What the next generation of HP Palmtops should have been.

Okay, it doesn't look like the 300LX series is going to measure up, and the 620LX, though color, is merely following the tradition of "larger is better."  So we're going to have to petition HP to improve the 400LX line. Or will they call it 400LX? After all, there's only 400 and 500LX until we get to 600LX... and that and 700LX have already been used. But now that they're using 20's (as in 320LX, and 620LX) perhaps we'll have some leeway. Oh, the uncertainty.

Here's some preliminary specifications:

• Same size and preferably same weight as the 200LX.
• At least 640x240 screen, fit into same space as 200LX. (But the screen could go all the way to the edges...I wouldn't mind...) 640x480, though desirable for VGA compatibility, would be too distorted to be practical, at the current size of the unit. So 640x240 would be decent. Perhaps if the screen vertically went all the way to the edges, and the horizontal part stayed at its current width, 640x350 would be acceptable. That would give us true EGA compatibility capabilities. At 640x240, we could emulate VGA, but it'd be a pain to switch back and forth between the top and bottom halves of the screen. Perhaps we'd better just settle for EGA. As for color...we should at least get 16 grey-scales, preferably 256. And the ultimate: color. How hard can it be to cram 640x350 color pixels in there, anyway? And of course a backlight is essential...and a touch screen is a good idea. Who wants to play solitaire using the arrow keys? And it should bend around like the OmniGo. Sometimes you want to write, sometimes you want to type. We need flexibility!
• A low-power 486 CPU would be appropriate. None of this fancy-shmancy-pants RISC crap. If it's not Intel-compatible, we don't want it. Because then it's not a palmtop computer, it's just a fancy digital diary. Like all Windows CE machines. Because unless you can run apps on the palmtop you can also run on the desktop...how can you call it a palmtop PC? Intel it is. (If, of course, they come out with a speedy RISC chip that is Intel compatible, or does a good job of speedily emulating an Intel CPU, that's a different matter. As long as it's extremely low-power...) In any case, it should be much snappier than the 186 we now have.
• We need more RAM, as well. Since the palmtop would now be capable of running more current (and larger) software, we'd need more space. 16 megs of ram for usage should be adequate, but with the small size of RAM today, we should be able to up it to 32 at least. For the lower-cost units, 8 megs could be substituted, but once again, it shouldn't be soldered in--perhaps a special socket.
• Storage space must be increased as well! All units should come with at least 500 megs of flash RAM internally. That's a minimum, mind you. If it sounds like overkill...it's not. We'll need it. Or at least some method of economical large-scale storage.
• Expandability is a good idea as well. The PCMCIA controller should be standardized, and another slot could be added on the other side, if the IR port and serial port were moved to the back.
• Sound functions should be improved. We should at least be able to record and playback digital sounds. Volume control should be improved as well--you don't want your palmtop beeping during an important meeting, or a test at school. Sound Blaster compatibility would be nice, but not necessary. Just something that has decent enough sound that voice-recognition software would work, since after all, if you could run that on the palmtop, it would greatly improve functionality.
• Some kind of wireless communication should be built in, and I'm not talking about local IR transfer, either. We're talking long-distance. Something along the lines of a transceiver for the RAM or ARDIS networks. Cellular might be nice, especially if the sound functions of the palmtop are decent enough to provide real-time full duplex conversations.
• As noted on Rob's pages, the keyboards should be switchable, for different applications. That would free up more space in the unit for other things, such as PCMCIA sockets, flash memory, etc., since the circuitry for the keyboards could be placed inside each keyboard package, and the keyboards could stick up a bit from the rest of the unit. One keyboard could have larger keys and no number pad, for word processing...another could have a larger number pad, and better arrow keys, for spreadsheets and other number-crunching apps. One that Rob's pages don't mention is a calculator keyboard. This would be good for the calculator app, with specialized keys instead of the meager selection of dual functions on the 200LX keyboard. Especially since there will be an app remarkably similar to the HP48...
• The main operating system? Hmm...GEOS seems to have a lot of pluses, but with the power we've got now, I suppose a new version of Windows CE (or another OS) that runs desktop apps would be fine. Of course, on a unit like the one we're designing here, you could run Windows 95 if you wanted to, or even Linux...
• The power is the most important part. AA's would be nice, but probably impractical to run such a powerful machine as this for very long. It'll probably take a much more powerful power source. A special high-density power pack? High-efficiency solar cells? A very small nuclear power pack?
• Let's not restrain ourselves. Here's a thought: why settle for a small CGA, EGA, or at best, VGA screen? If there was NO screen on the unit, we could make the whole thing into a keyboard and other hardware, effectively doubling our space. The screen could be transferred over to a pair of glasses, the likes of which you can see in electronics stores. Sure, the ones we have now aren't all that great, but in a few years, who knows? We could have a SuperVGA display laid onto an ordinary pair of glasses. That's better all the way around: a heads-up display would be easier to see and easier on the eyes than a tiny cramped screen, and there wouldn't be any problem with distortion. (Caused by the screen being wider than it is tall.) Full color would be possible! Of course, some people probably already have glasses. In that case, we would have three options: install hardware into their existing pair of glasses, supply prescription glasses with screens in them, or have anybody who wants to use the 400LX get contacts or eye surgery. The middle option is probably the best way to go. There's also the problem of having a bulky signal-transmitting cable run from the palmtop to the user's glasses. That could be problematic. Perhaps a modulated laser would be able to carry enough data fast enough for a decent screen...  

That's all for now. More will be added as I procrastinate important work to update this site...