Some of this difficulty is unavoidable when organizations are dealing with people connecting with a wide variety of equipment in a wide variety of circumstances. (Think Mac OS X and Apple’s limited hardware vs. Windows dealing with the vagaries of PC hardware.) But other technology challenges are easily remedied through decent design and forethought.

Case in point: connecting to the AT&T WiFi at Starbucks. I am an irregular Starbucks customer  —  but enough of one to have a pre-paid Starbucks card. Occasionally I use wireless provided by AT&T to check my email or do some reading while I drink my coffee and, in fact, the consistent existence of wireless is a reason I choose Starbucks.

(I suspect this use makes me pretty much the target audience for Starbucks, even if my technical knowledge  —  I am a former Web developer and systems administrator  —  perhaps gives me more familiarity with the underlying technology and design issues than some other customers.)

I have no trouble connecting to the 802.11 signal (“attwifi”). (My technology works fine.) It isn’t until I actually am connected that the trouble starts.

First, I’m greeted by a Starbucks screen with an orange AT&T bar at the top. There’s a box for a user name, a drop down confusingly labeled ”Make a Selection,” a password box, a checkbox for me to agree to term of service  —  and then two buttons, one labeled “Connect to the Internet” and the other labeled “Log In.”

OK, well, I’ve done this before (though it’s been a while), and I believe I already have an account. So even though what I really want to do is connect to the Internet, I’m going to try filling in my login info and clicking log in first.

So I put in my user name, then hit the dropdown box. There are 14 options, none of which say “Starbucks.” Hmm. Well, this is AT&T WiFi, so I guess I need to select that one. But wait, there’s three choices labeled AT&T! Which one am I? I suppose I’ll just try the first one, and see if it works.

So I put in my password (I think it’s the right password  —  it’s been a while since I last used the service), accept the terms, and click on the log in box (I hope that’s the right thing to do).

Instead of connecting to the Internet, I instead get a screen telling me that my email address hasn’t been validated, and I won’t be allowed to continue until it is. Find, it looks like I can send a validation email from this screen  —  but wait, I can’t connect to the Internet, so how am I supposed to see the email?

The screen also suggests that I must not have validated my email address properly before  —  but as far as I can remember, I’ve never been sent a validation email from AT&T anyway! So I guess, despite being a Starbucks customer, I don’t get to use the WiFi today, and won’t be able to until I get home and track down this validation email? What if it never comes? Do I have to return to Starbucks to try asking for another one?

Perhaps I logged into the wrong account. After all, I am a Starbucks customer, not an AT&T one, right? And I have a valid pre-paid Starbucks card (which I just used to buy my cooling coffee.) I’ll try going to Starbucks.com to see if there’s something about a Starbucks account or another way to login or access my account.

Starbucks.com only partly loads (why?!), but at the bottom there’s a link to “your account,” so I’ll try that.

Great, a login screen that’s branded with the store I’m actually in! My login works, but I’m not presented with much about WiFi. Still, WiFi is mentioned near “Account Management,” so I’ll try that.

OK, now on the left is “WiFi Info,” which sounds promising. Now I get a link to “create an AT&T account,” so I guess I’ll try that (do I already have one?) It says my prepaid card is my “ticket to free and easy Wireless access at Starbucks,” so that’s a good sign.

Unfortunately, the link appears to be broken. Hmm. I see two other laptops here online, so I guess I’ll try reloading. OK, that worked, now I get a screen “Starbucks Card Rewards with WiFi.” That sort of works. Now at least I have a “Sign Up” button to try.

This looks positive. I can sign in with my account on Starbucks.com. That sounds good, since I know that work, right? I just used it, after all.

Unfortunately, it doesn’t work. When I try to put my information in, I get the confusing response  —  presented in red as an error message  —  that I already have an account. Um, yeah, I know I already have a Starbucks account, that’s what I’m using! (Oh, you mean, I already have an AT&T account? Why don’t you say so! And that isn’t very helpful.)

Now what?

I give up and go home, not wanting to return to Starbucks again, and not interested in dealing with AT&T ever again.

In United States v. Ahrndt, No. 08-cr-468 (D. Ore. Jan. 28, 2010), a federal trial court held that a child pornography suspect had no constitutionally protected privacy right in the files found on his personal computer, stored in a shared iTunes folder fed by a Limewire account, accessible by a neighbor who was piggybacking on his unsecured wireless network.

via TechLaw: Court Finds Constitutional Significance in Defendant’s Failure to Password-Protect Home Wireless Network. (I recommend you read the whole piece by Thomas O’Toole.)

The neighbor connected to the open wireless account, opened iTunes, and found child pornography shared by the (now convicted) defendant. A police officer who responded to her call guided her through opening one shared file, and saw child pornography. The police them proceeded to gather details of exactly who’s network it was and obtained a warrant  —  in turn discovering more child pornography.

