That's based on the premise that the model T was less expensive than a horse (even after a few years of TCO) yet, they weren't... Consumers could have kept using horses, but chose to switch to cars in huge numbers because of other advantages (they could do things like travel farther distances, ignore daily maintenance, etc) that were not really obvious at the time. Sure, it's easy to look back and say "of course the car was popular, its *the car*" but that was not a sure statement in 1908, otherwise Ford wouldn't have been the only one in the USA doing it so cheaply/successfully for the better part of 10 years.

To be true the issue is that they were inadequately specific questions and of an inadequate variety for regression, on top of a heaping dose of selection bias. Since you didn't post the questions, only the answers, I will go ahead and Jeopardy! it... "What is your Age?", "what is your gender?", "do you prefer slide-keyboards or virtual keyboards?", "essay portion worth 2/3 of your final grade".

Then there's the premise in your comment that the survey was "seeking out respondents who had used both a phone with a slideout keyboard and a phone with a virtual keyboard" which tells me that your survey may very well have gone in front of 10,000 respondents and found the 49 that even knew what a slideout keyboard was, skipping past the 9,951 that had never used one and were quite happy with their virtual keyboards. This is a bit of selection bias which will skew your statistics to the point of worthlessness.

Don't use Mechanical Turk as a crutch; it's not that far from representative, and nonrepresentative samples are often just as useful, thanks to regression. The real problem is that you asked all the wrong questions. I suggest, if you want to gain *any* sort of ground on your quest to shake up the cell phone industry from the ground up by revealing what you think customers really want, is to read the Freakonomics books, and follow that up with a (well thought out) question to the authors. This sort of thing (mostly the situation where you insist on one thing via all available observations, when the opposite is true) is right up their alley. If you still think you are sitting on some sort of secret, start your own handset company, and get rich off of all the customers that are apparently being ignored.

The old Henry Ford saying goes (not that he necessarily said it) "If I had asked people what they wanted, they would have said faster horses". Point being, you never know what a consumer will do (even if you are that consumer) when presented with a new/different set of choices. Consumers are flocking away from physical keyboards when given the choice. Consumers overwhelmingly prefer thinner phones (since no matter how much more you charge, you can't get a slide out keyboard phone to be nearly as thin as one without) so when presented with the choice, they gladly give up the keyboard (if they ever wanted it) for a thinner phone.

Jesus this submission is so sad. Bennett, you overlooked the #1 rule of consumers... FIRST IMPRESSION IS EVERYTHING. If someone sees a lineup of ten phones in a showroom, nine of them thin and svelte and made of nice tightly constructed materials, while the tenth is twice as thick in order to accommodate the keyboard, they will immediately gravitate away from it. Yet, you overlooked this obvious decision point. Add to that the other rather obvious trend of smartphones: everyone wants to be Apple. The more your phone looks like an iPhone (to hell with what the courts think, amiright) the better it will sell. A slide out keyboard? Steve Jobs would come back to life as a zombie and have a personal sit-down to fire everyone at Apple if that ever happened. He would even skip eating their inferior, clunk-loving brains out of principal. So there you have it, please take a few more minutes to think through your next submission, and maybe you will actually have something insightful to say.

p.s. onscreen keyboards really do work great if you give them a chance. machine learning techniques by Google and Swype are getting pretty good at learning how and what you type, to allow for very fast and reliable input under even less than ideal conditions.

Access via ADB doesnt require root to get to the point where you can confirm/deny the existence of functioning memory. It looks like the Transformer line has an out of band update method by installing a microSD card with the flash zip, and doing a startup with certain buttons pressed. If it can be coaxed through this process (even with a dead screen) it would wipe any previous user data. Watching the device state via the USB port and ADB would be helpful to know if the device is likely to respond in that kind of scenario.

