Supercomputing improvements are nice, but I personally want to see them get the cash to profoundly increase their NEXRAD backhaul (the data lines connecting their radar sites to the outside world).
Right now, they're HORRIBLY backhaul-constrained. I believe most/all NEXRAD sites only have 256kbps frame relay to upload raw data to NOAA's datacenter for further processing & distribution to end users. As a result, they're forced to throw away data at the radar site to trim it down to size, and send it via UDP with little/no modern forward error correction. That's a major reason why glitches are common. In theory, the full-resolution data is archived to tape on site and CAN be mailed in if some major weather event happens that might merit future study, but the majority of collected data gets archived to tape, then unceremoniously overwritten a few days later. And most of the tapes that DO get sent in sit in storage for weeks or months before finally getting added to their near-line data archive.
The low backhaul bandwidth is made worse by the fact that the secondary radar products (level 3 radar, plus the derived products like TVS) get derived on site, and wedged into the SAME bandwidth-constrained data stream. That's part of the reason why level 3 data lags by 6-15 minutes... they send the raw level 2 data, and interleave the previous scan's level 3 data into the bandwidth that's left over. I believe the situation with TDWR sites is even worse... I think THEY actually have a single ISDN line, which is why level 2 data from them isn't available to the public at all.
As I understand it, they can't use lossless compression for two reasons -- since they have no error correction for the UDP stream, a glitch would take out a MUCH bigger chunk of data (possibly ruining the remainder of the tilt's data), and the error correction would defeat the size savings from the compression. Apparently, the processors at the site are pretty slow (by modern computer standards), so it would also add significant delay to getting the data out. When you're tracking a tornado running across the countryside at 50-60mph, 30 seconds matters.
If NWS had funding to increase their backhaul to at least T-1 speeds, they could also tweak their scan strategies a bit to make them more useful to others. For example, they could do more frequent tilt-1 scans (the lowest level, which is the one that usually affects people the most directly), and almost immediately upgrade all current NEXRAD sites to have 1-minute updates for tilt 1 (adding about a minute to the time it takes to do a full volume scan, but putting data more immediately useful to end users out much more frequently).
Going a step further, more bandwidth would open the door to a fairly cheap upgrade to the radar arrays themselves... they could mount a second antenna back-to-back with the current one with fixed tilt (ideally at 10cm, like the main one, but possibly 5cm like TWDR if 10cm spectrum isn't available, or a second dish of the proper size for 10cm wouldn't fit), and do some moderate hardware and software tweaks that would effectively increase their tilt-1 scanrate to one every 6-10 seconds (because every full rotation of the main antenna would give them a full tilt-1 rotation off the back). This means they could send out raw tilt-1 data with 6-10 second frequency. It's not quite realtime, but it would be a HUGE improvement over what we have now.
Unfortunately, NWS has lots of bureaucracy, and a slow funding pipeline. I think it's safe to say that the explosion in popularity of personal radar apps, combined with mobile broadband, almost totally caught them by surprise. Ten years ago, very few people outside NWS were calling for large-scale NEXRAD upgrades. Now, with abundant Android and IOS apps & 5mbps+ mobile data the norm, demand is surging.
That said, I hope they DON'T squander a chunk of cash on public datafeed bandwidth instead of upgrading their backhaul. I'd rather see them do the back-end upgrades that only THEY can do, and tell people who want reliable & frequent upgrades to get their data feed through a private mirror service (like allisonhouse or caprockweather) who can upgrade their own backhaul as needed, instead of having to put in funding requests years in advance.
You have it wrong. The third-world products are mostly UN-encumbered. It's the products meant for sale in the US, Japan, and Europe that end up gimped and crippled into proprietary uselessness.
Correction: be able to stream local content, not be locked to Google, and be both rootable & reflashable so it won't end up in landfills like the Revue and original Nexus Q did.
That's why double-gang divided NEMA boxes and conduit exist. Build the plastic box and conduit into the furniture, ship it with both ends covered by screw-on plates, and leave it up to the end users to wire it as they please (a pair of outlets, a single-gang outlet plus a low-voltage keystone, or whatever).
With a divided double-gang box fed by separate 1/2" conduits, you can run just about anything. At the other end of one conduit, they could put a 15A 120v RV-type "inlet", like this one: http://inverterservicecenter.com/Marinco-150BBIW.RV (if you wanted to wire both gangs for power, instead of using one for ethernet/fiber/whatever, you'd pull out the box divider at the outlet end and feed both from the same conduit). Carlon ENT conduit is perfect for this purpose ( http://www.tnb.com/ps/endeca/index.cgi?a=nav&N=3819+601+3818 ). Worst-case, they could use Hubbell's funky JLOAD single-gang multimedia outlets, which pair a single 120v power outlet with a pair of low-voltage keystones, designed for use with a special box that shields the high-voltage power away from the low-voltage wiring. ( http://www.cesco.com/b2c/product/447768 )
Re:funny comparing to "high speed rail" elsewhere
on
Amtrak Upgrades Wi-Fi
·
· Score: 4, Interesting
That's part of the reason why FEC (Florida East Coast Railroad) has never been actively *hostile* towards passenger trains, compared to railroads like CSX(*). FEC runs mile-long trains hauling limestone at 60mph on glass-smooth welded-rail tracks that are maintained to higher standards than some stretches of track in the NEC itself. FEC's one non-negotiable mandate for voluntary passenger service on their tracks has always been that someone else had to pay to lay down a second track, maintain it to FEC's no-compromise high standards, and equip every train that runs on them with in-cab signaling and the kind of automation rarely found outside of Japan(**).
Once Amtrak, Florida, and a federal funding act or two cleared the way for the feds to pay most of the bulk cost of double-tracking FEC from Jacksonville to Miami, FEC announced that Amtrak was welcome with open arms (Amtrak itself is still trying to scrape up funding for the trainsets themselves, or come up with a good way to split & join NY-Florida trains in Jacksonville so half can proceed straight down the east coast to Miami, and the other half can run to Orlando & Tampa (historically, Amtrak has always resisted splitting/joining trains anywhere besides an endpoint).
(*)About 15 years ago, FDOT approached CSX with a request to double-track it from Auburndale to Tampa for Tampa-Miami passenger rail. CSX refused. FDOT offered to TRIPLE-track it... and CSX still refused. Exasperated, FDOT offered to elevate a ~12 mile segment running through Lakeland, and CSX told them that the only way they'd voluntarily allow it is if FDOT agreed to let CSX refund the purchase price and demolish it at will if it later decided that the support columns or track structure were in the way of whatever they felt like doing. That was the turning point when FDOT decided that any future rail route between Orlando and Tampa simply *had* to run along I-4 instead of CSX... CSX was impossible to deal with in any sane way, and taking the corridor via eminent domain would have ended up costing more than building it down the middle of I-4 instead (I-4 was planned for complete reconstruction over the next 10-20 years anyway, and FDOT owned a fairly wide corridor that was straight and flat, so they just designed the empty space into the new road and bridges so it would be there when the day came to build the new tracks).
(**)FEC is a HUGE proponent of cross-training and automation, and because it operates entirely within a single state, it can get away with telling its union to go to hell over things that would get CSX crucified. For example, FEC requires all engineers and conductors to be cross-trained and capable of serving either role as needed (sensible and efficient, but *vehemently* opposed by railroad unions because it means the conductor can operate the train while the engineer takes a break, instead of having to staff a second engineer while the conductor twiddles his thumbs). I believe it also requires engineer-conductors to have college degrees.
I wonder whether Larry got vaccinated for rabies a month or two before the laryngeal paralysis became apparent.
There's growing evidence that indiscriminate annual rabies vaccinations commonly performed on dogs & cats might be the root cause of the growing number of cases of laryngeal paralysis seen among pets in the US.
At the macro level, the evidence is pretty compelling:
* Laryngeal paralysis was practically unheard of in cats, and rare in dogs, until about 10 years ago... right around the time vaccines that use live modified rabies virus started to be used in cats as an alternative to adjuvanted rabies vaccines made from inactivated rabies virus.
