Mountain View To Partially Replace Google Wi-Fi

itwbennett writes "Google launched the citywide Wi-Fi network with much fanfare in 2006 as a way for Mountain View residents and businesses to connect to the Internet at no cost. It covers most of the Silicon Valley city and worked well until last year, as Slashdot readers may recall, when connectivity got rapidly worse. As a result, Mountain View is installing new Wi-Fi hotspots in parts of the city to supplement the poorly performing network operated by Google. Both the city and Google have blamed the problems on the design of the network. Google, which is involved in several projects to provide Internet access in various parts of the world, said in a statement that it is 'actively in discussions with the Mountain View city staff to review several options for the future of the network.'"

oops

[slashmydots]

And yet, what is their plan to keep everyone on the wifi from being banned from everything everywhere? You get hundreds of people on one IP address from one gigantic wireless router and you've got a problem. One person does something stupid on slashdot, you're all IP banned. Last I heard, they don't send down individual outside IPs to everyone who connects. Even if they do, it'd shift around so much that it's basically the pay phone of the internet. You can commit any crime online and they'll never find y

Re:

[Chrisq]

And yet, what is their plan to keep everyone on the wifi from being banned from everything everywhere? You get hundreds of people on one IP address from one gigantic wireless router and you've got a problem. One person does something stupid on slashdot, you're all IP banned. Last I heard, they don't send down individual outside IPs to everyone who connects. Even if they do, it'd shift around so much that it's basically the pay phone of the internet. You can commit any crime online and they'll never find you because it's anonymous. Yeah, they can sort of track it but I can sort of fake my MAC address and laptop name too.

First of all there is no necessity to use a single IP address, they could route to a pool of addresses. When IPV6 comes along they could use individual addresses. Even if they use one address, they just won't ban it from Google services. For the reset "not our problem, contact the site admin"

Re:

[slashmydots]

About 90% of TOR exit nodes can't even use google at all without a verification code. About the same number can't access 4chan or Omegle or slashdot's discussion system or any major news site's comments box, etc x 100000. So that's real, actual proof that a single IP will get banned from just about anything because a handful of users will ruin it for everyone.

As for a pool of addresses, now it's like website ban roulette because you might hop on wifi and get an IP from the pool that's banned and you mig

Re:

[ONOIML8]

Wow! You would think that if this was a real problem it would have been reported, and killed the network, sometime since it was first deployed in 2006. How is it they've managed to operate all this time without being hindered by the problem you mention?

Re:

[slashmydots]

Who says it's working and people are happy? That's not what I read.

Re:

[gl4ss]

it doesn't quite work out like that in practice..

that is, plenty of places where you can get free wifi and only few countries in the world require even technically authenticating everyone who you give network access to.

anyways, just use vpn from the network. you should do that anyways. "but the vpn provider will be banned from everywhere1!!".. maybe, but again in practice doesn't work like that.

because you see not 100% of people are fucking idiots.

but apparently google is since they designed a network they

ADD -- Billionaire Edition

[rudy_wayne]

Google, which is involved in several projects to provide Internet access in various parts of the world,

There's a pattern here. They launch dozens of new products and then kill them a short time later. The roll out WiFi to their home town and then neglect it.

Advertising brings in 98% of their revenue. Everything else is just playthings for a company with too much money and no idea what to (usefully) do with it. Maybe instead of worrying about balloons in Africa you should work on the things in your own back yard..

Google is competing with local businesses by providing a multitude of services to its staff. Living in the shadow of the Googleplex is causing job losses and hurting rather than boosting the local economy.

Silicon Valley towns continue to suffer from terrible public schools and broken communities. East Palo Alto is a violent urban ghetto in every sense of the definition. And its smack-dab in the heart of Silicon Valley, right next to Facebook and Google.

Silicon Valley has all these "visionaries" saying they are "changing the world" yet they can't / won't change their own neighborhoods. The local schools should be showcases instead they are basket cases; the local communities should be healthy and thriving, but instead they are suffering from unemployment and all the other problems that communities across the country have to deal with on a daily basis.

What's the point in having these high tech giants in our midst when there is little or no advantage to the communities that surround them? They want special treatment; they want to pay little or no taxes; they would rather be a burden on their neighbors than ease the burden of others.

Re:ADD -- Billionaire Edition

[Albanach]

There's a pattern here. They launch dozens of new products and then kill them a short time later. The roll out WiFi to their home town and then neglect it.