The defendant argued it was an illegal search by the police and the warrant should be thrown out, since the initial finding by the officer violated his 4th amendment reasonable expectation of privacy.

The court said, no, if you leave your wireless router unsecured, your expectation of privacy is much lower. People routinely use unsecured networks of others, says the court, and setting a password to prevent this is clearly laid out in the instructions (in this case, Belkin’s). What’s more, if you have iTunes sharing turned on, you shouldn’t expect those files broadcast to everyone on your network to be private.

I agree with O’Toole that none of this breaks new Constitutional ground. It all seems perfectly reasonable to me, and it seems like the judge “gets it” with regard to technology (nice to see).

It does also imply that an open wireless network is not like a closed (but unlocked) door. Accessing it is not equivalent to breaking and entering. (I think this is the correct view of things, myself.)

O’Toole ends with a warning that you should password-protect your network if you want 4th Amendment protections. I would add that, if you choose to share your wireless network, then make sure you secure your computers within that network. The lack of a password for the network was only one factor  —  sharing iTunes files publicly also contributed.

Extending this argument, if you have private files, you should be able to get 4th Amendment protection by password-protecting them at any level (encryption is a practical protection, but shouldn’t be necessary for legal protection  —  if one extends from this ruling). Thus, you could share your iTunes library and  —  provided you password protect your financial documents  —  still have Constitutional protections for those documents (but not the music or videos you are sharing).

Seems common sensical, but that doesn’t mean that it’s necessarily “the law” everywhere! This was only a federal district court in Oregon, after all, and isn’t binding precedent (though it may be persuasive). But it’s an intelligent decision.

SSRN-The Failure of Public WiFi by Eric Fraser:

This short piece describes the failure of the widespread plans to provide public wireless internet access. It identifies three interrelated types of causes for the near-universal failure of these ambitious plans: regulatory, technical, and economic.

As the article points out, WiFi – while incredibly successful for constrained-area use, as in homes and businesses – has not so far proven very successful when deployed across larger areas. There was once talk of police agencies and similar being able to switch to lower cost, more flexible Wifi-based systems for use in cars, but that has generally not happened. Instead, many are using 3G cellular systems, which at least are also commercial, off-the-shelf sort of solutions (King County, Washington, for example, puts laptops with 3G wireless cards in its police cars).

Google’s experiment near its HQ is an exception, and I think “Free the Net” in San Francisco might qualify as well. Neither are municipally funded, however, and the model has not really been exportable across larger areas.

Ultimately, WiFi is an small-scale system, ideal for its purpose, and fun to extend to larger areas. WiMax and 3G are much more scalable, and do not need to rely on so many access points. The density required is simply too high to cover large areas – I have sometimes had trouble covering an entire house effectively, much less a city block!

On the supply side, the systems simply could not deliver what proponents promised. Because of WiFi’s technical and regulatory limitations on frequency and power output, blanketing a city proved to be prohibitively expensive. Outdoor areas and a few buildings could be wired for wireless access, but no one could deliver anywhere-internet using WiFi. Signals from streets could not penetrate large buildings, and property rights prevented municipalities from installing the required tens of thousands of access points inside private buildings throughout a municipality. As a result, public WiFi networks could be used indoors in only a few areas, or in many outdoor locations.

Sadly, 3G especially comes with costs that many had hoped WiFi might overcome. It is a “top-down” networking solution, provided by big companies, and comes with significant bandwidth charges. WiFi seemed like a wonderful, “bottom-up” approach that leveraged existing wired bandwidth without added additional wireless costs. It still has potential in many circumstances to be useful, whether it be for cafe settings, easily-deployed home networking, in rural areas without signal-blocking obstacles, or even unlicensed long-range point-to-point networks.

Fraser’s article provides more in-depth analysis, including an explanation of the physics at work, the business models, and more ideas about why public WiFi has not succeeded. I recommend it.

(Thanks to the Legal Theory Blog for pointing me to this.)

Related articles by Zemanta

• Sprint, Clearwire and WiMax
• Meraki “Free the Network” San Francisco Project Grows
• Two Approaches to Community WiFi Sharing

Webware – Starbucks ditches T-Mobile for AT&T:

Ubiquitous caffeine conglomerate Starbucks has ended its Wi-Fi partnership with T-Mobile in favor of one with AT&T.

Under the earlier plan with T-Mobile, Starbucks customers needed a paid subscription to access the in-store Wi-Fi service, and T-Mobile HotSpot subscribers will continue to have access to Starbucks Wi-Fi thanks to an agreement between AT&T and T-Mobile. But the new AT&T plan allows all customers 2 free hours per day, with a $3.99 fee for additional 2-hour chunks of time. Monthly subscriptions will cost $19.99 and will enable access to other AT&T hot-spot locations in addition to Starbucks.

