Ah, that's just because today's ultra-high-density platters require high-coercivity materials in order to adequately maintain the magnetization patterns that store the data. Low coercivity magnetic media just don't work at those densities.

Most Scots don't either - less than 2% speak any Scottish Gaelic.

It'll never quite stop being weird that Commander Keen, cartoony platformer, is what the guys behind Doom and Quake were working on in their early days....

It is a matter of taste; but the proliferation of 'widescreen' has really made multiple orientation setups more attractive. In particular, the ubiquitous 1920x1080 is cheap as dirt and nice and wide; but actually throws fewer vertical pixels than a nasty old 1280x1024 17' from about 2001. If you read or write a lot of text, or code with reasonably short lines, taking a cheapo 1920x1080 and rotating it gives you a 1080x1920: this is handy because it's still wider than 1024(so even old and horrible programs/layouts generally won't break, since anything that old and horrible probably expects 768 or 1024 pixel wide screens); but provides more vertical resolution than even substantially more expensive monitors in their native orientation.

I prefer my 'primary' monitor to be unrotated; but the amount of vertical resolution you can get for the money, without totally sacrificing width, from a rotated secondary monitor is pretty compelling.

In this case, you might want to go after the vets before the doctors...

It's not an accident that they were looking at agricultural workers (rather than, say, elementary school teachers, who would be seeing the worst of it from antibiotics-for-the-sniffles patients), nor is it an accident that there are 'livestock-associated' drug resistant strains.

Aside from price, which makes accepting multiple monitors rather compelling(you can get physically big ones for relatively small amounts of money, because of TVs; but if you want resolution the cost goes up fast and things really start to misbehave if you go high enough that DP MST or the like is required to drive the thing), it mostly comes down to how good your windowing system is at tiling and how well applications that expect 'full screen' can handle playing with others.

A good window manager makes carving up a single large monitor into chunks suitably sized for your various programs easy and painless. If you are enduring a less obliging one, it can be a fairly ugly business, actually less pleasant than getting some help from multiple physical displays, which are more widely respected even by poorly behaved programs.

That said, the 'two side by side, giant bezel in the middle' configuration is not my favorite. A larger primary screen, with ancillary screens on one or both sides gives you plenty of room for assorted lesser windows; but also avoids annoying bezels in the center of your field of view.

You don't choose between workspaces and physical screens, you just have multiple physical screens so that each workspace can be even larger and more pleasant to use...

You do eventually run into diminishing returns; but being able to display more than one monitor worth of stuff simultaneously definitely has its uses, and is something that being able to switch between workspaces, be the transition ever so elegant, cannot replace.

Given that 8 was the "Just because it's called 'Windows' doesn't mean it needs a functional windowing system!" release, It's pretty hard to argue with them.

Maybe some of that works on touchscreen laptops; but 'metro' is a tragicomedy on any monitor configuration worth using.

In addition, Tesla(whether or not you see this as an improvement is a distinct issue, it simply is so) sells cars much more like an enterprise IT hardware vendor sells hardware: at least within the warranty period, there is very much an ongoing interaction between the hardware and the vendor. System health information gets sent directly back, on site techs with specialized parts and firmware get sent out and so on. More traditional car companies are closer to buying a PC: the dealer will offer (often absurdly priced; but available) maintenance; and the vendor may become involved with certain warranty or recall cases; but they are otherwise largely out of the loop, with third parties handling the ongoing interaction with the hardware.

Something has to provide coordination and theatre-level intelligence for all deployed Highlander batteries, no?

We will, of course, run all your errands for you without gathering the data for marketing and other purposes; because doing that would be just too easy...

The assorted 'laws of war' are heavily leavened by what their framers suspect that they can actually get at least some people to agree to; but the overall theoretical foundation always seems to be an attempt to steer weapons in the direction of "Kills outright, or leaves a wound that, if treated, will heal with comparatively limited permanent damage."

It's not an easy standard to maintain(both in terms of convenience, mass-maiming is a hell of a shock to morale and logistics, and engineering, something that will kill if it hits you as designed will likely cause serious tissue damage and/or amputation if it scores a sub-par hit); but it's not really a terribly strange shared desire, from the perspective of the warring European powers of the 20th century that wrote most of them.

I'm fairly out of my depth with this stuff, so this is an honest inquiry: how do the magnetic nanoparticles fit into the equation?

I realize that, once coated with a suitably tailored binding protein, the particles will collect whatever target the binding protein was specified for (presumably this could even be tailored, for any target where a suitably tame binding compound is available), and probably fairly efficiently because of the absurd surface area of nanoparticles.

What I don't understand is the necessity of using the nanoparticles. It was my understanding that, outside of seriously immunocompromised victims, T-cells(and possibly other flavors of phagocytes, I'm fuzzy on the details) are extremely adept at engulfing and destroying foreign bodies, including 'clumps' produced by targets bound to the antigens produced by B-cells. This technique appears to be using a synthetic/introduced antigen(which makes sense if the immune system isn't producing the necessary antigen, or not ramping up production fast enough); but it also introduces the nanoparticles so that the antigen clumps can be magnetically scrubbed from the bloodstream, rather than cleaned up by the T Cells.

What is the peculiarity here that would make introducing the novel clump-scrubbing mechanism necessary and worthwhile?

The Protocol contains a loophole large enough to drive a truck through, never mind some photons:

"Article 3 Blinding as an incidental or collateral effect of the legitimate military employment of laser systems, including laser systems used against optical equipment, is not covered by the prohibition of this Protocol."

As long as the blinding is a side effect (mitigated by "all feasible precautions to avoid the incidence of permanent blindness to unenhanced vision") of a non-blinding purpose(setting things on fire, destroying machine vision/optical sensor gear, 'dazzling', and basically anything else you might feel like using a laser for, it's all legal. That is not exactly fertile ground for any sort of serious arms control, even if lasers weren't comparatively cheap and trivial to build, especially at the modest powers that will really boil your eyeballs but aren't subject to the engineering challenges of aspirational air-defense and antimissile systems.

It gives me no pleasure to say so; blinding is a pretty ugly thing to do; but the Protocol as written is about as effective as forbidding murder; but making it legal to put a bullet through any hat you see, regardless of whether it contains a head or not.

It's also an overcomplicated solution. OBD can get pretty nasty if you want access to esoteric stuff or manufacturer proprietary crap; but a basic, bluetooth-capable, OBD dongle that'll report the rough outlines of how a vehicle is being used is quite cheap indeed and not especially complex. I wouldn't necessarily want to try dead-reconing with nothing but that output; but answering "Am I driving right now?" is considerably less demanding.