Posts by Nigel 11 3191 publicly visible posts • joined 10 Jun 2009
LPG is liquid propane gas (could also be low pressure gas). It's not under high pressure, it's not a great hazard, the public can buy cans of it for blowtorches, camping cookers, etc. and in rural areas houses have large tanks of the stuff as their domestic gas supply delivered by even larger tanker lorries.
CNG is compressed natural gas, ie methane. To store a useful amount in a vehicle means a very high gas pressure.
Why no mention of compressed natural gas? True, it's still a fossil fuel. However, less CO2 emissions than diesel, and extremely clean-burning.
The barrier with respect to ordinary motorists is whether they be trusted with a high-pressure gas cylinder, and especially with the refilling thereof? That's less of an issue for professional drivers with strong vehicle inspection regimes already in force.
Another thing not mentioned is that cars with batteries employ regenerative braking. The greater the amount of start-stop driving, the greater the advantage of this technology, even if it's a hybrid with little electricity storage capacity. London taxis probably do more start-stop than any other sort of motorist. Why don't they take the drive system from a Prius and put it in a taxi? (as a solution that's available now)
Also in passing - why isn't there a van variant of a Prius? City centre vans are probably nearly as much start-stop as taxis but I'd guess lower annual mileage because more time stopped loading and unloading.
From what little I know of the taxi trade, the article seems rather pessimistic.
Don't taxi drivers spend a significant part of the day parked, either in taxi ranks or taking (legally mandated?) breaks? If these were equipped with fast charging technology, then a taxi could be recharged several times per day, not just once.
Equipping taxi-only locations with fast charging facilities would be easier than rolling them out to the general motoring population. Similarly battery-swap stations would be easier for a fleet of standardised London taxis.
Methinks firing big heavy slugs at aircraft and missiles is not an answer. One would have to hit a very small target with an unguided projectile. At long range - even at Mach 7 - the target will be able to see it coming and dodge. Especially so with something that by definition emits a very large low-frequency electromagnetic pulse when fired. Countermeasure - random automated jink on e-pulse detection. So stick to guided missiles for engaging aircraft at range, and rapid-fire machine-guns at short range.
Which leaves only ships and stationary targets which might be able to defend themselves against guided missiles and bombs.
The article doesn't mention the root-shoot balance which I read about elsewhere. It is an even more potent stabilising factor.
Plants need CO2 from the atmosphere and water and minerals from the soil. Leaves (shoots) collect the former, roots the latter. Plants, like all living organisms, self-optimize. More CO2 in the atmosphere means same leaves can collect more CO2, and so the plant's optimization shifts it towards growing a greater proportion of roots compared to shoots.
When the plant dies, its roots are most likely to trap carbon in the soil. The deeper the roots, the more likely it is that the carbon will be trapped. And I think it's likely to be plants in arid places that need to grow deep roots in search of water, that will benefit most. from rising CO2.
Maybe they'e suffering Macintosh envy, think that all there is to it is to squeeze the electronics into a flashy slimline case and charce an exorbitant price?
Or maybe it's a cynical pre-Xmas launch and the price will be halved within a couple of months?
Silly price aside, it looks rather nice.
I was wondering, why on earth don't people use a credit card instead? And why does *anyone* use a debit card?
Obviously, you need to engage brain first. You want one with perks. You set it up to settle monthly, automatically, in full, by direct debit. You choose one that doesn't charge an annual fee, unless the perks are still attractive after the fee is deducted.
Mine is an Egg Money card. The perk is 1% cashback on everything I spend. Mine is historically fee-free, but it would still be a decent deal with a ten-quid annual fee. I also get the legal benefits of the Consumer Credit Act (principally, if a supplier goes bust after I order, It's the card company's loss not mine).
I guess you should avoid credit cards if you aren't enough in control of your finances to be sure of repaying in full every month. But then, shouldn't such a person use cash, checking their bank balance at the ATM before withdrawing, so as not to go into the red and be hit by the bank charges extortion racket? Compared to which, paying 20% APR on a loan for a few weeks is positively painless.
