Waiting For Google's gPhone: What Will The Perfect Mobile Device Look Like?

Sep 25, 2007 (11:09 AM EDT)

Read the Original Article at

Until it was released in late June, Apple's iPhone was envisioned as the perfect mobile phone, a sleek, elegant device designed to deliver usability, performance, and versatility at the same time.

It came close, but it wasn't perfect: It cost too much, it only worked on AT&T's poky EDGE network, it wasn't open enough to please developers or would-be phone hackers, its fixed battery promised to be expensive and inconvenient to replace, and it threatened to generate outrageous data charges outside the United States, to name a few post-launch quibbles that were deal-breakers for some.

Several of these issues have been dealt with, at least temporarily, through a price cut and the release of software that untethers the iPhone from AT&T and software that will load iPhones with user-generated ring tones. But Apple's vision of mobile music and telephony remains a work in progress.

In the absence of perfection, the technology community awaits the Google phone, or gPhone. Though unlikely to be as aesthetically pleasing as Apple's first foray into phone design, the gPhone is expected to be more widely available and more affordable than the iPhone.

It will probably be built using some version of the open-source Linux operating system, a J2ME middleware layer, and a Flash/Ajax presentation layer or something similar based on the vector graphic technology developed by Skia, which Google acquired.

That particular software stack is what venture capitalist Simeon Simeonov described in his blog as the ideal mobile phone stack.

There are, of course, other mobile OS options. BlackBerry, Palm, Mac OS X, Symbian, and Windows Mobile come to mind. But the perfect phone should be as open as possible, as least inasmuch as it aspires to be part of the "freeconomy" that has developed around Google, where the dominant business model involves ads.

What's The Perfect Phone?

Let's be clear: The perfect phone is different for different people. "There is no one answer," said Forrester analyst Charles Golvin. "A lot of people who are BlackBerry devotees would say you couldn't improve on that much."

The perfect phone for telecom companies would come with a nontransferable mobile number to maximize customer lock-in and would require a percentage of the owner's annual income as a service charge, in addition to a continually escalating monthly fee.

Phone users and developers may have a different idea of perfect, but theirs remains likewise unfulfilled. "One overarching limitation to handset nirvana are the entrenched points of influence in the form of the handset guys, the operators, and multiple, competing software agendas, as well as network limitations that conspire against this becoming a reality in the near term," said John Jackson, an analyst with the Yankee Group.

But putting aside reality for a moment, let's imagine the perfect phone. At its core, it's either Linux or Mac OS X. Sorry, BlackBerry fans.

Linux, in fact, is expected to be the fastest growing smartphone OS over the next five years, according to ABI Research. The firm projects a compound annual growth rate of 75% for smartphones and predicts that Linux-based smartphones will account for about 31% of such devices by 2012. Shipments of all smartphones during this period are expected to total 331 million.

Then there's Symbian, the operating system that represents the lion's share of the smartphone market today, according to Forrester's Golvin, and is popular among developers.

But ABI Research sees Symbian's market share slipping because of competition from Linux and Windows Mobile. In March, the research firm said, "Symbian's strong position in the smartphone operating system market is under continued and increasing threat."

Operating-System Importance

Mac OS X will become more important, as iPhone sales continue. Estimates suggest Apple will ship 3 million iPhones this year, and the company stands to make a lot more money from its phones than its competitors. Beyond the software and media content that Apple will sell to iPhone users through iTunes, the company is likely to see significant revenue from carrier partners like AT&T. For example, O2, a subsidiary of Spain's Telefonica, is reportedly paying Apple as much as 40% of the telecom fees paid by iPhone customers.

As a matter of comparison, consider Windows Mobile. "The Windows Mobile guys have thus far failed to scale," said Jackson, noting that Microsoft expects to ship about 20 million Windows Mobile devices this year.

Apple is aiming to have shipped half that many iPhones with Mac OS X on board by the end of 2008.

Apple's relevance to this discussion isn't just a matter of the iPhone's booming popularity or its apparent profitability; it's also about the company's leverage with media companies. Apple, thanks to its ties with Disney and its near-monopoly on digital downloads with iTunes, can put content on its phones that won't be easy for Linux players to match.

But Apple's appeal to content owners may be diminishing as more and more companies, at least in the music industry, experiment with abandoning digital copy protection schemes, or digital rights management. Resentment of Apple is coming to rival fear of unauthorized copying among some of Apple's former iTunes partners. That resentment may translate into greater acceptance of, or at least resignation toward, Linux-based mobile devices, particularly if Google can extend some degree of copy protection to content on the gPhone.

