Talk:Hardware: Difference between revisions

From OLPC
Jump to navigation Jump to search
(Basic lemonade reciprs)
m (add ToC)
Line 1: Line 1:
<div>
__TOC__
</div>

== Older talk ==
== Older talk ==



Revision as of 10:55, 9 June 2010

Older talk

More discussion can be found in Talk:Hardware_specification, as Hardware was derived from Hardware specification.

Against the integrated pad (XO-3) industrial design

The original industrial design (of the XO-1) was splendid. I am curious how field experience has validated this opinion or not. How often were the different physical configurations used, and for how long? How quickly/often did the articulation mechanics break down from wear or abuse, especially in challenging environments? Was repair practical? Did software issues discourage use of the e-book configuration?


When one has a screen that needs protection, a basic clamshell design is needed. In principle, one could use a rigid detachable faceplate or case. But even adults might lose these, to say nothing of small children!


Because the XO-2 design kept the clamshell, including weight asymmetry which enabled a secure vertical landscape-AR screen ("laptop mode"), I could see some merit in it. What I didn't like was the idea of combining a display surface with one the user touches. Why? Keeping a touched screen clean without regular cleaning remains an UNSOLVED problem! See IPad Side Effect: Greasy Fingerprints. This challenge seems ESPECIALLY problematic when one projects use in areas with lots of dust, which provides abrasive grit to scratch the AR coating of the screen, if not the screen glass itself, as a side effect of frequently degreasing the screen of fingerprints with a cleaning rag. See Ask Slashdot: Rugged Laptop/Tablet Suggestions, 2010 Version?


If it REALLY was of great advantage to provide a large-area touch surface, then one should have made each of the two XO-2 sections provide either the display function, OR the touch function, but NOT both. Specialization would have also reduced cost and improved optimal design of the selected function.


And I also challenge the idea that a large touch area is usually worth the real estate. Long ago, we learned we could lift our mouse to make do with little surface area, not to mention software tricks like de/acceleration. And even if we allow that haptic feedback could provide keyboard finger guidance as good as passive (semi-)rigid surface sculpturing, the latter requires no energy or transducers.


Even if dirtying the screen were not an issue, there are problems with directly touching the screen. Vertical viewing surfaces reduce neck strain. But if that then means one has to hold one's arm out to touch the screen, one has traded neck strain for arm fatigue. I also notice that my hands sadly lack transparency. There are times when one wants to point at the screen and see all of it at the same time, which displacing the touch and viewing surfaces enables. One should also note that virtual hands and fingers on a screen can vary in size, unlike real hands and fingers, assisting delicate interaction. While software lets us do this whether we touch the viewing surface or another one, clearly having a "natural-sized" hand on the screen itself offers no advantage. Nature's way is not always best.


As you can imagine, if I view the move from the XO-1 to the XO-2 industrial design as a step backward, then I truly dislike the proposed XO-3 design of a tablet whose display integrates a touchscreen! Absent a case (which can easily be lost), there is no screen protection in transit and storage. The screen will need to be regularly swabbed clean of finger grease, which will offer lots of opportunities for dust to scratch it.


Surely if I have no tree or rock against which to prop up my tablet as I lie prone on the ground, then I can lay it on the ground or repose myself in a number of ways while I hold it in my hands like an old-fashioned codex - not too bad. But should I have a table and chair, it is a pity that the tablet cannot stand vertically on its own, leaving my hands free for work, or maybe just swatting flies.


And one day making the screen out of plastic, instead of glass, would make things EVEN WORSE. Glass is much more resistant to scratching than (soft) plastic (the potential AR coating issue aside). Talking about how, unlike a glass screen, a plastic screen is "unbreakable," only makes sense if one totally ignores the ease with which the latter will be slowly but surely destroyed by abrasive cleanings.


Is the current proposed OLPC design being driven by fickle consumer fashion in the United States, rather than by thoughtful ergonomic and maintenance issues, which was the case with the original design? Perhaps I am missing something and I can be educated. I will admit to doing little reflection prior to making these remarks.


- Docdtv 18:44, 3 June 2010 (UTC)


Mitigating woes of the integrated pad (XO-3) industrial design

Bromide: If one is stuck with lemons, one can still make lemonade. How might one mitigate the potential problems of the XO-3 industrial design outlined above?


Apple sells a carrying case for its new iPad, a view-mainly media device similar in design to the XO-3, and its rollout has inspired a third-party carrying case mini-industry, as the WSJ reports at iPad Fans Face New Dilemma With Tough-to-Tote Gadget.


Below are some rudimentary ideas for dealing with XO-3 issues. Could an organization like the D-Labdo much better, or maybe just help with design and optimal fabrication/integration? Or do groups like that dislike OLPC for various potential reasons?


The simplest design to protect the screen against impact damage is a rigid faceshield. If sealing can be provided as well, it can also protect against dirt and moisture.


Perhaps a snug-fitting shallow tray design might be useful. The XO-3 could lie face down in it during transport and face-up during use. The latter mode decreases the chance the detachable tray is left behind and lost due to oversight. The XO-3 should be clamped into the tray for its safety and the tray should feature a carrying handle.


An elaboration of the tray could add a means to use it as the foot of a stand which erects the XO-3 into an almost vertical orientation, tilted slighly back, a mode very useful when used as an e-book or an image screen.


Finally, one can imagine elaborating the tray so that it adds a keyboard and trackpad, basically remaking the XO-3 to try recovering the safety and ergonomic virtues of the XO-1 clamshell design. This assumes there is some means to cybernetically connect the tray with the XO-3, perhaps with wires, or perhaps without.


If one cannot add such an additional tactile input surface to the XO-3, other trouble mitigation strategies exist.


One can mandate all installed software segregate the screen into geographically consistent no-touch and touchable ghettos. The latter zone is much smaller than the former and hosts virtual keyboards and trackpads containing no subtle visual details, making both finger grease and the abrasive effects of removing it more tolerable. This touchable area should be restricted to one edge of the screen, or at very most, two.


The availability of a powerful microprocessor also encourages an interesting possibilty of potential utility to low-intensity-input applications. General speech recognition is hard, but recognition of a very small sound vector set, like one representing the few distinct keys of a keyboard, is much less challenging, both in doing speaker-specific training and in actual application. This can be made even more robust at the expense of speed by using the traditional crutch of a "phonetic alphabet," e.g. ALPHA - BRAVO - CHARLIE etc.


Finger grease issues aside, sometimes it would be helpful to do simple things like flip or zoom an e-book page with one's voice, allowing one's hands to do other useful work, including that which makes them especially dirty. This includes things like cooking, building, repairing, cleaning, etc.


- Docdtv 06:52, 9 June 2010 (UTC)