* Posts by IanDs

75 publicly visible posts • joined 16 Mar 2015


Record-breaking Aussie boffins send 44.2 terabits a second screaming down 75km of fiber from single chip


The issue isn't so much getting enough data in -- there are already switch chips with 25Tb/s capacity, with 50Tb/s coming in the near future -- as what happens to the optical signals. Optical transport networks rarely treat the whole fiber bandwidth (>40THz) as one point-to-point pipe (one massive "superchannel"), they use ROADMs (Reconfigurable Optical Add-Drop Multiplexers) to route different wavelengths to different destinations. The closely-packed signals of the "optical comb" approach can't be used for this because you need guard bands between the channels to allow them to be separated and recombined -- and also they don't all come from the same source.

There are parts of the OTN system where one fat pipe crammed with data is all that is needed (e.g. undersea links of several thousand km) but these are only a tiny fraction of the total installed hardware, far more is used for shorter links within countries and continents and these all use ROADMs.

There are also issues to do with yield and reliability when you integrate so many optical channels into a single device.

So it is an impressive technical achievement, but don't think it's going to completely transform the world ;-)

Nokia: Oops, financials aren't great. Never mind, 5G will solve our woes


Re: Nokia 5G

Assuming the rumours are true, it looks like a bad decision only with the benefit of hindsight...

Rewind 3 years or so (typical development time for chips like this) to when Nokia (or whoever...) had to decide on a silicon supplier, and Intel were promising that their 10nm technology would have better PPA and cost than anything else and was going to be available earlier too, so going with it instead of the obvious foundry process (TSMC) must have looked like the way to get an unassailable lead on the competition. I also know from a previous life that Intel were offering very attractive wafer prices to try and bring in new foundry business, which looked very tempting at the time.

And if Intel had delivered 10nm when they originally promised -- or even now, only a couple of years late -- this would all have worked, and Nokia (or whoever...) would have a product with no competition and cleaned up the market. But with no end in sight to the Intel 10nm fiasco which now doesn't look like it will be able to yield big chips (which I expect ReefShark is, even if not as big as Ice Lake) until well into 2020, this could indeed be a complete disaster for Nokia (or whoever...)


Nokia 5G

There are also rumours going around that Nokia's 5G programme -- which they're relying on to save the company -- will be badly hit because they decided to use Intel's 10nm process, which is hugely delayed (3 years now?) getting to yield good enough for mass production, and this could be fatal for the company.



Should probably be taken with a pinch of salt, but it's public information that their ReefShark 5G chipset uses Intel 10nm so I wouldn't bet on this rolling out any time soon.

And in current affairs: Rogue raccoon blacks out city power grid after shocking misstep


The first real-life test our UPS got where I used to work in Maidenhead was when a cat got into the local 33kV substation and took out the power to the entire industrial estate. At least they think it was a cat, it's not that easy to tell afterwards...

Tech rookie put decimal point in wrong place, cost insurer zillions


The other side of the coin...

On holiday in France pre-Euro, took 200 quid (in francs) out of cash machine. Got home a couple of weeks later to find multiple messages to contact bank, including "tour account is overdrawn, please deposit 18 grand immediately). Turns out whoever entered the daily exchange rate for the bank had set 1 franc as 9.5 quid instead of the other way round. And the stupid bank had then merrily debited this from our account, not noticing that taking this amount out of a cash machine was not exactly likely...

Cue bouncing mortgage payment, standing orders etc -- took us weeks to sort out. Never even got a proper apology or any kind of compensation either. Thank you Nationwide :-(

Intel’s first 10nm CPU is a twin-core i3 destined for a mid-range Lenovo


10nm is tough to do if you're Intel, but apparently not for TSMC doing 7nm (similar process) with multiple chips already in the hands of customers, some of which are *very* big chips for applications like networking where Intel-level yield would totally kill them...


If you have yield problems with your process like Intel do with 10nm, a big chip is a route to disaster, so you do a small one first (yield drops very rapidly above a critical chip size).

But a two-core CPU with no GPU in Intel's 10nm process (at least 2x density of their 14nm) is going to be tiny, around 8x smaller chip area than a 6-core CPU with GPU like i7-8770K which is the top of their current consumer range.

This suggests their yield problems are *really* bad, since even a quad-core CPU + GPU would still be 3x smaller area than an 8770 and they haven't even gone for this...