* Laryngeal paralysis is usually accompanied by rear hindquarters lameness, and other symptoms that are basically mild versions of the same symptoms on a rabies screening checklist.
* More to the point, if you were to read about the symptoms of rabies, then describe a hypothetical variant that's mild, chronic, and non-communicable... you'd almost *precisely* describe the symptoms observed in cats with laryngeal paralysis.
The live-vs-adjuvanted distinction is important, and part of the reason why the theory is so controversial. Adjuvanted vaccines have a long, well-documented history of inducing vaccination-site sarcomas (cancer) in cats. A small number, to be sure... but enough to be unmistakable and visible (it's hard to argue that a vaccine didn't induce a tumor when one spontaneously appears in the *exact spot* where a vaccine was injected a few months earlier, even if the precise ingredient of the injection that caused it is unknown). The arrival of non-adjuvanted rabies vaccine was hailed as a major step forward, because it's presumably safer for cats... unless, of course, a small number of cats exposed to it end up developing chronic lab-induced rabies.
Making matters worse, the revenue vets get from those vaccinations is a major part of their annual income. Because most vets do, in fact, love cats and dogs and want to do what's right for them, this presents them with a terrible choice they're forced to make. Some simply ignore the problem, blindly cite legal compliance as an excuse, and wash their hands of moral responsibility. Others will now provide annual waivers for elderly and debilitated cats who would be risky to vaccinate at all (illness appears to be a MAJOR risk factor in post-vaccination LP), and will offer clients the option of titer-tests instead in an effort to roll the dice fewer times and reduce the likelihood of VAS or LP in that manner.
In any case, there's already strong (if anecdotal) evidence that live rabies vaccines are a major (if not THE) risk factor behind feline laryngeal paralysis, and it's not inconceivable that something similar might happen to humans who are vaccinated with live, modified rabies virus as PEP.
I believe that in California or New York, if the driver of a car is arrested for DUI, the police can actually force PASSENGERS who are under 18 (but not 18-20) to submit to blood alcohol testing too, arrest them for alcohol possession if they exceed some threshold, and suspend THEIR licenses if they refuse -- even though they themselves weren't the ones behind the wheel driving.
I can attest to airbag injuries. I was in an accident 12 years ago, wearing a seatbelt, and had the airbag deploy. It pushed one of my hands up through the windshield (leaving me with so many laceration wounds, the doctors had to spend 10 minutes debating whether or not I needed skin grafts, and to this day I have no feeling across probably 15% of the back of my hand due to nerve damage), and left me with second-degree burns on both forearms.
The culprit? Hyundai designed the 2001 Tiburon for European and Asian airbags, then put bigger airbags into its US models without bothering to notice or care that a deployed airbag would actually push out the windshield... made worse by the fact that they were *notorious* for triggering in low-speed crashes that were WELL below the threshold set by the NTSB. Without the airbag, I would have walked away from the accident without meaningful injury, and my car would have had about $2,000 worth of damage. Instead, my hand got shredded, and my car was a total loss because airbag deployments are too expensive to cost-effectively fix (it's cheaper for insurance companies to declare the car a loss, pay the owner, then sell the car to someone who'll export it to Haiti and repair it *there* with the airbags omitted or disabled, unless you're talking about a very new, very expensive car).
The problem is that in the US, 99% of cars aren't actually *owned* by the driver... they're secured assets owned by a lender. If a state attempted to impose such a law, every lender in the state would *instantly* require "DWI forfeiture insurance", and every driver in the state would instantly see his annual car insurance rates get a few hundred dollars tacked on to the premium to cover it.
Are we talking about the same site? As badly as I want to feel sad, I can't even remember the last time that site had anything that seemed relevant, especially ever since Android arrived.
It's harsh, but as a practical matter, the "Linux Appliance" market is kind of in a state of both crisis and opportunity right now. Crisis, because anything x86 based with less computing power than a 2GHz+ Core2 can't really compete with a nameless Android USB-HDMI stick from Shenzhen... but those same nameless USB-HDMI Android sticks from Shenzhen are so cheap, you could practically connect them to port extenders, pot them in black epoxy, mount them inside an attractive case, and use them for things that would have been cost-prohibitive with a $599 slow Via system.
After reading about them, I can't help but wonder whether an interesting prosthetic compromise might be to somehow attach muscle fibers to implanted force-feedback strain gauges, then use the strain-gauge readings to control the hand itself (possibly in conjunction with all-electrical sensors used with other nerves to provide additional control. In other words, instead of pulling on bones, the muscles would pull on artificial ligaments cemented onto strain gauges attached to some kind of stretchy/springy plastic that itself is attached to a worm gear that moves the far point closer or farther from the muscle to alter the resistance.
For finer control (like individual fingers), it would take an idea from the way HTC's hybrid mechanical-capacitive switches used on their last few Windows Mobile and early Android phones worked. Basically, they used capacitive means to determine WHERE (approxiamtely) you were touching the phone, but used an actual switch triggered by a press anywhere in the region to determine when you actually intended to fire a switch event.
In a similar manner, the hand's controller could attempt to discern things like individual finger control by sensing the nerve bundle directly to come up with blunt motions, but sanity-check it in realtime against the muscle-triggered strain gauge, and use fuzzy logic to correlate patterns of nerve activity with specific variants of strain-gauge muscle tension to produce an intended action (so the actual finger, for instance, curled in direct response to the real-world forearm muscle pulling on the implanted strain gauge).
The problem with non-mechanical nerve interfaces is that they basically have the same kind of problems that capacitive touchscreens do... terrible signal to noise ratio and processing latency compared to real-world direct actuation. Nerves aren't like switch-triggered wires in a harness... they're more like a bundle of coax carrying multiplexed spread-spectrum QPSK signals with unbelievable amounts of background noise. By directly interfacing a few muscles with strain gauges, we can bypass the hundred (give or take) years of R&D it's going to take to get signal processing up to the point where it really needs to be, and just take advantage of the signal processing that the forearm muscles ALREADY HAVE to pick out the right signals and translate them into commands for the prosthetic hand.
Try dumping 2-4 packs of Sweet 'n Low into a diet Coke or diet Pepsi at a restaurant. Once the foam goes back down and it dissolves, you'll be amazed. It tastes almost exactly like regular Pepsi.
Diet Coke and Diet Pepsi weren't formulated to taste good, they were value-engineered to meet a price point. Back in the 80s, NutraSweet was still under patent, and quite expensive. So the used the least amount they could get away with and have it still be tolerable.
That's why Pepsi ONE came out in ~1999... the patent expired, so Pepsi could try to differentiate it at almost no extra cost to themselves by doubling the aspartame. The problem was, people who liked Pepsi ONE were much less tolerant of "diet drink" non-sweetness, and Pepsi ONE's improved taste went south REALLY fast as the aspartame broke down due to heat or age.They now blend aspartame and sucralose or AceK in the new limited-release Pepsi ONE to keep it halfway drinkable after the aspartame breaks down.
For a while, the bottling company for 2-liter bottles of Diet Mountain Dew in South Florida TOTALLY fsck'ed it up. It tasted fine in cans (presumably, from a different bottling company), but I threw away 2 liter bottle after 2-liter bottle for MONTHS before giving up on it and only buying cans for a while. The bad bottles first showed up at Wal Mart, then Target, then finally started to show up at Publix (Florida's main grocery store).
I suspect that the recent surge in 10-calorie diet drinks is partly due to the rising cost of corn due to ethanol demand... at this point, it's probably cheaper to replace the HFCS with a blend of aspartame and sucralose or AceK.
Is it just me, or does Coke Zero taste like what I vaguely remember Tab tasting like? I mean, seriously... I haven't had the misfortune of trying to drink Tab in literally decades (and that's coming from somebody who drinks NOTHING but Diet Mountain Dew, Pepsi One, and other diet drinks now), but the first time I tried Coke Zero... all I could think was, "Ewwwwww... Tab."
