It's only a metaphor, but holds surprisingly well. Worryingly well. So well that we, if we claim to be modern enlightened people, should have some kind of response.

But what? Switching operating systems - like switching religions - involves a lot of work if you do it properly. Unlike religion it is possible to "worship" two or more OSes, but many people find that an inefficient way to work. So how can we avoid unwarranted faith in our way of doing things, fighting between neighbouring factions, and all the other destructive forces that religions suffer from?

The Linux kernel does a good job of holding all the myriad Linuxes together: all need the kernel to evolve and improve, but none can afford to implement those changes alone. Android and iOS have opened peoples eyes to other ways of interacting with computers and rendered the Windows-Mac conflict less important.

Technology evolves, preventing us from stagnating and developing unchangeable "holy" rules. It's a natural human tendency to break into tribal factions, but it seems that technological progress puts a damper on this, forcing us to widen our horizons and helping us to work together. Suddenly progress seems more important than ever.

If you are a good scientist, then you will easily prove the specific claim "some black rocks attract one another". You will repeat the experiment, thereby showing that particular rocks consistently attract one another in controlled conditions. You will lend those rocks to other experimenters to test, proving beyond all doubt that certain rocks behave as you say. Then you will get on the cover of Prehistoric Nature magazine, and be given funding to find out why certain black rocks are special. Other researchers will scour the earth searching for more special black rocks, and reference your paper every time they succeed or fail. Eventually you will become a respected become professor in the department of black rocks.

Parapsychology fails because it cannot meet the requirements of the scientific method. Until it can offer a prediction which (a) differs from existing knowledge, and (b) can be tested in a reproducible experiment it is, by definition, not science. Science being the only method we have for establishing whether a claim is trustworthy, ologies which can't meet the requirements of science must be labelled untrustworthy.

Here's why string theory is different from crackpot hokum such as parapsychology:

• - Nobody is claiming ST has passed any complete scientific test. Parapsychology proponents put forward their claims as reality.
• - ST is compatible with known facts. Parapsychology is not.
• - ST is interesting because it might help current research to find the way forward. Parapsychology is disconnected from all scientific knowledge.

One should keep an open mind, but not gaping. Life is short, and we have to be critical. If we spent all our time re-testing already discredited theories we would have wasted our lives.

Parapsychology theories have been given every chance. They've been tested under proper laboratory conditions according to the scientific method. They've been tested again and again, over and over, given far more chances than any ordinary scientist might expect to be given. The tests were scientific, and the theories failed those tests.

Hanging on to disproven theories is what makes parapsychology a non-science. The -ology suffix is just a desperate attempt to associate with proven laws of nature. Science is right to give crazy ideas a chance, but also right to shun them when they are emphatically shown to be wrong.

And when the pseudo-scientists persist in dressing up their mumbo-jumbo, quackery, and bullshit as respectable ideas as, you can forgive the real scientists for getting a little bit cross.

For "unconventional science" read "mumbo jumbo". Parapsychology does not qualify as science, unconventional or otherwise. It only qualifies as quackery and bullshit.

An impressive balancing act, but the really interesting thing is to see how Atlas stumbles on objects which shift under its weight. It vividy shows the need for flexible feet on a biped. That's going to be an interesting engineering puzzle to solve.

True enough, but I don't think we're ready to design the next big accelerator anyway. The science is at an impasse, and these machines are too big and costly to build speculatively. What we need to do is keep developing various novel technologies, including laser and dual-beam wakefield acceleration as well as muon sources, and hope fervently that new discoveries at the LHC will soon show us the way we should go next. If the LHC does not discover anything beyond the Higgs, then I think particle physics may be in for some dark ages.

There are also ideas to build a circular muon collider. Muons are similar to electrons so give a nice clean signal in the detectors, but being 200 times heavier than electrons they lose much less energy as they circulate around a ring-shaped path.

The problem is muons are unstable, with a half-life of just 2 micro-seconds. But if you can collect them fast enough and accelerate them to near-light-speed, their lifetime increases due to time dilation. The nearer they get to light-speed the longer they last for, and it's thought that it would be feasible to get them going fast enough that they would last for a useful amount of time.

There are lots of advantages to circular accelerators: You can re-use the expensive accelerating sections thousands of times over by recirculating the beam; the beam itself is re-used over and over (only a tiny fraction of the particle are lost on each collision); and most importantly you can install more than one detector. Having two independent measurements is very important in establishing the reliability of any results.

Why? He stopped when the car said he should, got out, and was fine.

We have no idea how bad the impact was. It might not have felt very serious at the time, and presumably the battery did not actually start overheating until the second warning. In any case switching off the "ignition" would not have prevented the battery from catching fire.

Fair point. But even at 130 miles per 19 kWh, there's some way to go before you rival the energy efficiency of the best diesel cars. BTW Honda's own figures are much more pessimistic - any idea why that is?

It's not that I especially like diesel as a source of energy. The fumes are unpleasant and unhealthy, and as a cyclist I breath in more than my fair share. I just think it's important to be aware of the true impact of electric transport.

From your link: battery lifetime is 300-500 100% cycles, or 3750-4700 10% cycles.

Unreasoned belief ... hatred when confronted with other ideas ... do you have a special book too?

C'mon, one last chance. This thread is getting old now. The subject is "energy efficiency of electric cars", and to give you something to shoot for I've revised my calculation:
Since the Tesla makes so much of its ecological credentials, I'll compare it with a Toyota Prius T3. I'll also take into account Tesla's charger efficiency (80% reported by Tesla owners), and use better numbers for battery lifetime (500 full cycles) and power station efficiency (33% in the US).

The Tesla-S consumption at the socket is 814 kJ/km, and battery manufacture is 543 kJ/km. The heat energy required to generate that much electricity is 4.47 MJ/km, or 6.81 mBtu/mile.

Meanwhile the Pruis T3 claims 60.3 miles per US-gallon, which is equivalent to 1.40 MJ/km or 2.14 mBtu/mile. That's three times better than the EV!!!
(No, I don't include engine manufacture, same as I don't include the manufacture of the electric engine and regenerative braking system the EV uses.)

Off you go.

If you're interested in environmental science and would like to improve on my calculation then you're very welcome. However your fact-less attack smacks of religious fanaticism, and that's not ok.

Too many people are getting rich by making environmental claims that don't stack up. Resources are wasted on technologies that don't work, delaying progress towards solving our energy problems. We need less hype and rhetoric, and more careful analysis.

How you use the battery doesn't change its lifespan, because wear is proportional to use. If you drive 10% of the EV's range and subsequently recharge the battery by 10% of its capacity, the impact is equivalent to 10% of one full cycle. So you get 10,000 10% cycles instead of 1000 100% cycles, which corresponds to the same number of miles driven before you have to replace the battery.

Re. 30mph testing: I'm talking about normal driving circumstances for both cars. Running at 30mph continuous will dramatically improve the mileage of both cars, but it's a totally irrelevant scenario.

Re. fuel transport: Yes, that should be taken into account -- for the power stations also. They need to transport their fuel too. Extraction and infrastructure should also be accounted for. I'd love to see the figures.

Re. diesel mileage: Remember I'm using UK gallons. My own car, a 150 hp 1.9 ton 4WD, gets about 50-52 mpg (city driving, measured). There are plenty of more economical vehicles closer in size to the Tesla.

Re. battery manufacture: the data come from a Japanese study by Ishihara et al., and accounts for recycling.