Re: So, unsafe convictions?
That's a good point. Of course, you could only use that excuse for the first offence.
Sorry about the icon...
184 publicly visible posts • joined 25 Oct 2011
I wouldn't go for an iPhone because it doesn't really invite me to take it apart and see how it works, play with it, and use it for something completely different. Of course, I'd never steal anything (1), but I don't have the inclination to start with iPhones.
Do you think the thief took the 5S just to spite the fanbois? That would be another emotion for which I have no sympathy: I wish Apple users all the best, hopefully as they enjoy shiny new technology which largely works as expected.
Sorry: I sympathise with the poor chap who had his phone stolen. We have to live with finite risk in all areas, and it's not pleasant when we take a hit.
(1) What, *never* ?! Well, hardly ever..
The lab technician at Cambridge told me that Dolby was "thick as two short planks", but was a success because he surrounded himself with clever people. I think that is a bit harsh, but what he had observed was mediocre marks for write-ups of practicals, and he generally predicted people's exam based on such.
So Dolby must have been a visionary to carry it all forward, and the Dolby system must have been good enough to impress somebody (yes, I know you can pick that one apart).
Personally, I turned off Dolby on playback and turned down the treble a bit if required. I did experiment with a cassette recorder with Dolby C, and I used fairly good quality tapes (TDK SA), but it was a profound relief when digital recording for the masses arrived.
There's also several Philip K Dick stories to refer to:-
The Defenders - robots on the surface, pretending that the war is still raging, keep humans underground so that they do not engage in actual war
Second Variety - what we might now call "drones" are built to self-replicate and kill enemy soldiers, but develop so effectively that they kill both sides, and are so deceptive that they manage to find out where the last remnant of mankind is hiding on the moon
Probably more that I have forgotten...
Because I know the passwords are there, I have fallen into the error of thinking that everyone else knows too. If you, experienced as you are, did not know, then that helps me to see that it is more serious. But, of course, you probably treat an unlocked screen as a liability, so you are safe by other means. Therefore I recommend educating the masses to be much, much more wary.
C.f. the GP in the 1980's who thought that sending patient details down a phone line with a modem was secure because it was digital. He never thought that because the intended recipient could read it then so could any eavesdropper!
I have many books, tapes and DVDs that I have used once and then kept for sentimental reasons: fine. On the other hand, seeing that I re-use these items so rarely, perhaps I would be better served by cheap single-use media. Make it easy to pay, easy to obtain, easy to use, plays on anything, no DRM, cheap enough to put TPB out of business. That would be a disruptive business model!
And may I hope that the artists, publishers and suppliers would all get their cut?
If storage becomes cheap enough and convenient enough that old formats can be amassed and traded by those not associated with the creators, rather than invoking copyright to limit pirating, let the newer formats, which inevitably use up the bandwidth available for high-res content, still be the most desirable media, not precluded by price or DRM.
According to this model, we would want copyright limited to say 5 years or even less, so that, when the legitimate publishers etc. find that they cannot make money out of older works, we can still obtain anything from the amassed old formats.
All we need now is for people to behave sensibly and fairly. Sounds like a promising Dickens novel. Sigh.
everyone is competing for the most repulsive idea imaginable Quite.
It makes me consider rejecting any food I didn't process myself. Sorry: that could have been worded better.
But the issue is overriding the human instinct where we pick out the bits we don't like the look of. I expect that there is a lot of self preservation in this instinct, but it's not 100% reliable (examples, anyone?), and if we avoid Bad Science we should be able to improve our health.
So now it all turns into trusting scientists, politicians and the food industry, because I haven't the stomach for butchering, or biology, or, on a bad day, even cutting out the squidgy bits from vegetables.
I remember when our daughter was very young we had only just managed to coax her to start her dinner when our son said "She would have been all right, but she saw one of the carrots moving".
Paris, because she knows about squidgy bits versus pleasure.
The odds of a direct hit are perhaps not very high, and they'd probably build a moon base where very few impacts have happened, etc., but I agree that I would not relish the prospect of steeling my nerve even against a low probability.
Also, when the narrator said that the middle of March would be a good time to stay inside, I think mention should have been made of the specification of the bunker that can stand a direct hit like this: "inside" doesn't really cover it.
Never take information as sacrosanct which is verifiable from only one source. Quite. (1)
So now everyone believes that US spooks merely use commonplace techniques and that they congratulate themselves with the term "hack" when they RTFM? Is that what NSA wanted?
Which means that I can breathe a sigh of relief and put my correct DOB on Farcebork after all: or is that what they actually wanted?
How can I be so sure of my so-called scepticism?
(1) I'm sure I heard this somewhere else before, but don't just take my word for it.
Now lets move into the IBM universe. Quite.
The USENIX presentation on SD codes at https://www.usenix.org/conference/fast13/sd-codes-erasure-codes-designed-how-storage-systems-really-fail is very helpful (+1 for that post!) talking about erasure codes, but also lets us know where they are coming from: they really don't like using more disks than is theoretically needed to protect against their chosen failure mechanism.