The big problem was that Coke rammed it down people's throats and took away the old version completely. Had New Coke been rolled out alongside Classic Coke, Classic Coke TODAY would be a niche item available in 12oz 6-packs, maybe 2-liter bottles at liquor stores and large grocery stores, and if you could buy it at a gas station or convenience store at all, it would be as individual cans or retro-looking bottles in the original cane sugar formula at double the price of regular Coke.
By abolishing Classic Coke overnight and turning it into American holy water that people guarded with their lives, Coke managed to shoot themselves in the foot. As others have said, if Microsoft made Metro an optional part of Win8, and made it a free voluntary download for Win7, some people might have actually liked it. Instead, they've gotten people to hate Windows 8, and induced hatred of their phones and tablets, too. At this point, Metro has become so tainted & toxic, Microsoft is going to end up having to reinvent their phone and tablet UI from scratch yet again just to de-stigmatize it and undo the damage done by desktop Metro.
Is there any way to replace Windows' own search with one that lets you search for straight-up filename matches with minimal ceremony or additional UI gymnastics? Preferably, one that skips the index (or at least, jumps to the drive itself after grabbing the easy low-hanging fruit), and goes straight for the bare NTFS data structures?
I'm *so* sick of having to fight with Win7's brain-damaged searchbox every ****ing time I go to search for a file whose name I already know (enter name, hit return and wait to start the pointless search, click "search everywhere", wait for it to get its act together, click 'advanced', select 'show hidden/system/non-indexed', click 'browse', navigate to c:, and about 20 seconds later, finally start the search I actually WANT).
It depends. MySQL's championship role has always been lightning-fast reads from tables that change rarely and don't need much in the way of concurrency protection. I believe it was also the first free database to have a concept of partitioning.
The main problem with MySQL lately (past 5 years or so) has been the fact that it's really easy to get an inflated impression of its more "hardcore" abilities, and not discover until it's too late that MySQL's ability to handle things like failover, replication, partitions, and materialized views(*) aren't quite as capable as someone who's used to Oracle & reads MySQL's docs might initially think they are. I fell into that trap myself ~3 years ago, and got bitten pretty *hard* by harsh reality not living up to first impressions.
(*)'it's been a couple of years, but from what I remember, MySQL had a huge 'gotcha' with materialized views that rendered them mostly useless for me at the time... I think it involved having a function in the select statement that was a LOT more restrictive than the rules Oracle imposed at the time. I vaguely remember it was something like "Oracle couldn't index a MV if you had a function in the select clause, but MySQL didn't allow you to have a function *anywhere* in the select, even if it was merely transforming one value directly and deterministically into another." -- or something to that effect. I know that the whole matter of function-based indices is a problem in general, but I just remember thinking, "Oh, cool... MySQL supports them now!", then discovering later that MySQL's rules were much, much more rigid about what was and wasn't allowed than Oracle's (and Oracle's were pretty anal to begin with).
Anybody care to guess how many days it would take Verizon to change its mind about FiOS if Google showed up at the next Mantoloking city council meeting & offered to deploy Google Fiber there if the city paid the direct costs of laying the fiber itself? Oh, and offered to pay for the lawyers the city would need when Verizon and Comcast fought back the only way modern American corporations seem to be capable of competing -- by using the courts to block it, instead of trying to outdo them by offering better and faster service...
The big problem with Comcast is that it's unreliable, and makes no attempt at being otherwise. When a Comcast service area has a region-wide commercial power outage, Comcast goes down, and stays down, until commercial power is restored. Sometimes, it doesn't even take a regional outage. Last year, Tropical Storm Isaac took down Comcast for several hours across Dade & Broward counties. The failure's apparent cause? A power outage at a SINGLE LOCATION along their backhaul route. They talk about customers with backup power, but it only works if your house is literally the only one on the block without power. The moment the blackout extends to your neighborhood, or any point along the backhaul route between your house and their network center, your internet service is gone.
Ten years ago, a fragile market compromise existed between "the company that used to be the phone company", and "the company that used to be the cable company". The phone company had DSL. It wasn't as fast as cable, but it was built to run on carrier-grade infrastructure and enjoyed most of its same robust backup power and reliability. The cable company had DOCSIS. It was faster than DSL, but nowhere near as reliable. So, consumers kind of had a choice... slower service that rarely went down, or faster service that had daily interruptions & got knocked down for the count whenever ANYTHING significant happened to the local power grid. Then... our wonderful regulators decided to allow the phone company to neglect its infrastructure and allow it to degrade to the non-reliability of cable, and now we have lots of places where there really ISN'T anything that approaches the former reliability and robustness of landline phone service.
Wireless might be fast to bring back up after a storm, but buried copper with battery power at the central office didn't used to go down at all. There were people who came home to neighborhoods wiped away by Hurricane Andrew, and were greeted by phones making "off hook" noises from under the rubble. It was the lucky consequence of having mostly buried telecommunications infrastructure that was built to survive a 20-megaton nuclear bomb falling within 2 miles of downtown Miami, and another pair of bombs falling within a mile or two of Homestead Air Force Base. Today, U-verse isn't quite as bad as Comcast (AT&T still owns generators that were purchased years ago, though the state no longer requires them to maintain or use them), but the sad truth is that our nation's telecommunications infrastructure today is a pale, ghetto-fabulous shadow of what it used to be. Reliability and availability can no longer be taken for granted at any level unless you own your own fiber and have direct control over its endpoints so you can provide your own backup power. Even services like Frame Relay that were once protected by ironclad SLAs and stiff, automatic penalties with teeth, are now at risk of carriers just looking the other way and making a business decision to pay the fines and let the network go down because it's cheaper to take a hit for a few days now than to invest the capital into five-nines or better uptime.
Once again, it comes down to a question of scope and scale. Thousands of balkanized islands of photo and video data that are mostly offline are of little concern. The act of making that aggregate bulk data trivially searchable by anyone, anywhere, fundamentally changes its nature. Bars and stores might archive lots of data, but they keep it to themselves. When it comes to maintaining personal privacy, high barriers to entry, roadblocks to use, and restricted availability of personal data are a GOOD thing. Back when you had to go to the courthouse and fill out request forms to look at court documents, nobody cared about redacting social security numbers and similar private information from exhibits and evidence. The moment it became possible to automatically scrape court websites and do automated bulk OCR on the documents obtained from them, it became a HUGE problem.
As far as your condo complex goes, you might want to talk to your association's lawyer before holding onto the license plate data for TOO long, unless you want to start getting hit by subpoena after subpoena from divorcing couples doing discovery on each other at your association's expense. The cost of complying with those subpoenas can become expensive, and your ability to recoup those compliance costs by charging for the information might be limited or nonexistent. That's why most condo complexes intentionally destroy and discard such information after a while -- after a month or two, that information becomes a potential liability to the association.
Consumer Reports bases their opinions on values that don't necessarily align with yours. For example, they regard pet health insurance as being a bad deal. If you're somebody who'll euthanize a pet at the first hint of an expensive medical condition, they're probably right. If you'd give one of your own kidneys to your cat or dog to save his life if the vet told you it was his only hope for survival, they're absolutely wrong. Most people fall somewhere in between. If you pay $50/month for 10 years, and your pet dies from some accident without warning, the insurance company wins. If your 2 year old cat or dog gets cancer, the insurance company loses. If your 15 year old cat or dog develops some chronic illness that can be kept in check for years with aggressive care at the first hint of a flare-up (pancreatitis, asthma, HCM, etc), the insurance company loses *spectacularly* after briefly thinking it won... especially if he or she lives long enough to get cancer. The insurance companies partly keep their losses in check with somewhat high per-incident deductibles. If your kitty is running up $500-900/month medical bills month after month, even 20% deductibles are going to get really painful and weed out all but the most dedicated pet parents.
