A couple quick searches tell me that there are about 250,000 miles of railroad in the US and about 3,000,000 miles of paved road. Waze running on my smartphone seems to be able to warn me when I'm going 10 MPH over the limit pretty reliably. Given that the railroad network is less dense than auto roads it seems clear that it should be doable to create a similar device for trains that, at a minimum, warns the engineer when they are exceeding the speed limit for a given stretch of track. Throw in mobile data based updates of track conditions to handle track damage or construction work and it seems like something like this should improve the situation.

Building a perfect automated speed control system is not the only way to improved safety, we can move the needle a lot by providing by human operators with better data and warnings. Don't let the quest for the perfect solution block comparatively easy incremental progress.

Why do you believe Keurig is badly managed? They made a mistake and now seem to have realized their mistake (time will tell). Every company I've ever worked for (including the one that I owned) made mistakes and, in most cases, I think they learned from then and became better companies.

Also, why do you think the next company they work for will be managed better? If we hope to see a company die each time they make a mistake then the only companies left will be the ones that are still working on their first big mistake.

The point was, there was some justice in the failure of Arthur Andersen because their actions with Enron resulted in a lot of people losing a lot of money when Enron collapsed. That is not the case here, Keurig's DRM scheme may have driven up the price of a cup of coffee but nobody lost anything substantial as a result.

It never ceases to amaze me how unforgiving people can be. It might make you feel good to say Keurig should die but there are hundreds (thousands?) of people working at Keurig that had nothing to do with this decision and it seems kind of heartless to say they should all lose their jobs because a few executives made a bad decision later changed their mind.

I know a guy who worked at Arthur Andersen at the time they went out of business. The part of the company where he worked had no connection to Enron but they all lost their jobs in the fallout. He indicated one time that the part of the company that was involved with the Enron was something like 70 people out of 85,000. That's pretty harsh justice but at least in case of Arthur Anderson you can show that a lot of people were hurt when Enron went bankrupt so there is some justice in it.

In the case of Keurig nobody has lost their job or their retirement savings because of this, it's just a cup of coffee.

That IBM signed that contract is not nearly as crazy as it sounds (today) when you look at what was going on 50 years ago.

IBM was involved in an anti-trust fight over the way they excluded 3rd parties from making peripherals for their mainframes and they did not want to give the DOJ any additional ammo. At the time the real money in the computer industry was in the mainframe market (which IBM pretty much owned) and the PC market still looked a lot like a hobbyist thing.

IBM did what corporations and individuals almost always do, try to protect what you have now rather than risking that on an uncertain future thing.

I was in the Army back in the 80's when the first Humvees were being rolled out and they offer way better protection than the hold Jeep's and CUCV M1009s (essentially a tricked out Chevy K5 Blazer) that they replaced. They were designed for different a battlefield than the one we fought on in Iraq and Afghanistan, a battlefield were the threats were much more potent and a few pounds of Kevlar underbody armor was not going to make a bit of difference.

The Army _should_ have acted more quickly to upgrade or replace them with something that offered better protection against IEDs but that was not the fault of the people who designed the Humvee in the first place.

My wife and I (mostly my wife) home schooled our son through 5th grade and then transitioned him to a private school. My wife and I are both college educated, I have a job that pays well enough that my wife does not need to work outside the home and as such has chosen to be a home wife/mom.

The main drawbacks to home schooling that we encountered were...

1) It's a lot of work, the parent doing the schooling should consider it their job. If that parent also needs/wants to work outside the home I would find it difficult to recommend home schooling.

2) There is some expense (but a lot less than the private school he attends now).

3) It gets harder as they get older and the material become more difficult. My wife (who is not a math person) had no difficulty teaching all of the math up through 5th grade, but now that he is doing algebra, helping with math homework is something that falls to me.

4) Ignorant people wondered if we were in some sort of cult and were attempting to teach our son that the earth is flat, or that he would become some sort of sociopath because they figure we probably just lock him in a closet with a pile of text books.

