I got rid of the offending lights some time ago (the problem was mainly from a laptop which I now simply slide under some furniture). My question was one of getting to the bottom of why they bothered her and not me. It appears that I am be the odd one out here in apparently having the ability to snore right through the problem.

Can I just ask for some clarification? Are we here discussing LEDs which are so bright that you can see them at night even with your eyes closed or ones you can only see with them open?

The reason I ask is that my wife often complains about LEDs shining at night and in every case, no matter how much I let my eyes accustom to the dark, I can't see them with my eyes closed and since I don't sleep with my eyes open I don't find them a problem. My wife admits that she also can't see them with her eyes closed but still says that they make it hard for her to sleep but she can't explain why.

If someone can provide some sort of convincing explanation of why lights which you can only see with your eyes open make sleeping difficult I would be most appreciative.

I am afraid that I have never been persuaded by "civilisation will destroy itself" arguments, because (a) they have a poor definition of "destroy" and (b) options for further evolution don't seem to be well considered.

Expanding briefly if I may: Most cataclysmic events postulated don't seem cataclysmic enough. Suppose for example there was a huge nuclear war. Might that and the ensuing nuclear winter push humanity back to the dark ages? Well, very possibly it might, but we know from practical experience that getting from the dark ages to now takes about 1500 years or so, probably rather less if you have the smoking remains of the previous civilisation to get clues from. So, we get another go at being an advanced civilisation and presumably can repeat this depressing episode over and over again (see Azimov's excellent 1941 short story Nightfall).

For these cataclysmic events to actually make mankind extinct the population has to be reduced below a practical reproductive minimum (which clearly depends at least in part on how spread out the survivors are). We could imagine perhaps some sort of synthetic plague to which no-one is immune and which survives in the environment to such an extent that even small highly isolated populations are eventually infected. It sounds a bit unlikely to me, but again we know from experience that given a few million years our ape cousins will evolve to replace us. Of course, all primates could also be vulnerable to the disease, in which case we just have to wait even longer for an evolutionary replacement.

Conclusion: Short of managing to destroy all multicellular life forms, planets which evolve advanced life will have advanced civilisations from then on with possible gaps.

Yes, well maybe the aircraft's signature was too low for the threat system to engage them, but if you want to increase the signature of the stealthy aircraft there are lots of easy ways, such as:

1) Lower the undercarriage.

2) Many low signature aircraft have corner reflectors which either bolt on or are hidden behind doors and which greatly increase the radar returns. They are used to hide the true signature when flying somewhere where someone may try to measure your radar cross section. I have no idea if the F35 has such a feature, but I would be surprised if it doesn't.

3) Fit external stores. I don't know if the F35 supports this option.

So, a story about something that isn't a real problem and instead suggests a badly planned training exercise re-cast as an opportunity to say how great their aircraft are.

I am someone who does actually read the TOS for websites. I rarely like what I see and as a result, Slashdot is one of the very sites to which I subscribe.

However, the plain fact of the matter is that the vast majority of people don't read them and (here is the vital fact) almost always they don't subsequently feel that they have been disadvantaged as a result. For some strange reason, criminals and the generally dishonest are not setting up web sites, getting users to subscribe and then legally fleecing them. I am not suggesting silly things like First Born, but simple strategies like firstly including a clause saying you can unilaterally change the terms later (practically everyone does this) and then when you have a good few users change the rules to impose huge retrospective fees. Would this not work? I presume many people would challenge the bills in court and I have no idea what the courts would rule. Anyone know any case law?

It is clear to me that governments aren't interested either. Here in the UK, when you go into a shop you might often see a sign describing such things as their returns policy. At the bottom it will invariably say "Your statutory rights are not affected". This is because here consumers can't contract out of their basic consumer rights (e.g. if the product is faulty you are entitled to your money back and don't have to accept a voucher instead). There are some similar protections for buying things online (distance selling regulations) but none so far as I know that govern the contracts on web sites.