Meet TPU 3.0: Google teases world with latest math coprocessor for AI


Re: serial liquid cooling

I'll bet you that just out of shot are the other two ASICs per pod also connected in series, so 4 in series in total. Temperature difference between them will still only be a few degrees, water cooling is very effective at this, the main power limit is the radiator.

My PC is on fire! Can you back it up really, really fast?


Re: @Lukes scope

Heavy electrical lab at Cambridge. 3-phase AC motor-generator pair used for machine efficiency experiments, idea being that motor drew a couple of megawatts and generator returned a couple of megawatts minus a few tens of kilowatts back to mains -- each of these was about six feet long and four feet diameter in old money.

"Wonder what happens if I reverse the field coil polarity on the generator?"

<whump> lights go out in the engineering labs. Also the whole of west Cambridge...

Bill (oldish famously grumpy but *very* experienced lab manager) comes running in screaming "Which one of you f**cking morons reversed the field coils on the MG set?"

Not the first time this had happened was my guess...


Re: Fire

According to John D Clark in his fantastic book "Ignition", if you're unfortunate enough to be dealing with CTF (chlorine trifluoride or ClF3, a rather vigorous oxidiser -- or fluorinating agent, to be more accurate -- used occasionally in rockets by crazy people), sand won't be any good to put out a fire since CTF is perfectly capable of burning it -- along with metal, rock, asbestos, wood, test engineers...


Re: I recall even my mum (a bit like Dilmom) telling me a fire story

Nerd at school who shall remain nameless had father who worked at chemical works (Hickson & Welch in Castleford), asked him to bring a lump of sodium home to try this. "Lump" turned out to be a brick -- named due to size -- kept under oil. Nerd builds motorised Meccano conveyor belt the length of back garden, places large bucket filled with water at far end, places sodium brick in Tupperware container of oil on belt, starts belt, runs for cover. Cue large smoking crater in lawn, bits of Meccano and garden fence found several gardens away...


Re: I recall even my mum (a bit like Dilmom) telling me a fire story

Nitrogen tri-iodide sprinkled in front of blackboard. After various amusing bang-crackle-pop noises realises what's going on, wipes floor and bottom of shoes with dampened blackboard eraser -- you know, those ones with a strip of padded fabric in a wooden handle. By the next lesson (different chemistry teacher) eraser has dried out, said teacher has habit of banging eraser on board to attract attention...

Take-off crash 'n' burn didn't kill the Concorde, it was just too bloody expensive to maintain


Re: Love reading about this splendid machine

I think the author of the "Champagne..." book made quite a few posts as "Landlady" on the PPRuNe Concorde thread, some of which probably weren't fit for publication...


For anybody who's really interested in the technical background and history of Concorde, there's a *very* long thread (>100 pages) on PPRuNe with an enormous amount of inside information from lots of the people who designed, built, tested and flew it.


A fascinating way to spend an evening, if you're interested in that sort of thing :-)

Japan's Robo-Bartenders point to a golden future


As any fule know, the beer glasses -- at Narita, in Japan and in most of Europe -- are lined to leave room for the head, not to be filled to the brim. Anyone whining because their beer isn't filled to the top, go back to England where pubs often serve beer with a head in glasses designed to be filled to the brim...

MY GOD, IT'S FULL OF CARS: SpaceX parks a Tesla in orbit (just don't mention the barge)


Nah, that's not heavy lift. *This* is heavy lift...


"Super" Orion would have lifted 8M tons to Mars compared to 17 tons for Falcon Heavy ;-)

Electric cars to create new peak hour when they all need a charge


All the comments about the grid not having enough capacity and needing massive enlargement are just plain wrong -- the grid (and the generation behind it) has spare capacity almost all the time except at short power peaks, because if it didn't we'd get blackouts and power cuts all the time.

During the night there is massive spare capacity which goes unused, and means generators have to be shut down and restarted which is not good for overall efficiency -- this is where the main energy for charging electric cars would come from. Yes people may get home and plug their car in at 7pm, but mostly they don't then care when it gets charged so long as it's done by the next morning, so they could use cheap overnight power. Yes there will be exceptions to this, and they may have to pay more for power. Yes the batteries of cars which are plugged in and charged can be used as a resource to feed power into the grid at peak demand times when prices will be higher, and then get recharged when prices are lower.