The benefits that we enjoyed were...

1) Massive flexibility, during my son's 4th grade year he was studying American history so we incorporated a 4-week trip to the east coast into his program for the year. He was able to see in person many of the famous historical locations that he had, up to that point, only read about. My son probably did 10 times as many field trips by the end of 5th grade as I did in my entire 13 years of public K-12 education.

2) Class size, in every education funding debate I hear, one of the biggest things that teacher's are always asking for is a lower student:teacher ratio. Homeschooling is the absolute sweet spot there, the student teacher ratio in our home school was 1:1.

3) Tailored lesson plans, our son's teacher was the person in the world who knows his strengths & weaknesses the best. I don't think I can overstate this, no teacher will ever know your kids as well as you do or be able to customize the way they instruct them to the same degree. For example when he was younger our son had some vision issues that required special glasses and vision therapy, this caused him to struggle with reading. My wife was able to compensate for this by using more audio book in his lessons while we worked through the reading issues. I think it's pretty likely that without this ability to fit the teaching style to his needs he would have struggled badly in a typical classroom.

4) How many traditional schools let you come to class in your pajamas?

A couple other thoughts...

1) People often wonder why you would want to do yourself that a professional educator has been extensively trained to do. I think this sort of misses the mark, my wife and I were not doing what a professional teacher does (manage a classroom full of kids from many different family backgrounds with very different individual strengths/weaknesses and do it on a very tight schedule).

2) There are TONs of groups supporting home schoolers (at least in our area). There are several great Co-ops where kids can sign up for classes in area that parents feel they are not equipped to teach (e.g. our son was able to take classes like Archery and Robotics at the co-op that would not have been available at a traditional elementary school.

In the end I think the decision will depend a lot on your personal circumstances, if you don't feel good about your local public schools and cannot afford private school then home schooling is worth considering. Are your kids well outside the norms (either above or below) in some areas, if so then home schooling will allow you do adapt their education to help them catch up or allow them to excel. How does your wife get along with your kids? If they already butt heads over many things then making her mom and teacher may not be the best plan; as our son approached the teen years it was more and more difficult for her to act as both mom and teacher and that was a factor in our change from home school to traditional school.

That's my $0.02 worth.

This MIGHT not be quite as idiotic as it sounds initially. Grenade launchers come in two varieties, 40mm and 37mm.

The 40mm ones are what the military uses, a number of lethal rounds are manufactured for them (HE, HEDP, buckshot, etc.) they have rifled barrels to improve range and accuracy. They are classified as destructive devices and are very tightly controlled.

The 37mm ones are non-restricted (individuals can buy them in many states) and are classified as flare guns, only low-lethality rounds are manufactured in 37mm and they are smoothbore. They are used for crowd control and signaling. I believe private citizens can only purchase signaling rounds (smoke, flares) and others like teargas and rubber bullets are restricted to law enforcement only.

I'm not sure why a school district needs to fire teargas or rubber bullets (isn't that when you call the police) but it would at least be less idiotic that giving them grenade launchers capable of firing explosive rounds.

Yes, by passing a huge body of tax law that is selectively enforced against foreign companies (and some domestic companies that have the nerve to object to government policies) you can absolutely end all the tax evasion that you don't choose to ignore.

The Faraday cage idea sounds nice until you think it through. The devices on the flight desk are connected to miles of wire that runs to every end of the aircraft (I seem to recall reading that the 787 has nearly 100km of wire in it). Not all of this is part of the flight control systems but a significant chunk of it is and all that wire is an antenna potentially carrying RF signals back to the flight deck where cross talk can allow it to affect other systems.

This can all be improved by using things like fiber optics but as long as the gear on the flight deck has to talk to the engines, control surfaces, landing gear, etc. you are not going to solve this problem by just wrapping a Faraday cage around it.

