Posts by ibmalone

Egg on North Face: Wikipedia furious after glamp-wear giant swaps article pics for sneaky ad shots – and even brags about it in a video

Re: To play devil's advocate...

Not sure! https://foundation.wikimedia.org/wiki/Terms_of_Use/en doesn't really mention (in a quick search) advertising or promotion specifically, and navigating the policies is a nightmare, https://en.wikipedia.org/wiki/Wikipedia:What_Wikipedia_is_not#Wikipedia_is_not_a_soapbox_or_means_of_promotion might be closest, and though I don't think their approach explicitly violates any of this maybe explicitly adding the text would. Submissions should generally be CC-BY-SA, so the wikians would be free to strip out any such mark if they wanted (and you can understand why they might) so long as the metadata still contained the attribution.

A wikipedia page I occasionally contributed to about a guitarist once had all references to their equipment removed by a drive-by editor as "promotional", so attitudes vary. (Re-introducing the material was utterly uncontroversial among the regular maintainers...)

Re: To play devil's advocate...

Hi AC,

Yes, I suppose you could, provided it mainly demonstrated the subject being described and complied with the rest of the wikipedia policies. Which I'm having difficulty believing it would...

To play devil's advocate...

They didn't exactly deface either, they submitted professional photographs relevant to the article which the wikipedians are happy to use once the logos are removed. Yes, the goal was pretty insidious, but I'm not sure vandalism is the right analogy.

If they'd wanted positive press out of this, they could have made a big noise about having a project to create these pictures for wikipedia, without having to include their branding in the pictures themselves.

Germany mulls giving end-to-end chat app encryption das boot: Law requiring decrypted plain-text is in the works

Where there's a will there's a...

Have they spoken to Huawei?

Uber JUMPs at chance to dump load of electric bikes across Islington

Re: Chinese bikes already here

Lime and Uber are renting electric assisted bikes, Ofo and Mobike are also in London (or were, apparently Ofo are pulling out of London after previously pulling out of Oxford), but rent normal bikes.

Let's make laptops from radium. How's that for planned obsolescence?

Re: Stop it.

Go one better, and levy a disposal tax at the factory gate / port / border.

This tax would be based on the current cost of disassembly and responsible disposal.

Supposedly something along these lines is already done in WEEE, but I suppose it could be extended. (And perhaps made to actually work, for an individual in the UK who doesn't own a car WEEE disposal is pretty non-existent.)

Re: Stop it.

Only if you burn the plastic; oil made into plastic is oil not burned. Of course you may burn other oil to get the energy to run the process, just as you may burn oil to smelt aluminium or in farming and processing cotton, or in recycling.

Re: "the fantasy bliss of climate-change denial"

The Earth is now as warm as it has ever been in the history of human civilization and continuing to get warmer. This is simultaneous with there now being about 7.5 billion people, versus 1 billion in the early 1800s and less than 0.5 billion prior to the 14th century. Further changes in climate will disrupt the balance onto which we're just about hanging, we know from history what happens when there's a sudden scarcity of resources: wars, famine, mass movement of people. There is a well-understood mechanism by which the CO2 and other greenhouse gases that are being emitted can warm the planet, and a rather telling correspondence between their increasing emission due to industrialisation and the warming of the climate.

Given all this, pretending you've been asked to sit shivering naked in a cave and tarring those people who find the situation concerning as primordial slime, in the name of a 'scepticism' that often seems to start from the conclusion it wants to reach, seems a little... hubristic.

Re: Stop it.

The story on bag re-use and environmental impact is apparently complex, for cotton to be worthwhile you have to re-use it for along time (as you have). The most pressing crisis is CO2 really, so it's necessary to keep an eye on that over plastics pollution (the two don't necessarily have the same optimum approaches). E.g. https://ourworldindata.org/faq-on-plastics (just a quick find, not sure about reliability).

we should reward manufacturers who make products for longevity, not short lifespans and gimmicks.

But economic growth!

WikiLeaks boss Assange acted as a foreign spy, Uncle Sam exclaims in fresh rap sheet

Re: I was fine with the first indictment

This pretense that Trump isn't president, the UK doesn't have to leave the EU etc etc it's all just destroying democracy and polarising and entrenching positions. It helps nobody.

