The HP HW6945: Mobile Computing w/Telephony
I’ve carried a PDA since I got a Palm I in 1995 and programmable phones since my Palm-based Kyocera. Last year I wrote about my Samsung i730. I’ve since upgraded it to Windows Mobile 5 but it still a number of little problems.
I wanted an alternative and the HP HW-6934 caught my eye. I also liked the idea of having an unlocked GSM phone. At least I thought I was buying a phone but instead I have a device I can use as a phone.
Sometimes it’s the little things that make a big difference. I’ve been using mobile computing devices ever since my first palm and most recently I’ve The Samsung i730.
But the HP-HW6945 has a built in GPS. This made a big difference – even though the i730 could connect to external GPS the convenience of having the capability always available was enough to change my perception. I quickly found that WM5 allowed me to run a number of programs at once because it swaps them to the SD card.
I quickly loaded it up with applications including booth Google and Microsoft’s satellite map applications. WM5 allows multiple programs to share the GPS receiver so I can run them at the same time. Of course I’ve been experimenting with various navigation programs – having 2GB on my MiniSD card means I can keep the full map of the US with me all the time.
I’m going to write about navigation in more details separately because it’s a full topic in its own right. Knowing where I am is an enabling technology. I did write my own program to track my location and report it to my PC but later found programs that did just that. But the real value was getting experience in writing for the device and it’s actually quite easy.
The device hasn’t fully crossed the threshold. Swapping is slow and the battery doesn’t last long when you are using all of the radios. For now, though, I want to focus on what the device can do and the potential for such devices.
I did buy two car mounts from ProClip – one for my car and a second one for the road. It means that the phone is being charged while I drive. I can also charge it from any USB source which further reduces my dependency on battery life though I still need to worry about how hot it gets running all the applications.
What I find most interesting is that I don't really think of it as a phone – that’s just another application. In fact, just to play, I wrote a quick phone front end. I only used the simple interfaces so it dials and switches to the standard telephone application. In glancing at the APIs I see can probably write a very nice “telephone” application. In fact it doesn’t look too hard to write a program that can meld the cellular and IP-based voice and use each to its strength. The same goes for using SMS vs IP-based messaging. The phone also supports MMS or “multimedia” messaging but that’s just a lame subset of standard email.
Of course I do use the
Nice at it is the HW6945 is still limited by its parentage:
First, it’s treated as a telephone and sold through those distribution channels. The carriers have to be happy with the device to support it. A side-effect of this is the use of Bluetooth which is an abominable protocol. It has all the limitations of wires without the simplicity of wires. Unlike Wi-Fi which just provides transport Bluetooth is hard to extend. If the phone doesn’t support my headset I might not be able to work around it.
The Bluetooth supporters are often puzzled by my criticism of the protocol because they are excited about what it can do. But if you go beyond that narrow design point it’s unnecessarily frustrating. 802.11 is very similar but separates the radio from the policies so we can make it do more instead of being locked into to the initial design point.
This is the difference between the assumption that telephony is a service from providers and the idea that telephony is an application based on common protocols. The version we get with the device should be the starting point for innovation rather than all it could ever be.
Second, as a descendent of the PDA it was born as a portable extension of my desktop address book and calendar. That made sense in 1995 but today it should be treated as a full blown device in its own right and a peer. Microsoft’s Activesync is frustrating. I can only attach a single device at once and synchronization is built-in rather than just another application. I should be able to connect over any path and not be limited to a single primary synchronization point. Microsoft seems to treat Outlook as a basic part of the machine.
It doesn’t help that USB is such a flawed protocol – half the time I can’t even get the PC to accept the connection!
Windows Mobile is really Windows CE and is targeted towards embedded devices. But embedding the software deep within the device isn’t interesting unless it gives the user the capability to modify the behavior.
When Unix-based alternatives become available I may have to switch.
