Tethered

Today in Tedium: If you’re far from home in need of an internet connection, you most likely either use your phone or hunt down the closest coffee shop. But if you’re fancy, maybe you’ll tether your phone to your laptop, and connect that way. Maybe you’ll use your phone’s ad-hoc Wi-Fi network, or connect directly via a USB cable. While we’ve given it a more modern name, tethering, it’s not particularly far removed from the old dial-up modems of yore. And it’s saved my bacon quite a few times over the years. If you’re a nerd, it’s probably saved yours, too. So, how did tethering become a thing—and what did it look like when we first embraced the cellular internet life? Today’s Tedium talks tethering. — Ernie @ Tedium

Today’s GIF comes from a 2010 review of the iPhone version of PDANet, as featured on the delightfully-early-2010s YouTube channel ExtremeApps.

An image from the first patent filing that attempted to solve the problem of the cellular modem. (Google Patents)

We’ve been connecting cell networks to computers almost as long as we’ve had cell networks

Cell phones and similar mobile devices were still in their relative infancy in the mid-1980s. In the fall of 1983, the first commercial cellular service launched in earnest.

By the spring of 1984, companies were already hatching plans to connect modems to the network.

The first company credited with bringing networked computing capabilities to cellular networks was the Dallas-based firm Spectrum Cellular Communications. That firm first announced a device called the BRIDGE in 1984, and filed for a patent for said device in 1986. A 1984 piece in Infoworld described what the BRIDGE was capable of at the time.

Spectrum Cellular didn’t give me much to work with image-wise, but this handout image of their technology used in a car phone setup, appearing in a Stephen W. Gibson’s 1987 book Cellular Mobile Radiotelephones, may be as close as we get. (Internet Archive)

About the size of a VHS tape, the device allowed users to dial into external networks using a cellular phone, resolving the weaknesses of transmitting data over cellular networks of the era.

At the time, cellular networks were analog, rather than digital, and while it was technically possible to use them with a traditional dial-up modem, you wouldn’t want to. As noted in the patent filing, cellular networks were built for entirely different forms of error correction:

In the cellular telephone environment, numerous errors are induced into data transmission because of the problems associated with cellular telephone communication. Echo and fading problems cause multiple bit errors in the data stream, and such problems occur frequently with a moving vehicle. For example, the transmitted signal may hit a building or other obstacle and bounce erratically or fade as the vehicle is shielded from the cell antenna. This high frequency of error in the data stream transmitted by cellular transmission renders the error correction protocol present in conventional wire line modems unsuitable for cellular use. Errors occur so frequently in a cellular environment that the number of repeat requests becomes large and data transmission efficiency is reduced below an acceptable amount. In some instances, errors may occur so often that a correct packet may never be received. Thus, the error correction protocol present in conventional telephone modems is unable to cope with the problems presented in a cellular environment.

Now, to be clear, at the time, we were using cellular modems like dial-up modems, rather than the always-on packet-switched connections of the internet era. If you needed to contact a data connection with your cellular modem, you had to dial the number.

“It’s the same concept of using a telephone modem, but instead of phone lines, you use cellular air waves to send or receive data,” Spectrum president Dana C. Verrill said to InfoWorld.

All this stuff was expensive—the magazine noted that the company estimated people would spend $2,000 on the phone and another $1,000 on the portable computer. (In other words, the phone was more expensive than the computer.)

But as limited as this all sounded at the time, it allowed some futurists to do ambitious things. One of those people was digital nomad Steve Roberts, who created a device called the Winnebiko, which was essentially a Spectrum BRIDGE connected to a recumbent bike loaded with electronics. As Roberts wrote in Cellular Business in 1988, in a piece recreated on his website:

The bike’s five computers are networked together, allowing me to write, program and go on-line with remote computer services while riding. I type in binary at about 30 words per minute on a handlebar keyboard of eight buttons. There are more than three megabytes of on-board RAM, as well as a hard-disk-based file server that supports the various processors.

Communications gear includes ham radio equipment for all bands between 80 and 2 meters, with the addition this winter of an OSCAR Mode L satellite station (1.2 GHz uplink and 450MHz downlink). There is also a citizens band radio, a license-free AM broadcast station for information bulletins to passing drivers, a pair of packet TNCs and a 300MHz remote link with the security system. And, of course, there is an Oki 491 cellular phone, my essential conduit for business communications.

Yes, you read that right: He was able to bike and program simultaneously. (Take that, OpenAI.) Central to his abilities, he added, was the BRIDGE, which gave him access to a relatively high-speed online connection for the era.

With the Spectrum Cellular Span modem, I can sign on from campgrounds, roadside rest breaks or even while pedaling. Unpackaged, the unit is simply a circuit board inserted between the cradle and TRU, with an RS-232 link to the bike’s computer network. Power is derived from the phone itself, and CMOS circuitry keeps battery drain to a minimum.

Roberts’ excitement for cellular capabilities aside, it took a while for this capability to gain mainstream uptake even among businesses. Honestly, it didn’t really happen until around the time that GSM, a second-generation standard, had started to emerge, furthered along by the third-generation CDMA2000 standard in the United States.

Helping on this front was the rise of PCMCIA technology, which had started to reach maturity by the mid-1990s. Around that time, Motorola pushed radio modems into add-on cards, well before GSM was common in America. But even back then, it was still a matter of connecting to a service through a dial-up number.