Advertising brings in 98% of their revenue. Everything else is just playthings for a company with too much money and no idea what to (usefully) do with it. Maybe instead of worrying about balloons in Africa you should work on the things in your own back yard..

Well, doesn't the world just owe you?

Did you read the summary? They launched this network - presumably at the cost of significant time and money - seven years ago. It worked well for over five years. Given the lead time to design, implement and get approval for this network, it may well have been designed around 2004.

I very much doubt it has been neglected if it operated well for so long. It looks like demand is now beyond the capacity of the original network and that Google is addressing this.

How on earth does this make Google a burden? And why, precisely, should they stop caring about people dying in poorer parts of the world, so you can get a better free service?

Re:

[Anonymous Coward]

In defense of at least SOME of us libertarian wonks, we know all too well about corporate welfare and how that is, indeed, the real problem we face going forward. It's not because of the welfare queens, or the drug addicts. But we also know that someone who makes $100,000 a year isn't a 1%er, not by a longshot. The problem isn't even the 1%, it's the top 1% of the top 1% that skirt all tax and financial laws and become mega-ultra-super-filthy rich at the expense of everyday people. When people scream "r

Re:

[lymond01]

The problem isn't even the 1%, it's the top 1% of the top 1% that skirt all tax and financial laws and become mega-ultra-super-filthy rich at the expense of everyday people. When people scream "raise taxes on the rich" (which even as a libertarian, I agree with the sentiment...it's the definition that I disagree with), they mean some schlub and his wife who manage to pull in $120k working two jobs. Those people aren't the problem, it's the people pulling in $120M that aren't paying their fair share.

Sort of

Re:

[interval1066]

Well, doesn't the world just owe you?

I guess I'm of the old school, when I say I'll do something, I do my level best to do it. I guess I'm not down with the new norm of "I'll deliver what I promise as long as its convienient."

Re:ADD -- Billionaire Edition

[Sabbatic]

Minor point, but having lived in Mountain View all this time, I missed the part where "it worked well for over five years." It never worked very well, except sometimes in a few spots.

Re:

[pchan-]

As someone who lives in Mountain View, I'd like to second this post. Google WiFi has never worked well. In my experience it hardly ever worked at all. I'd be happy to be rid of it.

Re:ADD -- Billionaire Edition

[NatasRevol]

Shocking that a bunch of ridiculously rich 20-somethings are self centered.

Re:

[NatasRevol]

Yeah, guess who's in charge?

The still ridiculously rich people.

Re:

[JoeMerchant]

>What's the point in having these high tech giants in our midst when there is little or no advantage to the communities that surround them?

What a completely communist statement! Next you'll want them to start paying their fair share of taxes.

It never worked

[Anonymous Coward]

The summary concocts a wonderful fantasy. I live in mountain view. It doesn't work near my home, at the Caltrain station with nobody there, or on Castro. You connect and you're lucky to get an IP. If that actually works, you'll wait for a few minutes to get to your first page.

Re:

[Anonymous Coward]

same here, live in mountain view and can attest google wifi is a piece of garbage. give us google fiber, not this garbage wifi. here's a company that only knows how to take from the city, and never give back.

Re:

[bandy]

And what makes you think that Google Fiber will actually be any better? Yes, I've used the Google Free WiFi in Mountain View. It was crap. What's now known as 1X celluar data service was faster and more reliable.

Re:

[tji]

Yes, I would also like Google Fiber. But, to say they never give back is a bit harsh.. Google wireless was an attempt to give free WiFi. It sucks, and I never use it. But, it was a legitimate attempt. I'm sure they discovered that to do it right it would be prohibitively expensive, and so they let it languish. But, they're not a charity.

Re:

[Dishevel]

The free shit you gave me for a long time now is beginning too SUCK! All you do is take from me and the free shit you give me is no longer good enough!

I really feel bad when I see how selfish and self centered the American public has become in general.

Re:

[tji]

I can confirm. There is an AP on a light post just across the street from my house. When I first moved in in 2008, I tried to use it and found it to be very poor. Trouble getting an IP, trouble authenticating. It was slow at low times. In the evening, it was unusable - I'm guessing because of overload at some point in the network.

Re:

[vriemeister]

When I was in a one bedroom apartment in Mountain View I tried google and could never connect to it. I thought "What a pile of crap". Then I set up my own wifi router. I couldn't even connect to that from an adjacent room half the time. If you're in any sort of apartment complex all the (now old) 2GHz baby monitors and 2GHz phones absolutely killed any signal that has to go more than 30 feet or through a wall.