Ditto, for things like mobile phone insurance. If you're someone who buys dumbphones, or phones that are free/1-cent and would cost $200 or less if you bought them at full price, it's a bad deal. If you just bought a brand new Galaxy S4 with unsubsidized replacement cost of $600 or more & MINIMUM $200-300 charge for a cracked screen repair, you'd have to be positively insane to *not* buy the insurance. The key is knowing when it makes sense to discontinue it. Buy the insurance for the first 12-14 months, then drop it once the cost of buying a used replacement on eBay is less than double the deductible, and you'll come out ahead. Keep paying monthly premiums on a 2-3 year old phone, or spend more on the deductible than a used replacement on eBay would cost, and the insurance company wins. The main way phone insurance companies come out ahead is by customers continuing their policies LONG after the phone has ceased to be expensive enough to justify it. If you were required to insure the phone for 2 full years, phone insurance might be a bad deal. But since you can drop it at any time as of your next bill, it makes sense to get it with almost any new expensive phone & keep it for at least a few months.
Extended warranties are a tough call. 10+ years ago, stuff that didn't die within the first 3 months probably wouldn't spontaneously break on its own during years 3-5 anyway. Now, it's a total crapshoot. The $900 LCD TV you just bought might have a replacement cost of $600 in 3 years, but chances are that its warranty is only good for 1 year... and if the TV breaks out of warranty, you're probably fucked since modern LCD TVs are practically large-scale integrated circuits with almost nothing inside that can be cost-effectively repaired by an independent repairman. And unlike 10+ years ago, things like LCD TVs really DO spontaneously drop dead after 2-4 years for no obvious reason.
For cars, extended warranties are an equal crapshoot. If you keep the car long enough for the warranty to kick in and matter, it's basically a prepaid service plan that you'll probably break even on & avoid getting hit with a $1,000+ repair bill out of the blue long after the original warranty expired... as long as you don't get into a wreck before that point, and have the car totaled. The "car totaled after paying premiums for months or years, but before any expensive repairs" is really how they come out ahead, because almost ANY car that's more than 5 years old is eventually going to need a repair that's expensive enough to break even or come out ahead with the insurance. At the end of the day, enough cars get totaled for the insurance company to come out ahead, even after they've paid to replace the AC compressors, power steering pumps, alternators, and half the automatic trans
Yes, but the surveillance camera footage from those businesses doesn't get archived forever and published to the internet in a way that allows strangers to upload a scanned photo of your face and search through yottabytes of video footage for segments where your face was autodetected.
Most people have no real issues with being recorded by businesses, because real-world storage isn't free, and the overwhelming majority of that video footage gets autodeleted within hours, days, weeks, or at worst, months, unless something noteworthy happens and it's explicitly set aside for longer retention.
If Google archived Glasses video footage on the server for 72 hours absent a court order demanding longer retention, the only way to view it was on a device with DRM-enforcing videocard that protected it at least as well as Blu-Ray content (or Glasses that enforced similar end-to-end transport-layer encryption), Google robustly watermarked it & threatened site owners who allowed content with that watermark to exist on their servers (say, photographed off the monitor screen) with eternal banishment from Google search results unless the site owner actively scanned files for the watermark and deleted them), and its long-term searchability was extremely limited, people might not mind it as much. But when when you get into things like "eternal retention" and "casual, anonymous searchability via geotagged face-recognition", it absolutely DOES change the very nature of the video and the consequences of its existence.
If somebody breaks into your home in Florida, you can empty a semi-automatic rifle into his back as he runs to the door, and you're unlikely to face charges once investigators have established that 1) he was a bona-fide intruder and 2) you had even the slightest excuse to believe he was endangering you (like, "I feared he had a weapon left outside and was running out to retrieve it").
The main thing that makes Florida's law unique is that it extends beyond your home to all lawfully-occupied locations. So, if somebody pulls a knife on you and makes you give them your wallet, you could quite legitimately empty your gun into their back while they run away with it & claim they threatened to come back and injure you if they heard you call for help. Things would get messy if your own possession of the gun wasn't entirely lawful, or there was evidence that you yourself were somehow involved with an illegal act involving the dead alleged assailant (like a drug deal), but if the police investigation determines that you were the victim of a straightforward mugging that involved threats of violence or the use/brandishing of a weapon by the mugger & your own possession of the gun was lawful, that's pretty much the end of it. None of this is really *new* -- the main thing Florida's law did was codify the de-facto case-law norm into statute law, and provide automatic administrative procedures to nip prosecution in the bud before it ever got to the point of a real trial.
You're forgetting something important: the value to such a company of stopping people from producing food on their own is basically zero. OK, a little more than zero, but it's basically lost in the the margin of error. For every survivalist family of doomsday preppers growing their own organic food off-grid on a farm in Idaho, there are ~24,000 families that haven't eaten food that didn't come from a restaurant or a microwave oven within the past week, and a few thousand more whose idea of baking bread from scratch consists of tearing open a box of Krusteaz & dumping it into the bread machine with a cup of water before going to bed.
The sum total annual output of every organic farm in America would be hard-pressed to supply a few expensive restaurants in New York and Los Angeles. Have you SEEN the price of a flawless, unblemished organic tomato at a grocery store? The price is off the scale, because that one randomly-lucky bug-free tomato that ended up sealed in wax & cellophane at the grocery store had a few hundred neighbors that either got thrown away or made into spaghetti sauce because nobody would have bought them.
> This means indigenous American vegetation is not dependent on honeybees for fertilization.
And what percent of the food in an average American shopping cart is actually derived from indigenous American vegetation? You know, all those alien foods from places like Europe & Asia that we eat here... lettuce, tomatoes, potatoes, grapes, wheat, oats, rice, etc.
Man does not live by ethanol, high fructose corn syrup, and nacho chips alone, even if it IS possible to make it through a Saturday picnic consuming little else besides beef;-)
Congress needs to pass a law subjecting anyone who facilitates automated face recognition to joint and several liability for any torts that might arise from its use.
Nothing, and I mean NOTHING, will stop corporations like Google and Facebook in their tracks faster than making them jointly & severally liable for any torts ("tort" == "something you can be sued for under Anglo-American Common Law") that might arise from the use of their face-recognition data.
All the EULAs and disclaimers in the world won't save them from a lawsuit where a jury awards the plaintiff (say, a 19 year old college student who got tagged as a 'slut' by a fellow student's crowdsourced webapp built atop a hypothetical service like "Google Face Search") a million dollars, determines that the student is 99.999% at fault, and that Google is 0.001% at fault, because unless the student is a multi-millionaire, Google would end up on the hook for more or less the entire amount (that's what Joint & Several liability means).
Remember, under J&S Liability, it DOESN'T MATTER if the tortfeasor is violating Google's own TOS. Well, it matters to the extent that the overwhelming allocation of guilt would go to the guy who violated Google's TOS, but when he ends up being too poor to pay, the entire remainder would get dropped in Google's lap. If Google tried to spin off a subsidiary to run the service to protect themselves from such a lawsuit, that subsidiary would be bankrupted, put out of business, and shut down by the first lawsuit within weeks or months... and if Google kept re-spawning new entities that tried to inherit data from the previous one, eventually some state would allow a court to pierce the corporate veil and go straight for Google itself.
The net result is that any company that engages in face recognition would be forced, by threat of financial extermination, to guard the data with its corporate life and treat it like PII of the most sensitive kind. Small companies that went overboard would get sued into oblivion by victims before they had a chance to do much damage, and large companies' would be prohibited by their own shareholders from even *thinking* about making use of face recognition data, simply because they'd be put out of business if the data ever leaked out.
The best part about J&S liability is that it makes additional government regulation largely unnecessary. With the sword of J&S liability hanging over its head, Wall Street would swiftly stop face recognition in its tracks by any organization large enough to be a real threat to people's privacy.
Slashdot Mirror
Supercomputing improvements are nice, but I personally want to see them get the cash to profoundly increase their NEXRAD backhaul (the data lines connecting their radar sites to the outside world).