It’s about @*#&^!$ time! Finally, customers get some amount of wireless access without paying through the nose.

This may actually make me go to Starbucks again on occasion  —  at least I can check my email with my predictably boring coffee.

Some more details, from Wi-Fi Networking News:

AT&T says in their press release that all Starbucks Card holders, which is simply their value-storing swipe card system, will get two hours of free Wi-Fi a day. No purchase is needed: you just need an active card, I confirmed with the company. Walk in, buy a $5 value card, activate it, and you’re on for two hours a day from then forward. You can also use multiple devices with a single account, within reason, Starbucks told me. AT&T has also lowered the price for usage from T-Mobile’s somewhat egregious $6 per hour or $10 per day to $4 for a two-hour session. The monthly price, like the rest of AT&’s network, is $20 per month for outsiders, which also includes all 70,000 domestic and international locations in their worldwide roaming network.

But it’s on the point of ISP terms of service that I believe Schneier’s case falls. First, he expresses no concern whatsoever about others stealing his bandwidth from his open network  —  “pay it forward,” he suggests.

But bandwidth isn’t the only issue here, at least not from the ISP’s standpoint; it’s lost revenue. The reason ISP terms of service forbid customers from sharing bandwidth with neighbors is as much or more about the provider’s need to turn a buck as it is the finite nature of the product.

A big question here is: who’s bandwidth is being shared? The subscriber or the ISP’s? That puts it in “either-or” terms, of course, when really it isn’t so clear cut. Arguments cut both ways.

See also:

• Steal This Wi-Fi
• Is it OK to Share My WiFi? (Series)
• Why Provide Free WiFi?

Whenever I talk or write about my own security setup, the one thing that surprises people  —  and attracts the most criticism  —  is the fact that I run an open wireless network at home. There’s no password. There’s no encryption. Anyone with wireless capability who can see my network can use it to access the internet.

I spoke to several lawyers about this, and in their lawyerly way they outlined several other risks with leaving your network open.

While none thought you could be successfully prosecuted just because someone else used your network to commit a crime, any investigation could be time-consuming and expensive. You might have your computer equipment seized, and if you have any contraband of your own on your machine, it could be a delicate situation. Also, prosecutors aren’t always the most technically savvy bunch, and you might end up being charged despite your innocence. The lawyers I spoke with say most defense attorneys will advise you to reach a plea agreement rather than risk going to trial on child-pornography charges.

Personally, I don’t advocate this, but I do provide a Meraki-run SF “Free the Net” node that goes through my home Internet connection. It’s firewalled off from my main internal network, and it’s got QoS inflicted on it to avoid having neighbors using all my (limited…) cable bandwidth, but I too see providing some amount of wireless connectivity as a public good, and neighborly.

When Starbucks introduced for-pay Wi-Fi in 2002, it seemed like a great deal (especially for business customers who could expense it). But five years later, the model appears old and stale and ready for a complete overhaul. Prediction: Starbucks will start rolling out free Wi-Fi access within one year.

I wish. This seems to me like a no-brainer from the Starbucks perspective, but then again, I’ve been saying that since they first introduced it (it never really seemed like a “great deal” to me, even at the time—then again, I never could expense it…)

As I’ve written before, I think free Wi-Fi makes good business sense: it brings in customers and makes them happy. (Caveat: this doesn’t necessarily mean a free-for-all. It makes perfect sense to me to give limited-time coupons with a purchase, for example. But the point is to get customers in to your store with a relatively minimum investment.)

The Computerworld article suggests that Starbucks is competing with McDonalds, which has started rolling out free Wi-Fi. Perhaps, although personally, I would never choose between Starbucks and McDonalds on the basis of Wi-Fi access. I just don’t think they have the same target market, at least, not for anyone I know. Then again, the article suggests that increasing access via mobile devices like iPhones might be shaking this up, and that might well be true, as those devices have a completely different usage pattern (I imagine, though I don’t know that this has yet been proven) than laptops.

Regardless of the motivation, I might actually consider going to a Starbucks on occasion if they offered free (or even reasonably priced) Wi-Fi. (Or perhaps that would simply encourage more independents to offer Wi-Fi, something that’s been shockingly lacking in San Francisco, especially as compared to Seattle, for example).

Your San Francisco neighbors are building a free wireless network, and you can be a part of it! With a Meraki repeater, you can connect to the network, expand its reach, and provide access to others.

Right now, we’re giving Meraki Mini repeaters away to SF residents who can see the Free the Net signal, and the network will always be free.

Our network is growing at a healthy rate, with new communities lighting up each month. Even if Free the Net hasn’t made it to your neighborhood yet, we’ll be there soon, and we’d like to let you know when we arrive!

