Posts by JeffyPoooh 4286 publicly visible posts • joined 28 Jun 2013
Order a brand new identical replacement keyboard off eBay for $35, free shipping. Then, once it arrives, it takes just a few minutes with a simple screwdriver to swap it in, right? Budda boom, budda bing, done.
Toss the old one in the junk bin in case new keycaps are ever required.
What could possibly be wrong with this approach?
Oh, yeah. Apple. I forgot. Sorry.
"The [DVLA's] boilerplate response whilst in one area factually correct, seems to miss the point."
pal·ter - mid 16th century (in the sense ‘mumble or babble’): of unknown origin; no corresponding verb is known in any other language.
The word "Paltering" is being resurrected to describe such examples. Now we just need to have paltering made into a criminal offence.
No. Technology, oh yes. Software no; I struggle to avoid it.
I have written software before, years ago. 30-feet long and 'Mic Drop' perfect, walk away. More than once, so not just lucky.
These days it's all libraries and overly complex OSs, so even an individual effort is relying on hundreds of other idiots. It's now nearly impossible to achieve perfection, because modern programmers are all 'Standing on the Shoulders of Morons.'
It's hopeless. I want nothing to do with software. Beyond fiddling with Arduinos because they're so cute and mostly harmless.
PS. If I did work in the automotive industry, it'd presumably be Mercedes since that's what I drive (for almost 20 years now). That wouldn't be so embarrassing in this context, since their security is probably pretty good. They make fewer blunders than other brands.
PS2: People working in tech should know how it all works. Shouldn't sleep until it's understood, in the basics. Can't accept magic.
"...can pick up the signal from keys and copy them to the car, unlocking them and disabling the alarm system."
We need to clear about this. Especially considering system design as impacted by limited battery capacity.
Your car key isn't like a smartphone, needing to be recharged every night. So it cannot be emitting RF signals all the time. The tiny button cell needs to last a year.
Your key fob remote control (older RF technology) would be activated with a button press, and can thus have long range. So it can be captured surreptitiously from across the parking lot. It needs to have basic security like rolling codes (if somebody grabs the code out of the air, it's already stale). My ten year old Mercedes has rolling codes, just like the 1999 model I had 18+ years ago.
Note: The rolling code algorithm needs to be kept secret.
Keyless Car Starting is probably using RFID. That's much shorter range. The hackers need to get within a meter or two of your keys, so their can 'illuminate' it with enough RF to power it up. The system designers should include some handshaking, not just an easily copied serial number.
If the researchers have identified a make and model of vehicle where they're using RFID relying on just the tag's SN, then name names and inform the Insurance industry. Such cars would be immediately recalled and lawsuits would fly.
I detect that these researchers are exaggerating. There's no mention of the fact that many vehicles already have excellent security.
Those that don't should be punished via theft insurance premiums. It'll sort itself out quickly.
Researchers "...recently bought a car and they decided to see how difficult it would be to hack."
Step 1: Buy a car.
Hmmm...
In other words, comes with keys. Access to CAN bus is a given, since the Researchers can unlock the car door and lean inside. The vehicle owner isn't peeking out from behind the bedroom curtains, on the phone with the police.
Granted, some of this might also be applicable to a Rental car, but the detective is likely to check who rented the hacked vehicle the day before the deadly crash. So best to rent without showing any ID.
"Car theft rates....in the UK in the last two years....have risen 20 per cent. A direct causal link is difficult..."
If hacking were playing a significant role in the recent 20% rise, then the increase in thefts would be disproportionately focused on certain makes and models of vehicles, those that have the easiest to exploit weaknesses. Or those being targeted by a gang that bought a certain type of hacking gadget. It should be almost trivial to extract a correlation from the data.
Inferring a direct causal link needs only the Vehicle Theft data, a spreadsheet, and some common sense.
Is one of these lacking?
Ultrasonic Sensors (e.g. 40 kHz) are typically used for low speed Parking Assist / Obstacle Warning. On my Mercedes, they're only activated at 18 kmh or less. Above that speed, they're turned off.
