The Economist. Still worth reading.

E = (T2-T1) / T1

Everyone with an engineering degree knows this. Trying to extract much energy from low-grade heat at the output end of an engine is inefficient. This was figured out a long time ago. Here it is in The Manual of the Steam Engine. It's possible to increase steam engine efficiency by compounding, where the exhaust from each cylinder feeds a larger, lower pressure cylinder. This is cost-effective up to about 3 cylinders ("triple expansion"). Engines up to quintuple-expansion have been built, but the additional power from the last two cylinders in the chain isn't worth the trouble.

In 2005, this appeared in SF Weekly, about the gentrification of the Polk St. area of the Tenderloin:

That was back in 2005. Gentrification won.

Let me finish that sentence for you:

Option 1: "Let me just say that I laugh at your situation, secure in my knowledge that nothing like that can happen where I live."

Option 2: "Let me see this as a warning that despite rule of law, foundational documents, and all the trappings of representative government, this could still happen here. I will be especially on guard against those that try to subvert my country."

It's happening. First, take a look at a map of the Tenderloin, from "Areas to Avoid, San Francisco." Twitter HQ is in that area, between 9th and 10th on Market, and the long-standing "mid-Market area" around there is rapidly being rebuilt. In fact, just about everything south of McAlliister has been gentrified, except for parts of 6th St and a small section around 7th and the north side of Market. Rebuilding is underway along the Van Ness corridor too, and has more or less chopped a block off the Tenderloin on the west side. That's the old "Polk Gulch" area, once a gay rent-boy hangout.

So the SF Tenderloin is about half the size it was a few years ago. Progress continues.

After the first dot-com boom collapsed, about half the twentysomethings in SF left. After this one collapses, that will probably happen again. Face it, most of the useful things in "social" have been done.

For only $0.50, you can get this nicer toy microscope on Alibaba. People have been making microscopes from drops of water or glass beads since Leeuwenhoek invented the microscope. With tiny optics, the view is dim, but it works.

This is a training program, not a production process. They have a few people doing forging by hand, but not to make production parts. See the original article in the Japan Times. Toyota's process of continuous improvement of production requires that people working on assembly lines understand the process well enough to suggest improvements. They recognize that they've dumbed down the workforce too much.

Ford Motor funded the building of the Detroit TechShop for similar reasons. They need more people who have a good sense of how stuff is made. Who in the US gets a degree in production engineering any more?

We could do this in the US. By could I mean, if we changed tax regulations -- the system is mostly in place already. Wage income is deducted "pay-as-you-go" here, too. All of my interest, dividends and gains were already reported (but not deducted) by the entities that paid them. The IRS could have just sent me a bill for that with what they already know. Most of the data I put on my 1040 is redundant for the IRS. The biggest impediment -- other than changing the law -- would be that not claiming all your deductions could result in paying far more than you really should owe, especially if you have a mortgage, give to charity, or need to report other such deductions.

You can, but it requires human intervention from customer service and takes a few hours. Obviously, this would be a problem if it happened at the last minute

Tax Cut (H&R Block) online works with Linux browsers. Turbo Tax online complains but works anyway. And Tax Act online at least let me start without any warnings. There really seems to be little difference in the online versions of these services vs the installable Windows program, FWIW to you. I replaced my Mom's XP with Ubuntu and switched her over from Turbo Tax for Windows to Turbo Tax Online. Except for the (apparently bogus) warning when first starting, it worked fine, and she didn't really notice a difference in the experience from last year.

The paranoid might be concerned about filling out their taxes online, but the truly paranoid would note that an installable program could just as easily "phone home" with your tax info, anyway.

I used H&R Block online. Unlike Intuit's offering, it doesn't complain that I'm using Linux. (Turbo Tax seems to work anyway after ignoring the warning, though.)

Trading stocks and funds in a non-retirement account used to be a huge PITA at tax time. Good news on this year's 1040 is that you can consolidate all your capital gains (or losses) by short and long term and avoid entering a line for every single trade. This quite literally saved hours of work.

The JetCat isn't man-rated. It's for model aircraft.

A JetCat needs an overhaul every 50 hours of operation. Mean time to failure is maybe a few hundred hours. A commercial jetliner turbine needs an overhaul every 3500 to 5000 hours of operation. Mean time to failure is around 100,000 hours.

A Williams FJ44 is suitable for light aircraft, and could be used for a VTOL, but a pair of them costs over $1M.

Current status: "€140 raised of €2,250,000 goal".

The thing is, it's quite possible to build a flying car. The prototypes of the 1950s make that clear. The world needs some good small VTOL craft. But none of the people doing it seem to be able to bring it off.

Small jet engines cost too much but can make VTOL work. Wankel engines (the Moller embarassment) or electric motors and batteries (this thing) don'tt have the power/weight ratio needed to do it well. It's probably quite possible to build a battery powered VTOL today, but the flight time will be a few minutes, like quadcopters.