If optional air travel (that is: travel that is not dictated by work or family emergencies) drops radically as a result of enhanced interro... er... body scanning, it will still be enough of a hit to the economy that politicians will see a need to act. Remember, morality and ethics don't motivate politicians; votes do. If you can threaten their ability to get votes to re-elect them, they will move heaven and earth to get the problem solved.

There are different kinds of threats.

Bus drivers are probably more concerned about being attacked by passengers who are not terrorists.

In planes, I think we need a graduated policy. I think we need to have a very firm policy about shooting down planes that don't communicate and veer off of established flight plans toward heavily populated areas. Yes, that will kill many people (probably on the ground as well as in the plane), but that's something we need to have in place and make very clear to make hijacking a plane NOT worth it.

Also in planes, I think it makes sense to radically ramp up the security when the plane is full of fuel. A plane that takes off with just enough fuel to get from Boston to NY is a moderate threat, but one that takes off with enough fuel to fly non-stop from Boston to LA is another order of threat entirely.

If we could perform risk analysis and take appropriate measures to prevent the kinds of incidents that can harm thousands of people, while not making 95 year old ladies take off their diapers, perhaps we could go back to being a nation that enjoys travel.

I'm not flying either, and it's not because I'm afraid a plane will be blown up. Security lines were already onerous before 9/11 and oddly, they only got really awful in the past 3-5 years.

Until the U.S. can learn to do real risk assessment, I don't think there's any value in our attempts to curb risk.

A better comparison is you left your damned storage keys in your published machine image. This is a security blunder. That security key is, in simplistic terms, a password. Don't give out your passwords on the Web.

This has nothing to do with the security of Amazon's infrastructure or services.

I think what the GP was thinking of was the consequences of buffer overflow errors, which, while obviously not impacting to OS stability, directly, can have serious consequences outside of your application.

It's true that there's a time and a place for a low-level language like C and for a high-level language. This is why I suggest programming at both levels. I'd even suggest trying out Java, C++ and something crazily abstract like Haskell if you have a chance. The more languages you know well, the better job you can do determining what the right tool for the job is.

However, the reason to program in C once you have your bearings on the basics is to allow you to understand the consequences of your actions in those high level languages. You'll think twice about quite a few things in high level languages once you understand what the costs involved are. Sometimes the costs are worth it, but being able to make that assessment is one of the most important skills of an advanced programmer.

Python, Perl, Lisp, Scheme, BASIC, Java, JavaScript, shell... it doesn't matter. Pick up whatever feels most comfortable, or whatever tool is most readily available. Learn the basics of flow control, data structures, and the most fundamental algorithms. Learn how to debug, document, log, build, install and maintain.

Once you've mastered these basics, move immediately to C (not C++). Programming in C will teach you what it is that you're actually doing in that high level language, and force you to manage everything that you want to do. You don't just instantiate a new class to get a thread or forget about that variable's storage when you're done with it. You have to think about the details and understand them, and even if your entire career is spent writing Python, you still need to understand these things. I'd recommend assembly, but there's a lot more well documented sample code in C to work from, and getting larger projects done in C is still practical, if somewhat painful.

Once C is something you're comfortable with, trying out C++ is an excellent next step, but that's probably best determined by what you want to / have to work on.

Again, what value is a 20 year plan that you know will be rendered moot? Outlining a set of guiding principles and building the capabilities required to quickly adapt to change would seem to me to be more effective.

Also, remember that for about 50 years in the 20th century, there was only moderate change int he scope of business communications. Telephones revolutionized business, and caused nearly everything that businesses did to change. The ability to effectively manage organizational units over long distances as well as to instantly communicate with customers and vendors anywhere caused a major upheaval in the corporate landscape, but after that it was a relatively smooth increase (Faxes changed things a bit, but not all that much).

The introduction of the Internet and more specifically, the World Wide Web has been far more disruptive than the telephone, however. It didn't just change communication fundamentally. It also introduced vast new markets and opportunities directly. 5 of the 30 DJIA companies are pure computer/Internet companies today, none of whom were on that list (and in most cases, didn't exist) prior to the 1980s.

But that's just the problem, isn't it? IT isn't a primary business asset, and never has been, in most industries. And yet, it sucks down as much as a third of many large organizations' budgets!

I'll point out, though, that I was replying to your comments about long-term planning, not the merits of cloud computing (which I think is horribly over-simplified in most contexts).

No, it explicitly does not do that, which is what Slashdot got so wrong.

Marijuana, under this proposed law, would still be illegal at the Federal level, but the Federal Government would have no power to act in cases where there was no inter-state transport.

So, for example, if you grow pot and sell it to your neighbor, only the state would get involved. If the state wants to prosecute you for that, they can. If you sell it to your friend in the next state over, on the other hand, then you're still committing a feral crime.

