Isolated clusters of galaxies (such as the local group) are expected to have low total angular momentum (basically because the initial condition has low angular momentum, and in the absence of large mass anisotropy nearby, there is nothing to change this.) The mass of the local group is dominated by Andromeda and us, and hence so is the angular momentum. If the us/Andromeda pair has low angular momentum about their centre of mass (and given the pair is gravitationally bound), they will both pass close to that centre of mass - i.e., they will collide.

Of course, having an actual measurment is much more satisfying than having a theory.

Also - although they can be spectacular from outside, galactic collisions aren't expected to have bad results for life living on their planets. The biggest effect is that colliding dust clouds trigger a burst of star formation, so the night sky will be pretty.

It has been a few decades since I studied this, so I hope this is all accurate.

Another problem I've complained about in the past is rebus icons. I once used a source control system where the icon to commit a change had a document page with a tick mark and an arrow pointing at the page. I'd been using it for several years before I realized what it was supposed to represent - Americans call a tick mark a 'check', so this was the 'document check-in' button. At which point I also realized the same applied to an email client which had an icon with an ticked envelope - 'check mail'.

So, icons were supposed to be language independent, but instead in these examples they only made sense in one particular dialect.

I also have a problem with Swedish appliances (washing machines, ovens) which have indecypherable icons for the various modes, and the manual has invariably been lost years ago. If they just labeled the modes in Swedish, at least I'd be able to look up the meanings online.

In these judgement calls, I expect there is another piece to the equation, at least in the USA.

If you don't order a scan, and there was something bad which the scan might have picked up, you get sued.

If you do order a scan, find nothing, but scan results in the patient getting cancer 15 years from now, you don't get sued, as there is no way to know that it was your scan which triggered the cancer.

You say that like it is a long time. This is going at light speed compared to the SCO saga.

No, I'm worried about what happens AFTER the fuel dumps into the emergency storage tanks. It just keeps heating, unless something cools it.

OK, I've just done some research: http://www.osti.gov/bridge/servlets/purl/469120-avNXWz/webviewable/469120.pdf
It seems that they're constantly removing the fission products from the fuel, so there simply isn't enough 'after heat' to cause a problem for the cooling of the dump tanks.

I am worried that in normal operation there is a lot of very radioactive stuff being pumped around, controlled by valves, and chemically processed. I'm therefore expecting quite a lot of complex equipment in regions far too radioactive for human maintenance.

OK, I've looked it up [warning: large pdf], which I don't think you have. Show me the evidence that the much worse Chernobyl accident caused, or is yet to cause, more than a few hundred deaths.

The reaction then stops being self sustaining, and you just have to recover the containment units and repair the reactor.

At Fukushima, the reaction stopped being self sustaining seconds after the quake, and minutes before the tsunami. It didn't save them. You can't just wash your hands and say 'problem solved' when the chain reaction ceases. Fission products will keep generating large amounts of heat for months afterwards. If your 2-3 smaller tanks have no way to lose this heat, they will eventually melt.

I'm not saying that these new reactor designs can't deal with this, but you need much more evidence before can claim it's "literally idiot proof".

I was wondering the same.

Detecting gamma rays is pretty easy. Detecting within a few degrees which direction they came from is much harder. Lenses and mirrors won't work (at least, at any reasonable scale) to form an image. You could have two layers of detector, and measure the location of the gamma ray as it passes through both. You could look for Compton scattered electrons from the gamma ray, which would be easier to determine the direction of, but I don't think that would fit in something camera sized.

I'm also curious to know what exposure time the gamma ray camera needs - I'm guessing it will be pretty long - minutes, at least, maybe hours.

The probability is exponential in n, but for two planets, it is polynomial in p. I'd fixated on the second fact and missed the first. Given the context that we'd just changed n rather than p, I agree that 'exponentially' is more appropriate here.

I shall submit myself to the Committee for disciplinary action.

Early supernovae wouldn't help - the star is formed from the same material as the planets would, and the star demonstrably has almost no metals. Early supernovae would just mean that this star didn't exist (in its current chemistry), or that it is even younger than currently estimated, so as to form before the supernovae.

Interstellar captures are very difficult. Generally speaking, you need three gravitationally interacting bodies to allow a capture, as you need one to carry away some energy. Basically this requires the wanderer planet to turn up just when the star is passing close to another one, and even then to get really lucky. (Most often it is the lowest mass object of the three which gains energy, but we need the planet to lose energy.) Another possibility is you could lose that energy through tidal losses, but this requires the wanderer has very small positive energy initially, and passes very close to the star. Either way, the odds of such a capture are very low.

In addition, we have the fact that this star has two planets, which makes the odds against capture polynomially* smaller. Finally, if two planets were captured, we'd expect them to have different orbital planes. Given that they were detected by the 'wobble' method, I'd expect this could be measured, and would be mentioned if it had been so. However I can't promise that there aren't gravitational interactions which would bring the orbital planes into alignment over 13Gyr. Captures would also initially have highly elliptical orbits, which again the wobble method should notice, and again I don't know if 13Gyr is long enough to circularize the orbits by tidal effects or planet-planet interactions.

* This word brought to you by the Committee Against The Misuse Of The Word 'Exponentially'