Since Brexit has been dragged into this... the UK doesn't have to leave the EU, that's simply not how our parliamentary democracy works, and the reasons why are related to the way the vote was conducted being a terrible idea in the first place. With no plan on how the thing was to be implemented or timetable set out the outcome can only be regarded as a statement of intent. Now, having looked at what the finest minds of this country (!?) can actually deliver and how it stacks up against what was promised, if the democratic intent changes we don't need to leave and then re-enter as if there's some kind of democratic computational queue to be obeyed.

The rush to trigger article 50 by the eurosceptics in the tory party as if they were thieves trying to escape with the loot has actually undermined any effort to get a sensible resolution to the thing. Though the cynic would suggest their personal goals aren't necessarily aligned with the best interests of the country anyway.

Programmers' Question Time: Tiptoe through the tuples

Re: Fab

Eric Robson is spinning in his offline backup long term tape storage archvie.

I'd imagine he's particularly vexed about being stored before returning!

Audience: (applause, cheering)

Though no John Cushnie...

Honey, hive had it with this drone: Couple lived for years with thousands of bees in bedroom wall

Re: A little advice from an amateur beekeeper ...

And now I can't help but read that in his voice Elgar.

Re: Yet another bee story.

I can only think one person needs to brush up on their Latin... guess the veil was a bit heavy.

Not an 'apis crowd!

Yet another bee story.

Just a thinly veiled excuse for people to bombus with puns.

UK Space Agency cracks open its wallet, fishes out a paltry £2m for Brit plans to return to orbit

Maybe my favourite XKCD: https://what-if.xkcd.com/58/ "Space isn't like this:", "Space is like this:".

(Well, that or https://xkcd.com/979/)

Backup your files with CrashPlan! Except this file type. No, not that one either. Try again...

Re: Whatever happened to three copies ?

Those "whinging babies" may well have done all that and still be annoyed (because they're the kind of people who look after their data) that the people they pay for a backup service decided to delete backups without warning. They are also serving a useful purpose by doing (c) Warning others about this provider.

iPhone gyroscopes, of all things, can uniquely ID handsets on anything earlier than iOS 12.2

I'm not really sure why you couldn't use this to uniquely identify any device, calibrated or not, that wasn't adding noise.

They're using the discrete nature of the sensor output to let them work out the device-specific calibration values (integer sensor output to calibrated real numbers). For uncalibrated devices those values will not be unique (I'd guess batch-specific at best?). If you could find a way for the uncalibrated devices to secretly determine their correct calibration then I suppose that might be identifiable, but calibrating something is difficult enough, let alone without co-operation, and the calibration uncertainty is likely to be more than the spread of device characteristics anyway. Attacking calibrated devices you're really reading a set of stored floating point numbers that are the same every time you extract them.

It doesn't matter much either way. If it's stationary then you need enough sensor noise to get the jitter out, test mode on my phone seems to confirm there should be enough on a normal office table, though you still need to wait long enough to get the different combinations that will allow you to estimate the full calibration matrix. Because they're looking at changes between samples moving about is fine too, and they seem to do a trick with looking at the differences between pairs of outputs of similar values. There's not much of the work that is actually concerned with gait. There is a little discussion about bias correction in the javascript API causing a problem, waiting for the phone to be stationary is a suggested strategy for that. However the 100 samples they need is less than 1 second of data.

To crudely summarise, say you've got an accelerometer and it outputs integer values from 0 to 255. You have some calibration that turns that into physical units, say m/s^2. You still only have 256 possible values. Maybe they cover -20 to 20m/s^2, so they're about 0.156m/s^2 apart (-20m/s^2, -19.844m/s^2...). Look at the values you've collected over some set of samples and take the difference between close ones, you'll see +/0.156, +/-0.312 etc. On a calibrated device that 0.312 will not be exact, say it's 0.312445, this value will be different between phones. With the gyroscope you've got a slightly more complex scenario with 6 (I think, guessing it's symmetric) independent calibration values that can be extracted in a similar way. Because the devices are accelerometers the method works equally well stationary (values zero with noise), in steady motion (values zero with noise again) and constant acceleration (constant value with noise), but a bit of changing acceleration, unless incredibly jerky, is not a problem.