People who can't park next to a charging station (e.g. at home) will be at a disadvantage until charging at the kerb is rolled out -- but this isn't a reason to say electric cars won't be widely adopted by those for whom they work.

The simple fact is that petrol/diesel cars are cheap and convenient and quick to fill up and long range -- and also inefficient, polluting, and unsustainable as fossil fuels are phased out and renewable energy takes over, which we'd all better hope happens for the sake of the planet and our children.

Whinging about how you don't want to give up your petrolhead habits will soon sound about as convincing as complaining because you're no longer allowed to inflict your smoking habits on other people inside cars and buildings.

Meltdown, Spectre: The password theft bugs at the heart of Intel CPUs


Re: What I don't understand

Nope, AMD don't allow speculative execution at user level to access kernel level data, this is prevented by hardware. Intel do, as do some ARM CPUs.

The End of Abandondroid? Treble might rescue Google from OTA Hell


My OnePlus 3 has had regular OTA updates, most recent one was to Android O last week.

Tesla reveals a less-long-legged truck, but a bigger reservation price


"In efficiency terms though, it's a LOT cheaper to stick with diesel though given the infrastructure already exists and it's got a far higher energy density."

Nothing is cheaper or more compact or more convenient than digging/pumping fossil fuel of one sort or another out of the ground and burning it to drive vehicles. At least, so long as you ignore the environmental impact and indirect costs (global warming, pollution) of more and more people worldwide doing this, which many people (and governments) are realising is not sensible.

If the real costs of sticking with diesel were added to the current fuel costs, it wouldn't be cheap any more -- and it's likely to be politically unacceptable to keep doing this in most places in the not-too-distant future, which is likely to mean much heavier taxes to discourage people from using it.

And before anyone jumps in about guvmints taking away the freedoms of the people, that's precisely the job of guvmints if said people are doing something that kills a lot of other people, wipes out many species on the planet, and will lead to huge upheavals in society if not checked. Businesses won't do it voluntarily, all they're concerned about -- quite rightly, because it's what businesses are for -- is making money, not saving the planet or lives.

Thou shalt use our drone app, UK.gov to tell quadcopter pilots


The fact remains that drones like this that can be bought off-the-shelf and flown by an untrained (and uncaring) member of the public are a new thing, as opposed to enthusiast model aircraft owned by much more committed people. As such there are definite threats they pose to society -- real danger on one side (bringing down a plane, causing a car crash, dropping on people from a great height) and genuine invasion of privacy on the other (spying on nude back-garden sunbathers, small vulnerable children, people in general), which are going to need some type of legal control to stop idiots abusing them. Of course this will add some restrictions to people who use them sensibly, but this is true of pretty much everything. The difficult bit is getting the tradeoff right between onerous restrictions on sensible users and controlling stupid ones, but there's no right answer here -- the howls of anguish at restrictions from freedom-loving drone fans will be drowned out by press (and society) outrage the first time a plane crashes and kills people -- maybe lots of them -- due to a drone strike. And I wouldn't mind betting that Heathrow is one of the most likely places in the world for this to happen first, given the huge numbers of houses under the flightpaths...

Fresh chips from Intel (yay?) at 14nm (awww)


Re: 14nm, 10nm, 7nm

It's pure marketing nowadays, it doesn't even refer to the smallest feature -- at 7nm, literally nothing on the chip is this small. If we'd stuck to the traditional (and meaningful) "minimum metal width/spacing", 7nm processes would be called 20nm...

Which keys should I press to enable the CockUp feature?


Re: Optical mouse mat

We used Apollos too, in fact I think we had the first colour DN660 in the UK -- when Motorola were late delivering the 68000, Apollo built a bit-sliced clone of it out of 2900 series ECL logic. From memory the CPU filled 4 PCBs about 20" square each, in a massive card cage in a box the size of a washing machine that warmed up the office nicely but wasn't what you'd call reliable. Its one redeeming feature was that on top of the card cage was a space exactly the right size for sliding in a pizza to keep it hot ;-)

And we wrote our own layout tool to run on it, gesture-driven from a tablet like the old Applicons -- it was faster to use than anything commercially available, watching an experienced user drive it was like magic, them scribbling away on a pad while polygons literally flew around the screen...