The Washington Post article clarifies this (even though they still refer to it as "legalization" in the beginning of the story) along with pouring cold water on the idea that this bill is even a long-shot. It's not going to make it out of committee.

While the above is atrocious legal advice, it's more likely than not going to be the norm. It's just sad that we're living in a time between the dark ages of zero digital media and the advent of a true digital media culture. It's patently obvious that we're going to have to get to the point that the average person can exist in a known legal state with respect to their media collection, but that's practically impossible today unless you single-source your media from a behemoth distributor who will stand behind you in court. That's not satisfactory.

My collection consists of a hodgepodge of tracks I ripped on Linux laptops at the dawn of cdparanoia from disks that have long since been lost during moves or destroyed, iTunes rips, iTunes purchases, Amazon MP3 free downloads, Amazon MP3 purchases, downloads from Jamendo and free online distributors and netlables, downloads from my musician friends of their own tracks, podcasts, and so on and so forth.

There is no practical way for me to "prove" that all of my music tracks are licensed for my use. It's simply an impossibility. Then we get into the legal grey area of ripping friends' CDs... the labels would love to claim that this is not legitimate, and that you must purchase the CDs in question... but that may or may not be defensible in court.

It gets even worse: legitimate bootlegs (e.g. of bands who allow recording); mashups; rips Web sources; recordings that include background music; and so forth.

In an ideal world, I'd say that we need the Federal Government to step in and provide a set of consumer- and artist- friendly media usage laws, but the problem is that the U.S. Congress is for sale, and most world governments follow U.S. lead with respect to media legislation (for the most part, but with notable exceptions).

Sigh.

I get upset with Glen Beck for calling everyone he disagrees with a Nazi, so I'd be a hypocrite if I didn't admonish you for going hyperbolic on this one.

The U.S. plan passed by Obama is just insurance (well, that's the largest part of it). But, that's not what universal healthcare is in general. There are quite a few different systems out there. I think it was Frontline that recently did a pretty good job of dissecting the British, German, Swiss and Japanese systems to compare the different ways that they provide healthcare. It's enlightening to actually see what these countries do and what works well and does not, rather than just listening to U.S. politicians dismissively refer to all universal healthcare systems outside the U.S. as "socialism".

PS: Ah yes, in fact, here's their comparison: http://www.pbs.org/wgbh/pages/frontline/sickaroundtheworld/countries/

Here's the problem from VZW's perspective: the usage you just pointed out is rare, right now, and completely new in terms of how their network gets used (e.g. movies). So, what should they do when usage, and thus costs start to go up? Just raise rates to pay for the extra towers and upstream bandwidth? Sure, they can do that, but joe average who just wants a way to phone, text and occasionally check email is probably not going to be able to justify the extra expense when his plan goes up dramatically.

So, what they're trying to do is create tiers, within which similar usage patterns pay for each other.

To my mind, the problems are: 1) There needs to be a tier with truly unlimited usage at the top so that that app I installed which gets far greedier than I thought it would, doesn't cost me a few thousand! 2) These companies need to start truly competing which each other in order to balance against the temptation to increase margins with every plan change, which I think is going to require opening up more spectrum to more players. The feds have to get involved in that last item. I also think that mergers (including T-Mobil) should require forfeiting most of the spectrum owned by the smaller of the two companies, to be divided up among other, smaller players in the market. Why? Because otherwise encouraging competition only feeds the merger-mill.

Why can't we discuss the risk of a coming ice age without it having to be a refutation of an unstated argument about its impact on global warming? Can't we just discuss it on its own merits?

The phrase, "not entirely true," is a clever deceit, often used in debates and the press, when the assumption is that the reader/listener doesn't know enough to see the trap. The fact of the matter is that, since the responder cannot know the full truth for certain, any use of this phrase is immediately void.

Oh really?

What would a 20 year plan for IT have looked like 20 years ago when the ViolaWWW Hypermedia Browser and PGP had just been introduced?

I know that anything I laid out would have to be course-corrected several dozen times between then and now, and would probably have slowed any company down when it came to being able to adapt swiftly and adopt relevant technologies.

I don't think businesses have gotten immature. I think every business that couldn't adapt to radical and fundamental change in the technology landscape went out of business.

Doom! Doom, I tells ya!

Truth is that there's no data in most companies that's not handled by external entities. Your medical data? Handed out to a dozen companies before you even get to see it (if you get to see it).

Instead of running around yelling about risk, why not actually perform a solid risk assessment? Start by analyzing the risks your own IT department poses. Especially in companies where what IT does has little to do with their core competency, it's usually a safe bet that they haven't managed to hire the best and the brightest (it's just not in their toolbox to select the right folks), so that risk is probably going to be key.