex, and therefore has more margin than the individual access points that otherwise get sold. The wireless switch ties it into the space these vendors think they own--the wired network--and that is good for the vendors. And also, the enterprise is one sector of the market that still (so we are told at any rate) still has money to spend.
You want more evidence for the anti-gravity ability of Wi-Fi; its skill at remaining buoyant while every other aspect of IT is falling to earth? How about start-ups? Where else but in Wi-Fi could you expect to still see start-ups, funded by venture capital, such as Trapeze? And new service provider roll-outs, such as The Cloud, which plans to spread over the UK (as if our skies did not already have enough clouds).
VC-funded start-ups, studded with big-name executives, are a rarity these days. When I met Trapeze last month, I kept having to pinch myself to make sure I wasn't asleep, and dreaming of the halcyon days of the late 90s. But it seems that the idea of making Wi-Fi work in the enterprise, unlike any other idea in IT, is still hot enough to get VCs to fork over cash to back a business plan.
And service provider roll-outs are even more out-of-fashion. Anyone at the Cloud launch this week is going to be in a similar state of confused déjàvu. Isn't this a re-run of the old DSL, WAP or 3G service-provider launches? Surely the failures there have killed this kind of thing off?
With all this going on, it is tempting to treat Wi-Fi with the same kind of naïve optimism with which we greeted 3G, Web-based B2C, and all the other great bandwagons of recent years. But it is worth looking at all these developments with mature eyes.
There are differences. But do they show companies that are learning from the mistakes of the past? Or are these companies that would love to repeat the mistakes of the past, but don't have the money any more?
Take Hewlett-Packard. Its networking heritage all comes from the HP part of the company, not from Compaq, a company that--because the networking expertise it inherited from Digital Equipment had already been frittered away to nothing--was happy to sell network equipment from other vendors.
HP makes its own network switches, and has been taking market share from other vendors. Like the company's printers, they're cheap, reliable and functional (want a 24-port Layer 3 switch with two copper-Gigabit uplinks? HP just launched one for Ã,£467!). And unlike the printers, in networks you don't have to pay over the odds for consumables.
But with Wi-Fi, HP has had to go against this heritage, in order to get a product out quickly. With HP's heritage, you have to remind yourself to ask the question you automatically ask any network vendor: "Is this your product or just a reseller deal?" And in wireless networking, it turns out HP is doing the latter.
HP is selling Proxim's access points, and network switches from Vernier. There are good time-to-market reasons for this, and HP has been promising to do its own products for some time. However, it seems to be taking a long while, and the deal with Vernier suggests a long-term commitment to third-party products.
Now, this is different from Nortel, which also used partners, notably Symbol Technologies. But as it moved into wireless more forcefully recently, it appears to be using its own product. It certainly has been thinking about the architecture that wireless networks require, as it made a decision to break away from Symbol's "dumb-access-point" strategy, to come up with its own ideas.
It is quite easy to side with the company making its own technology here, but there are reasons behind the differences. Nortel has a history of covering all the network bases, while HP has always filled in the gaps with third parties -- it has a long-standing relationship with Foundry at the high end, one which HP's product are a few years' away from replacing.
Is HP finding it tricky to develop its own wireless networking products? And does that mean the "invent" part of HP's image is harder to maintain in a business environment predicated on speed to market and Dell-like delivery systems? Or did the company make a cool-headed decision that the wireless network market is never going to be as big as the wired-network market, so it should not get the lion's share of the company's limited R&D dollars?
This sort of pragmatism is everywhere in the Wi-Fi world, and I welcome it. The people at Trapeze may sound like the serial start-up jockeys that they actually are, but check one thing out. Unlike the hip start-ups of 2000, they actually have products to sell. They haven't come to the market with a proposition waiting to get bought up.
And while you can be sure that the Cloud's launch party may spread a haze of feel-good fervour, the bottom line will be worked out as closely as possible. Remember, that that service provider is based on broadband links to games machines in pubs. Three years ago, in the dot-com frenzy, it would have been talking of rolling out networks from scratch.
So let's all keep our heads shall we? Wi-Fi is fun. It is the Next Big Thing. But let's see to it that this revolution will not be oversold.