How to build a plane that never needs to land


Re: 2000 hour inspection cycle

Why are you assuming that the same reliability/certification/inspection regulations should apply to a lightweight unmanned drone as a passenger aircraft? If it fails it doesn't kill 200 people...

How to help a user who can't find the Start button or the keyboard?


Re: Stop being a smartarse who makes things worse

You think that's bad - my son's left-handed, and to prevent RSI he not only puts the mouse to the left of the keyboard (obvious) but swaps the mouse buttons over (not so obvious). And when working with test equipment like multichannel oscilloscopes -- with channel control buttons and knobs helpfully colour-coded to match the on-screen traces -- he changes the trace colours so he can distinguish them (he's also red-green colour blind). Watch people trying to use it when the red buttons now control the (some other colour) trace...

Cyber-terror: How real is the threat? Squirrels are more of a danger


A Russian guy I met in Israel told me he'd seen a nuclear reactor in the USSR controlled by a pneumatic computer, with all the gate functions realised by compressed air. Filled a room, noisy as hell, got round reliability problems using redundancy and majority voting to the point where a faulty module could be found and replaced while it was running the program without interruption.

Why bother? You don't get anything more rad-hard than something with no electronics in it at all...


Re: It's all about the blinky lights

A cat took out our local substation, which is when we learned how long our UPS would keep the servers running (answer: not long enough). Leastways they thought it was a cat, going by the crispy burnt bits left when it got into the 11kV busbars...

Now VW air-pollution cheatware 'found in Audis and Porsches'


Just how low the NOx emissions are with and without the cheat software is kind of irrelevant, the issue is that such devices are explicitly illegal and VW broke the law. Not liking the law or "everybody does it" isn't an excuse for deliberately breaking it or whining if you get caught, just like speeding -- if you get caught, 'fess up and pay up.

Every time I hear the term "rogue engineer" in this story it makes my blood boil. The chance of middle and upper VW management not knowing about and approving this type of defeat device is precisely zero, regardless of whether what started it off was an engineer saying "hey, I know how we can pass the emissions limits without urea injection" or a manager saying "come on guys, you gotta find me a way to pass the emissions tests without urea injection 'cos we've publicly said we don't need it and it's too expensive".

Finding documented proof of how far up the chain this went might be more difficult, this is the type of thing where people are often given verbal "don't put anything in writing" instructions so there can be some chance of deniability -- like the phone hacking scandals, nobody believes the denials of Rebekah Brooks and the like but if they say "I didn't know" or "I can't remember" enough times they get away with it.

The suspicion now must be that having found this in two completely different VW group products, it's in all of them -- after all, once you've found something that lets you pass the tests and save a shedload of money, why wouldn't you use it everywhere?

Whether other manufacturers are doing the same thing remains to be seen, but there's no proof yet that this is the case. Mind you, do you believe that VWs competitors didn't know what they were doing, given the way people move between companies? If they knew and weren't doing it themselves you'd have thought they'd have blown the whistle on VW, which suggests that they did know and kept quiet because they were doing it too...

TalkTalk: Hackers may have nicked personal, banking info on 4 million Brits


Re: He who buys Chinese must die by Chinese

At least Huawei kit is less likely to have backdoors built in for the NSA and GCHQ than kit from their USA/UK/EU competitors...

Self-driving vehicles might be autonomous but insurance pay-outs probably won't be


Re: Road Markings

You're still fixating on the relatively rare occasions where a *good* driver (not many of them about, in spite of what everyone thinks) might do better than an automated one.

These are massively outnumbered by all the rest of the time when the reverse will be true, because most drivers on the road today are *not* good -- and certainly not all the time. Automated cars don't get tired, inattentive, phone or text during driving, run red lights, ignore road signs -- all the things that people do all the time which cause most "accidents".

The statistics will be hugely in favour of automated cars, and it's this that will drive things like insurance and lawmaking, not the exceptions -- which will undoubtedly get rarer as the cars get better.


Re: Cleaning the car

So carry on driving your own car, and paying for it -- nobody forces you to use a car club, nobody will force you to drive an autonomous car -- at least, not for a long time. But expect it to get more expensive to insure in the short term, and maybe prohibitively so in the long term when <10% of the cars (manual) cause >90% of the "accidents".