Do you get the ADB device to register if you plug it in to a PC via USB and turn it on? That would be your only hope to wiping it assuming the screen is damaged but the SoC/flash still works to some extent. Also, have you tried opening it up? A similar thing happened to my Nexus device, and after popping the back cover off it turns out that the drop caused the battery to slide to one side, and come unplugged. Relocating the battery, adding a little more double sided tape, and snapping it all back together had it good as new in under 5 minutes.

Why would rates double as a result of putting into place a plan (and probably a few layers of communications systems on top of already existing infrastructure) to mitigate the problem before it starts? Oh right, because we would have to pay for a team at NASA, a team at FERC, a team at each of the regional ISO, etc. to all do the same thing? Ugh. Put NASA in charge, they got us to the moon damnit. If rocket scientists cant fix it, no one can.

Actually professor you might want to take a second look at those figures. A nuclear plant relies entirely on *already produced electricity* for safe operation. With a normally functioning grid, this is not an issue. Take that out of the picture (in a scenario like a CME hit) and it will have to fall back on site generators (the local turbine generation is likely to go down with the grid) which hopefully will have been isolated from the effects of the CME and can be instantly switched in to the site system to take over and shut the plant down. However, if any of those switching components went bad during the CME hit, it could be hours before they are repaired, which starts to push the cooling safety margins to the limit (the plant is, after all, still producing heat as if it had a job to do). There are certainly good disaster plans in effect at nuclear plants for situations similar to this, but do you really want to test them all at once? There are bound to be holes. Mushroom cloud style explosions are out of the question, but we know from experience with Fukushima that all kinds of bad things can happen (including lots of little explosions of errant hydrogen) when plants go dark and can't be shut down safely.

Generator-powered factories producing generators would suddenly be very very valuable.

The real question we should be asking is; why doesn't NASA have the authority to order a nationwide grid shutdown in the event that one of their several satellites dedicated strictly to predicting and identifying solar disruptions actually works and warns us before it happens? We have spent billions on this already, why not put that to use instead of fear mongering about how long it would take to manufacture a bunch of high voltage transformers?

The other key information is this: The SSA has 65,000 employees and is in charge of a staggering $736B per year (as of 2011, and it continues to rise). And we are here having a pissing match about all the reasons that $300M is too much to spend on the system that is supposed to make sense of over 300 million "customers" (1 dollar per customer?) One half of one percent of their annual budget is too much to get this right? Most corps spend upwards of 10% of their annual revenue on IT, and surely the SSA is not most corps but the scope of what they do is really impossible to underestimate so a project in the hundreds of millions shouldn't make anyone flinch.

The real missing key information is exactly why this kind of story is surprising, on any level, to anyone? My gut says it's the fake shock of someone who would protest anything that came out of the SSA.

Something tells me that with girlsgonewild.com, the "interaction" is mostly "client-side" so the, er, "workload" is actually minimal. And the use case count, I believe, still stands at 1, and they are at best appealing to exactly half of the US population. It's a bit different than a place like the Social Security Administration, an org that has taken on the unenviable task of managing retirement and disability insurance for *every goddamn american* which is a pretty ludicrous scope. If raw horsepower were the issue, yes bring in outside help. The real problem (or at least one of them) is that of all 65,000 employees, many of them have a specific task since the aforementioned scope is so grand. Try finding a way to economize when you are basically building a system for a small clerical office, and then doing it about 15,000 times with each iteration just different enough from the last to require constant rewrites.

"What was the experience of riding a bicycle has become the equivalent of traveling by jumbo jet; replete with the delays, inspections, limitations on personal choices, and sudden, unexplained cancellations — all at a significantly higher cost."

You can't exactly get everywhere you need to go via bicycle these days. Blame globalization.

Here is the form you were looking for: "are you or have you ever posted to Slashdot as Anonymous Coward? Ok next question: Are you or have you ever browsed slashdot at -1?"

we have a subversive on our hands!!!

And wind up with *super*smallpox? Good fucking plan, Einstein!

Actually, good fucking plan. Let's do it.