Actually, there are a number of issues with desalination as it presently exists. The cost of plant construction and operations puts the cost of water supply at roughly triple that of traditional methods. They also depend on fossil fuels and thus contribute to greenhouse gasses. They produce brine and boron contaminants that can not easily be disposed of on either land or sea without potentially significant impact of the wildlife exposed to them. They're kind of a mess, really, as a solution for fresh water when compared to a simple pipeline.

My guess would be that two groups, those that express the gene and those that do not have a 6 point difference in IQ on average, in favor of those with the gene.

Software updates (including the device OS) are pretty commonly done via network connections these days, be they wifi or some other other network (LTE perhaps). Now, I'm not saying that your hackability argument has no truth to it, since over the air updates are often tightly controlled as to the source of the update, so installing a non-approved update might be trickier. In principle however, this depends far more on the specifics of the controls the device has put into place to stop unauthorized firmware installation than it does with the means of connectivity employed to make it happen. Keep the port if you have to, but I haven't connected my iPhone to a computer via USB for data transfer of any kind in something like a year now.

It could also readily be argued that re-flashing with non-standard firmware and low-level debugging aren't consumer features and thus don't represent a substantial use case to drive the inclusion of the port for the sake of the percentage of users that would actually use it. I could see a JTAG connection for repair and debug for instance, but why a full USB port?

MHL has seen relatively light adoption. There are other ways to handle A/V out than a physical connection though. AirPlay comes to mind. The bandwidth required to send compressed audio and video at with multiple digital audio channels, high definition resolutions and high frame rates is well within the capabilities of WiFi. I stream 1080p video to my Roku from my Plex server all the time over WiFi.

I guess what I'm trying to say here is that you can argue that a physical connection is somehow better than a wireless one on the grounds of it being faster, more secure, requiring less power, or whatever you like. The real question though is is the experience of plugging in a physical cable and getting those things more compelling to the consumer than the experience of getting "good enough" capabilities without the need for a cable? Once you pass the "good enough" barrier, the cable vs wireless argument is pretty much a done deal for most people.

I'm sure there's definitely a whiz-bang factor at work here, but I think there's more to it than that.

Power is the last reason you need to connect a cable to most wireless devices now. Have low bandwidth data needs communicated at short distances (both a limitation and a feature)? There's NFC. Have one or two-way audio, or higher speed data transmission with the range of a room or two? There's Bluetooth. Need to communicate at greater range with much higher bandwidth? There's Wifi. Need to charge your device? There's Qi.

Why do I need a USB port anymore? My phone syncs over my WiFi network. It talks to my car audio system via Bluetooth. It talks to my car speaker phone or my headset via Bluetooth too. It just might, someday very soon, pay for my purchase via NFC as I swipe it at the checkout lane. Someday soon, you may even pair your device with Bluetooth accessories or join it to a WiFi network by passing it over a NFC pad. So I have to find the right cable and power adapter to charge it? Why should I have to do that when there's Qi?

Given that Qi can be combined with NFC, its possible that there is some hardware design synergy that makes the cost of implementing both together more palatable than implementing either alone. Honestly, if Apple were a member of the Wireless Power Consortium, I'd expect the new iPhone to have both NFC and Qi. Even without that membership, it just might anyway.

> Yes, Bacula is the only real solution

What a minute. Really?

OP is asking for a linux console application that can perform a backup over multiple block devices (in this case externally attached hot-plugable drives like USB), and Bacula is what you come up with as the *only* real solution? Obviously you've never heard of dump.

http://linux.about.com/od/commands/l/blcmdl8_dump.htm

I pulled a similar switch. I dropped TV service while bumping up the speed of my internet service, cutting my bill from about $120/month for both down to $40/month for the net. I picked up a Roku2 XD for $99 for the TV in the living room. I already had a PlayStation3 which could have handled most of what I had in mind, but I like the Roku. Then I added Netflix and Hulu Plus for about $8 each. I already had Amazon Prime for the free 2-day shipping at $79/year, but if you wanted it just for the streaming service that'd be about $6.58/month for that. Add in Crackle for a few more movies, Pandora for music, and a smattering of special interest channels like TED Talks, because I can. Those channels are free.