Hardware Perfection: OpenMoko

Until Google tips its hand and reveals the gPhone, the OpenMoko Neo 1973 is a suitable proxy for hardware perfection, or at least a starting point.

"The perfect mobile phone, I would argue, is something that becomes more than a phone," said Sean Moss-Pultz, president of OpenMoko, in an e-mail. "Not simply an appendage to carrier. Not merely a pod powering the matrix, but rather a new thing, a neologism; Neo for short. Eclipsing the phone means more than adding 'me, too' features or the bling of the moment; it means morphing the whole mobile market. It means moving the phone into productivity niches and creative niches that it hasn't served. This Neo will connect all forms of content anywhere, anytime. This Neo frees its owner to create content. It knows its owner's location, loves, and life. The Freed Phone weaves itself into its user's life and disappears, becoming as natural and invisible as the air we breathe.

"We don't know exactly how to get there. Only how it will begin. We need an army of Davids, a collective effort. And we need an open platform, based on Mobile FOSS [Free Open-Source Software], that changes and evolves as its owners' needs change. It becomes what it needs to be."

The OpenMoko Neo 1973 is built using a Samsung S3C2410AL-26 CPU, capable of running at up to 266 MHz, 64MB Samsung NAND flash, 128MB SDRAM, Texas Instruments' Calypso-based GSM modem, an AGPS module from Global Locate, GPRS analog baseband and RF transceiver chips, an 8GB Samsung microSD card, a TPO mobile LCD display and a Touch Screen controller, an audio subsystem, a vibration module, support for analog and Bluetooth headsets, a Phillips power management chip, and a Nokia BL5C battery, not to mention a stylus.

Being idealistic about this, though, we want a multi-touch screen (making the stylus optional), a more powerful processor (Via's Mobile-ITX board, perhaps), WiMax support (not to mention UMTS and HSDPA support), Unlicensed Mobile Access (currently being offered in the United States by T-Mobile), an 8 megapixel camera that can take video and transmit it in a live stream, optional programmable buttons, an accelerometer (like the iPhone), ambient noise sensors, a digital compass, a fingerprint sensor, an LED flashlight and laser pointer, FM receiver and transmitter, RFID read/write capability, an inductive charger, a solar recharging panel, and multiple SIM slots.

Platform For Programmers

Really, though, the hardware's not all that significant. What matters to developers is that they can write programs that interact with the hardware, without serious limitations. What matters to users is that the phone works, has enough internal storage, and can be outfitted with hardware modules that deliver the services they want. That, and it has to be fast. No AT&T EDGE network compromises. We're talking speedy 3G, as in, hopefully, next year's iPhone, or better.

While we're on the subject of hardware, let's insist that our perfect phone be designed to accept standard modules, preferably recessed ones that fit inside the existing case. Thus, a user could equip his or her phone with specialty hardware -- say, an altimeter, barometer, or cardiac sensor -- to suit special use cases. Not everyone would care about having a phone with a heart monitor, but having one that could be scripted to call 911 in the event the owner's heart stopped would be a killer app, so to speak, for some.

The perfect phone should also be adaptable to different modes of input. The iPhone's touch screen has a lot to recommend it, but it can't easily be operated using only one hand, as can be done with a Treo. It would be nice to have the option to customize tomorrow's iPhone, gPhone, or other object of technological desire with input controls suited to the owner's needs.

For example, the perfect phone would benefit greatly from a fold-out support arm on its back (unscrew it, and it becomes a stylus) and a front-mounted laser projection lens to both paint a full-size keyboard atop the table where the device is resting and to register keystrokes generated by typing on the virtual keyboard.

Doing all this without turning the perfect phone into a five-pound brick with battery that drains in 30 seconds: priceless. But what's good hardware engineering without some tough challenges?

Software Stack

"You could argue that the era of obvious innovation is gone," said Jackson from the Yankee Group. "The next wave of innovation is going to be a lot more subtle."

If you accept that argument, blame handset makers and telecom companies for a failure of imagination.

We can do better. Never mind the market, here's the way it should be: Phones should be small computers, capable of receiving and transmitting digital information using a variety of protocols.

In fact, phones are small computers, but getting information in or out of them isn't easy, apart from a few presanctioned methods. That's because mobile network operators like operating digital toll booths, charging for incoming or outgoing data, whether that's voice, data, or otherwise. They also don't have unlimited bandwidth and aren't eager to have customers flooding their networks with live video, for example.