Right now, they're HORRIBLY backhaul-constrained. I believe most/all NEXRAD sites only have 256kbps frame relay to upload raw data to NOAA's datacenter for further processing & distribution to end users. As a result, they're forced to throw away data at the radar site to trim it down to size, and send it via UDP with little/no modern forward error correction. That's a major reason why glitches are common. In theory, the full-resolution data is archived to tape on site and CAN be mailed in if some major weather event happens that might merit future study, but the majority of collected data gets archived to tape, then unceremoniously overwritten a few days later. And most of the tapes that DO get sent in sit in storage for weeks or months before finally getting added to their near-line data archive.
The low backhaul bandwidth is made worse by the fact that the secondary radar products (level 3 radar, plus the derived products like TVS) get derived on site, and wedged into the SAME bandwidth-constrained data stream. That's part of the reason why level 3 data lags by 6-15 minutes... they send the raw level 2 data, and interleave the previous scan's level 3 data into the bandwidth that's left over. I believe the situation with TDWR sites is even worse... I think THEY actually have a single ISDN line, which is why level 2 data from them isn't available to the public at all.
As I understand it, they can't use lossless compression for two reasons -- since they have no error correction for the UDP stream, a glitch would take out a MUCH bigger chunk of data (possibly ruining the remainder of the tilt's data), and the error correction would defeat the size savings from the compression. Apparently, the processors at the site are pretty slow (by modern computer standards), so it would also add significant delay to getting the data out. When you're tracking a tornado running across the countryside at 50-60mph, 30 seconds matters.
If NWS had funding to increase their backhaul to at least T-1 speeds, they could also tweak their scan strategies a bit to make them more useful to others. For example, they could do more frequent tilt-1 scans (the lowest level, which is the one that usually affects people the most directly), and almost immediately upgrade all current NEXRAD sites to have 1-minute updates for tilt 1 (adding about a minute to the time it takes to do a full volume scan, but putting data more immediately useful to end users out much more frequently).
Going a step further, more bandwidth would open the door to a fairly cheap upgrade to the radar arrays themselves... they could mount a second antenna back-to-back with the current one with fixed tilt (ideally at 10cm, like the main one, but possibly 5cm like TWDR if 10cm spectrum isn't available, or a second dish of the proper size for 10cm wouldn't fit), and do some moderate hardware and software tweaks that would effectively increase their tilt-1 scanrate to one every 6-10 seconds (because every full rotation of the main antenna would give them a full tilt-1 rotation off the back). This means they could send out raw tilt-1 data with 6-10 second frequency. It's not quite realtime, but it would be a HUGE improvement over what we have now.
Unfortunately, NWS has lots of bureaucracy, and a slow funding pipeline. I think it's safe to say that the explosion in popularity of personal radar apps, combined with mobile broadband, almost totally caught them by surprise. Ten years ago, very few people outside NWS were calling for large-scale NEXRAD upgrades. Now, with abundant Android and IOS apps & 5mbps+ mobile data the norm, demand is surging.
That said, I hope they DON'T squander a chunk of cash on public datafeed bandwidth instead of upgrading their backhaul. I'd rather see them do the back-end upgrades that only THEY can do, and tell people who want reliable & frequent upgrades to get their data feed through a private mirror service (like allisonhouse or caprockweather) who can upgrade their own backhaul as needed, instead of having to put in funding requests years in advance.
You have it wrong. The third-world products are mostly UN-encumbered. It's the products meant for sale in the US, Japan, and Europe that end up gimped and crippled into proprietary uselessness.
Correction: be able to stream local content, not be locked to Google, and be both rootable & reflashable so it won't end up in landfills like the Revue and original Nexus Q did.
That's why double-gang divided NEMA boxes and conduit exist. Build the plastic box and conduit into the furniture, ship it with both ends covered by screw-on plates, and leave it up to the end users to wire it as they please (a pair of outlets, a single-gang outlet plus a low-voltage keystone, or whatever).
With a divided double-gang box fed by separate 1/2" conduits, you can run just about anything. At the other end of one conduit, they could put a 15A 120v RV-type "inlet", like this one: http://inverterservicecenter.com/Marinco-150BBIW.RV (if you wanted to wire both gangs for power, instead of using one for ethernet/fiber/whatever, you'd pull out the box divider at the outlet end and feed both from the same conduit). Carlon ENT conduit is perfect for this purpose ( http://www.tnb.com/ps/endeca/index.cgi?a=nav&N=3819+601+3818 ). Worst-case, they could use Hubbell's funky JLOAD single-gang multimedia outlets, which pair a single 120v power outlet with a pair of low-voltage keystones, designed for use with a special box that shields the high-voltage power away from the low-voltage wiring. ( http://www.cesco.com/b2c/product/447768 )
That's part of the reason why FEC (Florida East Coast Railroad) has never been actively *hostile* towards passenger trains, compared to railroads like CSX(*). FEC runs mile-long trains hauling limestone at 60mph on glass-smooth welded-rail tracks that are maintained to higher standards than some stretches of track in the NEC itself. FEC's one non-negotiable mandate for voluntary passenger service on their tracks has always been that someone else had to pay to lay down a second track, maintain it to FEC's no-compromise high standards, and equip every train that runs on them with in-cab signaling and the kind of automation rarely found outside of Japan(**).
Once Amtrak, Florida, and a federal funding act or two cleared the way for the feds to pay most of the bulk cost of double-tracking FEC from Jacksonville to Miami, FEC announced that Amtrak was welcome with open arms (Amtrak itself is still trying to scrape up funding for the trainsets themselves, or come up with a good way to split & join NY-Florida trains in Jacksonville so half can proceed straight down the east coast to Miami, and the other half can run to Orlando & Tampa (historically, Amtrak has always resisted splitting/joining trains anywhere besides an endpoint).
(*)About 15 years ago, FDOT approached CSX with a request to double-track it from Auburndale to Tampa for Tampa-Miami passenger rail. CSX refused. FDOT offered to TRIPLE-track it... and CSX still refused. Exasperated, FDOT offered to elevate a ~12 mile segment running through Lakeland, and CSX told them that the only way they'd voluntarily allow it is if FDOT agreed to let CSX refund the purchase price and demolish it at will if it later decided that the support columns or track structure were in the way of whatever they felt like doing. That was the turning point when FDOT decided that any future rail route between Orlando and Tampa simply *had* to run along I-4 instead of CSX... CSX was impossible to deal with in any sane way, and taking the corridor via eminent domain would have ended up costing more than building it down the middle of I-4 instead (I-4 was planned for complete reconstruction over the next 10-20 years anyway, and FDOT owned a fairly wide corridor that was straight and flat, so they just designed the empty space into the new road and bridges so it would be there when the day came to build the new tracks).
(**)FEC is a HUGE proponent of cross-training and automation, and because it operates entirely within a single state, it can get away with telling its union to go to hell over things that would get CSX crucified. For example, FEC requires all engineers and conductors to be cross-trained and capable of serving either role as needed (sensible and efficient, but *vehemently* opposed by railroad unions because it means the conductor can operate the train while the engineer takes a break, instead of having to staff a second engineer while the conductor twiddles his thumbs). I believe it also requires engineer-conductors to have college degrees.
I wonder whether Larry got vaccinated for rabies a month or two before the laryngeal paralysis became apparent.
There's growing evidence that indiscriminate annual rabies vaccinations commonly performed on dogs & cats might be the root cause of the growing number of cases of laryngeal paralysis seen among pets in the US.
At the macro level, the evidence is pretty compelling:
* Laryngeal paralysis was practically unheard of in cats, and rare in dogs, until about 10 years ago... right around the time vaccines that use live modified rabies virus started to be used in cats as an alternative to adjuvanted rabies vaccines made from inactivated rabies virus.
* Laryngeal paralysis is usually accompanied by rear hindquarters lameness, and other symptoms that are basically mild versions of the same symptoms on a rabies screening checklist.
* More to the point, if you were to read about the symptoms of rabies, then describe a hypothetical variant that's mild, chronic, and non-communicable... you'd almost *precisely* describe the symptoms observed in cats with laryngeal paralysis.