One problem is that the machine has only 64 Megabytes of RAM and only 2GB (or 4GB for the i730) (The 4GB mini-SDHC wasn’t recognized) of “disk” (SD). Only? That’s still a huge machine – the mystery is why it seems to be so little space. One reason is that our appetites have grown.
Today I think nothing of putting a gigabyte of map data on the machine. When I view the satellite images I do get those images over the network on demand but I want the map data available because I can’t yet assume I’m always connected as when I’m trapped in an airplane. Because the airlines view everything as a profit center they are cutting back on their Wi-Fi connections and even their in-flight telephone services! To an Airline, connectivity is a distraction but we should strongly protest the forced isolation.
At least I can use my GPS (though typically a Bluetooth (alas) version because it isn’t feasible to keep the phone near the window) and watch my trip on a map as well as collect a record of path we took. In practice I often use my laptop instead when I can get power to my seat but those power connectors too are seen as frills. Hmm – maybe we can make that a security issue – better they supply the power than we be forced to pack a great deal of energy into each battery.
A third heritage is found the notion that the device stands alone. Of course being a telephone it has a radio but that radio is just used to call a tower. It does seem strange to have to align two IR beams when we can just send packets over the network. This too is a larger topic – if the device is connected then it can maintain relationships with anything else on the Internet.
Perhaps the most important aspect of these devices is that they give us a chance to experience real mobile computing or at least nibble at the possibilities. The experience is far from smooth but it is improving. Having the Opera Mobile browser I can make effective use of the web. The keyboard helps a lot and I like having a touch sensitive screen.
Having a real mobile device changed my perspective in the debate over municipal Wi-Fi. The debate is problematic because it’s unclear why we want it. People know the Internet is good but you aren’t going to be walking the streets using your laptop (though, with the Tablet PC, it’s almost feasible). It’s an expensive way to provide access to those who can’t afford it – the Digital Divide Rationale. It makes far more sense to share the existing broadband connection than build yet another infrastructure.
Mobile computing is about a lot more than just the applications and services provided by the carriers.
While waiting for Wi-Fi we can use the cellular radio as a data path though the prices range from relatively expensive to absurdly expensive.
Absurd is a relative term. If we consider SMS message to average 100 characters (r mr lk 10) and the price is 10�/minute that’s $1 for each 1000 characters which means each megabyte costs $1,000! Good thing I’ve got a data plan because I used 22 Megabytes last month.
The prices for the data plans vary as does the speed. For now the $15/month I pay T-Mobile for Edge is a good deal – at least compared with others. Verizon charges $45 and up for EVDO and Cingular prices are also high. But I now view them with bemusement because the whole idea is fundamentally flawed and can’t scale nor evolve quickly.
Far more interesting is taking advantage of available open access points via Wi-Fi. For now the coverage isn’t complete and we do need to improve the protocols to make it easier and safer to take advantage of them. Applications also need to be tolerant of areas where there isn’t coverage. By taking advantage of cellular and Wi-Fi radios we can start to understand mobile computing.
It is useful to know where you are. Unlike real estate location isn’t everything – just walk a few feet and you’re someplace else. We can now begin to explore what it does mean and not rely on those who look at location information merely as a way to choose what advertising to beam to us.
These devices put us in control. We may want to know what is available and even welcome suggestions but we don't need to justify the devices as little more than receptacles for LBS ads.
Remember, your modem wasn’t very useful until we had the web and the web wasn’t very interesting until people could experiment and discover what was possible. Some of the original ideas look quaintly na�ve, if not stupid, but others changed the world.
No matter how many SMS messages we send, crackberry messages we receive nor how much music we cram into an iPod we won’t discover the future unless user/developers can create and share. The HW6945 gives a hint of what is possible. We should build on it and demand more opportunity and fewer built in assumptions and favors.
I hope that within a few months my enthusiasm for the HW6945 will wane as I find new devices that are much more capable but for now it gives me a chance to explore the possibilities of mobile computing and the importance pervasive connectivity.