A Sprint PCS Vision Sierra Wireless PCMCIA card, which, despite the name, uses a CDMA network. (Wikimedia Commons)

None of these options were cheap. A 1996 Network World piece describes a $519 PC Card Type III (a.k.a. a rechristened PCMCIA) that connected to a cell phone with a $139 cable. (Those numbers, with inflation, are equivalent to $1,029.30 and $275.67, respectively.) People would cry bloody murder if they had to deal with those kinds of prices just for network connectivity today.

It wasn’t until 3G hit the scene in the early 2000s that the internet was deeply integrated into the mobile experience, albeit in fits and starts. (If you were using a cellular modem in the pre-3G days, God help you.)

At first, such connections were made with the help of PC cards with built-in cellular modems, external wireless modems, or PCs with built-in cellular connections. But as the years moved on, it became much more desirable to just plug in the device you already have in your pocket and tap into its internet connection.

There was one problem: Mobile providers shuffled their feet in a big way.

The iPhone 3GS, which you might’ve been hard-pressed to tether from. (Apple press photo)

Why was it so hard to get tethering on early iPhones? Spoiler: It wasn’t Apple’s fault

The concept of tethering was a known entity by the mid-2000s, when the first iPhone came out. In fact, numerous smartphones, such as the various BlackBerry devices, offered it as an option.

Not that you would know it if you were an iPhone user. For the first couple of years, it wasn’t even an option. And it wasn’t because Apple was being stingy about support, either. It was an AT&T problem.

The first example I can find of a high-speed tethering-enabled phone with vendor buy-in dates to around 2005, when Verizon announced the Blackberry 7130e had tethered internet capabilities on its network in some major cities. At the time, you could pay $15 on top of your $79.99 voice plan to get access to speeds that averaged 400 to 700 kbps—slower than your average cable modem of the era, but close to DSL. In today's money, that's $23.88 on top of a $127.33 bill.

As noted by Computerworld in 2006, other providers were often more gun-shy than Verizon, or wanted users to rely on their dedicated mobile hotspots. As David Strom wrote:

I have been saying for a long time that data is still a four-letter word for the carriers, and that they treat people who want a data plan as third- or even fourth-class citizens. Many of the data plans require you to also have a voice plan or are $20 cheaper a month if you have a voice plan. Some require two-year contracts, just like the voice plans, to get the best rates on the plans and equipment. And the major carrier Web sites are a maze with many dead ends.

Oh, the lengths we would go to get a tethered internet connection.

Of course, as you might imagine, there were always people who cleverly worked around these limitations. One of the earliest and most prominent examples, called PDANet, is one of the few apps to claim a home on PalmOS, Windows Mobile, BackBerry, Android, and iPhone, which should give you an idea of how long the concept was in the ether. Way back in 2004, Popular Science writer Andrew Zolli described PDANet as “my favorite Treo app, hands down.” (Later, it became a big hit on Android.)

But eventually, the iPhone forced the conversation—and highlighted how behind the curve AT&T was on the data front. No matter if you were running on EDGE or 3G, your fancy phone was unable to share its connection with other devices in a vendor-sanctioned way. This phone was in the midst of changing your life, but good luck trying to siphon the Jesus phone’s connection for your personal needs.

This was in part AT&T’s strategy for avoiding undue stress on its network, which was being tested by data-hungry iPhone users. It was frustrating as an iPhone user. But it was also a sign that the telecom giant underestimated the power of data.

At one point, AT&T made it clear that iPhones would soon support data tethering. But more than a year in, there was no movement, as Harry McCracken noted. And this wasn’t an Apple thing, either. After a tethering app was removed from the App Store during the summer of 2008, a Gizmodo reader emailed Steve Jobs to ask why AT&T offered BlackBerry phones with tethering capabilities, but not iPhones.

“We agree, and are discussing it with ATT,” Jobs wrote back, in a characteristically terse reply.

It took forever for it to happen, and when it did, the added cost was significant—an additional $20 on top of your standard plan. People did what people do: They worked around the provider-created shackles. They stealth-released tethering apps in unrelated apps. They relied on jailbreaking. And they counted the days until the iPhone exited its exclusivity with AT&T.

And when Verizon showed up, it cut the B.S. and gave the nerds what they wanted.

A CLEAR device, which was what I used to get around AT&T’s notoriously gun-shy stance on tethering. (Bytemarks/Flickr)

AT&T’s tendency to get in the way of consumers likely led to an exodus after it lost iPhone exclusivity and contracts ran out. It also pushed some users, like myself, to other options. For a while, I was rocking a CLEAR device after I decided that my tethering use as a blogger with a coffee habit was such that I was likely to piss off AT&T if I kept using a jailbroken app. In D.C., CLEAR worked well, but once you got out of the city limits, it left a lot to be desired. As soon as I got my shot, I moved to Verizon for a few years.

But Verizon got cold feet about unlimited data, too. I remember paying full price for an iPhone 5 just to keep my grandfathered unlimited plan. (By the time I upgraded again, T-Mobile had iPhones and reasonable unlimited plans, so I went with that.)

I think the mobile industry gradually realized it was mathing the math all wrong during this era, favoring minutes over data—and adjusted its strategies accordingly. (Example: Verizon, after infamously dropping unlimited plans in 2011, brought them back in 2017, complete with 10 gigs of tethering data.)

These days, pulling up a personal hotspot on your laptop is no big thing. You hit a button, set a password, and let your laptop log into said phone—whether over USB, Bluetooth, or Wi-Fi. (You should probably plug in your smartphone when doing this, as it’s known to gobble up battery life.)

But it was pretty messy for a while there, and it had nothing to do with the devices. It was basically the providers getting in the way. Fortunately, in our data-hungry world, the providers have mostly gotten out of the way when it comes to tethering.