The live-vs-adjuvanted distinction is important, and part of the reason why the theory is so controversial. Adjuvanted vaccines have a long, well-documented history of inducing vaccination-site sarcomas (cancer) in cats. A small number, to be sure... but enough to be unmistakable and visible (it's hard to argue that a vaccine didn't induce a tumor when one spontaneously appears in the *exact spot* where a vaccine was injected a few months earlier, even if the precise ingredient of the injection that caused it is unknown). The arrival of non-adjuvanted rabies vaccine was hailed as a major step forward, because it's presumably safer for cats... unless, of course, a small number of cats exposed to it end up developing chronic lab-induced rabies.
Making matters worse, the revenue vets get from those vaccinations is a major part of their annual income. Because most vets do, in fact, love cats and dogs and want to do what's right for them, this presents them with a terrible choice they're forced to make. Some simply ignore the problem, blindly cite legal compliance as an excuse, and wash their hands of moral responsibility. Others will now provide annual waivers for elderly and debilitated cats who would be risky to vaccinate at all (illness appears to be a MAJOR risk factor in post-vaccination LP), and will offer clients the option of titer-tests instead in an effort to roll the dice fewer times and reduce the likelihood of VAS or LP in that manner.
In any case, there's already strong (if anecdotal) evidence that live rabies vaccines are a major (if not THE) risk factor behind feline laryngeal paralysis, and it's not inconceivable that something similar might happen to humans who are vaccinated with live, modified rabies virus as PEP.
> I don't think Florida has figured out that one yet.
Whoops: http://abcnews.go.com/blogs/headlines/2012/09/drunken-man-on-horseback-leads-police-on-chase-across-florida-town/
Florida courts have even upheld DUI convictions for people riding a BIKE: http://www.jacksonvillecriminallawyerblog.com/2008/11/in_florida_you_can_get_a_dui_f.html
I believe that in California or New York, if the driver of a car is arrested for DUI, the police can actually force PASSENGERS who are under 18 (but not 18-20) to submit to blood alcohol testing too, arrest them for alcohol possession if they exceed some threshold, and suspend THEIR licenses if they refuse -- even though they themselves weren't the ones behind the wheel driving.
I can attest to airbag injuries. I was in an accident 12 years ago, wearing a seatbelt, and had the airbag deploy. It pushed one of my hands up through the windshield (leaving me with so many laceration wounds, the doctors had to spend 10 minutes debating whether or not I needed skin grafts, and to this day I have no feeling across probably 15% of the back of my hand due to nerve damage), and left me with second-degree burns on both forearms.
The culprit? Hyundai designed the 2001 Tiburon for European and Asian airbags, then put bigger airbags into its US models without bothering to notice or care that a deployed airbag would actually push out the windshield... made worse by the fact that they were *notorious* for triggering in low-speed crashes that were WELL below the threshold set by the NTSB. Without the airbag, I would have walked away from the accident without meaningful injury, and my car would have had about $2,000 worth of damage. Instead, my hand got shredded, and my car was a total loss because airbag deployments are too expensive to cost-effectively fix (it's cheaper for insurance companies to declare the car a loss, pay the owner, then sell the car to someone who'll export it to Haiti and repair it *there* with the airbags omitted or disabled, unless you're talking about a very new, very expensive car).
> Drive drunk, lose the car.
The problem is that in the US, 99% of cars aren't actually *owned* by the driver... they're secured assets owned by a lender. If a state attempted to impose such a law, every lender in the state would *instantly* require "DWI forfeiture insurance", and every driver in the state would instantly see his annual car insurance rates get a few hundred dollars tacked on to the premium to cover it.
> Heavily industrialized states like Texas and Florida
WTF, *FLORIDA*?!? Florida has no meaningful industry to speak of besides tourism and building construction arising from the state's own growth.
Are we talking about the same site? As badly as I want to feel sad, I can't even remember the last time that site had anything that seemed relevant, especially ever since Android arrived.
It's harsh, but as a practical matter, the "Linux Appliance" market is kind of in a state of both crisis and opportunity right now. Crisis, because anything x86 based with less computing power than a 2GHz+ Core2 can't really compete with a nameless Android USB-HDMI stick from Shenzhen... but those same nameless USB-HDMI Android sticks from Shenzhen are so cheap, you could practically connect them to port extenders, pot them in black epoxy, mount them inside an attractive case, and use them for things that would have been cost-prohibitive with a $599 slow Via system.
After reading about them, I can't help but wonder whether an interesting prosthetic compromise might be to somehow attach muscle fibers to implanted force-feedback strain gauges, then use the strain-gauge readings to control the hand itself (possibly in conjunction with all-electrical sensors used with other nerves to provide additional control. In other words, instead of pulling on bones, the muscles would pull on artificial ligaments cemented onto strain gauges attached to some kind of stretchy/springy plastic that itself is attached to a worm gear that moves the far point closer or farther from the muscle to alter the resistance.
For finer control (like individual fingers), it would take an idea from the way HTC's hybrid mechanical-capacitive switches used on their last few Windows Mobile and early Android phones worked. Basically, they used capacitive means to determine WHERE (approxiamtely) you were touching the phone, but used an actual switch triggered by a press anywhere in the region to determine when you actually intended to fire a switch event.
In a similar manner, the hand's controller could attempt to discern things like individual finger control by sensing the nerve bundle directly to come up with blunt motions, but sanity-check it in realtime against the muscle-triggered strain gauge, and use fuzzy logic to correlate patterns of nerve activity with specific variants of strain-gauge muscle tension to produce an intended action (so the actual finger, for instance, curled in direct response to the real-world forearm muscle pulling on the implanted strain gauge).
The problem with non-mechanical nerve interfaces is that they basically have the same kind of problems that capacitive touchscreens do... terrible signal to noise ratio and processing latency compared to real-world direct actuation. Nerves aren't like switch-triggered wires in a harness... they're more like a bundle of coax carrying multiplexed spread-spectrum QPSK signals with unbelievable amounts of background noise. By directly interfacing a few muscles with strain gauges, we can bypass the hundred (give or take) years of R&D it's going to take to get signal processing up to the point where it really needs to be, and just take advantage of the signal processing that the forearm muscles ALREADY HAVE to pick out the right signals and translate them into commands for the prosthetic hand.
Try dumping 2-4 packs of Sweet 'n Low into a diet Coke or diet Pepsi at a restaurant. Once the foam goes back down and it dissolves, you'll be amazed. It tastes almost exactly like regular Pepsi.
Diet Coke and Diet Pepsi weren't formulated to taste good, they were value-engineered to meet a price point. Back in the 80s, NutraSweet was still under patent, and quite expensive. So the used the least amount they could get away with and have it still be tolerable.
That's why Pepsi ONE came out in ~1999... the patent expired, so Pepsi could try to differentiate it at almost no extra cost to themselves by doubling the aspartame. The problem was, people who liked Pepsi ONE were much less tolerant of "diet drink" non-sweetness, and Pepsi ONE's improved taste went south REALLY fast as the aspartame broke down due to heat or age.They now blend aspartame and sucralose or AceK in the new limited-release Pepsi ONE to keep it halfway drinkable after the aspartame breaks down.
For a while, the bottling company for 2-liter bottles of Diet Mountain Dew in South Florida TOTALLY fsck'ed it up. It tasted fine in cans (presumably, from a different bottling company), but I threw away 2 liter bottle after 2-liter bottle for MONTHS before giving up on it and only buying cans for a while. The bad bottles first showed up at Wal Mart, then Target, then finally started to show up at Publix (Florida's main grocery store).
I suspect that the recent surge in 10-calorie diet drinks is partly due to the rising cost of corn due to ethanol demand... at this point, it's probably cheaper to replace the HFCS with a blend of aspartame and sucralose or AceK.
Is it just me, or does Coke Zero taste like what I vaguely remember Tab tasting like? I mean, seriously... I haven't had the misfortune of trying to drink Tab in literally decades (and that's coming from somebody who drinks NOTHING but Diet Mountain Dew, Pepsi One, and other diet drinks now), but the first time I tried Coke Zero... all I could think was, "Ewwwwww... Tab."