So far, everything was incredibly simple. I just plugged it in, and it worked. (Of course, I did a little QoS work on my router, just in case’s Meraki’s rate limiting wasn’t good enough—I really don’t want to have visitors maxing out my upstream bandwidth and consequently taking out my VoIP service.) I’m not yet on Meraki’s map, but hopefully I will be soon. Perhaps eventually we can get some Meraki-provided DSL lines around here too…
See also:

• Meraki “Free the Network” San Francisco Project Grows

• Is it OK to Share my WiFi? (Series)
• Why Provide Free WiFi?

EarthLink said late Wednesday that it is bailing out of a contract to build San Francisco’s free Wi-Fi service.

Earlier on Wednesday, the city of Houston announced that EarthLink had agreed to pay a $5 million penalty to the city for not meeting its first deadline for building its wireless network. EarthLink has nine months to start construction or figure out a way to get out of the contract altogether.

And now, the company has also dissolved its contract with San Francisco, which was approved in January but was awaiting final approval from San Francisco’s Board of Supervisors.

From c|net.

At least Meraki‘s independent push to provide wireless access to San Francisco residents is growing, perhaps because of much greater grass-roots, citizen-based involvement?

From a legal standpoint (contract law), it seems possible that Earthlink might just escape the contract entirely (maybe, but there might be details I don’t have) because the Board of Supervisors never really accepted their offer (or never generated a final offer for Earthlink, depending on the details of the situation).

So I think the San Francisco mayor might possible be right to blame the Board of Supervisors, but then again, the Board might be right to say they were right to be leery, given Earthlink’s meltdown:

San Francisco Mayor Gavin Newsom, who had stated publicly that he felt the current contract was sufficient, blamed the Board of Supervisors for dragging its feet and blowing the deal.

Supervisor Ross Mirkarimi said the mayor was completely wrong in his assignment of blame.

“The mayor wanted us to rush into a deal that was half-baked,” he said. “And now he’s trying to cover his tracks instead of looking at the real reason this deal fell through which is the fact that EarthLink is having a complete financial meltdown.”

They brag about the service’s simplicity: just plug it in and it works. In fact, that’s exactly how it worked for me. (Actually, I cheated, and signed up at a mall kiosk. The only extra work they did, though, was to log in once the device was connected to a computer to activate it. No calling remote technicians, no wiring, no fiddling with MAC addresses. You can watch their installation video to see for yourself.)

I signed up for the “Premium” service, which promises “up to 1.5 Mbs” down, and “up to 256 kbps up.” I could have opted for the “Value” service, which offers slower downloads for about $10/month less.

Here in Honolulu, on the 27th floor, the Cleawire modem claims full signal strength (note that I couldn’t reliably get any WiFi signals to work due to high packet loss, even with a dd-wrt powered router with an after-market antenna), and I see no packet loss. A “ping” to Google looks like this:

PING www.l.google.com (72.14.253.103): 56 data bytes
64 bytes from 72.14.253.103: icmp_seq=0 ttl=243 time=171.751 ms
64 bytes from 72.14.253.103: icmp_seq=1 ttl=243 time=158.453 ms
64 bytes from 72.14.253.103: icmp_seq=2 ttl=243 time=171.066 ms
64 bytes from 72.14.253.103: icmp_seq=3 ttl=243 time=143.770 ms

4 packets transmitted, 4 packets received, 0% packet loss
round-trip min/avg/max/stddev = 143.770/161.260/171.751/11.402 ms

Decent, although 161 ms average round-trip time isn’t great. It might make some games, for example, lag a bit (but I don’t play online games). In contrast, my Comcast service averages around 55 ms. (And note that the fact that I was using Clearwire in Hawaii might have an impact on this, of course! So YMMV.)

Speedtest.net gives this using a server in San Francisco:

Downloads do seem to be quite snappy, but I do notice a slight delay while browsing sometimes—not terrible, but noticeable.

Although line-of-sight is not required, Clearwire does recommend putting the wireless modem in your window, and, if necessary, using an 802.11 WiFi router to share the connection. (This is what I’m doing right now.)

Mobile Use

One intriguing idea suggested by Clearwire is to take the wireless modem with you. Provided Clearwire serves the location you’re visiting, this gives you mobile access while away from home. Of course, I would call this more of a “luggable” solution than truly a mobile one, since the modem, while lights, isn’t tiny, and does require a power outlet to function.

To test this, I took my laptop and the Cleawire modem up to the roof of my building (the 31st floor). Everything worked normally, with full signal strength. Next, I tried an outdoor hotel bar by the beach. Other than the fact that (1) beach bars in hotels are expensive and (2) functioning power outlets can be hard to come by, it worked great!

See also:

• Can I Share My WiFi? (Clearwire)