At the moment drivers get sued for this -- or their insurance companies do, at least in the UK -- but the general view is "well, accidents happen". When this is no longer the case, you can expect the blame (and cost) to shift much more heavily onto the driver for being stupid enough to endanger lives by driving a manual two-ton killing machine.


Re: Mark Twain

I might be depressing, but it's exactly what will happen. In the end it is all down to risk -- try telling someone whose relative was killed by someone having "fun" in a car that it's OK to act stupid while piloting a couple of tons of metal at high speed -- the one taking the risk not being the one who got killed.

Nowadays people who to have major fun in a car go to track days, where they can burn rubber and crash to their heart's content without damaging anyone else. In future maybe this will be the only option unless you can afford stupidly high insurance premiums -- and even then think of the resulting outcry when (for example) a Saudi prince "having fun" runs over and kills a toddler (aww, think of the little kiddies!). Maybe anybody driving manually then becomes a social pariah, just like smoking over said toddler is seem as today...


Re: Forget the techies, ask professional drivers

Lots of accidents and fatalities -- often to other people -- are caused by drivers "having fun" on the public roads. If you want fun in a car, do it on a track or on a road with no other people to hit or kill when it all goes horribly wrong. I'm not against fun, just when it damages other people...


Re: Road Markings

So you think you can tell the difference between one shade of white and another better than a computer hooked to a high-res camera? Bet you can't...


Re: Road Markings

Why would it take so long, it's a relatively simple problem in physics to do with grip, slip angles, coefficient of friction, spotting ruts and so on? For sure, most drivers can't behave correctly in these conditions either going by the number of 4x4s you see in ditches every time it snows...


"Driver's intuition" is nothing more than careful observation and correlating this with what you know happened in the past, which is the kind of thing that autonomous cars ought to get pretty good at pretty quickly -- after all it's just extracting data and spotting trends/correlations, software can be pretty good at this.

But you're still worrying about the wrong thing -- even if we couldn't give them "drivers intuition", for every accident caused by this they would probably avoid a hundred due to not driving like a twat, net gain 99 fewer accidents. Bad drivers driving like idiots cause way more accidents than careful observant drivers avoid with intuition...


It's difficult to see how any decent autonomous car could be as bad at driving as the moronic meatbags behind the wheels of most of today's cars, they certainly wouldn't have as many "accidents" -- which is not what the emergency services call them nowadays because they rarely happen due to unforeseeable circumstances, "incidents" are what drivers cause by behaving like inattentive tired phoning texting looking at the girl on the pavement humans.

Everyone who objects to them on the "what happens if they have an accident under these unusual circumstances?" grounds is ignoring the fact that they won't have most of the "accidents" that people do -- yes there will be cases where an accident happens because the software goes wrong or it can't cope with something, but probably at least 10x fewer than with people behind the wheel.

Face it, most drivers are crap and most "accidents" are caused by drivers -- it would be pretty difficult to come up with an autonomous car as bad as this...

Minicab-hailing app Uber is lawful – UK High Court


Re: What am I missing

Going by the judge's remarks, legally it wouldn't matter if Uber did meter the journey in real time, so long as it did it by the smartphone sending data back to the server to calculate the cost.

The issue is not just the definition of a taximeter, but that the vehicle has to be *equipped* with it -- which means it's physically attached to the vehicle and legally associated with it, for example the vehicle owner is liable if it reads the wrong fare, not the driver (unless they're the same person).

Uber may be not exactly playing the game in areas like taxation, liability, minimum wage and so on, but if people (and governments) don't like this they have to find a *legal* way of making them step into line -- if necessary, by changing the law.

The judge's clearly explained judgement is that according to the current legal definitions of black cabs/PHVs/taximeters Uber are not breaking the law. The fact that they might destroy existing businesses is an unfortunate consequence of the fact that they've come up with a new business model and the old taxi forms haven't -- if the government wants to protect the old jobs (which let's face it, is unlikely since they'll be replaced by Uber ones) they'll have to pass a law to make it happen -- but as the judge said, it's not easy to define a law which stops Uber but doesn't also stop PHVs.


Then use this system. Or use Uber. Or phone your favourite PHV. Or hail a black cab.

