CNET también está disponible en español.

Ir a español

Don't show this again


Will phones get smart before pigs can fly?

Well before that, according to Microsoft's Ya-Qin Zhang, whose task it is to put real intelligence into smart phones.

After years of working to crack the handheld and cell phone market, Microsoft is counting on Ya-Qin Zhang to help build products that will finally appeal to the masses.

There are lots of exciting features and technologies in the PC that have not been delivered by the phones yet.
That shouldn't be too much of a challenge for Zhang, who is used to taking on daunting tasks. He joined the company late in 1998 and built Microsoft's research lab in China from scratch into an organization of more than 500 people.

But in taking on the role as head of development for Windows CE and Windows Mobile, Zhang must help turn a comparatively tiny and still unprofitable unit into something worthy of Redmond's attention.

Zhang's challenge is not only a technical one. Cell phone makers and others have been hesitant to let Microsoft onto their devices, fearing a repeat of what transpired in the PC market. The company's latest tactic: allowing device makers to revise the Windows CE source code without having to share it with competitors--or Microsoft. CNET recently spoke with Zhang about smart phones' future.

Q: There were problems getting the first release of Microsoft's Smartphone to market. What are some of the things Microsoft learned, and what are you doing to address those issues?
A: In the smart phone or connected device, you have to work very closely with operators--especially for the converged device. It doesn't mean the previous model had any problems, but this is a new industry that operators, OEMs (original equipment manufacturers) and ourselves--we all learn things.

To be frank, there were some quality issues--things that in the handheld world or the embedded-computing world might have been acceptable. But in the carrier world, they might generate an unacceptable number of support calls.
We spent a tremendous amount of engineering resources in the last few months to make quality and stability the highest priority. A device--especially one that has a complex operating system with multithreading and multitasking and that has to work with a new type of network with which the operator does not have a lot of experience--does require multiple iterations. In the meantime, we do have to put more emphasis on testing and on quality.

When you look out two to three years, will there still be a distinction between handhelds and cell phones?
If you look at the hardware, you are going to see probably half a gigahertz- to 1GHz-embedded processors.

The whole smart-phone platform--the industry is still in a very early stage.
That is very powerful. Moore's Law continues--but obviously with constraints.

The constraint is battery power, which is not progressing at a rate beyond 10 percent. The processing power, the storage and the communications bandwidth are going to continue to follow Moore's Law. The more exciting part is really the software and the connectivity.

I don't know if it's exactly two years or three years, but let's say in the foreseeable future that we will have seamless mobile computing. The first thing is really seamless connectivity, really making the transition from a single radio to multiple radios. It could be Wi-Fi, BlueTooth, WiMax, UWB (ultra wideband) and also other cellular radios. The important thing is to make sure that there is seamless roaming and handover and a consistent experience. That is a critical technology that we need to enable.

When do we get to the point where the device can move between these networks without users having to manually change the settings?
In our next release, we will be able to do Wi-Fi and GPRS (General Packet Radio Services). For seamless connectivity roaming on all those radios, it is going to take many years. It is going to be an evolutionary process.

You mention that some of these capabilities are going to come in the next version of Windows CE. Microsoft offered a technology preview of Windows CE 5.0. When are we going to see that come into products?
There are two things. One is CE 5.0 we are going to release in the summer. The other is (the next versions of) Pocket PC and Windows Mobile we are going to release sometime next year. I cannot give you any details right now. Seamless roaming is a key part of that, but roaming is by no means complete. You can only roam between two different radios.

When you look at Windows Mobile today, what are the biggest priorities, and where are the areas where you see the biggest need for improvement?
The whole smart phone platform--the industry is still in a very early stage. If you look at Windows Mobile, our primary priority is to optimize the experience for what we call mobile information workers.

There is a lot of technology there already. We can make a lot of improvements in working better together with the PC and the server. I'm not just talking Windows Mobile but the whole smart-phone category. There are lots of exciting features and technologies in the PC that have not been delivered by the phones yet.

Obviously, there are technical challenges, because the smart phone is very constrained. It's constrained by the real estate, the battery power, the footprint, the screen size and the way you interact with it--you don't have a huge keyboard.

A smart phone has to make phone calls in a way that is transparent to users. It should be transparent to users. Usability: I think we have a very nice user interface, but when you put in a lot of features, you want to create a very easy-to-navigate experience. I think that we can improve in all of these things. The features are there, but we need more.