The redoubtable Carl Reiner asked Mel Brooks’ 2,000 Year Old Man what the greatest invention of all time was. Brooks replied, “Saran Wrap.” It’s clear, you can wrap a sandwich in it and see if it’s still good …
When, in protest, Reiner asked about space travel and modern medicine, Brooks answered, “Oh, those are nice too.”
That pretty much describes my level of excitement for 802.11, the impending standard for wireless LANs Chad Dickerson writes about so enthusiastically. It isn’t that 802.11 is a bad thing. I just don’t see this as being worth a CTO’s time and energy.
LANs are infrastructure. Infrastructure is important but it isn’t strategic from a business perspective unless it enables a fundamentally new way of working. For example, 802.11 would be business strategic if a company used it to organize around ad hoc teams instead of a fixed organizational chart. Just imagine every employee with a golf cart instead of a cubicle, equipped with an 802.11-capable laptop and a wireless IP-telephony-enabled telephone. Instead of remodeling every time the company reorganizes, employees could park next to today’s manager and co-workers.
Infrastructure also isn’t strategic from an internal IT perspective, unless, that is, it lets you shift a significant share of the budget from maintenance to creating new business value. But even if 802.11 were fast and sturdy enough that it could eliminate LAN wiring altogether, it wouldn’t have a big enough impact to qualify.
And it isn’t. 802.11b, the more practical of the two (it costs less and has a 100 meter range) runs at 11 Mbps. 802.11a runs faster — 54 Mbps is possible under ideal conditions — but is limited to 20 meters.
Since the payload we deliver to the desktop continues to become bulkier, the 100 Mbps switched, dedicated, noise-resistant bandwidth that’s now what you get at dirt-cheap pricing in the wired world still looks a lot more attractive than wireless, especially since the placement of wireless transceivers is as much black art as it is engineering.
Don’t get me wrong. 802.11 has its place — in niche situations like trade shows where fast set-up and take-down are the dominant requirements for a network, and home networking where pulling cable means punching holes through walls and ceilings. And one of these years, a successor to 802.11 just might perform well enough to become the dominant network topology in most companies.
Even if it does, it will still be infrastructure. Most CTOs should have bigger fish … stored, of course, in Saran Wrap(R) … to fry.