The big problem was that Coke rammed it down people's throats and took away the old version completely. Had New Coke been rolled out alongside Classic Coke, Classic Coke TODAY would be a niche item available in 12oz 6-packs, maybe 2-liter bottles at liquor stores and large grocery stores, and if you could buy it at a gas station or convenience store at all, it would be as individual cans or retro-looking bottles in the original cane sugar formula at double the price of regular Coke.
By abolishing Classic Coke overnight and turning it into American holy water that people guarded with their lives, Coke managed to shoot themselves in the foot. As others have said, if Microsoft made Metro an optional part of Win8, and made it a free voluntary download for Win7, some people might have actually liked it. Instead, they've gotten people to hate Windows 8, and induced hatred of their phones and tablets, too. At this point, Metro has become so tainted & toxic, Microsoft is going to end up having to reinvent their phone and tablet UI from scratch yet again just to de-stigmatize it and undo the damage done by desktop Metro.
Is there any way to replace Windows' own search with one that lets you search for straight-up filename matches with minimal ceremony or additional UI gymnastics? Preferably, one that skips the index (or at least, jumps to the drive itself after grabbing the easy low-hanging fruit), and goes straight for the bare NTFS data structures?
I'm *so* sick of having to fight with Win7's brain-damaged searchbox every ****ing time I go to search for a file whose name I already know (enter name, hit return and wait to start the pointless search, click "search everywhere", wait for it to get its act together, click 'advanced', select 'show hidden/system/non-indexed', click 'browse', navigate to c:, and about 20 seconds later, finally start the search I actually WANT).
It depends. MySQL's championship role has always been lightning-fast reads from tables that change rarely and don't need much in the way of concurrency protection. I believe it was also the first free database to have a concept of partitioning.
The main problem with MySQL lately (past 5 years or so) has been the fact that it's really easy to get an inflated impression of its more "hardcore" abilities, and not discover until it's too late that MySQL's ability to handle things like failover, replication, partitions, and materialized views(*) aren't quite as capable as someone who's used to Oracle & reads MySQL's docs might initially think they are. I fell into that trap myself ~3 years ago, and got bitten pretty *hard* by harsh reality not living up to first impressions.
(*)'it's been a couple of years, but from what I remember, MySQL had a huge 'gotcha' with materialized views that rendered them mostly useless for me at the time... I think it involved having a function in the select statement that was a LOT more restrictive than the rules Oracle imposed at the time. I vaguely remember it was something like "Oracle couldn't index a MV if you had a function in the select clause, but MySQL didn't allow you to have a function *anywhere* in the select, even if it was merely transforming one value directly and deterministically into another." -- or something to that effect. I know that the whole matter of function-based indices is a problem in general, but I just remember thinking, "Oh, cool... MySQL supports them now!", then discovering later that MySQL's rules were much, much more rigid about what was and wasn't allowed than Oracle's (and Oracle's were pretty anal to begin with).
Anybody care to guess how many days it would take Verizon to change its mind about FiOS if Google showed up at the next Mantoloking city council meeting & offered to deploy Google Fiber there if the city paid the direct costs of laying the fiber itself? Oh, and offered to pay for the lawyers the city would need when Verizon and Comcast fought back the only way modern American corporations seem to be capable of competing -- by using the courts to block it, instead of trying to outdo them by offering better and faster service...
The big problem with Comcast is that it's unreliable, and makes no attempt at being otherwise. When a Comcast service area has a region-wide commercial power outage, Comcast goes down, and stays down, until commercial power is restored. Sometimes, it doesn't even take a regional outage. Last year, Tropical Storm Isaac took down Comcast for several hours across Dade & Broward counties. The failure's apparent cause? A power outage at a SINGLE LOCATION along their backhaul route. They talk about customers with backup power, but it only works if your house is literally the only one on the block without power. The moment the blackout extends to your neighborhood, or any point along the backhaul route between your house and their network center, your internet service is gone.
Ten years ago, a fragile market compromise existed between "the company that used to be the phone company", and "the company that used to be the cable company". The phone company had DSL. It wasn't as fast as cable, but it was built to run on carrier-grade infrastructure and enjoyed most of its same robust backup power and reliability. The cable company had DOCSIS. It was faster than DSL, but nowhere near as reliable. So, consumers kind of had a choice... slower service that rarely went down, or faster service that had daily interruptions & got knocked down for the count whenever ANYTHING significant happened to the local power grid. Then... our wonderful regulators decided to allow the phone company to neglect its infrastructure and allow it to degrade to the non-reliability of cable, and now we have lots of places where there really ISN'T anything that approaches the former reliability and robustness of landline phone service.
Wireless might be fast to bring back up after a storm, but buried copper with battery power at the central office didn't used to go down at all. There were people who came home to neighborhoods wiped away by Hurricane Andrew, and were greeted by phones making "off hook" noises from under the rubble. It was the lucky consequence of having mostly buried telecommunications infrastructure that was built to survive a 20-megaton nuclear bomb falling within 2 miles of downtown Miami, and another pair of bombs falling within a mile or two of Homestead Air Force Base. Today, U-verse isn't quite as bad as Comcast (AT&T still owns generators that were purchased years ago, though the state no longer requires them to maintain or use them), but the sad truth is that our nation's telecommunications infrastructure today is a pale, ghetto-fabulous shadow of what it used to be. Reliability and availability can no longer be taken for granted at any level unless you own your own fiber and have direct control over its endpoints so you can provide your own backup power. Even services like Frame Relay that were once protected by ironclad SLAs and stiff, automatic penalties with teeth, are now at risk of carriers just looking the other way and making a business decision to pay the fines and let the network go down because it's cheaper to take a hit for a few days now than to invest the capital into five-nines or better uptime.
Once again, it comes down to a question of scope and scale. Thousands of balkanized islands of photo and video data that are mostly offline are of little concern. The act of making that aggregate bulk data trivially searchable by anyone, anywhere, fundamentally changes its nature. Bars and stores might archive lots of data, but they keep it to themselves. When it comes to maintaining personal privacy, high barriers to entry, roadblocks to use, and restricted availability of personal data are a GOOD thing. Back when you had to go to the courthouse and fill out request forms to look at court documents, nobody cared about redacting social security numbers and similar private information from exhibits and evidence. The moment it became possible to automatically scrape court websites and do automated bulk OCR on the documents obtained from them, it became a HUGE problem.
As far as your condo complex goes, you might want to talk to your association's lawyer before holding onto the license plate data for TOO long, unless you want to start getting hit by subpoena after subpoena from divorcing couples doing discovery on each other at your association's expense. The cost of complying with those subpoenas can become expensive, and your ability to recoup those compliance costs by charging for the information might be limited or nonexistent. That's why most condo complexes intentionally destroy and discard such information after a while -- after a month or two, that information becomes a potential liability to the association.
Consumer Reports bases their opinions on values that don't necessarily align with yours. For example, they regard pet health insurance as being a bad deal. If you're somebody who'll euthanize a pet at the first hint of an expensive medical condition, they're probably right. If you'd give one of your own kidneys to your cat or dog to save his life if the vet told you it was his only hope for survival, they're absolutely wrong. Most people fall somewhere in between. If you pay $50/month for 10 years, and your pet dies from some accident without warning, the insurance company wins. If your 2 year old cat or dog gets cancer, the insurance company loses. If your 15 year old cat or dog develops some chronic illness that can be kept in check for years with aggressive care at the first hint of a flare-up (pancreatitis, asthma, HCM, etc), the insurance company loses *spectacularly* after briefly thinking it won... especially if he or she lives long enough to get cancer. The insurance companies partly keep their losses in check with somewhat high per-incident deductibles. If your kitty is running up $500-900/month medical bills month after month, even 20% deductibles are going to get really painful and weed out all but the most dedicated pet parents.
