The Mindlessness of the Telcom World

I just got some interesting feedback on the 3G 'revolution' from Maynard which seems worth posting. Any more takers? Incidentally, the biggest problem seems to me to be that the people responsible for releasing these devices either don't read weblogs, or don't listen to end users, or both.

The absolute mindlessness of the telcom world continues to amaze me. Here in the US I have a device called a Treo 300 which is a PalmOS/cellphone device. It is about the size of any other PalmOS device, so slightly larger than a cell phone, but not uncomfortably larger. This is nicer than a cell phone in two ways.

(1) It is a better phone. It has a larger enough color screen that it can display large buttons showing you options of what to do for things like speaker phone or 3-way calling, or manipulating your address book. (2) It has a built-in web browser (which works very well for reading blogs). What is frustrating is that this is probably the best such package put together in the world, and it is still so OBVIOUSLY badly designed.

The one thing that's right about the package is that the cell plan provides for unlimited data at $10 a month. Anything that charges by the byte is, be real, dead in the water. Who wants to be hostage to per KB charges, or a plan capped at so many MB per month when these are so unpredictable and difficult to track---far more so than one's voice minutes. But that's all that's well designed.

For example---the source code to the web browser is not open source. The web browser is slow and does some things poorly---fact of life---but these idiots persist in the belief that it is a source of competitive advantage. Open the source, let people replace the bits that suck with bits grabbed from various other open source browsers and come up with something 3x faster and with a better UI. Another example---the obvious way to connect the web with cell-phones in a way that is really useful is through location information---eg "I feel like a Jamba juice, tell me where the nearest one is". We don't need GPS---heck info as simple as the cell tower handling the signal would be good enough. But is anyone providing information like that in a way that's useful to 3rd party internet service authors? Not that I know of in the US.

A third example. As I said, blogs work really well on such a device because they are mostly text and because most blog software is smart enough to generate HTML that displays well on such devices (even pages that you'd imagine being a disaster like Brad deLong's work out well ---- the columns of stuff on the side just don't appear and what one sees is the useful content nicely formatted to the screen size. But the only commercial web site that seems to have any clue about these devices is Google, which formats both its search page and news page appropriately. Dealing with the types of sites that you'd expect to be perfect matches for such a device---newspapers, yellow pages, maps and driving directions, weather---is like pulling teeth.

So what could be done? The telco's could encourage (perhaps through some fraction of the revenue stream) some of the more obvious such sites to get their acts together.The browser software (as I said above) could be smarter in terms of handling such pages (but that requires CPU which is limited in these devices) or: the telcos could setup proxy servers that would preprocess the HTML, use various heuristics to reshape it into something usefully formatted for small-screen devices, and send on the rewritten HTML. Done right, this could be hooked up with a compression scheme that would substantially improve throughput, while allowing the browser on the limited-CPU device to work a whole lot faster. But no telco appears to be even thinking this way. This is, BTW, not hard, nor is it sci-fi. Various individuals have set up just such proxy servers on their home machines, but they are generally hardwired to manipulate particular sites of interest to those individuals, and are not widely advertised for the obvious reasons that these individuals can't support the entire US mobile population hitting their home server.

A final example. In a world of ethernet, cable modems and instant-on; in a system where the technology is intrinsically packet based, the hardware/software on these machines is marching proudly into the 1980's using modem-emulation, UARTs and POP. It's freaking unbelievable how primitive the networking HW/SW is that they've hooked up to their so sophisticated cell-phone infrastructure. The bottom line is that not one of these companies deserves to make money, because not one of them appears to have a clue what year they are living in. They appear to know nothing about modern networking (packets, proxy servers etc), open source, or the fact that the web succeeded because of open standards accessible to everyone. The lesson they appear to have taken away from the failure of WAP was that they didn't control it enough and so they need to be even more dictatorial next time round. Even MS, which you'd like to think is slightly smarter than these guys, appears to be incapable of doing anything useful in this space. I think our only hope is that Steve Jobs has some top secret team somewhere working on an iPhone-enabled iPod. (Such a device is actually not too silly. You need a larger color screen, but apart from that using Bluetooth to speak to a small headset means that the size/weight may not be an issue.)