It's because they have fairly short effective range. Ideal for Parking Assist, useless for Automatic Braking.
Ultrasonic would be a very poor choice of technology for Automatic Braking, as it would inherently require something like a 2m diameter dish and sophisticated coding to provide enough system gain and effective range to be even the slightest bit useful for Automatic Braking purposes. Even then, not really.
I'd be surprised if Mazda were so silly as to try to use Ultrasonics to provide control of Automatic Braking functions.
This isn't the only example in this column of Cirlig seemingly being somewhat unfamiliar with how such modern vehicle technology and systems actually work.
TB correctly noted, "...sizeable 'unless' [someone shared link]..."
But it's also multiplied by the small fractional odds of anyone noticing the shared link, those noticing actually viewing, noticing the live private data, capturing said private data, and then misusing the data. It's not surprising that nothing came of it.
Still, the initial observation is ultimately correct. Just not any significant risk, unless repeated endlessly. So corrective action to prevent recurrence is A Very Good Thing.
"...the video stream was only available to those who had a direct link to the video, and was not searchable via YouTube or accessible without the URL."
Unless somebody involved shared the link in a public forum, then it's actually extremely secure. At least effectively so.
Tom Scott has a video about "Will YouTube run out of addresses" (+/-?), where he explains the YouTube URL addressing scheme and its relative 'address space'. He touches on the utter sparsity of it.
So, why don't you try to find Tom Scott's video by randomly trying various YouTube URL strings until you stumble across it?
Actually, try to find ANY YouTube video by random YouTube URL string. See if you can randomly get a hit.
Good luck!!
Locations, Locations, Locations.
The story goes... Kmart, the failed retail corporation, had been devalued to just a few hundreds of millions of dollars valuation. Somebody suddenly bought it. Why? Because although it was nearly worthless as a retail chain, its real estate holdings, including Leases, was worth billions.
I recently replaced the battery pack buried deep within my Surface RT 2. Successfully!!
It was a frightening combination of Watch Repair and Bomb Disposal. It was many times more difficult than the iPhone 3GS I did years ago. More dangerous too, huge cells glued in.
"...what I do have are a very particular set of skills, skills I have acquired over a very long career. ..."
If such a repair had been attempted by anyone without the complete set of skills and knowledge, they'd have had a flaming lithium battery burning holes in their dining room table.
Such 'Combination Watch Repair / Bomb Disposal' battery replacements shouldn't be widely encouraged.
It'd be nice if they were designed like Cordless Telephones. One minute job.
JJ was quoted, "Smartphone sales are starting to decline at an accelerating rate."
Unclear. Starting? The decline or the acceleration?
Parsing options.
1) Existing decline in sales, starting to accelerate.
2) Just starting to decline and immediately accelerating.
3) A slight decline in sales, the rest is overblown hype.
Hmmm. Um... #3?
It might be the situation where the customer's hard disk has been corrupted, and the IT service actually being provided is File Recovery (where the recovery software could be searching the entire drive looking for file fragments for reassembly).
Or it might be something more sinister. But that's far less likely, I think.
There's a nice picture on the Interweb showing the Earth with the entire side more or less pure Pacific Ocean, plus on minus the camera not being infinitely far away and thus being ever so slightly misleading.
So nearly 50% odds it'll go splash in the Pacific. Plus, rumors indicate that there may be some additional oceans on the other side. Gotta be about 65% wet in total in that latitude band.
So, most likely it'll just disappear while nobody is looking.
Good luck to those in the At Risk latitude band.
JS19 suggested that the environment under discussion was "...roughly < 1/32 Sea Level pressure...".
At orbital velocity, roughly 1/32 Atmosphere would cause any spacecraft to be quickly incinerated. This is one reason why there are not very many satellites orbiting at 80,000 feet above the Earth.
Ratio of 'Age of Universe' to 'Age of the Earth' = about 3.
Not a million-billion. Just '3'.
