> But what about cell phones and roaming? [...]
I suspect that this works only because most cellphones are not roaming
at any given moment. (To those who know - can the cellphone system
handle having more than a small fraction of the cellphones roaming?)
I also note that this works only because each cellphone *has* a "home".
Individual portable phone numbers might work if handled similarly, but
it would make service for an individual's phone number dependent on a
switch that the individual may not have any particular relation to at
all (and, depending on how numbers are assigned, may never have had).
If I lose service when the switch down the street is knocked out, I
find that a lot more tolerable than if, say, I lose service whenever
some switch off in Germany, which I've never lived near, dies.
Of course, PSTN facilities are supposed to have five nines of uptime.
Five minutes a year is probably livable, even if you square it (ten
minutes a year) because two mostly unrelated hops (caller -> home
switch -> current switch) are involved.
/~\ The ASCII der Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents.montreal.qc.ca
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B