Re: Error correction isn't good enough nowadays.
So if you really don't ever want to lose data, you have to write and read the entire medium before you use it for real.
Amen to that.
2782 posts • joined 10 Jun 2009
So if you really don't ever want to lose data, you have to write and read the entire medium before you use it for real.
Amen to that.
HDD systems typically fail from mechanical failures but the underlying data is maintained and you can usually get someone to haul the data off the platters for enough money.
If "spend enough money" is not an option, always give ddrescue a chance! In my experience many drives fail soft: they report so many errors to an operating system that they look dead, but a utility that tries and tries over and over again in an intelligent way (as ddrescue does) will eventually retrieve all the data, or all but a very small fraction of the data.
Of course, you need backup, for the times that the drive does turn into a brick, just like that.
Also always keep an eye on the drives' SMART counters. In my experience many failures are flagged a long time in advance by an exponentially increasing number of reallocated blocks. I pre-emptively replace such drives, long before the error count hits the SMART failure threshold. It's the rate of increase that's the give-away, not the absolute number. I have one drive with a few hundred reallocated blocks from the first time it was written to, but that number hasn't increased by as much as one since then.
This is why you put extra checksums in things that really matter.
ZFS, for example, does end-to-end checksumming, so it'll spot when a disk drive has corrupted data, and recover using redundantly recorded data where that is available (always, for metadata). Because the ZFS checksums are done at the CPU, it'll also spot errors between the disk drives and the CPU ( a failing disk drive controller, a flaky memory controller).
If your filestore can't be relied upon to protect you, store SHA256 sums of the data in your files. You can't correct using an SHA256, but the chances of a data error getting past it is truly infinitessimal.
Mercedes briefly marketed a car that you unlocked with your fingers (in SA). That was until carjackers started amputating the driver's finger when they robbed him of his car ....
I am never going to allow a part of my body to be a key to an object of value!
I guess there may be some ingnoramuses out there that think the best place to find information is by guessing domain names. (Good luck to them, they need all they can get.) The rest of us type the key words into Google etc. and in instants find (for example) http://heylady.net/2010/08/06/why-i-hate-amazon-and-will-never-ever-ever-buy-from-them-again/
Except for one thing, the internet could easily dispense with names and go back to numeric addresses. That one thing is the extra level of indirection, the advantages of which I don't need to explain to technical readers. Heck, if domain names were just a dressed-up integer incremented by seven for each new one, domain-mistype-squatting would become impossible as a pleasant side-effect. Tinyurl.com proves (by existing) that many people actually prefer to use short if meaningless names.
Why on earth does amazon think a .amazon domain is even worth arguing about?
I type "amaz" into my browser and it auto-completes. (Like "ther" for this site, "goo" for google, etc.) And if I don't know a web address, "goo" will find it for me pdq.
And want to do SERIOUS serious work - get a desktop. (25 inch screen, or maybe two, proper full-depth keyboard).
There's nothing stopping Intel from taking out an ARM License. Their process technology would mean that Intel-fabbed Arm would be the best of class.
However, they'd regard that as losing because they'd have to split the profit with ARM. They'd rather keep 100% of the profit, if they can.
It's a VHS / Betamax battle on a far greater scale. I don't care to pick a winner at present.
No, they'd just give police cars a new number-plate on a weekly basis. Or give the police permission to clone someone else's plates, like they've been caught cloning dead babies' identities.
What's really stupid is the assumption that criminals will obey the rules. In this case, that they'll display the valid number plate for the car in which they are travelling.
Those with something to hide will clone a plate from another car of the same make, model and colour. Or they'll register the car to a fake identity at a fake address. Or they'll buy a hire-car company so that they can travel in a "borrowed" hire-car and create a false paper trail after the event should they need to.
So mining the data won't catch anyone that we really want caught, unless they're stupid, in which case it'll just be an "evolutionary" pressure to breed smarter criminals. Yet the data is still stored, waiting for a malign individual to use it for criminal purposes, or for a malign government to use it for genocide.