THREE.
The numerical equivalent of a penny. Hardly worth walking over to pick it up off the floor.
Chemistry 101 Rule of Thumb: Chemistry slows down by about half for each 10°C drop. Space is generally cold. Cold might be an advantage for certain delicate molecules, but most often it just slows things down.
Warm puddles are generally faster. Much faster than frozen asteroids.
So, those proposing any of these sorts of Panspermia ideas need to clearly explain why cold space is significantly more conducive to making these life precursors than all the endless variety of environments found on the nascent Earth. Next multiply the (dis)advantage of cold space by the likelihood of the necessary proto goo then traveling through space and reaching the Earth. Finally consider the total quantity landed on the sterile Earth, and the resultant odds of it being put to use. The net final answer is that this Panspermia idea is either more, or less, likely than the canonical 'warm puddle' on a nascent Earth.
Until they address the above point, it's all just useless noise.
The answer appears clear, but maybe there's some subtle advantage to ultracold chemistry that needs to be explained. I remain open minded, but highly skeptical.
RF questioned, "...elliptic-curve cryptography is not affected..."
The history of math(s) is littered with sudden and unexpected connections popping up where one field is suddenly made precisely equivalent to a completely "unrelated" field.
So when Dr. Christian Szell asks you, "Is it safe?", the future-proof correct answer is, "Presently yes. It is very safe. But in the longer term, probably not. So be very cautious."
Because it is widely assumed that self-driving cars will be Perfectly Safe™, then there's really no need to impose pesky speed limits on them. Allow them to drive just as fast as their wee feisty silicon brains allow.
Speed limits are for fallible humans, not Perfectly Safe™ self-driving cars. Right?
If not, then we'd better revisit the assumption.
Pbswinger attempted, "...in German cars helplessly spinning... [and something suspiciously VW or Audi-ish]..."
Reportedly most, yes *most*, Mercedes sold in Canada are now the AWD 4Matic. Combined with Nokia studded ice tires, they're only stopped when the wheels leave the ground in foot-plus deep snow. Then you dig out and carry on.
Our house is down an icy hill, sometimes pure sheet of glossy ice. We commute daily up and down.
This year it's only snowed three times, and only needed plowing *once* (so far).
NakaSC asked: "how hard is it to change from summer to winter tyres?"
Canada here. We have some experience with this.
Buy another set of four rims. Even if they're mediumly-expensive aftermarket alloys. Then you can change over to winter tires in your driveway.
The point is not to save money in the long run (although that's almost more-or-less true). The point is that you can change to winter tires on your own schedule. Like when they announce a major snow storm is suddenly and unexpectedly arriving 'early' on the 10th of November, then the garages will be booked up until mid-December. The actual expense comes when you crash because you're forced to drive on icy roads with your useless summer tires. It's just easier to do it yourself. Takes maybe an hour.
So it's best to purchase four additional rims, four studded ice tires (e.g. Nokian Hakkawhatevertheyrecalled), and an AWD car. Then you can go out in the middle of an ice storm and harass the SUV drivers by blowing them off at the lights. Or drive into and out of foot deep snow banks, in a comedically-capable AWD car (e.g. Mercedes 4Matic AWD). Or stop in the middle of an ice cover hill, and then accelerate away hard, up the icy hill.
It's actually perfectly normal to have to drive through 6 or 8 inches of snow once in a while. More dangerous is the ice, because crashing is much worse than merely being a bit stuck.
Some people have tumultuous lives, with Stinking Piles of Ruin and Damnation™ from early morning to late-late at night. They may be thinking - no that's too strong a word - ...hoping that by buying and possessing a £999 smartphone, then that device would at least provide some small 5.8-inch space where everything is Just Perfect™. A refuge from imperfection. And so it is. Very satisfying, while the chaotic world swirls around them.
Of course what actually happens is that they immediately drop it, and crack the screen. And of course they failed to purchase any insurance or protection, and they can't afford the repair. So now they're stuck looking at Perfection™ through a broken window.