According to the researchers' site apple have patched. Not read the paper in detail yet but it's a nice idea, "Fig. 1 (a) presents the raw gyroscope measurements collected by the two devices. From the figure, we can clearly observe the quantization. This is because the outputs of the gyroscope ADC are integers. Taking the difference between two sensor readings directly reveals the gain of the sensor. According to Equation 2, the difference between two measurements, can be calculated as".

Then goes on to:

~ΔA = round(G_0^-1 ΔO)

Where ~ΔA is their estimate for the changes in sensor readout, G_0 is an initial guess of the gain matrix and ΔO are the changes in reported gyroscope output. You can then start a recursive estimation of the actual calibration matrix G.

You may not need to add much fuzz to defeat this version of the attack if it relies on the quantisation of the output data (effectively, measuring the smallest distance between non-identical outputs), though it does take a least squares approach to its repeated fitting, but a more statistical variation of the same may be able to overcome that. Apple's fix apparently follows the authors' suggestion to apply random noise uniformly distributed over the discrete step width, this means the straightforward rounding step that lets you into ~ΔA and from there to estimate the gyroscope calibration G is defeated, but clever use of something like a Kalman filter to get at the device motion might let you start to average out that noise (since it's at the sampling frequency and motion changes are unlikely to be).

Edit: adding uniform noise at the level of the readout steps isn't going to negatively impact accuracy very much (it will a little). However if you can use statistical techniques to beat added noise then it may get into a trade-off for obfuscation versus calibration accuracy.

Pushed around and kicked around, always a lonely boy: Run Huawei, Google Play, turns away, from Huawei... turns away

Even if they stick with AOSP for the time being and provide their own alternatives for play store, email, maps, that will put Huawei in the position of being able to legitimately hoover the data Google currently hoovers. Whether you'd think ironic or counter-productive is the more apt phrase might depend on whether you take into account western governments' demands to build back doors into devices and software.

Believe the earlier poster just made a typo, if you re-read the post as google play not being used for the phones in China it makes more sense. As for that: https://www.techradar.com/uk/news/huawei-suffers-major-android-setback-as-google-pulls-access-to-core-apps-and-services "However, the loss of the Google Play Store – which for users outside of China is the key source of Android apps" is the closest I can quickly find to the play store not being used in China, but reporting on the radio last night mentioned it too, so I think it's the case that Huawei ships without play store there, or ships with an alternative that gets actual use (as opposed to the Samsung store on my last phone...).

Let's check in with our friends in England and, oh good, bloke fined after hiding face from police mug-recog cam

Re: Only in Nany State Britain...

A former martial arts instructor of mine once pointed out you can walk down the street carrying any number of seemingly innocuous but dangerous objects. Apparently this realisation came to him while carrying a vacuum cleaner through town, a retracted telescopic rod makes an excellent improvised club. A sharp edge is certainly more easy for the casual assailant, but plenty of items will make a convenient stabbing weapon for those willing to put the effort in. I'll leave it to the reader's imagination rather than enumerate them. Needless to say, the list of things that will need to be specially licensed rapidly becomes very long. The industrious criminal will also find ways even with small blades: how did people off each other before the invention of metal? Well, the Aztec Macuahuitl is one answer https://en.wikipedia.org/wiki/Macuahuitl Pacific Islanders used similar weapons made by embedding sharp rocks or shells into a frame.

Yet, we put up with these things? Why? The answer is that the items we use in everyday life, including knives, have a utility that far outweighs the small risk that somebody will use them for evil, and that restricting them does not much reduce the options for harming others, but does greatly reduce people's ability to carry out everyday tasks. This is why despite it being possible to run people over with a car we don't place any great restrictions on buying them. For knives, there are uses of knives outside the kitchen, and the few alternatives you've suggested don't even cover the possible things you might want them for in the kitchen, while introducing a whole lot more appliances. (Asides: what is the environmental impact of that and the resulting harm? What's the resultant harm from encouraging people to eat ready-meals more frequently?).

If you're ever lost on the Moon, Ordnance Survey now has you covered for Apollo 11 anniversary

Re: Courtesy of the British tax payer

They're a government-owned company and they sell things, very successfully. Maybe the British tax payer should thank them? https://www.ordnancesurvey.co.uk/about/governance/annual-report.html

Japan's mission to mine Mars' moon is cleared – now they've filled out the right paperwork on alien world contamination

Re: Too much fuss about contamination

Complicated life "only" started about 1 billion years ago (depending what you count as complicated, but plants date from around then), while single celled organisms have been around about 4 billion years. There's a big gap before things kicked off. If Mars is less hospitable then there's a chance there is life there that hasn't developed beyond single cells.