Enzymes (catalytic proteins) are extremely specific to particular electronic configurations of particular elements and molecules. It's not just possible to just substitute one element for another, not even similar ones like Phosphorus and Arsenic. The enzymes and proteins have to be altered so that the substitute becomes useful, rather than a poison. The electronic balance is so finely-tuned that living organisms even select isotopes, rejecting the heavy isotope of Oxygen, Nitrogen, Carbon and especially Hydrogen. The energy levels are very slightly wrong, and enzymes incorporating the heavy isotopes don't always work properly.
Take haemoglobin (the red iron-based oxygen transport in our blood). There are three similar substances found in other living things. Insects use a copper-based molecule. Some marine worms use a cobalt-based one. Chlorophyll, which has a different role, is a rather similar molecule based on Magnesium.
Haemoglobin is the best oxygen transport molecule, and Iron is as abundant as Copper (maybe more so) on dry land. So why haven't insects evolved to utilize Iron for their oxygen transport?
I imagine that the sequence of mutations necessary is unlikely, and the intermediate forms that would be necessary never found a niche where they could survive and evolve, due to competition from un-mutated insects. Or perhaps the intermediate is such a poor oxygen transport, that the mutation is only viable in a tiny marine common ancestor. Also, although copper-based blood works less well than iron-based blood, it's probably less of an issue than body design for insects (with exoskeletons, which rules out lungs, restricts oxygen supply, and places a size limit of a few inches on all insects). In the case of life with lungs, haemoglobin would offer a much greater advantage, making large life-forms, and even large flying life-forms, possible.
Birds, by the way, have a subtly different haemoglobin protein to the mammalian version. Also iron-based, but the big protein framework surrounding the Iron atom is subtly different. That's how some geese can fly at the altitudes where we need a pressurized environment to survive.
There are several routers that are open hardware, on which the manufacturer's firmware can be replaced by Linux (or by a different Linux kernel). Some are designed to be open, some have been cracked by enthusiasts. Linux has supported IPV6 for years. So the cost problem will solve itself just as soon as there is a mass market for IPV6 routers.
Yes, it does have to be over £100 (and under £30K)
As for terms and conditions, they can't do that. The Consumer Credit Act is the law of the land, and any term or condition that attempts to usurp your legal rights is invalid. They sometimes try, on the basis that most of their customers won't know that such a condition isn't worth the paper it is written on. They also try random misinformation, misdirection, losing your letters, and any other trick that they think might make you give up. One has to persist.
I'm thinking interesting things about the consumer credit act at this point. If you *did* pay up, you could then hold the credit card company liable for the consequences of the non-delivery of the unscramble key and the consequential losses. Because they are jointly liable in law (with a blackmailer!), and it's already been proved in court that the CCA joint liability extends overseas.
I think someone suing their bank for the cost of all consequential damages ought at least result in some heavy pressure being exerted to find the culprits!
Coffee is not a patch on any sort of Cola drink for destroying electronics (and teeth)
Most laptops can be dropped down a drain these days, so if you sling yours onto the seat inside your car, make sure you retrieve it before someone opens the door.
Dropping your backup disk (USB, 3.5", aluminium case) a meter or so onto the laptop's keyboard is a good one. Especially if it lands corner down, dead centred on the laptop's hard disk.
A can of propane makes quite a decent kaboom, if it leaks out in a fairly confined space and mixes with air in the right explosive ratio. I remember reading somewhere that dustcarts are actually designed with this mishap in mind, so that the top of the cart gives way rather than its sides, channelling the blast and the garbage-shrapnel upwards (hopefully harmlessly, unless the cart is under a bridge at the time).
Yes, that would be the way to go. Also un-used enriched Uranium fuel rods aren't a serious hazard. Enriched Uranium is only a few times more radioactive than natural Uranium, and Uranium oxide pellets in Zirconium tubes are pretty robust. We'll be OK just as long as a well-used reactor never re-enters, with all its accumulated fission by-products. Dump used fuel rods into the Sun?
