Covering the entire collection area with cheap concentrators and a fraction of the area with expensive, high-efficiency converters might be more economical than covering the entire area with cheap, low-efficiency converters.

In other words, the metric to maximize is kW h/$, not % conversion efficiency.

Unlike Gates' claim that "The problems with today's reactors, such as the risk of accidents, can be solved through innovation.", that is a problem that might actually be solvable with innovation, like charcoal filters, or heat-exchanging house ventilation.

Did you notice that title, "The Force Awakens"? The Force itself is asserting sentience, and acting through Rey.

The deal about "how the Force works" has been altered. It is the filmmakers' prerogative to alter it further.

As others have answered, "several" in this context means a vague small number.

I'd like to point out that when writing up a news article, the author should have taken several minutes to look up exactly how long ago the X32 ABI was introduced.

"It was just six years ago..."

I think the author mixed sales numbers for the full year 2017 with a list of top sellers for the first half of 2018.

I found a page that has 2017's top sellers as
1. Ford F-Series 896,764
2. Chevy Silverado 585,864
3. Ram Truck 500,723
4. Toyota RAV4 407,594
5. Nissan Rogue 403,465
6. Toyota Camry 387,081
7. Honda CR-V 377,895
8. Honda Civic 377,286
9. Toyota Corolla 329,196
10. Honda Accord 322,655
11. Ford Escape 308,296
12. Chevy Equinox 290,458

Those numbers match the ones on http://carsalesbase.com/us-car..., cited as the source of numbers in the article.

According to that list, the "leading American SUV" is the Ford Escape (not the Chevy Equinox). In 2017 Ford (not Chevy) sold 79,000 (not 100,000) fewer than Toyota sold Camrys. To me, that doesn't change the gist of the article. Ford and GM are dropping cars from their lineups to focus on more profitable trucks and SUVs, while Toyota and Honda are still selling plenty of cars, while the Nissan Rogue, Toyota RAV4, and Honda CR-V are handily outselling Ford Escape and Chevy Equinox.

Next time the price of gas goes up it'll be bailout time once more for Detroit.

I don't know why I care--I'll be long gone--but I like to think there's a slight probability that a lucky (maybe!) few humans leave on generational ships.

Just those two make a very good start. Also see the FidoNet rules:

DESQview might have been a better choice. Several colleagues and I used it in much the same way as your two devs, with Brief as the text editor instead of Multi-Edit.

AWS and Azure each do over $20B a year (and growing), adding $1B a year more to either one will not create a monopoly.

I know this is Slashdot and nowadays people don't even read the summary, let alone the article, much less the actual paper to which the article links, but there's an awful lot of straw littering the floor here.

The chaff bugs are inserted by an automated tool, they will not have to be written or worked around by people writing the actual code.

The authors are aware that the chaff bugs must not be exploitable.

The authors are aware that they will face automated tools for finding bugs and determining their exploitability.

The authors are aware that this is just a proof-of-concept and would require refinement to be useful in the real world.

They did in fact test their chaff bugs against a fuzzer, which found them. They took steps to ensure that the bad behavior is actually harmless (for example, overwrites go to areas of memory that aren't actually used).

How do you see this making software harder to maintain? The chaff bugs are inserted by an automated tool, maintainers looking at the source code will never see them.

For testing, run the test cases against chaffed and unchaffed builds. Test cases that fail in both are probably real bugs. Test cases that fail only in chaffed are probably chaff bugs. Test cases that fail only in unchaffed are probably serious trouble.

Oddly enough, the authors of the paper explain this in section IV-B Ensuring Non-Exploitability

What's with the hostility toward academia? Lots of great ideas have come out of academia. No academic myself, just decades of writing real-world software, during which I've learned not to judge other people, especially without reading their work.

The chaff bugs would be inserted by an automated tool. No need for programmers to make up anything.

They would not be indistinguishable from the real thing to a skilled hacker; the idea is that they would be so numerous as to make the effort of finding the actual exploitable bugs uneconomical.