Re:

[operagost]

Use the 5 GHz wireless N band. Yes, this probably means buying a new router (if it says it's more than 300 Mbps, it must support 5 GHz) and WLAN card. I bought a new laptop just last year, and someone saved $1 by putting a card that only supported 2.4GHz 802.11n in it.

It used to work for me

[fredz]

One downside of free WiFi is that there is no one to call when it stops working. I live in Mountain View, and when Google WiFi first came out it was good enough that I dropped my DSL. I was very happy with it for about a year, but then it got intermittent at my house, working well on some days and not at all on others. After a few more months it wouldn't work at all from early evening until after midnight, presumably because everyone else was also trying to use it at that time. After that I switched to Com

Wi-FI

[hjf]

Large-scale wi-fi? I guess they should call that "that big ugly thing that just does not work". WiFi is difficult to make work properly in a big house, and they tried to make it work in a city-wide scale. I just never understoond the point of making public wi fi hotspots available for free. In 2006? Maybe. In 2013? Who uses that? 3G and 4G services have the potential to provide a much better service than a half-assed wi-fi.

And i'm just guessing here, but I'm pretty sure this is a "mesh" of AP. Back in the early 00s, WiFI meshes, or WDS, were supposed to solve all connectivity problems. People would just have a big mesh of APs that would cover from one house to another and some internet pipes here and there. Guess what? Meshes just don't work. They aren't reliable. If they are reliable, they're also unusably slow, and don't scale with hundreds of users.

If a city wants to give free internet acces, why bother deploying wi-fi? Wouldn't it be a good idea to just subsidize 3G/4G access and make it free for everyone? Before you all jump at me claiming that wi-fi is free: remember it's not. Your $30 AP can't work in outdoor conditions. Cases are expensive, heat makes the AP crash. You need to wire some of them at least. And heat also kills the APs in just a few years (just look at the Cisco outdoor APs which look nothing like toy APs. And cost thousands each). Deploying and maintaining a Wi-Fi network made of cheap APs in a mesh probably costs the same in the long run as using very expensive APs. And why throw away so much money into that? I know Mountain View city has money to spare but i'm sure there are other things where the money could be used, rather than providing "free wifi".

Re:

[laie_techie]

I just never understoond the point of making public wi fi hotspots available for free. In 2006? Maybe. In 2013? Who uses that? 3G and 4G services have the potential to provide a much better service than a half-assed wi-fi.

Most Americans have data-caps on their wireless plans. Our connections get faster and faster, while our quotas shrink.

Re:

[interkin3tic]

Physics question: is the part of the spectrum we use for 3 and 4G inherently better at long distances than the part we use for wifi? It would make sense, but I could also see the problems you suggest being due to economies of scale and lack of R&D rather than an insurmountable problem.

(not to imply that the problems with wifi need to be insurmountable for it to make more sense just to go with 3 and 4G.)

Re:

[xtal]

3G systems use licensed spectrum. The power levels are much higher, they're typically in good spots for coverage, and the receive antennas are high gain (and very expensive). As nobody else is using the bands, the noise floor is very low. This represents much of their advantage.

The bruhaha over 700MHz is that it does travel very effectively through things, but this isn't always what you want for high bandwidth. More APs is better as this enables you to reuse spectrum. Further propagation increases the probl

Re:

[tech.kyle]

An accurate reply. Lower frequencies do push through physical obstruction easily, but require a large bandwidth to get any real throughput. I've dealt with some 900 MHz equipment and have had trouble getting anything more than 2 Mbps through it in in ideal situations.

Re:

[hjf]

Ubiquiti user here. NanoStation M900. 10mbps on a link 1km at 9m off the floor, in a city with 3-4 8-story buildings in the middle. Not bad at all.

Re:

[tech.kyle]

As a Ubiquiti user as well, we haven't tried their 900 MHz offerings, but I'm starting to wish we did. I'll have to make a better push to get some in.

Re:

[hjf]

No, please. Stay off 900MHz... i like my noise floor at -105dB ;)

Re:

[chuckinator]

Yes, it is. Wifi operates on the 2.4 GHz and 5 GHz unlicensed Industrial/Science/Medicine (ISM). The higher the frequency, the closer the radio horizon and the lower the tendency of the signal to penetrate through solid materials. Cellular radio tech operates on 700, 800, 850, 1700, 1900, and 2500 MHz. The higher frequency cellular tech is typically used for higher speed data links (3G, 4G), so it degrades some at farther distances than the voice links but will go back to older operating modes when nece

I live in Mountain View...