I can see that disks are cheap, and I agree that recovering from a failure should not detract from normal operation more than is acceptable. Discussion then follows the "unfortunately .. fortunately .. unfortunately there was a pitchfork in the haystack ..." route, where people pay their money and make their choice. So I observe:
- some systems need higher spec disks which cost more
- more disks means more power dissipation, and we are seeing customers realising that they need to look for lower dissipation equipment
- you'll get much better performance from a hardware implementation of error correction, but it costs money, and it is still not likely to run so fast that the rest of the system does not notice it
Start with the idea that adding a parity bit on the end enables you to detect one bit in error, but not two bits in error. Now trust me that adding more "parity" bits, each calculated using XOR of a subset of the data bits, can not only detect more errors but also provide enough information to work out the original data if there were few enough errors. For a trivial example, sending one bit of data and two parity bits, each of which is merely a copy of the data, allow you to correct up to one error in each set of three bits. But that is a very poor ratio of data to parity, so a lot of work has been done to find "codes", which are described by the "subsets of bits to XOR" mentioned earlier, which offer a much better data to parity ratio.
The idea is to apply this to disks by taking a block of data that we want to write to disk, add parity bits to it, and share out the data and parity bits across the RAIDed disks. If one disk becomes unreliable, when we read back a set of data and parity bits we can reconstruct the data, but only if there are not too many errors. So research is done to find "codes" that work with a specific number of disks and can cope with one disk failure. Or two disks, if there are enough parity bits, and so on.
Mr Blaum is trying for sufficient parity bits to cope with two failed disks, which is what RAID 6 claims. He has only presented one aspect of the codes he considers, called the parity check matrix. Whereas this matrix does indeed imply multiplying the data+parity bits by said matrix, there are always methods of avoiding doing exactly that but getting the required answer. The point is that the parity check matrix is comparatively easy to analyse.
If you want to understand the detail, start by reading about Hamming codes.
Also, read about "finite fields". GF(2) means Galois Field order 2, which is a very posh way of saying "binary, and use XOR to add and AND to multiply". GF(anything else) is a more complicated arrangement where we redefine ADD and MULTIPLY to behave themselves, even though the numbers don't behave quite like familiar integers and certainly don't look like them. See e.g. Reed-Soloman codes.
That's enough for today...
HTH
1. There was an experiment set up on the mezzanine floor of Cambridge University Engineering Department which had pitch flowing down an inclined channel (gradient of about 1 in 3?). I saw it in the early eighties, and it had flowed a few inches. Does anyone know when it started?
2. I heard reports that the Metallurgy Department at Cambridge had its entrance foyer redecorated, but the architect specified a lead mural, which began to flow. The dept. was quite annoyed to have this disaster attributed them. Or is this all just a myth?
What a very interesting chain of events.
Personally, I would have no temptation to look at other search engines if Google did everything I wanted:
Disable Google Instant Results
yes: I can do that
Specify search term meta data
yes: "mandatory_term" -exclude_this_term site:only.look.on.this.site,
no: "SpecifyCaseButNotMandatory"
Speed up searching by omitting things I didn't search for
That would include ads, so I do not expect satisfaction!
Show all hits regardless of proprietary considerations
All a matter of trust. I will be as loyal as a puppy until the day I can prove they have betrayed me.
After all, I switched from altavista for two reasons
1. Google seemed to be quick, and found many relevant hits
2. Altavista promised me an email account for life viz a.n.other@altavista.net and then changed it to a.n.other@cheerful.com: I dare you to tell me you wouldn't be embarrassed to use it!
So why wouldn't Google wipe out the competition simply trying to be the best?
least common denominator
Whereas I sympathise that LCD is poetically correct, I feel the need to point out on a techie forum that HCF is mathematically what you probably had in mind. Using HCF would force people to think about what you wrote, but surely that's the idea here.
Given Windows is a Target for Miscreants and as a result Third Parties deliver stuff to prevent Borking then rather than Mistakenly Borking the Computah and Locking it into Multiple Reboot Syndrome themselves can't the Third Parties be given or create some 'Whoops Space' during Reboot that says "Perhaps We Have Borked Your 'Putah'. Click Yes to Go Back to the Time Before We Fucked it For You'.
That over-arching program would need testing too, which would be even more tricky than testing the virus update program.
(Sorry if I've made a serious comment about a joke, in which case perhaps both can stand.)
I can hardly think straight about the misuse of "unlimited", "infinity", "ultimate", etc due to the mind-fogging anger (hence icon). Perhaps I can take refuge in satire.