I seem to recall that the nuclear reactor can actually be used *as* shielding *for* the crew.
The ship has a pretty much unshielded nuclear reactor and VASIMR drive at one end, with a long pole connecting it to the crew module at the other. Inverse squares deals with the neutron flux from the reactor. The real danger for the crew is a solar storm. In which case, shut down the drive and orient the ship so the reactor is precisely between the crew module and the incoming radiation.
Incidentally the pole can be pretty flimsy by terrestrial standards. It doesn't ever get subjected to as much as a milligee!
If a quantum computer with a large number of quantum bits can be constructed, it means that all cryptography based on a very large number with just two large prime factors becomes useless. This, because a quantum computer can simultaneously try all potential prime factors in a single operation (boggle), and (like magic) reveal the true ones. The impact of this on the internet as we know and use it should be obvious.
Quantum factorization has been demonstrated, with four qubits, revealing that 15 = 5x3 or 14 = 2x7. There are huge practical difficulties for increashing the number of qubits. There *may* also be some unknown physical constraint. If so, it represents new physics beyond what is known today.
Therefore if such a computer cannot be constructed or fails to work as predicted, this may tell us something extemely interesting about physics. I'd bet that this is the case. The alternative - that quantum computers of arbitrary complexity can actually work - is just too wierd for me to countenance.
By the way, if anyone reading this is working on quantum computing and actually makes the 1024-bit breakthrough which I suspect is physically impossible, you must immediately spam all details to everywhere you can think of. Otherwise you have put a billion-quid price tag on your head. That's what it would be worth to a government to keep it secret, or even just to suppress it. You'd be dead hours after they found out, if the secret was not well and truly out of the bag by then.
Maybe they don't think long telephoto lenses are fair? (ie taking photos of people without them knowing about it). Don't forget, making images of any living thing is forbidden in the Koran, so they probably feel much more strongly than we do about photography without the subject's consent.
PS if you think this is totally daft, ask someone who has found a picture of themselves in what they thought was a private place, adorning page 5 (or even page 1) of a tabloid newspaper. Paris doubtless knows all about that.
I didn't take photos, but I was once called in to fix a failed server (one of the old 130W TDP CPUs). Diagnosed a blown PSU, but smelt very odd when I took the cover off. Didn't take long to find the well-cooked mouse that had inserted its neck into the CPU fan. Hope the poor little squeaker died quickly.
Never worked out how it got in. I know that a mouse can squeeze through a 1cm slot, but there wasn't any aperture vaguely that size.
Did you really get there before Knuth published the definitive work on algorithms? Bubble sort (bleugh), Shell sort (OK-ish for a simple code and limits on dataset size), several O(ln2 X) sorts (best, unless you forget to randomize first and get unlucky, or get thermodynamically unlucky with "randomize").
Pity you didn't publish.
Yes, there should definitely have been mention of Donald Knuth and TeX (and MetaFont). The proof of how good it is, is that it's still in use today. AFAIK its ability for typesetting mathematics remains unsurpassed. And it was perfectly usable on an Intel 80286 PC with 1Mb of RAM.
However, it's not quite what many would call desktop publishing. It's better for producing books and theses, than newspapers and flyers. It's better for thoughtful authors, than for journalists facing a deadline measured in minutes.
Today, it has become the core of many programs which make the learning curve less steep (especially for non-programmers). For example LyX is now fairly mature. It's freeware. It's probably best described as a WYSIWYM program - What You See Is What You Mean - in other words you instantly see a graphical representation of your markup and your text, but it gets re-rendered (or reflowed) when you click to create and view PS or PDF.
The world's interesting real spoken languages are Creoles. These are languages that were invented out of necessity, when history jammed two (or more) peoples speaking totally different languages together in the same society. The first generation speak a pidgin mish-mash plus their "own" language. Soon, in the second or third generation, the pidgin is refined into a creole, and both parent languages soon die out. Nobody plans any of this - it just happens.
