Go read this website. Dvorak really isn't that great compared to other layouts. If you're going to learn a new layout, pick one that performs better. Dvorak was a good idea, but it was designed in the days before computers and modern statistical analysis.

That's it! Exactly the one I was thinking of. It looks like "QGMLWB" is actually the best layout overall, according to his statistical work.

Renewable energy and "sustainable transportation" were largely tried in the 19th century and abandoned because they were too limiting.

Wind power is considerably older than that. It's actually considered to be the first form of non muscle based power used by himans.
Also "renewable" and "sustainable" have reached the point of being politcial "weasel words" more often that sensible descriptions.

Dvorak is good, but Colemak is also a very good alternative that's probably a lot easier for Qwerty typists to adapt to. Also, some guy did a bunch of research and made a website (wish I had a link, sorry) about different keyboard layouts and found that Dvorak was actually eclipsed in some metrics by both Colemak and another layout he created.

>I think they do manual point-to-point wiring on the switches. But if you look at the sculpted shape of a Maltron, they don't lend themselves to conventional PCBs.

This sounds like an application for flexible circuits boards. Point-to-point wiring is far too labor- and time-consuming.

Is this the wisest choice? In my opinion, MicroUSB is actually a pretty crappy connector, and doesn't have very good retention. MiniUSB and regular USB-B ports are far better and sturdier choices for a corded item on my desk which gets bumped around a lot. The only really good thing about MicroUSB is the thinness, but that's only important on mobile phones, not large items like keyboards.

I don't get the bit about weight. Aluminum has a very high strength-to-weight ratio; you're not going to get a product with the same weight and durability with wood (e.g. you could use balsa, which is extremely lightweight, but it also have low strength and absolutely terrible hardness).

In high quantities, aluminum should be pretty economical; you can just use a big press to stamp it. The big cost here is the tooling, but after that the per-unit cost is cheap. Milling is far, far more expensive than pressing/stamping, and only really makes sense if 1) the quantities are really low and/or 2) the product cost is really high and 3) it's really needed for some reason. For enclosures and the like, stamping is usually sufficient.

To keep it from looking like an Apple product, there's something really cool you can do with aluminum called anodizing. Anodize it black and give it some corners and sharp lines and it won't look anything like an Apple product. Anodizing is better than paint since it's more durable. For a high-end product as this is sure to be, it should be within budget.

...but only if Samsung proved its viability first ;)

Saving even a couple of hundred dollars per unit might be a drop in the bucket when compared with the peripheral costs. Yes, IT departments everywhere might be able to save a little money on the purchase of each computer by buying all of their parts from NewEgg and installing Linux on the computer that they cobble together from parts. Still, it ends up being cheaper, when you add up all the peripheral costs, to buy ready-made computers from Dell with Windows pre-installed.

Except that OEM preinstalls tend to be of very little vaule in "enterprise" environments. Even those fairly free of "crapware".

Ok thanks for the clarification. Terrorists indeed claimed to have stolen a BUK system from local Donbass military base, on 29/06: https://twitter.com/kram_ua/st... Here's another claim - they're saying they fixed it: https://twitter.com/Dbnmjr/sta...
UA army claimed all BUKs were crippled beyond repair, so it is also likely that specialists and materials needed for repair (if it indeed happened) came from Russia.

Or what was stolen was rather less "crippled" than the Ukrainian Government would like to admit.
It isn't as if politicians tend to be the most honest of people, even in peacetime.

You mean something like http://www.mortarinvestments.e...

I wonder how much of that has to do with the tendency among Android app publishers to make their products exclusive to Google Play Store, which is (legally) available only on devices with preloaded Android OS. This makes AOSP distributions undesirable to end users.

On the other hand any kind of "app store" in an "enterprise environment" where installing any kind of software is about the last thing "end users" should be doing.

I agree with your comments on this as well. What would we do with all the prison space currently housed by drug-related occupants? That would put a heavy dent in the income of the organizations that manage the prison systems (which are mostly cronies of the politicians).

Could always put politicians (and their cronies) in there :)

As far as the US is concerned, I don't think use of substances like crystal meth or pcp would be an issue if more enjoyable/safe drugs were equally available. What makes meth so cheap and available is an artificial inflation of other substances caused by Federal drug policy.

One side effect of prohibition is that of "legal hights". Which whilst not illegal often turn out to be considerably more toxic then what's been previously banned.

One thing I can caution you about, as a med student, is to question where the guidelines for the "correct" cholesterol came from. Hint: it's probably an echo chamber, and not tied to scientific evidence.

IIRC the numbers were the lowest the drug companies could get away with.

Anyway, what I'm saying is not that statins are bad, but merely to question what your goal cholesterol really should be. Your brain has a lot of cholesterol (myelin is high in cholesterol) and cholesterol modulates your cell wall plasticity. Too low would be bad.

One very intresting thing is pattern B LDL particles contain less cholesterol than pattern A LDL particles. Which rather means that breaking lipoproteins apart and looking at their lipid "cargo" is likely to be less useful than looking at the lipoproteins themselves.