Portrait of a Linux iPhone-killer wannabe
In the race to be the first “iPhone killer,” the most unlikely but perhaps most intriguing candidate is based on a new Linux platform with the peculiar name OpenMoko.
Major mobile phone vendors are tripping over themselves to release devices to compete with Apple’s iPhone. LG Electronics has its Prada, High Tech Computer has the Touch, and Samsung Group will release its Ultra Smart F700.
However, OpenMoko comes from a low-visibility Taipei-based company, First International Computer (FIC), which is best known for manufacturing laptops for vendors such as Hewlett-Packard.
The first phone based on the OpenMoko platform, FIC’s Neo 1973, is currently available to developers, with wide release expected this autumn. Can a relatively unknown vendor using a new Linux platform succeed in such a highly competitive market?
“The likelihood of this product becoming mainstream is very low,” said Ken Dulaney, an analyst at Gartner in Stamford, Conn.
However, Sean Moss-Pultz, primary architect of OpenMoko, disagrees. In an interview, he didn’t refer to the Neo 1973 as an “iPhone-killer” — the media and bloggers have been doing that. However, he did say he expects the device, with its iPhone-like touch screen, to be a hit. That’s because application developers will have complete access to the system.
“Most of the [Linux] consumer devices don’t give developers access to low-level hardware stuff,” Moss-Pultz said. “We want [developers] involved in the most fundamental parts, such as the kernel and device drivers.”
That, in turn, is leading to a flurry of development that will make the device so feature-rich and customizable that it will be compelling for both consumers and enterprises, Moss-Pultz said.
The first attempt
Even without a flock of busy developers, the Neo 1973, with its expected price of $300 for an unlocked version, should be attractive, Moss-Pultz said.
He said the device will work over Global System for Mobile Communications (GSM) cellular networks (AT&T and T-Mobile in the U.S.), although the first version will support only older, modem-speed General Packet Radio Service (GPRS) cellular data access. By contrast, the iPhone has been criticized for only supporting enhanced data rates for GSM evolution (EDGE) cellular data technology, which is significantly faster than GPRS but slower than 3G.
“Initially, data speeds won’t be anything to write home about,” Moss-Pultz acknowledged. “But we’re working on 3G versions.” Beyond that, the Neo 1973 should be an eye-grabber with its 2.8-in., 640-by-480-pixel resolution and a touch-screen display, Moss-Pultz said.
“It’ll be like reading something printed on paper,” Moss-Pultz said. The device will be fast, sporting a 400-MHz processor, and it will have significant graphics acceleration for gaming and video. It also will support Wi-Fi and Bluetooth. In addition, the device will be suitably diminutive at about 4.7 by 2.6 x 0.7 in., although it will weigh a somewhat beefy 6.5 ounces.
Enterprise and consumer appeal
While he said the initial release will be attractive, Moss-Pultz acknowledged that the public release of the Neo 1973 in October is just the first step.
“October will be a 1.0 release. It’ll take some revisions before it gets real good,” he said. That will occur as more applications become available, he predicted. Some developers are creating consumer applications, such as those for collecting and playing media, Moss-Pultz said. Other developers are creating applications for enterprises, such as the ability to synchronize data with server-based applications like Microsoft Exchange.
In fact, the most intriguing possibilities are in the enterprise, where Linux servers and applications are common, he stressed.
“The device uses the same libraries as server and desktop Linux,” Moss-Pultz said. In particular, OpenMoko uses the GNU C library, the X window system and the GTK+ tool kit. “If you recompile the [existing enterprise] applications, they’ll work.” He said there has been strong interest from corporations, particularly because the phone can be customized.
“Pretty much all the big enterprises have contacted us and are interested,” he said. “Enterprises have scores of IT staff who can customize and maintain Linux applications. With this phone, the company can customize it exactly the way they want it for their employees.” By contrast, most cell phones are notoriously uncustomizable.
Another big appeal to the enterprise will be security, Moss-Pultz said.
“You just don’t get viruses on Linux,” he said. “And it supports all the existing VPNs [virtual private networks], things like Open SSL, all the connectors. All the things you’d want for the enterprise, you’ve got.
When the initial public version of the device is released this fall, at least one large, well-known company will announce it is committing to the device, Moss-Pultz said. He refused to provide the name of the company.
OpenMoko’s architecture, its ability to be customized and the wealth of applications means that phones based on the system will become more valuable over time, Moss-Pultz said. And that’s something that enterprises will appreciate.
“When you buy a PC, it has almost zero value,” he said. “You start adding applications, and it has more value. Usually, though, buying a phone is like buying a new car; you buy it, and a few months later, its value drops.”
Will carriers care?
To succeed, mobile phones typically need the support of cellular operators. After all, particularly in North America, most users buy their phones from their cellular carriers.
That’s the only option with Code Division Multiple Access (CDMA) carriers such as Verizon Wireless and Sprint Nextel And while subscribers to GSM operators such as AT&T and T-Mobile are freer to buy unlocked phones through retail channels, few do. And cellular operators don’t seem excited by OpenMoko or the Neo 1973.
“On the carrier side, they’ve been lukewarm,” Moss-Pultz acknowledged. He said he has had talks with AT&T and T-Mobile, with the latter expressing some interest. But a deal is not imminent.
Even more distressing for a phone that will likely depend on word of mouth for its success, industry analysts are cool toward the device.
“Without a carrier behind it, it can’t be an iPhone killer,” said Craig Mathias, an analyst at Farpoint Group and a Computerworld columnist. Another problem is that developing applications for OpenMoko is not necessarily the best approach for consumers and enterprises, he said. Instead, a more attractive approach is to build browser-based Web service applications, the approach taken by Apple with the iPhone, Mathias said. Such applications are more secure and can be used by any Web-accessible device, he said.
Avi Greengart, an analyst for mobile devices at Current Analysis, also doubted that the system or the phone will succeed. He noted that there are a lot of Linux mobile platforms, some from high-visibility vendors. Palm is porting its operating system to Linux, Nokia based its N800 Web Tablet on Linux and Samsung, Motorola, Panasonic Corporation of North America and others are working to create a standard mobile Linux platform.
“If you’re an application developer, your goal is to sell applications, so you need the largest possible installed base,” Greengart said. And as a new Linux platform, OpenMoko won’t have a large enough user base to excite commercial developers, he said.
Gartner’s Dulaney was skeptical of Moss-Pultz’s expectation that enterprise developers will pick up the slack.
“Enterprises need to hold someone accountable for the platform,” he said. “Enterprises need names they know and trust,” Dulaney said. “In this case, they have never heard of this company. The likelihood that they can prove to enterprises that they have what enterprises need is low.”
However, Moss-Pultz said he is confident that the extreme openness of the OpenMoko platform will attract developers and that the developers’ applications will attract users.
“Initially, it’ll be for savvy users and people who have used open-source products like Firefox,” he said. “But I think we’re just six months away from being mainstream. When you allow the user to put exactly what they want on the device, you exponentially increase the device’s value. I’m confident we can do it.”