[HockeyPuck]

It's never really worked. The main downtown area, Castro Street, service is abysmal. If I can get an IP address, page load times are terrible. On the other hand, turn off the wireless on my phone and go with 4G and surfing is nice and snappy.

This was a PR stunt by Google, they were never serious about this. Or the old addage, "you get what you pay for". Given all you needed to access it was a google account, of course they're going to do the bare minimum. Conversely, with Vz/ATT, you're paying for service.

Re:

[Anonymous Coward]

My mother lives in Mountain View and I agree - Google Wi-Fi was never usable when I visited. I've tried outdoors in the middle of a park and always gave up in frustration. And people should know that it definitely doesn't work for you indoors. The gear that they recommended for getting the signal indoors was pricey too. I doubt that any businesses (E.g. cafes) invested in that, because their customers would be really frustrated trying to use it.

Or

[ProudParanoid]

This was a PR stunt by Google, they were never serious about this. Or the old addage [sic], "you get what you pay for".
.

How about: large Internet providers losing business to free try to undermine free with an extended DOS attack. Or the old adage, "follow the money".

Re:

[Sabbatic]

There's also the old adage, which too many people never manage to understand, that just because it doesn't cost the user money, that doesn't mean it's free.

Re:

[thetoadwarrior]

A lot of Google products are half baked and don't work that well, imo. I'm guess that's why the like calling things beta because they don't have to do a great job.

Re:

[Ksevio]

I used it around 2009 and it was great for walking around with an iPod touch. It worked best around downtown and the train stations, but even out a bit more I got good signal. Wasn't the fastest out there, but it was enough for basic browsing and google maps

Early design that was never updated

[chipperdog]

It looks like their system was based on Topos MetroMesh in-band mesh backhaul, so someone using 1Mbps on one access point is tying up 1Mbps on multiple APs. This was at a time when 256 Kbps could have been considered "high speed". Not sure how many APs are connected to a fiber/copper/microwave backhaul, I guessing just a few to cut costs, but it would seem to me that as many APs as possible should have an out of band backhaul (ideally fiber to every AP)...Also APs antenna/RF might have to be tuned down to reduce coverage and a higher density of APs is likely needed with the increased usage - or multiple APs with sector antennas ( 3 x 120 degree, or 6 x 60 degree) are placed at locations instead of single AP with omnis - I'm thinking something LIKE the Ubiquiti NanoStations [ubnt.com] could be placed around a light pole to act as combination AP and 60 degree sector antenna and be more ascetically pleasing than the box with whips attached found right now, electronics to feed the NSs could fit inside most poles at the base, or in a small box near the AP/Antennas.

Re:

[tech.kyle]

Sectors on a tower going down to specialized hardware for the client is the typical WISP setup. NanoStations don't have much of an E-plane and must be carefully aligned, not to mention the integrated antenna has relatively poor attenuation (a NanoStation Loco shoots a 40-60 degree cone, but has even worse attenuation). A more common setup would be a Ubiquiti Rocket with an external sector going to 802.11 compatible clients (be it another high-power, long-range 802.11 device or a regular ol' laptop/phone).

Re:

[chipperdog]

That's why I emphasized the word like...Although the antenna performance is better, the rocket plus traditional sector antenna might be an overwhelming size for the typical street light, the NS and LOCO are a size where neighbors might not complain about the "junk" on the pole....I did see someone present once on a saucer shaped access point unit [xirrus.com] that was actually 12 (?) access points with tight sectors...If they have an outdoor version that could sit on the pole top...

WISP

[tech.kyle]

As an employee of a WISP, the 2.4 GHz public band is horridly congested. 5.8 GHz isn't as bad, but you need near-perfect line of sight. Next one up is 24GHz which is too expensive to implement at this time.

Mesh networks suffer from throughput issues as each packet needs to take up extra air time to be retransmitted X number of times. Even the good 5.8 GHz implementations peak out at about 300 Mbps half-duplex (~150 Mbps actual IP transport), assuming very high RSSI (read: expensive equipment), shared among

Re:

[chipperdog]

Have you every used 3.65 GHz? Licensing is pretty easy (when the FCC is open again), and one would think it would be less congested,

Re:

[tech.kyle]

We use it for important links, in congested (urban) areas, or, of course, when we need a really good SNR to keep modulation high. I'm not involved in the licensing process, but I don't know why we don't use it more often.