|
WINDOWS & .NET WATCH:
Thinking Outside the Aux
By Larry O'Brien
 December 15, 2004 — Microsoft has begun talking about auxiliary displays for laptops. My least-favorite explanation of these “aux” displays is a comparison to the time-telling LCDs that grace the outside of many clamshell-style cell phones and that provide all the inconvenience of a pocket watch and none of the style.
Does anyone want, much less need, a display whose flexibility can be out-matched by a Happy Meal giveaway? Of course not. So 30 seconds into any discussion of aux displays, you’re talking about bitmapped displays of at least several kilopixels—which undermines the premise that aux displays will have no discernible effect on the cost of a laptop.
So a laptop with an aux display will have an incremental cost. Which means that aux displays will be optional. Which means that developers can’t count on their presence. Which means that virtually the only code that will target aux displays will be from Microsoft (which can issue an internal fiat) and from niche developers.
But put aside the display temporarily. Consider a fundamental disparity between a powered-down laptop and a cell phone or PDA in standby: The laptop requires intervention to become active, while the devices can wake themselves. The laptop is in a coma, while the devices are in REM sleep.
Now, imagine that your laptop had a new power mode that was sustainable for days on end, a REM mode that woke occasionally, glanced at the system clock, groggily processed some important thought, and then drifted back to sleep. That is worth getting excited about.
Every conversation with Microsoft about mobile computing involves battery life. Battery technology may be advancing but hardly along the geometrical curves of transistor density or storage capacity.
Meanwhile, people want brighter and more pixel-dense displays, hard drives capable of holding video collections, and WiFi access everywhere, all in thinner, lighter and cooler packages.
It is an intractable problem. Keeping your laptop alive throughout a transcontinental flight involves darkened screens, radio silence and healthy doses of prayer.
Changes in software can affect battery life significantly. Microsoftians boast of the win their SPOT watches got from nothing-but-software changes, while Intel talks as much about power management as performance optimization (well, almost as much).
Everything from spin-waits to video playback, wireless network traffic to drive access—all of these things can drain batteries quickly. Yet there are no tools for programming for battery longevity: Until there’s a “power profiler” that gives the developer the fundamental feedback necessary to optimize for mobility, the situation can only get worse.
Returning to the subject of displays, it seems to me that one of the greatest gambles of Avalon (the new display stack that was a “Pillar of Longhorn” until recently being decoupled from the operating system) is its energy efficiency.
One can make a case for betting either way on its fundamental efficiency: worse (involving, as it does, managed code) or better (by improving context passing and switching).
It’s a no-brainer that the high-gloss demo interfaces we’ve seen, with video and animated transition effects, aren’t going to be friendly to those reliant on batteries. The display is, without a doubt, the great battery hog of a laptop. That’s why the key to the aux display concept is to decouple it from the display.
Inherently, aux display scenarios must involve something less than full laptop computational and display capabilities.
I can imagine a REM-mode application that woke up every minute, queried a Bluetooth GPS, and went back to sleep if the location were within a mile of a precalculated track. Similarly, I can imagine REM-mode applications that looked for upcoming appointments, checked for e-mail (toggling the WiFi radio off before returning to sleep) and so forth.
In every scenario I can imagine, I want one of two things to happen vis-à-vis the display: Either I want an audio cue, or I want the laptop to come fully awake or prepare to come fully awake. And in either case, I want to trigger an event on the auxiliary display that I already carry—my cell phone.
Every single person who will buy the high-end laptops that will support the aux display is already carrying a cell phone. The aux display SDK should assume nothing more about the display than a publish/subscribe service to which Bluetooth phones and media players can attach.
If one of the subscribers is an on-board cell-sized display, so be it, but if Microsoft is to attract corporate developers, it would be foolish to emphasize bad display capabilities over the attractiveness of REM-mode applications.
Larry O’Brien is a technology consultant and analyst, and the founding editor of Software Development Magazine.
Larry O’Brien is an independent technology consultant and analyst, and the founding editor of Software Development Magazine. |
|
Advertisement
Click here for a complete listing of Windows Watch Columns.
Click here to see a complete Column Archive. |
Back to Top
|