English is (possibly) the grand-daddy of them all. It's been evolving for nearly a thousand years, since Norman and Saxon communities started to merge. Compare Chaucer to Shakespeare. Development has slowed in more recent centuries, but English is still a fast-moving language compared to some. In particular there's a trend towards jettisoning what fragments remain of formal grammar after the incompatibilities of Norman and Saxon destroyed much of it.
A much more recent arrival is Tok Pisin, the official language of New Guinea, that was once known as Pidgin English. (Tok Pisin = Talk Pidgin = Talk Business). The name of the language reflects its birth, out of the need of local traders to talk business with their colonial masters. But in a country fragmented by hundreds of native languages and dialects, it took deeper root and developed into a fully-fledged language that continues to evolve (and to diverge from its original English roots).
I've known for a long time that a system set up to dual-boot using Grub from the MBR will randomly stop working and need re-GRUBbing from a stand-alone LInux CD or USB. I'd always assumed it was MS borking the MBR because they thought they owned it and didn't check. Or maybe malware.
The way that avoids this (using XP) is BOOTPART http://www.winimage.com/bootpart.htm, and install GRUB into the first sector of the linux partition instead of the MBR. Then you can boot Linux via Windows MBR and BOOT.INI. Some day I'll find out how to do the equivalent with Windows 7 (or has MS made it impossible to boot Linux via the MS boot loader? Wouldn't surprise me).
A rootkit on the hard disk can be detected by a scanner that is not handicapped by operating from within the compromised O/S. One that boots off a CD or DVD for example. Theoretically, a perfect rootkit cannot be detected from inside by any means once the O/S it infects has booted.
Freeware: get one of the LInux rescue kits such as Trinity Rescue Kit or Recovery is Possible. Shut down windows, boot, update the ClamAV definitions off the net, and scan your hard disk. The commercial AV vendors ought to encourage offline scanning, but maybe it presents problems in how they protect their revenue stream.
For anyone who cares about their personal CO2 output, pullovers all round and turning the central heating down to 18C is a good way to proceed. (I'm assuming their home is insulated to the maximum practical extent). Obviously one's partner has to be on the same wavelength. Kids won't even notice, it's always grown-ups that notice chilly extremities first.
And there's a case for spreading this know-how, just as much as the know-how for micropower servers. Trouble is, us geeks are the last people who'd be listened to. Imagine how much energy would be saved if the fashionistas made chunky knitwear de rigeur!
The trouble with the current Atom CPU design is that it doesn't use very much less power when it's idle. Something like 20W busy, 17W idle. And it's not really quite fast enough if you want to use it as your everyday graphics station rather than just as a no-display server.
The Core i3 drawback is that with 73W to dissipate when it's busy you'll need active cooling.
Intel should get as much of the Core's energy usage optimisation into an Atom asap. 20W peak, 5W idle ought to be do-able. Maybe 30-40W peak better, just as long as the thing has thermal throttling done in such a way that the CPU restricts itself to 20W and still offers decent performance once it's heated a passive 20W sink to the maximum acceptable temperature. 5W plus 40W bursts would average under 20W for many usage patterns most of the time.
(Why do I like passive heatsinks so much? Silence, reliability, and another minor waste of energy eliminated)
"30W is more than low enough to come into the "electrical noise" or even "cable loss" category of any modern home (£3 a month-ish? £30 a year?)"
Rubbish - if you regard this as research. One person finding out how to save 30W now can become 30MW if a million people do. the same. That's quite possible if the know-how is spread and the technology becomes mainstream.
But before it happens, we need better micro-power servers than a ShevaPlug. Something like a MATX board with all the connectors one would expect to find on it, but an ARM processor.
The energy-wasting villain of the piece is Microsoft, doing everything it can to prevent any Windows-incompatible hardware being made available to the masses.
"What they need is an alternative model whereby you can just buy a DVD set and be done with it until the next major release.... like Windows."