It's your choice, nobody is forcing you to use Uber. Of course if lots of people use Uber and it's successful some of the other options might disappear, but that's how business works -- unsuccessful ones disappear. I'm sure the hand-loom weavers had the same view about power looms as black cab drivers and PHVs do about Uber, but that didn't stop them losing all their business to the upstarts.


How exactly are Uber ripping PHV operator's phones out? They're just offering another (very convenient and easy and reliable) way of booking and paying for a PHV, if you don't like it don't use them.

If other PHVs offer similar electronic booking/tracking systems, good for them. Uber don't have a monopoly, anyone who wants to (including PHVs) can offer a similar service. The problem is that to do this takes modern IT knowledge and infrastructure, and most PHV firms consist of a few people who basically know how to use a phone and write your details down wrongly...

I was in China recently and *everyone* uses a taxi booking/tracking service a bit like Uber -- you can pick a car on your phone, see how far away the car is and when it will get to you and track it, phone the driver if they're delayed and so on. But the fare is argued about with the driver when the cab arrives, which is fine if you're Chinese and know how to negotiate, not so fine if you're a tourist who accepts the drivers first (high) bid. Great if you like haggling, but Uber seems a better system than this since at least there's some estimation/regulation of fares.


Re: Couple of things that bother me

I don't know where you live, but in and around London PHVs are *way* cheaper than black cabs.

The difference between Uber and a conventional PHV booked in advance is just the charging mechanism, which is just a contract, and there's no law that says *how* payment has to be decided on so long as it is agreed with the customer -- fixed fare in advance (normal PHV) is one way, fare depending on time/distance/surge (Uber) is another way, fares depending on astrology and the phases of the moon would be perfectly legal so long as it was agreed that this would be done. You're free to agree (and use Uber) or not (and use a black cab or PHV).

If you buy air tickets the price varies hugely depending on demand and availability, Uber is bringing some of this to cab fares. But PHVs (and black cabs) do this anyway with higher fares at odd hours/holidays/Xmas, it's just another way of doing the same thing.


Re: Taxi Meters are "Outlawed for private hire vehicles"

Correct -- outside London a PHV can either be fitted with a taximeter (in which case the same rules apply as to black cabs in London, the meter must be used and shows the fare and that's what you pay) or not (in which case the same rules apply as to PHV in London).

So outside London PHVs can either be like black cabs (in London) or PHVs (in London), both of who hate Uber for the reasons I said -- the position is exactly the same but with different labels.


The judge's explanation was very clear. The basic difference between PHVs and taxicabs (in London) is that legally taxicabs *must be equipped with* a taximeter to calculate fares based on time and distance, and PHVs *must not be equipped with* such a taximeter.

Regardless of their disruptive business model or how they calculate the fares (e.g. surge pricing), Uber cabs are not *equipped with* taximeters, and are therefore legal PHVs with a novel method of booking/paying/charging -- which is not illegal according to the law, no matter how much black cab and conventional PHV drivers dislike the fact, and how many of them might be put out of work as a consequence.

The judge can only rule on the law, and its clear that Uber are not breaking it. If this has terrible unforeseen social consequences (job losses, downgrading of well-paid full-time jobs to badly-paid part-time ones) then it's up to the government to change the law to stop this. But as the judge points out, this is almost impossible to do without any law designed to stop Uber also hitting existing PHVs who use some other way of booking and calculating fares. So long as Uber drivers and cars are controlled in the same way as other PHVs (licensing, vehicle checks) there's no way to legally separate them out and stop them taking business from black cabs and other PHVs.

Basically the black cab drivers are annoyed because they are locked in to an old expensive non-GPS business model by the black cab regulations, and the existing PHV drivers are annoyed because Uber have a much more convenient/safe/reliable way of booking and paying for a PHV than they do. They've both been blindsided by new technology and there's nothing they can do about it, apart from trying to persuade the authorities that Uber is illegal -- which they have singularly failed to do.

Even if people dislike Uber as a company or because of their business model or surge pricing or job losses or for whatever reason, that isn't a legal reason to stop them operating -- what's going on now is the equivalent of the hand-loom weavers smashing mechanical spinning/weaving machines a couple of hundred years ago, and just as likely to succeed...

Top boffin Freeman Dyson on climate change, interstellar travel, fusion, and more


And your point is what, exactly?

The numbers (95% and 5%) are taken from international reviews of the opinions of experts in the field of climate science. If you'd rather believe pseudo-scientific people who aren't experts that's your prerogative, I'd rather listen to people who are likely to know what they're talking about.