I strongly suspect that most smaller organisations don't even read their own TOS and simply copy them from someone else. I have often felt that with the vast majority of websites for which one might need to sign up being basically the same, it would be a good idea for the government to create three or four boilerplate TOSs to cover say 90% of cases. Web sites could then simply have a sign saying "Our web site is governed by UK Gov TOS 3" (I am sure a catchier title could be invented). Consumers wouldn't need to read the TOS because they were all the same and had been carefully checked, but web site owners would also benefit by knowing that their TOS had been well written (at someone else's expense) and would therefore be more likely to stand up in court than one they copied from another similar site and then got their nephew doing law at high school to tweak.

So far as I can see, these articles express the view that a society based entirely on objective decision making wouldn't be perfect and therefore shouldn't be considered. Well, Duh! Surely it is completely obvious that it wouldn't be perfect, not least because there are large areas of the human condition not amenable to the scientific approach.

But, surely the question is not whether such a society would be perfect, but whether it would be better - on average - than other arrangements currently on offer. I have no idea what the answer to that question is, but may I submit that if one is to postulate such a society then that is precisely the question which needs to be asked.

Perhaps I should add for those overseas that very few of the current members of the House of Lords got there through birth. The vast majority these days are "Life Peers" who have been elevated to the Lords by the government including both Baroness Stowell and Baroness Shields.

Firstly, the Prime Minister can be a member of the House of Lords, although that hasn't happened in modern times. Lord Salisbury was the last Lords PM (1886 to 1892). He had previously been an elected Member of Parliament but had been elevated to the Lords (1868) before becoming PM. Cabinet Ministers can be from the Lords, although the only current full such cabinet member from is Baroness Stowell of Beeston who is the leader of the Lords. There are however several current 'deputy' ministers from the Lords including Baroness Joanna Shields who is the Minister for Internet Safety and Security (I think that means internet censorship).

Also, the Prime Minister doesn't have to be the leader of the largest party, but in practice they always are since Parliament can throw out any Prime Minister they don't like and clearly the biggest party will like their leader best.

So, the big problem with Chip+PIN is that you have to keep the card in for the duration of the transaction? Seriously? Good grief people in the USA must be short of things to be inconvenienced by!

I have to say that I didn't quite understand all of your explanation, but fortunately as I never to the the USA I don't need to (Phew!). Do I however deduce that before long mag stripes will be disappearing from your cards and the rest of us can then give them up as well?

BTW, why doesn't the candy store put up a sign saying "No card transactions below $5". Plenty of shops in the UK do, but perhaps you have a law (or more likely hundreds of different laws) against it.

I can confirm that the switch to Chip and PIN caused very few problems here in the UK. At least not that I as a consumer noticed, it might have been a pain for the shop owners.

Therein lies the problem. Here in Europe (and practically all of the rest of the world) we have switched to CHIP and PIN which allegedly makes skimming much more difficult. Unfortunately, this technology appears to be too complex for Americans to understand so we all have to have mag stripes on our cards as well just in case we ever go there. I never go to the USA, so the mag stripes on my cards are entirely useless other than for skimmers.

Does anyone know of any UK banks which offer a "I am never going to go to North America so please send me a card with a blank mag stripe" service or even a "I sometimes go to North America so please send me two cards, one with mag and one without" service?

Yes, Microsoft may have an incentive to use things like UEFI to make it harder and harder to run anything other than their latest OS on new hardware. They may even have the gall to try to move to a subscription model so that you have to keep paying even if you don't want to upgrade (didn't I read that Adobe did something like that?).

Fortunately, devices like the Raspberry Pi are very hard for MS to control and the latest versions are getting fast enough to use as a normal Linux PC.

As I understand it, it was always perfectly clear that the vote was not binding and merely a mechanism to collect suggestions. David Attenborough did get a substantial number of votes and those who voted for that name were probably rather more interested in suitably naming a polar research vessel than just having a laugh.

