It's even less hard to believe when you look at what much smaller companies are doing in this department. Rovio (maker of Angry Birds) already provides free Wifi to a larger area of downtown Helsinki than Google is rolling out here, and they're a company that is 1/300 the size of Google (measured by revenue). That hasn't exactly bankrupted Rovio, so it's plausible that, to a first-order approximation, Google could provide free wifi to about 300 U.S. downtowns, if they really wanted to, without stretching themselves that far.

I think it's more reasonable to say that it's worth it for Google to put free wifi in parks if a few of their engineers get work done because of it.

They already buy us mobile broadband, so we can do work without needing to be near WiFi :)

What's the point of having wifi on those buses then?

Convenience, I guess?

Have you forgotten Google's history? i.e. StreetView cars guzzling down Wifi access point data.

Is that wrong? It allows them to offer location based services without making people use GPS. This is a service that you can disable, by the way.

I don't see anything wrong with collecting SSIDs that are being broadcasted onto public roads.

But to get back on topic- Google is doing a good thing here and should be commended for it.

Have you forgotten that everyone does this as it's the cheapest way to provide accurate location service on mobile devices even if you have AGPS?

If you don't want your SSID or Wifi public, disable SSID broadcast in your AP and turn on WPA2. The biggest threat to you isn't Google slurping up your SSID/Mac Address, it's some dude standing stationary outside your house capturing everything you do.

If you want SSID broadcast but don't want to be indexed, add "_nomap" to your SSID and Google will ignore it. But Skyhook, Apple, and others probably won't.

Just did the math (based on 2012 revenue and reported # of employees), and yup, $600k is what Google could easily afford for a single employee (with a good $400k left over for business expenses, per employee).

That's amazing.

It's not a tight bound, I used a general rule of thumb that a fully loaded employee is 2x their base compensation. A good google engineer can easily earn $200+k per year in salary + bonus + stock. Obviously the 2x rule of thumb doesn't scale as health care, insurance, and other costs don't scale up, but I'd estimate that given the perks at Google, there's at least another $50-100k overhead. So realistically, it's probably 2 seasoned engineers, or 2 seasoned engineers plus a fresh graduate.