The following text is copyright 2003 by Network World, permission is hearby given for reproduction, as long as attribution is given and this notice is included.

Building the new on the bones of the old

By Scott Bradner

This is a belated comment on a short article in Fortune back in January about a Bell Canada trial of turning old pay phones into WiFi access points. Definitely, a Duh moment. Boy does that idea make sense, well, maybe it makes sense, but it sure should make sense.

The idea is to swap out one of those mostly-idle-because-everyone-has-a-cell-phone public pay phones for a WiFi access point and give folks loitering nearby a chance to get on-line. So far, the Bell Canada trial is offering free Internet access through the converted pay phones but I expect that will not last, Bell Canada is a phone company after all and someone has to pay for the swap.

This idea has almost everything going for it. (I wonder how many patent applications have already been filed on this idea.) Bell Canada already has the locations and permission to use them for communications. Pay phones are already in all the best places (and quite a few of the worst). The locations already have phone lines that can be used to support DSL for reasonable speed Internet access. You can even do DSL link sharing and keep the phone (to support people who forget to charge their cell phones). Some of them will already have power.

Bell Canada's web page on the trial (http://www.bell.ca/accesszone) shows a picture of a quite neat looking access point, that they call a "hot spot box," that is protected by a transparent plastic shield. I cannot tell from the picture how big the thing is but it does look like it might resist the attentions of vandals -- up to a point.

About the only thing wrong with the idea is that the phones belong to a telephone company. Bell Canada seems somewhat cluefull when it comes to data networking in general (they have worked with the research data networks in Canada for years) but it remains to be seen if that translates to being cluefull enough for this type of network. Imagine, if you will, your local Baby Bell (shesh, those babies are getting big -- and some of them are getting broke in a big way) trying to install, price and run such a service.

And what happens when people start getting mobile phones with WiFi interfaces and VoIP support? Or, when people start using the VoIP stacks that are already in most laptops? You can bet that the phone company will suddenly realize that they are offering a service where a main use is telephone company bypass. Thinks good get a bit funky when that realization happens, maybe they will try to pull a Panama and try to block the VoIP applications. (See A Panamanian platypus - nww 11/18/02)

The basic idea should be a real good one. Instant, well almost instant, ability to deploy WiFi to many hundreds of thousands of places, with most if not all the infrastructure already existing. There is just this nagging worry about the clue density of the owners.

disclaimer: The US News report that was just published implied good things about clue density at Harvard but the above reflects my own not the university's.