It's not a valid comparison to compare the price of 1kg of rice to the bulk price...

For example, looking at WalMart Canada, they'll sell you 8kg of rice for C$1.06 per kilo, or they'll sell you 900g for C$2.52 per kilo.

Of course, the 900g price is decently lower than the prices that Numbeo is quoting, but the basic premise holds true: food has different prices in different places. For one thing, the ability and willingness of people to pay more can drive prices up. For another thing, the cost of transport can too. And I can tell you that people in a country with a per-capita GDP of $6,985 are probably not willing to pay as much for stuff as people in a country with a per-capita GDP of $50,577.

It really doesn't. One kilo of rice costs C$3.67 in Montreal. One kilo of rice costs C$0.98 in Havana. Not everything is cheaper, but many things are.

You can see the direct comparison here:

http://www.numbeo.com/cost-of-...

Satellite internet should work fine for Netflix. It's not latency sensitive, and while Satellite is typically not blazingly fast, it works fine at lower speeds, and the dynamic scaling is pretty seamless.

Could you talk to why a more traditional fire suppression system (such as sprinklers) wouldn't work? It seems like building something into the ship itself, which would take up little space compared to a big bulky robot that needs to wander the ship, would be an enormously simpler problem to solve. I realize that fire in a warship is going to often be accompanied by structural damage (while in a building the structural damage would probably be a result of the fire rather than the cause of it), but you'd think that sufficient redundancy and resiliency built into the system could accommodate for that.

My S95 didn't do much in the way of bokeh in the first place, so the limited amount of depth of field permitted by the smartphone isn't much less than the S95. Of course, for most people, limiting depth of field is not really something they care about (or even necessarily want).

Obviously cameras with large sensors are going to be much more capable, my argument is more that smartphone cameras have passed a threshold of "good enough", the point where the pictures that they take are sufficiently good to essentially eliminate the P&S market. Cameras like the RX100 are great, but at $500-600 you're not going to find many people buying them. IIRC the entire global market for all mirrorless cameras is only like 3.5 million a year, whereas smartphones are nearing 400 million. And the mirrorless figure is dropping while the smartphone figure is increasing.

Then clearly the camera market is beyond saving, destined to be relegated to the prosumer and professional niche (a tiny fraction the size of the former market), and the smart companies will do what Sony has done and get into manufacturing mobile camera stacks for smartphones and tablets.

Smartphone cameras are already on par with or better than P&S cameras were when smartphones started supplanting them in the first place*, and they're only going to keep getting better. They'll obviously never match the much larger sensors on DSLRs, since they're still improving too, but the best smartphone cameras today (which will be the mainstream smartphone cameras of tomorrow) have already passed the point where nearly everybody doesn't care.

*: My last P&S camera, which cost as much as an entry-level DSLR, was a Canon S95. It's a 2010 camera that you can still buy today. It was enormously better than my 2009 smartphone, but it's now inferior to my 2014 smartphone camera in nearly every way. My smartphone has better low-light sensitivity due to big advancements in sensors (BSI was the big jump between 2009 and 2014) and ever growing smartphone sensor size (smartphone sensors are now almost as big as the S95's), and on-sensor phase detect means my smartphone focuses faster and better to boot. Not to mention my smartphone is much better at video, doing higher resolution at higher framerates for longer durations. About the only way the S95 is still better is form factor (and even then it took an after-market grip to get the S95 feeling really comfortable), but as you said, the best camera is the one you have with you, and my smartphone is always in my pocket, while any camera never would be. It'd be in a backpack at best, and why dig for a P&S in my backpack when the phone in my pocket takes better pictures?

There will always be a market for professionals and prosumers, but the problem is that their products are generally priced high enough that they form a barrier to entry for more casual users. Casual users are generally happy with their smartphone cameras, and they're not going to make the jump to a dedicated camera unless they can get something that is a significant improvement at a reasonable price. DSLRs are generally still $400ish, and mirrorless are typically even more than that. That's just not enough to convert people with a casual interest. If they sold something like the Rebel SL1/EOS 100D for $200, they might get people who are curious, but they're not.

What's the cost to make one of these things really like? Because it would seem that advancements in manufacturing technology should have driven the cost down dramatically over time, and it doesn't seem like that's happened. Are the camera manufacturers just unwilling to undercut themselves, to accept lower margins? The problem is that the effective cost of a smartphone camera is $0 for most people, and that's definitely undercutting standalone cameras...

I don't. I like predictably scheduled releases. Ubuntu's release strategy particularly pleases me, with predictable releases every 6 months, and long term support releases every 2 years, with support for upgrading either from regular release to regular release, or from LTS release to LTS release.

Of course, I don't run Linux as a desktop platform, so Ubuntu still works nicely for me in a server environment. I tend to run only LTS releases on important servers (typically waiting until 6 months after an LTS release before upgrading to it, and regular releases on unimportant servers (like my home server).

It's backwards compatible, but not necessarily forward compatible. The RPi1 and RPi2 don't use the same instruction set, moving from ARMv6 to ARMv7, and adding the NEON instruction set extensions.

I did try that the most recent time, it didn't help. It could see the 2.4GHz wireless network, but not the 5GHz wireless network (both networks are broadcast from the same router).

Because you should trust your server provider not to mess with your traffic more than you should trust Verizon? Who cares about the NSA, if they want to get your data they're going to get it. Meanwhile, Verizon is actively MODIFYING your traffic...

Key exchange is also really not a problem, the entire point of a secure key exchange is that the keys are never transmitted in the clear. You don't need physical media.

Yeah, so I pay $10 a month for my own VPS, and run my own VPN on it. I have full control over what exactly gets logged, and data retention laws don't apply.

KVM has poor host platform support (it runs on Linux and nothing else). KVM has poor compatibility with host hardware, requiring CPUs with certain features. KVM has a somewhat involved installation process. KVM has limited graphical support, relying on SPICE remoting which (at least currently) lacks any real hardware acceleration support for either 2D or 3D graphics.

KVM is fine for a server environment, but it's extremely limited when compared to even the free version of VMWare.

Spend $5 or $10 a month on a VPN or a VPS and encrypt all your web traffic. As soon as your ISP is actively inspecting and modifying your traffic, it can't be trusted.

You shouldn't have to do this, true, but it's a solution to the present problem.

Neither did I, in the past. I vaguely recall having had wifi-related issues when the notebook was newer, but before I put Yosemite on it, I don't think I had any wifi issues any time recently before that. It was only after yosemite that I started having issues.