Plan B. Stop all intrusive advertizing. Work with Google so if I want to find out about your product, I can. Work on your product, so happy customers will recommend you to their friends. In particular, make sure that your post-sales sustomer support is A1. Nothing makes me more likely to buy than hearing from a trusted third party that when something went wrong, it was put right with an absolute minimum of hassle!
My philosophy is always to be a buyer, never to be a sellee. Any attempt to pressurize me into buying just annoys me. Charities that employ chuggers get written out of my will, if they were ever mentioned. Spam of any sort gets your organisation added to my buy-last list. And so on. You ought to be happy I can use Adblock-plus. If I had to mentally filter those adverts, a lot more of you would be on my mental do-not-touch-with-a-bargepole list!
I can think of an organisation that espouses most if not all of the above. It's called John Lewis. It's rather successful.
""anti-business value system". I rather think he means open-source. If Adblock-plus didn't exist, I'd have to write it. If Mozilla didn't support plug-ins, I'd have to fork it.
If someone pasted adverts on your garden wall, you'd be right to be annoyed and the fly-poster would be breaking the law. Why is pasting adverts all over my screen any different? (Apart from some of them being malware-insertion attempts ... akin to pasting with toxin-laced glue? )
Once, someone wrote an app to sign up a spammer's home address to every source of physical junk snail-mail the algorithm could find. About a hundredweight per day! Not sure about the legalities, but burying the bastard in his own effluent is a lovely thought.
Light isn't really so fast. Old electronic engineering approximation is a foot per nanosecond (or 30cm). So for a 100GHz clock, that's 0.3mm per clock.
In practice with things like this you don't normally measure the exact arrival time of a pulse, you measure the phase of a modulation of a carrier wave. As someone noted above, this scheme has a lot in common with GPS and might be a very interesting way to observe and test general relativity.
BTW is Mars the best place for such an experiment? I would have thought that the (cold) dark side of Mercury might be more useful, because it's deeper in the Sun's gravity well and moving a lot faster. Maybe Mercury next?
The other basic safety precaution that must have been lacking, was a mains supply wired though an RCCB!
Suggestion to the UK gomernment: that they scrap all the absurd PAT testing regulations for anything other than appliances that are ported on a frequent basis (vacuum cleaners etc). IF (and only if) all mains outlets in the area are wired through RCCBs. Or even make them compulsory, in exchange for scrapping the vastly more wasteful business of PAT testing every PC, printer, wall-wart PSU, charger, mains cable ....
I'm sure someone in Microsoft has Windows activation statistics. Although it does appear as if the top-level management is refusing to look at them.
In days of yore, the couriers drew straws to decide who would bring the latest bad news to the attention of the tyrant. It was a dangerous job. In those days the messenger might well be literally shot or otherwise executed. The worst that can happen these days is rather less, but losing your job for being accurate about the Emperor's new clothes is still a possibility under the worst sort of management. Safer to stay quiet until asked?
(Lu-Tze quote about leaders, which applies equally to managers. "The second-best leader is respected, and the third-best is feared. The worst is hated. When the best leader's job is done, the people say 'we did it ourselves' " ).
There's a good chance that if YOU wrote it, you can run it under WINE on Linux. WINE has improved of late. Certainly worth a test. The problems seem to arise with things written by MS, or by big software houses with privileged access to interfaces that MS doesn't publish.
The question should be "how is content going to be created ...."
That's things as humble as pages of text, web pages, spreadsheets, entering the data to a 'base.. You don't do any of those things well with a tablet. If your employer insists, you'll soon be visiting your doctor with RSI and hiring a lawyer.
Putting an interface optimised for tablets (I'm being charitable) on desktop computers cannot but reduce productivity. Then there's the retraining costs. Why are they surprised it's not selling well, and that most of the places it's "selling" to are "downgrading" as soon as they unpack the hardware?
And the thing to add to that, is that as a person experienced in using both Windows XP/7 and Linux, I can move quite easily and not unhappily to an Apple iMac. It's about as easy as moving from one conventionally operated car to another. A bit of initial irritation with the minor controls, but nothing completely infuriating and no need for a formal re-education.