Hi! It looks like you're working on a marketing strategy for a product nowhere near release! Would you like help?

I *redacted* you, but *redacted*

Another remote-code execution hole in top database engine SQLite: How it works, and why not to totally freak out

Re: Stop using C

Thought process: "How does one exploit SQL injection on SQLite, which is most often used for an application's private storage?"

Follows link.

"For Google, at least, the library backs Chrome's WebSQL database API."

Oh, as you were then.

Nvidia, King's College train robot overlords to spot oddities on radiology scans

Re: Old News?

From earlier experience of what this group is working on, you are feeding in images alongside their medical records. Things like radiological annotations are already very regularised, it's not really necessary to teach the machine to converse with you.

One big elephant is selection of images; the imaging you have for a particular patient depend on what was requested on the basis of previous consultation, if you're not careful you are training it that people with wrist x-rays largely have broken wrists.

Re: Curious

Ethical approval, for research involving people you need to get ethical approval from a research ethics committee https://www.hra.nhs.uk/about-us/committees-and-services/res-and-recs/. It's more NHS-speak than edu-speak though. Also see "having ethics for" and "covered by/under ethics" .

Portal to 'HELL' cracks open in street – oh sorry, it's just another pothole

Re: Not any more.

And to fill out the picture: local roads are generally paid for by the council, while motorways and A roads come from central government (e.g. Highways England in England). In London TFL looks after major roads ("red routes") and the London councils do the rest.

'Software delivered to Boeing' now blamed for 737 Max warning fiasco

Re: does this affect more aircraft than the ill fated 737 max?

My understanding from spending too much of my childhood playing flight simulators is that AoA is not the same as artificial horizon. AoA tells you the angle between your nose and airflow (where you think you're going and where you're going). On fighter aircraft that are performing extreme manoeuvres the two may be wildly out of alignment, at a more sedate pace it will largely depend on the lift you're getting. If you're not going that fast then having the nose above the horizon may still only result in level flight.

We regret to inform you the massive asteroid NASA's all excited about probably won't hit Earth

This is the dawning of the age of austerity.

Take your pick: 0/1/* ... but beware – your click could tank an entire edition of a century-old newspaper

This is avoidable; you use the password itself to reversibly scramble the stored key: new password simply re-encrypt the key. The protection on the key is admittedly only as strong as the password, but as the key itself should be random it is difficult to brute-force since the only check is to attempt to decrypt the data with the test key.

I did read the post, maybe 'salt' carries the wrong implication, if you really wanted to be secure you would not use such a function, you would use both the password and the stored key to generate the actual key for decryption. This way there is no explicit password check, only a generated key.

Oddly I had always thought you would salt the encryption key with the password somehow, rather than just do "is this the password?", as the later leaves the keys on the device somewhere, reducing the problem to one of getting hold of them (as in this example).

Surprising absolutely no one at all, Samsung's folding-screen phones knackered within days

Re: Why would a layer you aren't supposed to remove

Stainless steel doesn't flex well, most metals are subject to work-hardening (and finally cracking), a hinge of course is something different (and also subject to wear, but much longer than the lifetime of the average smartphone).

I think spider silk might be another candidate here. Just need to teach them to spin much faster.

Google readies Pixel for the masses, but are the masses ready for Pixel?

Re: I can't say that I'm trusting of Google hardware

Don't have a Pixel, but on android I certainly never find data or bluetooth on when I've turned them off (and I generally do turn them off when not using). The scanning wifi when wifi is off thing is a bit pernicious and the control for it is well hidden.

As for data, many apps will happily communicate in the background with it on, and rather than trusting them not to I'd prefer to have it off, also allows you to better control your use of allowance and increase battery life if they're unable to talk. Similarly for bluetooth I turn it off if not in use as it's one less thing to spend power on. Since the update to Pie I've found I get up to four days on one charge with light use, not quite as good as my last Nokia, but tolerable.

