Data General made this OS (and derivative) for their 16- and 32-bit minicomputers in the early '80s. My high school had one and it's what I learned Pascal on. Since DG seems to have dropped the line in 1988 and switched to concentrating on their Unix derivative before finally crashing and burning in 1999, and it never ran on machines that most hobbyists would be familiar with, I suspect that there are few orphaned installations out there...but apparently I'm wrong.

Maybe I missed something. Isn't this just a perfectly sensible extension of "innocent until proven guilty"? If I call you a thief and you sue me for libel, why should the burden of proof be on *you*, exactly?

If you are accusing me of libel, why should the burden of proof be on me? You are the one making the legal accusation, not me. If you were accusing me of anything else, the burden of proof would be on you to prove that you have legal redress and that I did what I was accused of. The burden of proof is on you, not me. Why should it be different for libel?

In the US, libel typically has several elements which must be proven, including (a) it was published, (b) it was defamatory, and (c) it is false. Where the whole "defendant has the burden of proof" bit comes in is that if the defendant can show the truth of the supposedly libelous statement, then the plaintiff's case fails on (c), and it is often easier to rebut the accusation of falsehood than the rest of the case.

In the UK, the "falsehood" element is missing; a true statement can be considered libelous. This makes life much harder for the defendant.

How about, you want to study very high speed cars. On a daily basis people are caught by cops going over 100 mph all over the usa. But the odds of putting a camera up on any old street corner and seeing a 100 mph car are very low and at best you might see one in a zillion years. Like cosmic rays.

Not quite. Using your analogy, it's like you want to study the handling and aerodynamics of a car going 75. There are millions of cars on the road, but the best you can do to see cars going 75 on the public highways is via road-side cameras, 100 feet away from the road, and you don't know when the cars are going to drive by, and they could be doing anything from 60 to 120mph and you don't know what in advance. One time one of your colleagues saw a car zip past at 300mph, but it hasn't happened again. You see 75mph cars all the time, but your pictures are not all that good.

However, on the track you can set up an observing station that is 10 feet wide, has pressure and strain sensors embedded in the roadbed, has air pressure meters at close intervals, has high-speed, high resolution video recording from both sides, above, below, and at various angles. And you know, to the fraction of a second, when a car is going to go through your sensor. And you have dozens of expert drivers who send a car through the observing station at 30 second intervals at exactly 75mph.

The hundreds of inexpensive road-side cameras you and your colleagues have deployed see more cars doing 75+ than you will see on your track, but you know much more about how the cars handle from your data.

What the LHC does is slam hadrons -- large collections of quarks bound together by strong nuclear forces -- into other hadrons at high energy. The LHC uses the hadrons it does not because there is anything special about them but because it's somewhat easier to get the energies they want to study using the hadrons they choose. They also chose the energies they use for the collision for convenience more than anything special. Ideally, they want the most energetic range they can accurately control. If they could build a bigger collider, capable of higher energy collisions, they would, but these things are complicated, big, and expensive.

Cosmic rays are a mixture of fast particles, including hadrons of various sizes, traveling at very high speeds. Many cosmic rays are bare protons, the same as used in the LHC. The energy range of cosmic rays is wide, ranging to many more orders of magnitude higher than the LHC. A collision between a proton from space at 100TeV and a proton in an oxygen atom in the upper atmosphere of the earth is very similar to a proton-proton collision in the LHC, but much higher energy.

If I am interpreting a graph on Wikipedia correctly, cosmic rays with an energy of over 1000 TeV impact the Earth at a rate of about 1 per square meter per year. Given the size of the Earth, that's 14 million/second. So 14 million collisions hundreds of times more energetic than the LHC can do happen in the Earth's atmosphere every second. And there appears to be a power scaling going on. 10TeV cosmic rays are thousands of times more frequent than 1000TeV cosmic rays.

The difference, and why the LHC was built, is location. Looking at cosmic ray collisions tells us what the end result is going to be, but it doesn't tell us what happens partway through. If you look at a car crash on the side of the road, you know that the car got squished and the driver was injured. If you look at a car crash in a lab with cameras and crash dummies, you can tell that the driver hits the windshield before the crumplezones absorb all the energy.

The same sort of thing with the LHC. If the LHC will create Higgs Bosons, they are being created all the time in the upper atmosphere. But Higgs Bosons are expected to last an incredibly short amount of time, and all we see is what's left after they decay into other particles. We can't see cosmic ray collisions clearly enough to see if the decay particles come from Higgs or from other processes we understand well.

Sir Pterry isn't a Knight Commander (which is a title within various British Orders), but a Knight Bachelor (which is a title outside the Order system). Formally, there are no initials he can add to his name as a Knight Bachelor, but many add Kt. So he could be styled "Sir Terry Pratchett, OBE" (Officer of the Order of the British Empire), but not "Sir Terry Pratchet, KBE" (Knight Commander...).

In the non-embedded world, people rely on the operating system to handle the time, DST and all.

Linux and MacOS systems are configured to use a second count from an epoch internally, which it converts to local time using a timezone database. The timezone database is designed to be updated, and updates are routinely sent out whenever a timezone (or DST rule) changes worldwide. The timezone db for Linux changed within days of the law being changed. The DB is historical, so it'll properly report times in the past, even though the rules have changed since then.

The main problem with Linux embedded applications is updating the DB when the rules change.

Windows... I dunno how Windows handles it.