That's the board's job. Or the shareholders'. I doubt they'll do the deed until it's too late to make any difference. They're all PHBs like him (apart from the PH, anyway). I'd guess about a year after they've ceased supporting Windows 7, i.e. two years too late.
IBM, if you still want revenge for OS/2, your time to move is fast approaching.
Or "the names and the faces change, but the assholes stay the same".
If it was Win 8, I'd agree. The right Linux distro and UI is quite sprightly on a truly ancient PC, so I'd expect it to run happily on a Chromebook. ISTRR Linus uses a Linux'ed Chromebook!
Netbooks also weren't helped by the fact that we had to stick with the same 1024x600 resolution and 1GB RAM spec for years.
That's part of how Microsoft killed them. It had to have that spec, or Microsoft wouldn't allow it to be sold with Windows XP (after XP stopped being generally available). I think SSDs were also banned by MS.
Windows 7 was too much of a resource hog to run on less than 4Gb, and Atom CPUs were architecturally limited to 2Gb. So there was a period during which a Windows 7 Netbook was impossible - low-power i3 CPUs were too expensive. By the time the tech advanced at the low end (courtesy of AMD), Windows 8 was about to become the only retail chioce.
I once (for fun) installed Win 7 on an Eee PC upgraded to 2Gb RAM and an SSD. It was usable, if a bit "sticky". Biggest problem was that the Win 7 UI really wants 1366x768 minimum.
Netbooks sold like hot cakes before Microsoft found ways to kill the format, because they were cheap enough and light enough to cart around without worrying yourself about theft or damage, and good enough to do most of what you wanted to do on the move.
Microsoft killed the Netbook but the market niche never went away. Now the Chromebook is filling the niche. Quelle surprise. Another Microsoft own goal.
BTW you can get cheap Android tablets for the same price and some of them are doubtless occupying the same niche. But for many purposes, you really do want a keyboard!
Flea-brained gadget? Pea-brained gadget? Any connection?
So is it not actually true that the plague bacillus can survive many centuries to infect someone who digs up a plague-pit? (Bubonic / Pneumonic Plague is a bacillus, not a virus, though some of the plagues might not really have been Plague. The symptoms also match a killer flu like the 1919 one, so far as one can tell from such contemporary medical description as has survived).
I had read that plague graves were best undisturbed for milennia. That should someone accidentally dig one up, they should immediately re-bury it, cease work, keep a VERY close watch on their health for the next couple of weeks, and change the building plans.
After you've exhausted the TLAs move on to the ETLAs and the GBFLAs.
Your best upgrade is LibreOffice, before anyone retrains themselves to use ribbons and tiles and all the other crap that Microsoft has stuffed into the releases after 2003.
Openoffice / Libreoffice? It's less of a fork than Office 2003 to Office 2007 was / is.
Anyway, now Oracle have got out of the way, there's a chance that the fork will be re-joined over the next year or two.
When there's 3 GPS systems: believe the two that agree?
I would have thought that civilian-grade accuracy was plenty with respect to dropping a nuke near enough to its target. Military precision GPS has more to do with flying a conventional or cruise missile through a single specified door or window!
Someone saw a way to get all their ancient PCs replaced with brand new ones.
My guess: a spherically symmetrical collapse wouldn't.
Two neutron stars or black holes orbiting each other will eventually merge. They lose energy by gravitational radiation ( and probably by drag induced by infalling gas, etc.), and get closer, which increases the radiation, until there should be a realy intense pulse just as the two stars or holes merge. Bigger, I think, than any single-body collapse could accomplish. (Caveat: "obvious" answers in GR, often aren't)
A magnetar's 10^10 tesla field, by contrast, has an energy density of 4.0×10^25 J/m3, with an E/c2 mass density >10^4 times that of lead.
remarkable things happen within a magnetic field of magnetar strength. "X-ray photons readily split in two or merge together. The vacuum itself is polarized, becoming strongly birefringent, like a calcite crystal. Atoms are deformed into long cylinders thinner than the quantum-relativistic de Broglie wavelength of an electron." ... At 10^10 teslas, a hydrogen atom becomes a spindle 200 times narrower than its normal diameter.
