Without replacing these poles, the city would be unable to use the municipal services from which it still plans to save $3.5m over the 10-year contract life, and thus it would be pennywise and pound foolish to leave the status quo.

Today's announcement sees Meraki nicely footing the bill for extending their service into neighborhood affordable housing, municipal-speak for low-income housing that's subsidized typically through government efforts and funds. Meraki will install networks at 12 buildings in the Tenderloin, known as San Francisco's roughest neighborhood, now going on many decades with that designation.

Meraki claims a "presence" in 42 of 52 major neighborhoods in the city, although their map tells a very different story about how usage is clustered in areas in which it would make perfect sense that usage was seen.

Meraki has engaged in a very interesting public project, and likes the imprimatur of San Francisco, even as they don't really need the city; the city, in contrast, needs them (or Newsom particularly) to salvage something from years of planning that blew up in their faces.

Anyway, SF's EarthLink network would never have been built; or, having been underway, would never have been completed.

Forgive my snark tone and cynicism: Meraki has put a lot of resources into building a publicly accessible network across a hunk of SF that wouldn't otherwise exist.

This is a critical distinction. Telecoms are covered under the Telecom Act of 1996 that requires non-discriminatory access to utility poles to avoid incumbent local exchange carriers (ILECs) and utilities from being gatekeepers that prevent competitive service from emerging. There are a series of tests in the law and local qualifications, too, that allow a firm to be a registered telecom. An FCC decision last year ruled that companies that mix telecom and unregulated information services on the same wires aren't disqualified from getting the Telecom Act deal, however.

But E-Path seems to meet none of the criteria except their desire to pay $10 instead of $50 per year per pole. Utility poles have held up many other municipal networks. We're not hearing more about them these days because such networks are now being built on a smaller scale for different purposes, where the number of nodes and their placement is rather different than networks built with the intent of providing indoor coverage.

Cablevision, by the way, qualifies as a telecom, this article states, which helps them in placing nodes for their planned $300m network across their coverage territory. They can also mount nodes in-line with their cable lines, using power from their cable plant on the lines already.

E-Path appears to have a variety of communication problems as well. The article notes, "Tortoretti said his Washington, D.C., attorneys disagree with LIPA's interpretation. But the attorney Tortoretti said represents E-Path, Charles Rohe, said he couldn't speak about the company or the dispute."

Later, E-Path's "chief executive said he hopes the county will help with his LIPA dispute." But an aide to the Suffolk County executive said, "That's not really our issue. That's out of our control."

Correspondent Craig Plunkett, quoted near the end, points out that if the counties were to change their minds and want to buy services on the network, the proposal would have to be rebid (appears as the sound-alike "rebuild" by accident in the online article at this moment).

What's been clear is that after spending more than $3m, the city couldn't acquire more than a few hundred regular subscribers, about 10 percent of the point they'd need to pay expenses and pay down capital outlay. But it turns out that the backend was as important as their network deployment, IDG News Service reports.

The latest city network administrator brought in Aptilo Networks for backend authentication and session processing, opened the network to 15-minute free trials, and started accepted ad hoc payment. The new network guru also let outsourced contracts expire and brought customer support and other services back in house to reduce expenses and improve the feedback loop. He discovered their existing authentication system was licensed for 500 users, so that might have explained their failure to grow, too.

The city now has 1,000 regular users at all levels, from pay-as-you-go to monthly household subscriptions. They've revised breakeven down to 2,000 subscribers, and say they are breakeven for expenses.

The other problem Lompoc had, by the way, is that the cable and telephone companies didn't sit still. I exaggerate, but when Lomopoc was planning its network, it had very poor coverage for its 42,000 residents for DSL and cable modem service. When the Wi-Fi network was announced, the incumbents started pulling copper, coax, and fiber, and dramatically improved network coverage. The $3m wasn't entirely ill spent so far: it was a kind of reverse incentive to the private companies to get their act together.

The city would like to avoid congestion pricing and tolls to manage traffic better. The system would allow parking pricing and durations to change dynamically. San Francisco is investing nearly $100m in an overall congestion reducing program, SFpark. This article cites an expert who estimates 30 percent of core business district traffic is from folks searching for a parking spot.

An embedded device with a 5-to-10-year battery lifespan relies information about parking availability and traffic speed through a mesh network.

It's unfortunate that such applications weren't in place when San Francisco was thinking about Wi-Fi public access. The intelligent integration of necessary city services that require a wireless backhaul with a public access Wi-Fi network could be a viable model. But early RFPs were focused entirely on public access and SF's contract with EarthLink excluded any linkage between the public Wi-Fi network and any municipal business.

One of the projects leaders uses terms that should warm every New Yorker's heart, if he or she knew what they meant. IT head Paul Cosgrave says the system will overcome silos, an often disparaging term for the separation of resources across groups that can only expensively be overcome. It's the government and business equivalent of the academic problem of a lack of cross-discipline focus.

One of the first applications allows sanitation workforce managers a frighteningly precise amount of knowledge about routes, activities, and behavior of trucks in their territory. Let's hope that's not misused! Efficiency is one thing; micro-management is another.

Another project is testing wireless water-meter reading. The city hopes to spend $90 per meter for the upgrade and shed part of a $12.2m contract with Con Edison that covers 850,000 units. What should be useful about this is that problems can be detected by monitoring waterflow patterns, which in turn allows the often huge problems that take months to notice (occurring underground or in basements where rivers formerly flowed) to be stopped before they turn into multi-million-dollar problems for property owners or the city. Anytime anything happens in Manhattan, it's a multi-million dollar problem.

Oakland County, Mich., project officially "on hold": For "on hold," read, "never going to be built." The pilot area in seven communities has been turned off, and MichTel has been unable to obtain the $70-odd million they project needed to build out the county-wide service. The state's ongoing reliance on the automotive industry makes it a hard sell to commit public dollars in advance of a return on those dollars, too.

Virgin Mobile buys Helio: The last vestiges of EarthLink's three-pronged approach to fighting the wireline monopoly appears to be at an end. EarthLink pushed its 50-50 partnership with SK Telecom in mobile virtual network operator (MVNO) Helio as one prong; its municipal Wi-Fi division as another; and its DSL business as a third. The muni division is nearly out of operation, and DSL lines continue to fall in quantity quarter over quarter. Dial-up is still their cash cow. Helio lost hundreds of millions to obtain just 170,000 subscribers (that number down from 200,000 at the start of 2008). EarthLink will receive a pittance for its investment, part of the $39 million in stock that Virgin will pay for Helio; SK Telecom will invest in Virgin Mobile to obtain a total 17 percent state. Virgin itself makes just a very tiny sliver of profit. MVNOs buy minutes and data from carriers, and Virgin Mobile involves Sprint as a partner, making it the only tolerably successful MVNO.