Loose Women woman's IR35 win deals another high-profile blow to UK taxman's grip on rules

Re: Avoidance vs Evasion

I would say the issue is that they use tricks that are not available to individuals and smaller companies. The corporations that do this are not absolved by simply following the rules like good citizens; there is nothing simple about it, they actively seek out the loopholes and have the money to buy the expertise to do so, expertise that is as good as that available to the governments who write the laws. And yes, our desire as consumers to pay less contributes to this competition, they would do it anyway as it would mean bigger profits, but equally Amazon, for example, can sacrifice more profit and run on thinner margins than competitors that don't have the scope to shuffle their profits to countries that will take less of them.

A quick cup of coffee leaves production manager in fits and a cleaner in tears

Excellent, had not seen that before!

Might have to subtitle it with proper units of force though...

Nothing to do with me, but a certain MR scanner in London, a few years ago, inhaled a vacuum cleaner rod down its bore. A "mere" 3T, but they are big and ramping it down and back up is slow and very expensive. Some consultation with the manufacturer was necessary, I think in the end it was decided it could be carefully (but forcefully) hauled out. The worry at that stage is that however much force you are putting on the object to get it out is also being exerted on the magnet coils and their supports, which you definitely do not want to damage in any way.

Much earlier, and with a much more forgiving (and, I'll stress, non-clinical) magnet, I learned that modern copper coins are in fact steel. Just a gentle tug in the pocket (hang on, is it SFTWE already?), but after that I was much less blasé about what metals might or might not be magnetic, better to play it safe instead. (In my defence, we did have a couple of small magnets to test objects, it turns out 'copper' coins are only weakly ferromagnetic.)

For the benefit of other readers... that's "FatallyFlawed" campaign, not a campaign that is fatally flawed.

And yes, ironically, the kids removing them would be improving safety.

Google Fiber experiment ends with Choc Factory paying Louisville $3.8m to clean up its mess

Re: Scorched Earth

The real story though is 'disruptive internet company discovers what 100 years of civil engineering could have told them', with the subtitle 'having a cool name doesn't mean it works'.

So you've 'seen' the black hole. Now for the interesting bit – how all that raw data was stored

Re: More?

An important point I missed (if you're feeling generous you can say 'lensing' covered it): the bright ring is due to accretion, but more fundamentally it's the exterior of the photon sphere https://astronomy.stackexchange.com/questions/30317/m87-black-hole-why-can-we-see-the-blackness/30360#30360. That is to say, we are actually looking at the effect of light orbiting a black hole, basically confirming the Schwarzschild or Kerr equation holds outside the event horizon (there's a bit in one of the papers about testing the Kerr nature of M87).

Re: More?

It's a good question. The data is going to be mined for years I'm sure, to test different theories on black holes and accretion discs. For one thing, just the shape of that halo confirms theories about how accretion discs behave around black holes, before we just had models, now we know they are right (sometimes), it also allows confirmation of the mass of the black hole, again this was previously only known indirectly, so we can be more confident about those methods. One thing that's gone largely unmentioned, they've got images from four days (nights?), showing the structure is stable over that time period. The ring structure confirms a lot of things about gravitational lensing and how light behaves near the event horizon.

The horse's mouth is a good place to start: https://eventhorizontelescope.org/. If you really want to get into the details, the first few papers are at IoP (open access). Papers V and VI are the most interesting for what they can actually tell from these images.

Like, why do they keep referring to it as a "shadow" and not an actual view into the gaping maw of eternity itself..

In the simple sense we can never see into a black hole, because no light can escape. There is the phenomenon of Hawking radiation, but it's produced outside the event horizon. So all you can see (if that's the right word), is nothing, a shadow. There's slightly more to it though, the shadow, the region from which you don't see light, is distorted by lensing effects and is bigger than the horizon, if the black hole was rotating significantly it would be noticeably oblate. Slightly disappointingly, M87 doesn't spin much, but they can at least confirm its shadow looks as expected.

But the simplest bullet point? Ever since they were first hypothesized, people have thought black holes couldn't exist for one reason or another, it's only really in the last three decades they became widely accepted, and now you're looking at one. (Quasars were discovered about 1960, but it was a long time before black holes were accepted as a possible explanation.)