Like when schools decided that "fail" was a dirty word, because it might discourage pupils. I formulated a new set of driving test results that specifically avoided the dreaded "fail" word:
a) You have passed your driving test with Pass Level "Driver" = pass
b) You have passed your driving test with Pass Level "Passenger" = fail
c) You have passed your driving test with Pass Level "Pedestrian" ;->
Let's try the same for broadband:
* Unlimited = Subject to limitations
** Truly Unlimited = All you can eat (assuming you only eat fluffy cats)
*** Ultimate Truly Unlimited = <shrug> You only get what you pay for </shrug>
Quite. It is a miserable choice between the policy originally complained about and the unacceptable replacement. I can see why the activist suggests voting for the former, because it is the stick that should hurt FB most when they use it. I can hardly bring myself to vote either way.
What the slaves need is a Get Out of Gaol Free card: how about an app. that backs up all your FB postings etc. with a non-proprietary standard (sorry for the pleonasm) data format? As with all backup methods, this must be as simple as possible so that people who have only just about learnt enough to use FB can "vote with their feet".
Eh, let's go an put t' breeze up t' Saxons...
Well, it is one way to undermine the military capability of other nations. Too cynical?
(Icon chosen for its non-reliance on high tech, but, hey, technology is always going to be blamed because in each age it is simply "what people are trying to use". I can just imagine someone once muttering about the new flint arrowheads not staying on the arrow shafts properly.)
I thought ISO9000 certification was supposed to avoid the use of counterfeit components, so where is this going wrong? (i.e. why are so many counterfeit devices being used)
I see from http://en.wikipedia.org/wiki/ISO_9000 that in 2009 China far outstripped any other country for number of ISO 9000 certificates held: are many of those suppliers not keeping their promises? But surely their certificates would get revoked? (Tell me if I need a reality check.)
Otherwise, I can see that people wanting to keep down costs would choose non-certified suppliers. This would apply to DELL and many others, but I can't see the US military cutting corners that way. (Again, tell me if I'm wrong.)
I had thought buying the DVD would be a reliable way to avoid adverts, but I find all the trailers, animated logos, and exhortations to buy rather than steal, take so long that I put the DVD on with the sound down and go off to make a cup of tea. I have a sickening prescience of such extras developing the need for me to nudge them out of an endless loop (1), or adverts being reinserted in the main content.
Is it my fault because I buy cheap DVDs? Tell me there are up-market DVDs that have less intrusion. I somehow doubt it.
Ad skipping could translate to copying the DVD's main content. Naturally I resent the cost of a DVD-R and the time taken. No doubt that is already illegal (in the UK, but not with USA's fair use policy?), and I now predict measures taken by the DVD manufacturers to make this awkward.
(1) Sorry: perhaps I should not have put that idea in writing
I remember GPS being very much to do with US military, and that only they could make use of the more accurate position information: everyone else had to make do with the less accurate version, which was dithered I believe. About the time people started making headway with averaging techniques to get more accuracy out of the low accuracy information, the US military removed the encryption on the high accuracy information.
How do I reconcile a) assumed military reliance on GPS with b) GPS's painfully low jamming margin? They must have sufficient diversity planned. I'd expect training with sextants and morse code and all sorts of other low(er)-tech options, but perhaps I am unreasonably hopeful.
People not having a backup plan seems risky, a bit like not having backups of essential files ;-)
BTW, people might like to look into any resurgence of LORAN. Whereas many LORAN transmitters have been turned off, I predict they will be reinstated. One great thing in it's favour is that it takes a lot of effort to jam it!
I wonder if the CCL scheme (http://www.ccli.co.uk/) could be adapted from Church Use to Private Use. Currently, a church pays for a yearly licence, and a good proportion of this is passed on to the copyright holders according to the periodical reports from churches saying what material has been used.
1. Looking at the price list for licences for a whole church to copy and perform copyright material, I'm sure people would want lower prices per person per year. Would people pay something comparable to the Television Licence fee, or would they only accept something lower?
2. I see that all the copyright holders need to subscribe to the scheme, and that will take a lot of negotiating. Greedy copyright holders might easily spoil things, but it may be a case of "join in or miss out". (Yes, I know that implies people will infringe copyright: I'm trying hard to be realistic ;-) )
3. Perhaps it is too much to ask people to report every song, film and book they have heard, seen and read each month (the CCL model). Perhaps people would report each download? Perhaps the downloads could be logged by the suppliers? I can see problems with all of these.
OK, I'm definitely not presenting something workable yet, but we need something better than the government trying to prop-up the present system.
Thanks for reading this far. I know it's my own fault for posting four days after a hot article ;-)
Mine's the one with last year's newspapers in...
They said he "was in the process of being taken to a police station"? Does this take so long that they cannot be sure it is completed?
Mine's the one with the ... oh, no: that one doesn't go with my ... er ... on the other hand, it's a bit nippy out there ... yes, I'm coming! I'm coming! Be quiet and let me think! ... please take these cuffs off so that I can change my coat ... now take the cuffs off the coat stand ... you're not going to arrest me wearing that?! I can't be arrested wearing the same coat as you!