No wonder some people are so cranky.
"UPS" you scream.
No, I'm referring to the power failure caused by the UPS catching fire, ...again.
A well designed database would have journaling at the transaction later, and more journaling again at the FS level. Oh, sorry. SAP.
My buddy runs the IT for a company. He tells me that the server can have its power cord yanked out, and the backup server in his basement at home will complete the transactions, transparent to the users. They run in parallel and his done something clever at the networking level.
"...could potentially see supercapacitors achieving energy densities of up to 180whr/kg - greater than lithium ion batteries."
^^ Too Obvious Propaganda Alert ^^
The PR person writing the news release chose the number 180 only because it's slightly "greater than lithium ion batteries."
To remain honest, the word "potentially" was included.
It's been reported to me the 'Parky Priuses' (Toyotas parked at airports for a week or so) go dead because the keyless entry reader fails to go to sleep, chatters away endlessly, and runs the smaller 12v battery flat. Then the car won't start, and you'll need a jump. System design mistake?
(2nd hand Info Alert)
"Fuzzing a few pixels here and there can fool convolutional neural nets."
If I may reclaim the phrase "convolutional neural nets" to be reserved for the human brain, then it's worth pointing out that such biological neural nets are not so 'brittle'.
These findings of the new artificial neural nets being 'brittle' is strong evidence that they're doing it wrong.
A hand reaches up and resets the 'Strong A.I. Clock' away from only a few minutes to midnight, adjusting it back to read 'Real Soon Now', just like it's been for much of the past 40 years.
El Reg noted, "...today's quantum computers' main application is as experimental platforms to verify that quantum, rather than classical, processes are at work."
So, it is unclear.
Crikey, bang out a benchmark test that's aligned to quantum computation, make a comparison to classical, and then tick YES or NO. Then go for lunch. Repeat the test in the afternoon.
One of my kidiots (age 14) has somehow picked up an extraordinarily wide ranging vocabulary, primarily from playing video games and watching YouTube. At times he uses obscure words (correctly) that cause our jaws to hang open in disbelief.
Q: "Where did THAT word come from?"
A: "Video games."
DougS noted, "...Its winter here, if they make my PC output a few watts more heat..."
It seems clear to me that we are very close to the point in time when simple electric baseboard heaters will be replaced with server farms in the same form factor. Even today, there's only about one order of magnitude price gap between a 1kw bar heater and 1kw worth of CPU. Considering the value to humanity of this concept, the gap needn't be zero. So this is much closer than one may believe.
Obviously it'll save money on both sides if the cost of heat and the cost of CPU cycles are split.
Somebody needs to start setting up the CPU Cycles Auction website so that when my wall thermostat calls for heat, the banks of Heaters/CPU Farms in my house will have something to do beside folding proteins for charity.
Excuse me, YAWN.
The essential points in this news item could have been written in the 1970s or early 1980s.
There has been endless work done on various approaches to randomly searching the nearly-infinite solution space, desperately looking for the global optimum.
Climbing peaks, setting aside a fraction of the next generation of child proto-solutions to explore more distant options, while the others continue to climb the local hill. "Ooh, let's make the binary string into conceptual DNA, and perform string sex!" There's still a good chance that you've settled on a local peak, missing the tiny width but towering height global optimum just over there.
Yawn. Welcome to ancient history. This is all 1970s or 1980s tech, just a frosting of Moore's Law.
If you want to impress me, then how about using these machine learning approaches to determine the optimum algorithm to manage the machine learning itself? Last I've heard, they're still essentially hand coding the algorithms that manage the machine learning itself.
If anyone had bothered to apply some recursive conceptualization (just one example), then A.I. wouldn't be stuck in the disco era forever.
2018 GOTO 1978
/rant
Phuzz suggested, "...countries have been doing that since written history became a thing."
That's either an amazing coincidence in timing; or perhaps there's another explanation for this apparently common starting point for A) the written history of such activities, and B) ah, er, hmmm... written history itself.