I think that they tried that back in Red hat 7-8-9 days and it wasn't profitable. What can they offer for peanuts that you can't have for free with CentOS or Scientific Linux?
With open source you can add a lot of perceived value and charge a lot to a few (RH), you can add less perceived value and charge less (various other Linux support companies supporting CentOS or Fedora) or you can provide the results of volunteers rebuilding the sources without any guarantees but at zero cost (Centos).
Microsoft can do what it does because Windoze source is secret and no-one else can compete. What it does is more akin to running an illegal drugs "business" than an honest one. The first fix is free (comes with your shiny new PC) but you'll be paying for the rest of your life (locked in to Microsoft's proprietary formats). They've even locked just about every hardware vendor into selling Windoze with every system whether the customer wants it or not ... and worse, they get away with it because "there's no demand" for Linux!
If you want Enterprise, you probably aren't very bothered about the cost. If you don't want that level of support, you'll probably be happy with CentOS 6, presumably due out in a month or two, and pay someone else for the software support if you don't have it in-house.
"*** Desktop users *** running RHEL/Centos have been struggling with ancient application versions, and in some cases (chrome, etc) - no ability to run the application at all due to glibc versions."
That's because it's a server distro where stability and long-term support is valued and the bleeding edge is something to stay well away from. I'd suggest that if you want rup to date desktop, then run Fedora (for Red Hat flavour), or Ubuntu. Red Hat don't ship frequent releases, because they are committed to continue to provide support for several years.
I find swapping to and from between CentOS servers and Fedora desktops very little trouble. Ubuntu rather more painful (because all the management stuff is unfamiliar, rather than because of anything to do with quality). I even put Fedora on a server once, when Red Hat lacked the necessary hardware support. It was easy to migrate it to Centos once the Red Hat support caught up. However, wouldn't recommend this where 24x7x365 is required!
The government should have an independant agency that estimates the time and cost of completing every form that the government creates. Further, there should be an appeals procedure through which a business can claim compensation if it can show the estimate to be grossly inaccurate.
The governmeent and its civil servants will then be unable to claim that they don't know the cost of their red tape, and cost-benefit analysis will be possible. It'll still be their benefit, our cost ... but at least there will be a reasonable factual basis for argument.
You're right that the desktop of the near will be better equipped with an 80Gb SSD not a 250Gb HD. Hard disk manufacturers can see this coming. If the hard disk is to have any future, it's at the high-capacity end of the market.
As for who needs (say) 100Tb drives ... if they were on the market today at £100 each, we'd be buying. Give our researchers any amount of CPU power and storage, and they'll fill it to capacity within a few weeks. Molecular modelling, if you were wondering. Others I know something about the IT needs: particle physicists, astronomers, DNA, Oil prospecting. Doubtless lots of other things that I've never found out about.
Sensible uses I can think of:
Batch-duplicating USB data sticks.
Uploading data from multiple devices, for example portable standalone barcode scanners
Testing software RAID with 16-disk configurations (where the disks are probably memory sticks)
Or even just as a charging station for umpteen USB-chargeable devices. I've got an 8-port USB hub doing just this. The hub has never been connected to a computer, just to its mains "brick".
Does seem a bit pricey, though. One can get three six-port hubs for considerably less.
Not a stunted throwback nor revolutionary - just a boring cheap notebook for use on-the-move.
Many of our staff have a netbook for giving presentations and network browsing away from home. They're adequate. They're cheap enough to throw in a case or bag without worrying too much about breaking it or losing it, and they are less heavy than a standard notebook.
The remarkably simple formula for success: just about fast enough, low cost, light weight and decent battery life.
Sales dropping? I'm not surprised, everyone who wants one has got one. Also they aren't really adequate for Windows 7 but you can't get XP unless you retro-install it on a corporate license. Intel will probably ship an Atom fast enough for "7" soon and then the market might pick up a bit.