Ditto, for things like mobile phone insurance. If you're someone who buys dumbphones, or phones that are free/1-cent and would cost $200 or less if you bought them at full price, it's a bad deal. If you just bought a brand new Galaxy S4 with unsubsidized replacement cost of $600 or more & MINIMUM $200-300 charge for a cracked screen repair, you'd have to be positively insane to *not* buy the insurance. The key is knowing when it makes sense to discontinue it. Buy the insurance for the first 12-14 months, then drop it once the cost of buying a used replacement on eBay is less than double the deductible, and you'll come out ahead. Keep paying monthly premiums on a 2-3 year old phone, or spend more on the deductible than a used replacement on eBay would cost, and the insurance company wins. The main way phone insurance companies come out ahead is by customers continuing their policies LONG after the phone has ceased to be expensive enough to justify it. If you were required to insure the phone for 2 full years, phone insurance might be a bad deal. But since you can drop it at any time as of your next bill, it makes sense to get it with almost any new expensive phone & keep it for at least a few months.
Extended warranties are a tough call. 10+ years ago, stuff that didn't die within the first 3 months probably wouldn't spontaneously break on its own during years 3-5 anyway. Now, it's a total crapshoot. The $900 LCD TV you just bought might have a replacement cost of $600 in 3 years, but chances are that its warranty is only good for 1 year... and if the TV breaks out of warranty, you're probably fucked since modern LCD TVs are practically large-scale integrated circuits with almost nothing inside that can be cost-effectively repaired by an independent repairman. And unlike 10+ years ago, things like LCD TVs really DO spontaneously drop dead after 2-4 years for no obvious reason.
For cars, extended warranties are an equal crapshoot. If you keep the car long enough for the warranty to kick in and matter, it's basically a prepaid service plan that you'll probably break even on & avoid getting hit with a $1,000+ repair bill out of the blue long after the original warranty expired... as long as you don't get into a wreck before that point, and have the car totaled. The "car totaled after paying premiums for months or years, but before any expensive repairs" is really how they come out ahead, because almost ANY car that's more than 5 years old is eventually going to need a repair that's expensive enough to break even or come out ahead with the insurance. At the end of the day, enough cars get totaled for the insurance company to come out ahead, even after they've paid to replace the AC compressors, power steering pumps, alternators, and half the automatic trans
Yes, but the surveillance camera footage from those businesses doesn't get archived forever and published to the internet in a way that allows strangers to upload a scanned photo of your face and search through yottabytes of video footage for segments where your face was autodetected.
Most people have no real issues with being recorded by businesses, because real-world storage isn't free, and the overwhelming majority of that video footage gets autodeleted within hours, days, weeks, or at worst, months, unless something noteworthy happens and it's explicitly set aside for longer retention.
If Google archived Glasses video footage on the server for 72 hours absent a court order demanding longer retention, the only way to view it was on a device with DRM-enforcing videocard that protected it at least as well as Blu-Ray content (or Glasses that enforced similar end-to-end transport-layer encryption), Google robustly watermarked it & threatened site owners who allowed content with that watermark to exist on their servers (say, photographed off the monitor screen) with eternal banishment from Google search results unless the site owner actively scanned files for the watermark and deleted them), and its long-term searchability was extremely limited, people might not mind it as much. But when when you get into things like "eternal retention" and "casual, anonymous searchability via geotagged face-recognition", it absolutely DOES change the very nature of the video and the consequences of its existence.
> Just because he broke into your house doesn't mean you can shoot him.
In Florida, for all intents and purposes, it does. http://en.wikipedia.org/wiki/Stand_your_ground
If somebody breaks into your home in Florida, you can empty a semi-automatic rifle into his back as he runs to the door, and you're unlikely to face charges once investigators have established that 1) he was a bona-fide intruder and 2) you had even the slightest excuse to believe he was endangering you (like, "I feared he had a weapon left outside and was running out to retrieve it").
The main thing that makes Florida's law unique is that it extends beyond your home to all lawfully-occupied locations. So, if somebody pulls a knife on you and makes you give them your wallet, you could quite legitimately empty your gun into their back while they run away with it & claim they threatened to come back and injure you if they heard you call for help. Things would get messy if your own possession of the gun wasn't entirely lawful, or there was evidence that you yourself were somehow involved with an illegal act involving the dead alleged assailant (like a drug deal), but if the police investigation determines that you were the victim of a straightforward mugging that involved threats of violence or the use/brandishing of a weapon by the mugger & your own possession of the gun was lawful, that's pretty much the end of it. None of this is really *new* -- the main thing Florida's law did was codify the de-facto case-law norm into statute law, and provide automatic administrative procedures to nip prosecution in the bud before it ever got to the point of a real trial.
You're forgetting something important: the value to such a company of stopping people from producing food on their own is basically zero. OK, a little more than zero, but it's basically lost in the the margin of error. For every survivalist family of doomsday preppers growing their own organic food off-grid on a farm in Idaho, there are ~24,000 families that haven't eaten food that didn't come from a restaurant or a microwave oven within the past week, and a few thousand more whose idea of baking bread from scratch consists of tearing open a box of Krusteaz & dumping it into the bread machine with a cup of water before going to bed.
The sum total annual output of every organic farm in America would be hard-pressed to supply a few expensive restaurants in New York and Los Angeles. Have you SEEN the price of a flawless, unblemished organic tomato at a grocery store? The price is off the scale, because that one randomly-lucky bug-free tomato that ended up sealed in wax & cellophane at the grocery store had a few hundred neighbors that either got thrown away or made into spaghetti sauce because nobody would have bought them.
> This means indigenous American vegetation is not dependent on honeybees for fertilization.
And what percent of the food in an average American shopping cart is actually derived from indigenous American vegetation? You know, all those alien foods from places like Europe & Asia that we eat here... lettuce, tomatoes, potatoes, grapes, wheat, oats, rice, etc.
Man does not live by ethanol, high fructose corn syrup, and nacho chips alone, even if it IS possible to make it through a Saturday picnic consuming little else besides beef ;-)
Congress needs to pass a law subjecting anyone who facilitates automated face recognition to joint and several liability for any torts that might arise from its use.
Nothing, and I mean NOTHING, will stop corporations like Google and Facebook in their tracks faster than making them jointly & severally liable for any torts ("tort" == "something you can be sued for under Anglo-American Common Law") that might arise from the use of their face-recognition data.
All the EULAs and disclaimers in the world won't save them from a lawsuit where a jury awards the plaintiff (say, a 19 year old college student who got tagged as a 'slut' by a fellow student's crowdsourced webapp built atop a hypothetical service like "Google Face Search") a million dollars, determines that the student is 99.999% at fault, and that Google is 0.001% at fault, because unless the student is a multi-millionaire, Google would end up on the hook for more or less the entire amount (that's what Joint & Several liability means).
Remember, under J&S Liability, it DOESN'T MATTER if the tortfeasor is violating Google's own TOS. Well, it matters to the extent that the overwhelming allocation of guilt would go to the guy who violated Google's TOS, but when he ends up being too poor to pay, the entire remainder would get dropped in Google's lap. If Google tried to spin off a subsidiary to run the service to protect themselves from such a lawsuit, that subsidiary would be bankrupted, put out of business, and shut down by the first lawsuit within weeks or months... and if Google kept re-spawning new entities that tried to inherit data from the previous one, eventually some state would allow a court to pierce the corporate veil and go straight for Google itself.
The net result is that any company that engages in face recognition would be forced, by threat of financial extermination, to guard the data with its corporate life and treat it like PII of the most sensitive kind. Small companies that went overboard would get sued into oblivion by victims before they had a chance to do much damage, and large companies' would be prohibited by their own shareholders from even *thinking* about making use of face recognition data, simply because they'd be put out of business if the data ever leaked out.
The best part about J&S liability is that it makes additional government regulation largely unnecessary. With the sword of J&S liability hanging over its head, Wall Street would swiftly stop face recognition in its tracks by any organization large enough to be a real threat to people's privacy.