Everything about it is essentially on-demand. I can even purchase pay-per-view-like timed rentals on Amazon or the like if I really want.

To make it all even better, I installed Plex Media Server on a Mac that I use as a server. On that, I can load any video, audio, or photo files that I want. The Roku has a Plex channel that can see the media server and play its content, including any channels that are available on Plex but not on Roku. Further, I've got the Plex app on my iPhone, which can display any of Plex's content over the net from home so I can watch it anywhere. This means that if I want, I can download content via iTunes or some similar service. Now, if I were the sort to be into downloading less than legal content, I could also set up Sickbeard and Couchpotato on the Mac so that Plex would have all sorts of content.

Total price, if you ignore the value of Amazon Prime outside of its streaming service and discount the up front investment in the Roku, is roughly $62.58/month, or a touch over half of what I used to pay. What I'm missing is sports and pay channels like HBO. Although there are some sports channels on Roku and Plex, I suspect that sports fans may find the selection lacking. HBO and the like I can live without, but if I couldn't I've read right here on Slashdot that most of those shows have become the internet download darlings on bit torrent and the like.

Next project? Installing Asterisk and converting the house to VIOP.

Blocking 30% of visible light doesn't mean obfuscating what's on the other side. It means simply letting less light through. Think lightly (30% is lightly) tinted windows, not soap scum on the shower door.

OK, so here's the process as it is supposed to work:

• Step 1: Decide what it costs to create the product.
• Step 2: Project sales.
• Step 3: Price the product based on costs vs. sales required to profit to the required margin within the time required by your business plan.
• Step 4: Adjust prices and/or improve infrastructure to maintain target margin after the profitability goals are reached.

This normally means that prices have downward pressure as your infrastructure gets paid for. In opposition to this is the upward pressure of the cost of investing profit into improved infrastructure. Generally, people do a good enough job at forecasting adoption rates, infrastructure costs, and other product costs to have prices drop over time. Failure to do so basically means you've screwed up pricing your product.

What they're claiming is that they have to limit the bandwidth because they don't have the infrastructure to support the level of average utilization they're seeing because they have the "problem" of too fast an adoption of their product (bandwidth). But the reality is this falls under the heading, "nice problems to have." They *should* be able to simply scale up the infrastructure to handle it, but they have not invested the required money and man-hours to do so fast enough to meet demand.

The difficulty is that what they sold people was "unlimited" data usage. Data usage is just bandwidth over time. You can't call it unlimited if your response to too much demand is to throttle what's supposedly unlimited. Hint: throttling is also called limiting bandwidth. Unlimited data usage != limited bandwidth.

Failure to plan and invest properly on their part does not free them from the obligations they've made to their customers.

Even more correctly, an equation "describes". The accuracy of the description provided by the equation relative to measure, may merely be an estimate, or may be perfect. They may describe the limits within which values must fall, or the values themselves. Measurements themselves are after all, only estimates withing a certain degree of precision. Playing semantics isn't really the point though. What you have here is a theory that accurately predicts and explains the behavior with an accompanying equation that describes how the variables contribute said same behavior.

Everyone knows that's why the Mayan calendar stops there. Windows 8 comes out. The world ends. Apocalypse explained.

I would have to second this recommendation and extend it. Roger Penrose , is an incredible mathematical physicist and yet managed to pen a highly readable and even entertaining book that asks more deep questions than a teenager can shake a stick at. Reading this book as a teen made me want to be a physicist, a computer scientist, a philosopher and a mathematician all at the same time. He also wrote two follow on works, Shadows of the Mind: A Search for the Missing Science of Consciousness, and The Large, the Small and the Human Mind. Again, this is not only about math. Instead it can be seen as a book about many things to which math is a gateway that just happens to talk about the math involved along the way.