The magnetic field of a magnetar would be lethal even at a distance of 1000 km due to the strong magnetic field distorting the electron clouds of the subject's constituent atoms, rendering the chemistry of life impossible
Does anyone know what is the maximum magnetic field strength that could exist in even wierder circumstances? Is it where the engergy density of the field would cause collapse of space-time into a black hole, or is there some other limitation?
so Maxwell's equations have the full symmetry that would result from it:
Feel free to write the equations with full symmetry. Then there's a supplementary equation that you use to generate the more useful but less symmetric forms, because as far as we know magnetic monopoles do not exist.
There's actually a proof that in a universe with spherical topology, at least two magnetic monopoles must exist. It's the same as the one that says you can't get all the hairs on a (perfect, seamless) tennis ball to lie parallel to the surface at once. So either the universe has a toroidal or more complex topology, or magnetic monopoles do exist somewhere out there. (Possibly, outside the observable universe, at which point things start to get a bit philosophical. )
I read a paper about giant waves once. (I found that the maths was nearly as heavy as the seas it described). Basically, once sea waves get very large, they cease to behave in line with simple theory, and start to attract each other. This means that the distribution of size of ocean waves has a long (and very dangerous) tail. Where the ocean is shallower than a good few times the height of the wave, things get even worse.
"Freak wave" seems a reasonable term for something that a mariner will see no more than once or twice in a lifetime of seafaring.
400Hz is also MIL spec. It's used on ships and planes because 400Hz motors can be smaller and lighter. It's not used on utility grids becaujse long-distance transmission losses are prohibitively high. 400Hz pre-switched-mode power supplies were also lighter, and needed less huge electrolytic capacitors.
Did CDC sell a lot of kit to the US Military? I'm guessing that they did.
@AC 6th July 09:39 think about the then-acceptable fuel (in)efficiency of a 1968 gas-guzzler, and consider that one can greatly increase the range of any electric car by adding more batteries. However, that's at the expense of performance (more weight) and efficiency (more energy wasted speeding up and slowing down the extra battery mass).
I suspect that the top speed of 50mph had quite a bit to do with the weight of the batteries needed to get that range. A lot of progress HAS since been made. Today an electric car can be competitive on most fronts except range and capital cost, and perhaps the biggest thing holding e-cars back is the lack of a standardised national recharging infrastructure.
Didn't English evolve from a Germanic language anyway? So this is just it coming full circle :-)
In general, no. English is actually a relatively recent new language (compared to Greek, German or French).
It was formed by the amalgamation of the Germanic language spoken by the English Saxons, with the Norman-French spoken by the 1066 invaders. As the communities merged, a creole (technical linguistic term) developed. To see what was happening, get a copy of Chaucer's "Canterbury Tales" that has the original on one side of the fold and a modern English translation on the other. Chaucer was near the start of the process. The Saxon proto-English used by the peasants, and the French proto-English used by the nobles, are still quite different, but coming together (his pilgrims understood each other without translators). It's a good read, by the way. Then watch a play by Shakespeare (who perfected the unified English Language, or maybe even invented it). You shouldn't need any translation.
The process slowed down after Shakespeare, but hasn't stopped. In particular, the grammars of Norman French and Saxon were incompatible, and English has been and is progressively jettisoning its grammar. It is quite possible English will evolve into a pure placement-positional language over the next few centuries (more like Chinese in structure, than anything else of Indo-European origins). The collision between two languages may also be the reason why English has voraciously assimilated words it needed to plug gaps real or imagined in its own vocabulary, from any source, or by neologistic invention. (Is neologistic a word? Do I care? )
Back to "shitstorm", it's no surprise at all that both parts of the word are of Germanic origin. English has preserved a distinction between "polite" words of French (noble) origins, and "rude" ones of Saxon (peasant) origins, which are synonyms or almost so. (e.g. "tempest" vs "storm", "execrement" vs "shit"). Many languages (including, I'm told, Gaelic and Arabic) don't have any rude words, and one has to employ florid combinations if one wishes to offend. "May the fleas of a thousand camels infest your armpits" and suchlike.
so a Scheisssturm would mean "a really good storm", the opposite of the intention.