So you're saying, take an OS and application model designed around single screen touch based devices and port that to multi-monitor PCs that don't have touch screen but have mice and keyboards instead; and you think that will be a user experience that people will love?

Didn't some other software company try that recently and it sucked?

Just sayin'

No, not assuming that; assuming that map data is _one_ of the inputs to the system and as such the system will less capable operating in areas for which map data is not available which may increase the chances that the system will need some sort of human input to determine the right course of action.

It seems pretty clear that there is going to be a transition period where autonomous vehicles will absolutely need to have the ability to let the driver takeover for situations like:

1) Driving in places where you usually are not supposed to due to road work or an accident.
2) Driving in places that no map data is available for yet.
3) Getting a vehicle onto a lift at the repair shop for servicing.
4) Pulling a trailer, this adds an entire new level of difficulty that I suspect autonomous car makers will not tackle in the first wave of vehicles.

Even in the long run it seems like there will be edge cases when you will need to do things that you would typically want the computer to prevent, like:

1) Drive into a lake or river, people do this when launching/recovering boats or when going ice fishing in certain areas.
2) Driving off-road on private property.
3) Intentionally drive into obstacles, I've sometime had to drive my truck through brush to get were I needed to be.

As long as a vehicle has manual overrides to allow for these sort of exceptions then a drivers license should be required.

I think the idea that organizations may want to keep control of things that are highly sensitive makes a lot of sense (from a security perspective) but from an reliability perspective I don't know that I buy it. People seem to have this built in sense that "I'm safer when I'm in control" but that is not always the case, or perhaps it's more a case of "you never actually have as much control as you would like to believe".

Example: If you look at deaths per billion kilometers traveled; Air, Bus & Rail (modes of transportation where control has been handed over to someone else) are all substantially safer than travel by car (where you are in control). I know that you can slice and dice these numbers different ways (e.g. deaths/journey or deaths/hour) and get somewhat different results but even when looked at in those ways bus and rail are still _always_ safer than car travel so, in this case at least, being in control does not improve safety.

I'm not saying that the cloud is the right solution for everything but I would really like to see more data on how up-time for cloud based services compares to on premise solutions before jumping to any conclusions.

In the early days of the mainframe there were many vendors (IBM, Univac, Burroughs, CDC, Honeywell, GE, RCA); IBM was not the first but by the 70s they owned over 70% of the market and today they own 90% of the market (but almost nobody cares because mainframe sales were long ago overtaken by PC sales.).

In the early days of the PC there were many OS vendors (Microsoft, Apple, Tandy, Digital Research, Commodore, Atari), Microsoft was not the first but by the mid-90s Microsoft owned over 90% of the market and still does today (but fewer and fewer people care because PCs have been overtaken by mobile device sales).

Today there are many mobile OS vendors (Google, Apple, Microsoft, RIM, Ubuntu), Google was not the first but today they own 70% of the market, in ten years ???

In each generation there has been one dominant platform and the owner of that platform has used their market dominance to attempt to exclude others. Why exactly do you think things are going to play out differently this time around? I'm not saying that this is desirable but absent some basic change in the way the industry works it seems somewhat inevitable.

"Those who cannot remember the past are condemned to repeat it".
- George Santayana

If the information that validates the boot loader can be easily updated after the OS install then it can also be easily updated by malware as part of it's own install process.

But "NO", you say. It would require that your authenticate as an admin user to do that.

Unfortunate fact, if you write a piece of software that pops up a dialog that says "Please enter your admin password so Slimeware can install something really important that you can't live without!" a lot of people are going to enter their password (which is probably 1234 anyway). Existing malware has demonstrated this time and time again.

Your suggestion for a trivial solution may work fine for your average Slashdotter but will be ineffective in the world at large for the same reason that you are safer traveling by commercial airline than you are traveling by car. An airline pilot does not have to be perfect for safety to be better on average, he just has to be sufficiently "above average" when compared to the typical driver to tip the scales.