Let's assume for a moment that they're serious about deporting people.

What's the cost if they get through, and have to be tracked down by traditional methods? What's the cost of putting more people there to achieve the same level of effectiveness? What's the cost of flying conventional aircraft to do the job?

When pitted against those methods by comparison, $28,000 might actually not be all that bad.

8chan. 4chan is kill.

8chan gets SJWs as well, and it even lets them have their own boards. That's doing it the right way – counter speech you don't like with more speech.

His career might, but he doesn't want to tank the party's hopes. He can take certain unpopular opinion on himself, but only so much and only in certain areas.

But barring brain trauma or breaking their necks or something, that time element means these athletes have time to transition into other careers. Some do, and not all of them go into sports media or coaching (although many do). Some buy car dealerships or other ordinary businesses, with greater or lesser degrees of success. Others may found ill-fated video game companies with large government loans, and crater after a couple years. The point is that the smart ones will use the money gained during their athletic careers to bootstrap something even bigger afterward -- or something just big enough to keep them happy without having to go back to crab fishing or herding sheep or whatever it is their family did before they got lucky.

Maybe TFA is bad at conveying what they're doing then, because the impression I got from it was "we have a way to electroplate multiple metals selectively by adjusting the voltage. Doing this enough times can make the bulk material much stronger." If laying down a plating layer nanometers thick is now "manipulating materials on a molecular level", then I can do that in my kitchen with less than $100 in equipment. I believe the thickness of the plating I typically lay down in a single pass is on the scale of four or five atoms, but I make hundreds of small passes, stop, clean, and make hundreds more. If I tank plate rather than brush plate, I don't have as much control, but it could still be done if I had an assistant (robot) to move the parts around for me.

Even if I grant both possibilities in full, how does this make them integral to each other? They could electroplate thousands of thin layers before, it just required moving parts between vats. This is obviously impractical from a human labor perspective, but it may not be substantial at all with machine labor, so it seems they have reduced a cost. They haven't solved a fundamental problem. Would you care to explain, with citations from TFA, how exactly I am clueless?

I know perfectly well how the metal ions deplete from the solution, changing the voltage and time required to get an effective coat. Once they drop below a certain level, it just stops working and throwing more power at it doesn't help. The mixture of solutes would have to be carefully monitored and controlled to prevent this from becoming an issue, at which point it seems simpler to me to use one bath per metal. You can recharge the solutions on a pretty regular basis that way, and not have to do much monitoring at all. You can step up voltage (to a point) to accommodate a weaker solution, without fearing that you're going to attract a different metal. Possibly most important for economy of scale, you can more easily recover the residuals of the spent solution for reprocessing if they haven't all been mixed together.

Having done electroplating myself, though only on a small scale, I have noticed that sometimes applying an excessively high voltage doesn't make the solute metal ions stop attaching, it just makes them bring along some "scum" along with them. Most notably, throwing too much voltage at silver solution produces a black scum which must be cleaned off before anything else (including more silver) will stick. I have to imagine other metals have the same problem.

If one ion truly prefers a given voltage and sticks to the surface preferentially, this might block other ions that want to form the "scum", but I still think this would significantly limit the number and type of ions they can have in any given solution.

It also seems that the technology to do this is simpler than they are billing it -- pull the parts and dip them in each bath as needed. Rinse in between. We have robots to do this now, it's not like someone has to stand there and watch.

That's what I meant when I said the safety profile is well known. Some people are hypersensitive to light, whether it's their eyes or their skin. They already know it, so this won't sneak up on them. For everyone else, it's not going to hurt them to have a high frequency signal modulated onto their light bulbs.

Upsides: Unlicensed spectrum. Pretty much unenforceable even if it was licensed. Little or no bleeding over from desired coverage areas, at least indoors. Plenty of bandwidth to go around. We know the safety profile of this sort of radiation quite well also.

Downsides: Line-of-sight only, so an AP in every room would pretty much be required (or equivalently, fiber from a central AP to every room). Probably can be degraded by "noisy" light-emitting devices, but spread-spectrum will probably get around that pretty well.

It sounds a little like using fiber optics for the last-mile problem, only in this case it's the last-meter problem and possibly without a fiber.

Oh, incidentally, the thread I posted to may be of considerable interest here.

http://deskthority.net/keyboar...

Ironically, it's the tale of a guy who goes to the completely opposite extreme and then wonders why some people would dare to disagree with him.

Go deep, not wide. Offer Fn layers and dedicated keys, but put those dedicated keys in back, not off to the side. If you do expand to the side, expand left, not right.

You may want a Tipro MID, though those are hard to come by in the U.S.

Cherry MX Black switches (heavy linear), relegendable keys on the top three rows. You can get them in either a matrix layout or a staggered ANSI or ISO layout for the bottom four rows. The top four rows are always a matrix. If you want more keys, they come in various sizes and bolt together.

Be aware of the significant problem that the programming software requires Windows. It does not run on anything else. The PS/2 connected versions not only require Windows, but 32-bit Windows. (The USB versions will accept 64-bit.) While PS/2 to USB (and vice-versa) conversion works, it does not allow programming. They must be programmed on their native interface. However, once programmed, they stay programmed and can be used on any system or any OS, and either interface type.

Another option is the Cherry boards I posted elsewhere, though they don't have nearly as much customization ability as the Tipro.

There are simple, off the shelf answers out there, you just need to look at the point-of-sale market. This means you may end up with an unnecessary credit card reader attached to your keyboard, but otherwise there is no real issue. (Besides, wouldn't being able to swipe a card, even a magstripe, be a nice second factor for login?)

As I posted to Deskthority just yesterday:

If that was the case, putting a blob of material on the power supply chip (and nothing else) wouldn't remedy the problem – but it does (see the last post on this page.)

Filing a false police report is criminal in and of itself, even if it doesn't result in an expensive, resource-wasting, and potentially injurious or deadly response from the police.

Do it once, maybe you get away with it. Keep doing it, and you can [i]expect[/i] to get caught.

I think this is a solid indication that you have gone insane in the membrane.

No matter how good your "live view" screen is, it won't be of resolution comparable to a matte glass screen. This may eventually become indistinguishable. However, there will always be a little bit of latency and flicker, no matter how good it gets. The "latency" of a mirror box is and always has been well below detection thresholds for humans.

The only "mirrorless" I'd be interested in at this pint is more accurately a half-silvered mirror. Some of the light goes to the detector, some of it to the focusing screen, and nothing moves. Unfortunately, you sacrifice half your light sensitivity (one stop) for this.