I don't know about "opposite". A really good Storm of Shit ... for the observers. Somewhat different for the participants. There's a word for that ... oh yes, Schadenfreude.
Where do German verbs go? It depends whether you're trying to be formal / pompous / incomprehensible or not! Informally, after the subject and before he object, like English. Formally (and almost always so in writing) at the end of the sentence. You can skip ahead with your eyes, but not with your ears.
I too was wondering how a combination of two English words both obviously of common origin to the German, could be called an Anglicism? I'd have expected it to be almost immediately back-translated into German. Or would a native German speaker find that "Shitstorm" trips off the tongue more easily than "Scheissturm"?
I think magnetic storage will hold on at the large data end. When one combines bit-patterned media and heat-assisted recording (BPM and HAMR) one can anticipate drive capacites measured in tens of Terabytes. This is technology that's already working in the lab. I should think that the question at present, is whether anyone will ever want single drives that size? Having Terabyte flash caches up front may well drive demand.
I don't think power consumption is too much of a problem for the forseeable future. An SSD uses something like one watt spread out over maybe 40 square centimeters. (And note, some of that is the controller not the flash chips). A CPU or other highly active logic chip uses maybe 100 watts originating in maybe one square centimeter, but spread out over a heatsink footprint of maybe the same 40 square centimeters. So provided there's good thermal connectivity between stacked chips, one could stack a hundred of them (with a decent heatsink on top) without exceeding the thermal loadings that are commonplace in desktop PC CPUs and high-end graphics cards. (I doubt we'll be seeing 100-deep stacks any time soon, but thermal loading isn't the reason).
Price/Gbyte? Well, it'll be interesting to see just how far and how fast economies of (vast!) scale can drive down the price of manufacturing flash chips.
Presumably Apple buy LEDs from companies in (say) China that aren't paying royalties. If they were buying royalty-paid proucts from (say) Cree, there would be no case.
I'd like to know whether Apple got a cease-and-desist letter some years ago, or not.
But even if not, they're using a patented technology on which royalties have not been paid. I think the way the law works, the aggrieved party sues Apple and Apple can sue their supplier in turn. (Except, the supplier is in China? Well, that's the choice they made).
Do patents apply to personal use?
In the UK, that's a definite no. One of the purposes of the patent system is to place inventions in the public domain, so that everyone else may attempt to create further advances on the technologies so disclosed. To profit thereby you'll have to license the original patent, or await its expiry. Infringement is manufacturing or using a patented device or technology for profit.
In the days before patents existed, inventors kept the methods by which they manufactured things secret, and the method of manufacture often died with its inventor. This was not in the public good.
Not all patents are bad. It's the stretching of patent law into areas for which it was never intended that's bad, along with USA-style law and lawyers. Software should be covered by copyright, not by patent. The imitation of business methiods should be permitted - first-mover advantage should be sufficient reward. And so on.
It's certainly not trolling, if it's a genuine patent on a genuine invention. If Philips, Cree etc. are paying royalties on the patents, that suggests to me that this is the case. (In the case of Cree, that's a royalty on a large percentage of everything they make). These are large companies that are perfectly capable of out-lawyering a university if they feel they have a case (and probably even if they didn't, but I'll assume that they are honorable and/or care about the PR).
And at Halloween? Obvious, but a big yellow cat's eye would be good all the same.
giving the thousands who pass by on the M1 motorway something to ponder: “What the bloody heck is that for?”
I was one of them. Thanks!
CAD pads? I can't swear to how common, but I saw one being used for CAD a couple of days ago. Wacom still has a business making them.
The track-ball was probably better, at least until the modern optical mouse (no moving parts) arrived. You can still buy them. They cost a bit more.
For driving a CAD or artistic-design system, track-pads(*) with pens or wands remain commonly in use. Absolute rather than relative position control. Again, better, but definitely a lot more expensive.
The mouse is definitely the best idea if cost is factored in.
(*) not to be confused with the finger-sensing track-pad on a laptop.