Excuse the pedantry, humour opportunity. :-)
A.I. boffins, shocked and surprised by latest observations, effectively announce that they actually don't really understand what's going on inside neural networks.
The phrase "haven't got the first clue yet" springs to mind.
Is there an explanation for their apparent dough-headedness?
PhilipN was worried, "...it could split in two."
From a Project Management point of view, this is trivial to avoid. Open DOORS database, insert new Requirement 'Structural#1837': BY THE WAY, IT SHALL NOT SPLIT IN TWO. Save, close file, go for lunch.
Absolutely trivial. Takes only a minute.
phuzz wrote the inherently self-contradictory, "You'd be surprised... ...Not much..." <- :-)
"...it also give you complete freedom in orbital inclination, and launch times and locations."
I understand that launch location and orbital inclination, if optimised for fuel, are somewhat linked. The Saturn V famously used the "Roll Program" to tune the orbital inclination to match the Moon's.
Launch times would be impacted by weather, both the rocket's limits and now this added aircraft's limits. So it unfortunately remains lacking in "complete freedom". Perhaps they could move to a different airport the night before the storm (if that's even practical, what with having to prepare a rocket).
Yes, the slab itself is very capable, but tedious.
Physical I/O, a term which perhaps I just made up, refers to the conceptual 'edges' of the device (sensors being merely one category), where can interface to the real-world. It's a design space where there's lots of room for interesting growth.
-Dual high quality microphones to capture binaural audio.
-Software Defined Radio chipset dedicated to the user (distinct from the data modem). Enabled tuning into DC to L band (up to low GHz).
-IR temperature sensor (directional type)
These sorts of things...
Justice Alito said, "...move [the data] around at will."
Worse than merely "move". They could quite easily smear the data files across storage servers in a dozen countries, hashing it up so that nothing would make sense without reassembling (say) 80% to 90% of the data (to allow for failures). It could very easily become effectively impossible to subpoena data, if the companies chose to play games like that.
Sorry to be the bearer of bad news, but the rules do need to change to be directed at a point of control.
I expect downvotes because some people won't like this.
But it's still obviously true.
Sorry.
The only things different (radio-wise) about the Moon compared to the Earth are as follows.
It should be radio-quieter (lack of people).
No ionosphere.
No atmosphere, no moisture.
Horizon would be much closer (size of Moon).
In general, modulation schemes ("waveforms") are getting more clever and more optimal each generation. That's simply a function of human intelligence.
Other than that, it's system design and trade-offs.
"QPSK modulation might work well on the Moon"
I would hope so. If it didn't, then that would be a very interesting observation. :-)
jonfr suggested, "Since on Earth the best 1800Mhz can give is 25 km."
MHz
And why is 1800 MHz limited to 25 km? It's certainly not an inherent limitation of the band, as there is no such thing. Perhaps given a certain EIRP and 'waveform' technology, it might be a rough rule-of-thumb guideline. But it'd be wrong all the time.
The client wanders in, looks around for a queue or any sort of Take-A-Number organization. Finding none, stands there looking confused.
After 26 minutes of being simultaneously bumped and ignored, releases belt and drops trousers to ankles. Within seconds, additional service staff absail from the ceiling and one of them inquires if they can offer any assistance, and by the way, would you mind pulling up your trousers?
No wonder the staff require mental health counselling. Dealing with so many customers standing around in their pants.
Perhaps they could install a queue concept.
On the same point...
My present Asus Zenfone 3 supposedly has "Hi Res" 24-bit/192kHz audio. What it doesn't seem to have is enough low-impedance oomph (current output) to drive my better headphones. When I switch from my $25 headphones (higher Z) to my $200 headphone (lower Z), the bass response gets worse. In spite of the better headphones being spec'ed to 3Hz.
On my other gadgets, the expensive headphones are wonderful. For example, my several old iPhones had excellent analog outputs (and a headphone socket) and they could make the big headphone actually vibrate with the infrasonic bass.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
