Or, to put it differently, there are cities where anyone with a fistful of dollars can sign some papers in the morning and begin tearing up the streets in the afternoon. That is not the case here. Roberts noted that Google has talked about "experimental construction" practices. Well, we don't exactly have an Experimental Construction Permit here.
"If Google is going to set this up as a contest for whoever makes it easiest for them, I don't know how we'll fare in that," admits Roberts.
On the other hand, Roberts said that San Francisco's rules and regulations might actually be a point in the city's favor. Google has expressed at least an ostensible desire that the city it works with not bend over backwards to accommodate it. If Google has to cross its t's and dot its i's, then the company would go a long way toward proving the feasibility of constructing fiber networks instead of the status quo.
One point against San Francisco, however, is it's too damn big. Google's guidelines called for a municipality of 50,000 to 500,000 residents (a city more like St. Louis, for example). So, rather than constructing a citywide network, Google's hypothetical San Francisco venture would be a neighborhood system. Roberts speculated that the southeast of the city -- the Excelsior or Bayview -- or portions of the Richmond and Sunset might be most applicable.
Finally, San Francisco would be unwilling for Google to build a system -- even for free -- and then own it privately. Roberts proposed an agreement, much like the one provided to City College, in which the city owns the actual physical system, but signs a long-term use agreement with Google.
When asked how much cost the city might incur, Roberts said it's too early to tell. That was also his answer when asked how much cost would be acceptable.
Photo (and Lego work) | Gayle Laakmann