Security would be a challenge, and it seems unlikely that tomorrow's open phone would be any more secure than today's PC. But that's where telecom companies could make money. They could provide security services to keep bad applications off phones and to limit the kind of damage bad applications could do. The perfect phone would have to make some concessions here, but they'd benefit everyone. No one, apart from a few cyber criminals, wants mobile zombies driving denial of service attacks.

But security shouldn't be an excuse to hobble phone functions or prevent phone-related data from being legitimately accessed and manipulated. There's no technical reason why the iPhone, the gPhone, or our perfect phone couldn't, for example, be configured to record live audio to an MP3 file stored on a remote server chosen by the user. The hardware is in place. But the software doesn't yet support this because those who control the handsets don't see a market for that kind of software.

The Internet offers proof that individual developers and entrepreneurs can create new markets where none existed before. So imagine for a moment how a programmable phone might really work. It would be just like your computer, with only a few preinstalled applications, like the iPhone (for those who just want a no-frills mobile phone), or with a wealth of user-loaded programs for manipulating audio, video, and text data. It would be conversant in e-mail, IM, SMS, and MMS messaging, not to mention VoIP, POTS, WiFi, and WiMax.

Sharable Applications

More significantly, its many hardware modules would be accessible from software applications. Owners of the perfect phone could, for example, make their GPS coordinates available, and nearby friends with phones could be notified through a variety of modes and applications. Or they could record their own ring tones and assign them to contacts in their address list or play them live during a conversation. Or they could program their phone to route calls somewhere else or play a prerecorded message at certain hours. They could have calls from certain numbers trigger e-mail messages to the caller or somewhere else. The phones could be programmed to support mesh networking and avoid carrier networks. You name it, the perfect phone can do it, without a fee, preferably.

These kinds of innovative features are starting to be offered by a variety of online sites, including GrandCentral, which Google recently acquired. But more can clearly be done with Internet services and software that runs on phone hardware. Allowing carriers to monopolize phone services lets companies like AT&T get away with, for example, charging $5 per month and a $7.50 activation fee for a residential service like priority ringing. Priority ringing assigns a distinctive ring tone to certain numbers. Can you imagine being charged that much by Microsoft to have Outlook display incoming e-mail from colleagues in red rather than black? Well, maybe you can, but even Microsoft would never have the hubris to actually do that.

Some might argue that being able to access Web-based services, as can be done using the iPhone's Safari browser, is enough. But many online tasks cry out for a dedicated application that has been engineered for small-screen interaction. For instance, booking a flight on Expedia, with its myriad checkboxes and text boxes and other inputs, is just too much of a hassle on the iPhone. It can be done, but not easily. And in order for the general public to give a damn, it has to be easy.

Forrester's Golvin said that his perfect phone would capture the BlackBerry e-mail experience and the iPhone's Safari Web browser, it would have a 3G radio, and it would be able to tune in broadcast TV.

Opening The Network

The perfect phone may remain unrealized for years because the telecom companies don't see a benefit in releasing their stranglehold on phone development. Apple has begun the process of trying to make phones more functional, but to date the iPhone's services remain highly bandwidth dependent. And AT&T's EDGE network just doesn't cut it. Perhaps the gPhone will push the envelope further.

That's not to say that the likes of Nokia, Motorola, and RIM aren't innovating. They are, but they're working within the system rather than trying to fundamentally alter the way the business works.

"What's going to compel an operator to open its network to any device?" asked Jackson. "We don't know yet."

That event might be Sprint Nextel's 4G wireless broadband initiative with Intel, which the company expects will enable it to offer high-speed WiMax coverage to about 100,000 million people in the country by the end of 2008. Or it might be Google winning the FCC's 700 MHz spectrum auction in January and erecting a new wireless network that accommodates an open phone platform. (Verizon's recent lawsuit to overturn the FCC's rule that the spectrum winner has to let consumers connect any device or software shows how much the telecom companies fear open access.)

If and when mobile operators open their networks, Jackson believes it will be the result of some industry disruption rather than some sudden fit of altruism.

The perfect phone is the perfect storm, and the weather conditions look just about right.

But such storms take time. "One of the things that people miss both about Sprint's WiMax deployment or a Google-allied consortium building out a 700 MHz network is that building of those networks will take time," said Golvin. "People care a lot today about coverage. Even if those networks are developed and do embrace this open Internet architecture, [phones created to take advantage of those networks] are going to carry all this baggage from the past, just so you can get on all the other networks."

In the meantime, why not try an imperfect phone? And if you aren't entirely satisfied, console yourself with the knowledge that you can always buy an improved imperfect phone next year. In fact, that sounds like a pretty good business model.