Re: Tape

This is why I was looking at the data rate and talking about multiple tape drives, you can actually get decent bandwidth. Back of the envelope a tape drive has about half the write bandwidth of an HD (is there a longer write latency? anyone care to chip in?), compression can bring it up, but I'm guessing raw observation data is not very compressible. From the description they've had to do work to make sure the data is being distributed across the discs, it's certainly more than one or two could handle (while you could RAID-0 it instead that's unattractive because one missing disc would corrupt all the measurements), but it looks like the resulting bandwidth of the system is more than they need. There's two things determining the size of the disc array, the total capacity needed and the bandwidth needed, the difference with HD is adding capacity automatically adds bandwidth, with tapes you need to increase the number of more expensive drives.

Correction to what I posted previously though, the bandwidth is 8GB/s (64Gb/s from the article) I've just realised why going from 350TB/day to 4.4GB/s is missing a factor of two... The 32 drives in one of those DBE should be able to handle something like 2-3 times that (may well be bottlenecks elsewhere), but it changes the numbers for my back of the envelope for tape as that's based on the minimum number of drives you can get away with, 24 would only just manage on paper, so 32 tape drives should have the bandwidth, but would cost more than the 128 HD for a site. Hadn't accounted for the read at the other end, though you don't need the same bandwidth there. The cost calculation may change for longer observation periods at inaccessible sites; the longer you need to keep the raw data before you can ship it the more storage you need, while your bandwidth requirements stay the same.

By bizarre coincidence, the SKA was one of the projects I applied to do a PhD on, but didn't get that project (maybe just as well, hadn't realised they still haven't built it yet!).

Re: Tape

Tape capacity is catching up again, and will probably pass HD soon as HD are at the limits of what's physically possible while tape spent a long time not being developed much, LTO-8 can take 12TB, though in 2017 they'd have likely had to go with LTO-7 and 6TB, or one of the proprietary formats which can take more. Bandwidth is comparable, though not as good, about 350MB/s (uncompressed, tape figures often quote compressed due to use as backup), vs about twice that for 12TB helium drives. Drives are expensive, but individual tapes are a lot cheaper than really big HDs of the same size.

However I'd guess the need for a drive makes it less attractive for this application where you want to be able to maintain high bandwidth and continuous writing (also there's the wraps thing which I don't know how it plays out in practice, the wraps are written in alternating directions, but not sure if there's a delay at the change-over). Likely cheaper overall to have an array of HDs than an array of tape drives. That 350TB/day is about 4GB/s, so you'd need over 10 tape drives with two changes (three tapes) a day with LTO-8 to keep up (probably more, since you'd only just keep up), while about 30*12TB HDs will provide the capacity and easily have the bandwidth. A 12TB helium drive is about £400, a basic LTO-8 drive is about £2000 (tapes for it are ~£60), so the break-even cost-wise is about 6 tapes to one drive, which obviously has much less bandwidth. (Doesn't account for the extra hardware to host the HDs though, which will start to balance it out once you've got more drives than can connect to a single server.)

Re: Helium filled at high altitude.

Commercially available sealed He drive (and they tend to be the high capacity ones) vs building, maintaining and running a pressurised cabinet at every site. You'd need to get quotes, but I'll bet on the first option being cheaper.

Re: He filled hard disks

I'm not sure that's entirely true, the drives are being moved, so backing them up prior to that would escape any damage caused in transit. Also simultaneously writing to two sets of drives would avoid the fail-on-read problem (and also any fail on write issues, well, not avoid, but do (1-p)^2 on the failure rate). But with this amount of data some of it will be missing or corrupted anyway, so their algorithm for combining it likely has to deal with the missing observations and a corrupt hard-disc in petabytes of data may well just be another source of noise.

Motion detectors: say hello, wave goodbye and… flushhhhhh

The Barbican Centre has an interesting collection of plumbing conveniences from the late modern period. My favourite remain the (original I think) hands-free taps. The place was designed in the 70s and finished in 1982. I would ask those unfamiliar with the conveniences there to guess what late 70s technology these taps employ. Early infra-red? Some kind of ultrasound sensor?

They are operated by foot pedal.

Not only this, but the control mechanism is effectively on/off. As is the way with the era's public plumbing, some of them produce a trickle, some will produce a deluge that misses the sink and sprays you with water. Which is which? You're going to have to get close enough to step on that pedal to find out...

Astronomer slams sexists trying to tear down black hole researcher's rep

Re: Let's hoist one

Minor correction, Here's to Dr Bouman and the rest of the black hole spelunkers.