Someone has to be *on the plane* to register the phone with the clone. So he's still a suicide bomber, albeit with a remote-control bomb in hold baggage now under someone else's control. I thought the big worry was a phone-bomb loaded as *freight*, and a non-suicide-bomber able to detonate it from the ground. (Possibly also GPS-track it).
You could also make my plane-cell base station de-register any phone as soon as it is turned off. Don't know how phones work well enough to say if a phone and its clone both on at the same time could be detected. I'd hope so, if only as some sort of unusual protocol error which would be reason enough to immediately deregister both the phone and its clone.
And as someone suggested above, phone-suppressors in the hold, faraday-shielded from the cabin by the metal floor, sounds like such a good idea they ought ti implement it immediately. If cabin floors are not solid metal, then they should henceforward be designed with a metal foil screen (which could be retrofitted during maintenance).
It ought to be possible to design an airliner cell, which doesn't connect to any phone that hasn't been explicitly registered. Registration could be self-service. After the flight takes off, a passenger wanting to use his or her phone turns it on and takes to the registration point. This would be a metal box (faraday cage, radio screen) which reads the phone's ID from inside, once the box's door is closed. Once registered, the phone would work normally from any seat in the cabin, but (automatically) only until the plane next landed.
So any passenger on the plane could use their phone, but a phone in the luggage compartment would be unable to register itself even it it had been cunningly reprogrammed to try. Wouldn't stop suicide bombers, but would defeat the non-suicide variety (or at least force them to use less precise triggering devices such as timers or air pressure switches).
Wi-fi notebooks, effectvely ditto.
""It takes that long for a desktop operating environment to get so long in the tooth that no matter what you do you can't keep it going any longer. And we'll keep it that long because the costs of getting a working desktop in the first place are so absurdly high that doing anything else is completely irresponsible,"
What he means, is it takes that long for Microsoft to launch the replacement, to issue the final service pack, to issue a number of security patches some of which appear designed to degrade your systems' performance, to cease all support, and finally for someone to "discover" a still-unfixed and now unfixable day-zero bug.
If it were open source you could keep it running for as long as you wanted to.
If the files it processed were all fully open standards then there really wouldn't be an issue about whether you used a decades-old desktop or the very latest one. It's only Microsoft that hardwires the file format to the processing program to the desktop O/S, for their own commercial gain and everyone else's pain.
(Not sure if it's relevant, but I recently booted Win 98 SE in a VM under Linux. Took all of two seconds to boot and was a vastly more responsive environment than Win 7 native on the same hardware.)
Well, clearly both the USA and the Chinese have learned from the back-doors which Microsoft once shipped (and for all we know still ship? ) with Windoze.
The right non-protectionist answer would be to insist on open source, complete with everything necessary to compile it from the source, as shipped or with patches. (And of course, manufacturer patches to be supplied as source). As long as someone actually vets the source and rebuilds the binaries, there's far less likelyhood of anything nasty lurking therein.
Which just leaves the problem that these days it's quite feasible to implement secret backdoors in the hardware itself.
Probably not the only reason. They'd have to have good ores as well.
Simple economics: if someone paid x can mine quantity y in one shift, and someone paid 10x can mine 11y in one shift, then the men earning 10x can put the other mine out of business. (All other factors being equal of course). Higher productivity relates strongly to ore richness and to how much other rock has to be removed to get at the ore, although high-quality mechanisation and automation can also be significant.. For bulk materials like iron or copper ore, transport costs are very significant, but for small amounts of valuable product, they're not.
Just in case, does "scrap tin" originating from the DRC have a different isotopic fingerprint of its valuable impurities compared to ores from other parts of the world? A smuggler would find it quite easy to change the Ta:Nb ratio or other elemental ratios, but if the ratios of the isotopes of these two elements fingerprinted the metal, it would be far harder to disguise where it came from.
I'm guessing not - with the exceptions of materials originating in radioactive decay and materials of biological origin, isotopic ratios are pretty close to globaly constant because the chemistry (especially of heavy elements) is not isotope-sensitive.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