Having said that, allowing a completely open vote online rather than allowing people to choose from a short list was clearly a daft idea and asking for trouble. They were lucky that the most popular name was at least repeatable in polite company.

No true!

I fear that you have entirely failed to grasp the point I was making. It is true that the transmit signal is many orders of magnitude stronger than the receive signal, but one cannot fix that entirely with the circulator, no matter how good it is. Time for circulator and antenna 101!

I typical ferrite circulator has three ports (let's call them A, B and C). Energy put into port A comes out of B, energy into B and out C and in C to out A. You get the idea. Now, as with everything in life, circulators aren't perfect and they have a parameter called 'isolation'. I typical value for a modern circulator is 20dB (or a power factor of 100). This means that if I for example put 100W into port A, then 99W will come out of port B and 1W will go the wrong way and come out of port C (in practice a little bit of the power will be lost internally as heat). Supposing that I connect A to the transmitter, B to the antenna and C to the receiver. In my example I will get 1W flowing into the receiver which could be 100dB (10^10) more than the intended receive signal. Clearly something else needs to be done, but making the circulator better won't help. Why?

Because of reason (b) in my comment - the return loss of the antenna. Antennas also aren't perfect and they have a parameter called return loss. An ideal antenna will take all the power from the transmitter and convert it into electro-magnetic waves propagating away. Real antennas however have imperfections and some of the power from the transmitter goes into the antenna and bounces back out again. A typical value for a good antenna is 20dB. Really good narrow band waveguide antennas (e.g. a decent radar) might manage 30dB, but the antenna on you mobile phone or Wi-Fi base station may well only manage 10dB. So, where does that leave us?

Returning to my example. 100W comes out of the transmitter and 99W goes to the antenna. If it has a 20dB return loss (if we are lucky) then 1W (give or take) will bounce back into circulator port B and nearly all of that will emerge from port C and go into the receiver. So, the receiver is getting 1W due to circulator imperfections and about 1W due to antenna imperfections. We can improve our circulator until the cows come home and the most we will do is reduce the power into the receiver from 2W to 1W which isn't going to save the day.

As I said in my comment, the real cleverness here is not the design of the circulator (which is probably as good as it needs to be), but the amazing performance claimed for the subsequent (and not very well described) canceller.

The gist of what is clever here is the canceller which removes the transmitted signal from the receiver. Circulators have been around for donkey's years (not just in military systems) but they are bulky (especially at lower frequencies such as those for mobile comms). The are often used to allow a single antenna to operate at both transmit and receive either alternately (e.g. radar) or on different frequencies (e.g. satcom). Making a solid state one is clever, but this isn't the first one.

However, some of your transmit signal will always end up in the receiver for three reasons; (a) the circulator isn't perfect, (b) the antenna doesn't have a perfect match so some of the transmit energy sent to it bounces back again and (c) energy can reflect back from the immediate environment. Cancelling schemes exist, and invariably consist of some mechanism for sampling the transmitted signal and feeding just the right amount back into the receiver exactly out of phase. In theory this works, but in most practical circumstances the extremely high level of cancellation needed requires a completely unachievable precision.

For added pain, the solution tends to be very narrow band and the cancellor's settings have to be continually updated as the transmit interference changes (particularly in a mobile environment due to (c)).

If they have managed to make this work in a practical and useful way then it will be very impressive, but I would need to see some real world experiments to be convinced of its practicality.

The point I am making (Derp) is that location isn't enough. Even if I could enter the coordinate the car will still need to know about such niceties as one way roads and local traffic rules. Example: Sometimes I go and park at another site where some areas are shared with aircraft movements which involves a load of specific driving rules.

Now, on the public highway all complexities like one way systems can be gathered and entered into the system, but for private land this sounds a lot more difficult. It seems to me that a far simpler solution is some simple driving controls allowing you to manoeuvre the vehicle at least at slower speeds.