But what I'm getting at is that these sorts of things that I've listed off are coming standard now with many of the latest hardware gadgets. Not all gadgets have all those things, of course. And the Pandora has a lot more going for it than any other gadget out there (and the cost reflects that of course). But my point is that in another couple of years, I suspect a lot of us will think it's almost quaint not to have built-in GPS, accelerometer, magnetometer, skype, cell-phone sim card slot (yes making it a real cell phone), wireless N, WiMax, etc. That's because so many gadgets coming out today have that stuff. Imagine what two years from now will bring.
There's a lot of integration and system-on-a-chip designs going on right now. It's making it a lot easier to design gadgets with these things. It's how the Pandora console exists even.
Of course you didn't mention that the devices you're comparing the Pandora against are mostly smartphones, which can play games but are designed to fill not quite the same market niche as the Pandora. Let's have a look at your features.
GPS – Found in some or most new smartphones. Not found in most non-smartphone handhelds. The only one I can think of was the disastrous Gizmondo. Could be nice for the Pandora but those people who want to use it with their Pandora will most likely be happy with a dongle and those who need instantly-available GPS will already own a smartphone sporting it.
Accelerometer – Found in some handhelds and consoles. Could be neat but only if they can fit it in without sacrificing any current capabilities. Very few games use accelerometer input well and fewer do so on handhelds. Yes, that includes the iPhone. I own an iPod touch. Most accelerometer games for it would've been better with conventional inputs.
Magnetometer – I think a few smartphones have it. Only useful in addition to a GPS module and/or an accelerometer. Makes more sense in a GPS/compass USB dongle than onboard.
Skype – Not a hardware feature; availability is determined by whether Skype Ltd. decides to provide a Pandora version. Launch availability probably would've cost too much money.
SIM card slot and cellular modem – Horrible idea. The functionality is orthogonal to the Pandora's current main purposes, the modem would shorten the battery runtime and between the modem, regulations, required tests and distribution issues (not to mention the fact that you'd need half a dozen different hardware versions for GSM/CDMA, UMTS/CDMA2k, different frequencies etc.) it would've pushed the proce of the Pandora way past what's acceptable. Also, the Pandora's form factor is really horrible for a telephone. The N-Gage has taught us that if you want to make a gaming phone it has to be a phone first and a console second. This would've meant throwing away just about everything the Pandora is actually about.
Of course you could use the modem for data only but then you're still stuck with all the issues like hardware versions (at the very least you need different versions for UMTS, CDMA2k, LTE and EV-DO), cellphone regulations and the like. The cost would still be prohibitive.
Wireless N – Would've been nice indeed.
WiMax – Virtually the same as a 4G cellular modem, with similar drawbacks except for the form factor as it seems to be data-only.
Okay, regarding "instant on", what I'm referring to is the idea that you can press the power button to turn it on, and it instantly responds with a login prompt and/or your desktop. No need to go through a boot-up procedure. Just remember the state of the system at last power-down. Oh, and power-down should be as instantaneous.
[...]
"Instant on" is a feature that everyone strives for, but most never come close. Ten seconds is close. A couple seconds would be much better. A minute is pretty far away.
There are three ways of doing this:
- Back in the stone age, home computers simply had their entire OS in ROM and didn't perform anything like hardware initialization at all. This is not easily possible with *nix.
- You use non-volatile RAM. Not feasible because NVRAM is neither dense nor fast nor cheap enough right now.
- You don't power down in regular usage at all but merely enter a low-power mode. This is what virtually all ARM-based devices do and what the Pandora is also expected to do. A bootup time of one minute becomes irrelevant if you only reboot the device when you upgrade the firmware. Waking up from low-power mode should be virtually instantaneous. (For a great example see the iPhone. The thing takes ages to boot but you literally only reboot it when upgrading.)
So ARM CPU is to a car as x86 CPU is to an airplane? Not exactly. It's more like ARM CPU is to a Diesel engine as x86 CPU is to a gasoline engine. Both engines can pretty much do the same things. But at least here in the U.S., gasoline is cheaper and more readily available than Diesel. My point is that x86 has a lot of advantages.
So does ARM. I'll point out that in Europe, diesel is cheaper, more efficient and just as easily available as gasoline. Both are suited to different tasks. ARM tends to be good at low-power computing, a market x86 has had a hard time penetrating.
Besides, if having a nice architecture was worth anything SPARC would've displaced x86 back in the 90s. Even if Intels new low-power x86s should be the Betamax to ARM's VHS, well, I still expect them to be Betamax: Interesting for a few niches but shunned by the consumer market because of a rather mediocre first offering.
And 3D viewing is the new rage. It could be fad, but I suspect it will have staying power. Especially in hand-held gaming consoles. Definitely nothing like car vs. airplane, at least in my opinion.
I expect it to be constrained to a niche due to ergonomics concerns. 3D displays tend to suffer from biological compatibility issues, which is inherently market-limiting. Nintendo's new handheld might work but if I designed a new handheld I wouldn't sink money into a 3D display without a few further years of experience and technological refinement.