There are still a few climate change deniers -- many of them scientists with links to fossil fuel and allied industries, which doesn't increase my confidence that they're impartial -- but the numbers are falling year by year as the evidence against them builds up.

Even if you're one of them and convinced that you're right and all the others are wrong, you still need to think about the consequences of you being wrong (mass starvation etc.) compared to the consequences of them being wrong (reduced oil company profits, no more SUVs to look big in).

Do you feel lucky, punk? Well, do ya?


Just because somebody is a world expert in [thing A] doesn't mean they're any more of an expert in [thing B] than a random passerby. Of course they might know a lot about [thing B] or they might be spouting utter drivel, there's no way to be sure.

Yes it's possible that more than 95% of the scientists on the planet who actually understand all the issues about climate change are wrong about global warming. But it's 20x more likely that the other 5% who say it's not caused by man are wrong, and this ratio is going up year by year.

If we believe the 95% and decide to seriously do something about it, the worst that can happen if they're wrong is some fossil fuel companies will make less money and lifestyles will have to change to reduce energy consumption.

If we believe the 5% and hope the problem will go away and they're wrong, the worst that can happen is massive flooding, food shortages leading to starvation, wildlife extinction, and massive disruption of the world economy and life as we know it.

Given the odds and the consequences, any rational person -- even a gambler with no opinion either way on who was right but just considering the odds -- would conclude that it's much better to assume that the >95% are right and the <5% are wrong.

Google's .bro file format changed to .br after gender bother


So if someone wanted to use a file extension of ".pro" for files with a professional purpose, would they get told off for the same reason as someone earlier who suggested ".ho" ?

Boeing builds British Airways 787 Dreamliner in 4 minutes


787 isn't as pleasant to fly in long-haul as 777/747 or Airbus -- one reason is that something that sounds both cool and a good idea -- LCD "window blinds" instead of pull-down plastic ones -- isn't because they still let some light through when "closed". Try going to sleep (to get onto destination time) when the sun's up outside and then see how good you think they are...

VW’s case of NOxious emissions: a tale of SMOKE and MIRRORS?


Urea injection (Adblue, Bluetec etc.) does reduce NOx emissions both during emissions tests and normal driving, which is why most manufacturers use it. VW decided to save money by not fitting it, so the only way they could then meet the standards -- especially in the USA -- was by cheating.

Other manufacturers may have done similar things but there's no evidence yet, and no need for them to do it -- if you've gone to the expense of fitting urea injection you can pass the tests without cheating.

Whether all diesels emit more pollutants in real driving than during tests is a separate issue and not unlikely, in the same way as they don't deliver the same economy in real life as in tests. But that's nothing to do with cheating like VW did, it's due to the tests being completely unrealistic and easy to get too-optimistic results with various fiddles which are within the rules.

Full duplex! Bristol boffins demo Tx and Rx on the same frequency AT THE SAME TIME


Does this work even with time-varying propagation conditions, including handset movement, local reflections from moving objects and so on? It's not so difficult to get good cancellation in a static demo system where you have lots of time to estimate the response accurately and cancel it out, it's a lot more difficult -- maybe impossible? -- in a time-varying mobile radio system.

EU graciously lets Dutch splurge €33m on 'leccy car charger network


Re: The ignorance spewed by ACs is laughable

AC, try looking at some reliable verified numbers instead of rehashing Internet twaddle. Even allowing for the entire lifecycle of electric cars (including construction and disposal/recycling), their CO2 burden is a lot lower than IC cars -- for example, look at the real numbers in SEWTHO. Recharging and battery fires (including in accidents) are a hazard, but a much smaller one than using petrol. The well-to-wheel economy of electric cars is a lot better than IC engines, and even more so compared to the stupidly inefficient hydrogen/fule cell proposals. Even if the electricity comes from fossil fuels, the overall pollution of EVs is a lot lower than IC engines, and is better filtered and not on the street.

Yes they're still expensive, have more limited range, take longer to refill even at supercharger stations, and would put a big load on the grid if everyone used them. These disadvantages have to be compared to the CO2 and street pollution savings, quietness, and performance. They're not for everybody, but for some the advantages are compelling, and the number of people who this is true for will only go up in the future.