both the previous day crew that survived and the ones that died diagnosed properly the problem: something is fucking moving the control surfaces. What neither did, was understanding the underlying cause was MCAS command.
Hey I diagnosed the problem: the plane fell out of the air. Future crews should keep the plane in the air. I am very smart.
The Lion Air pilots couldn't figure out why the plane was fucking around on them, and a pilot riding along diagnosed the issue.
The next day, a Lion Air plane crashed due to the same issue.
An indicator that suggests a particular issue which looks like other issues but isn't would tell the pilots immediately that this particular issue is happening, thus they could take immediate action instead of flailing around confused.
Glycerin and propylene glycol are irritants. They're bigger irritants than tobacco smoke.
They're also well-known low-toxicity. They don't cause the damage that smoke causes. These compounds can absorb through the lungs into the blood without producing cytotoxicity in the doses consumed.
PG is actually less-toxic than VG, but VG is less-irritating. VG is natural, and PG is an organic compound.
I think the health effects are pretty well expected to be less-damaging than the health effects of huffing charred plant aerosol.
Be that as it may, why the fuck would you put the plane and crew back in the air the next day without an investigation? As soon as this irregularity occurred, the plane and crew should have been grounded. Determining the cause, the response, and the obvious (not likely, but observable without projecting other possibilities) outcome without the irregular situation which corrected the issue, an immediate action would become visible.
That action may be preventative (fix the problem) or contingency (ensure all pilots have training for familiarity with the situation and its appropriate response).
When an irregularity occurs as such, you immediately don't know that the individual plane is unsafe. You may discover by investigation that the entire fleet is unsafe; but you don't know if the plane or its pilots are unique to the failure. You ground that whole damned set of factors, examine the situation, quickly identify the contributing parameters, and decide if you need to ground all planes, issue new training, or just deal with the specific set involved.
That's just basic risk management. It has nothing to do with airplanes.
And at the resonant frequencies of the various physical structures, or at sufficiently close subharmonics thereof, at a microscopic level, non-ionizing radiation can also cause enough flexing and stretching to break the covalent bonds that hold together strands of DNA. That was what the study on THz EM concluded.
Oh for...really?
At subharmonics, much of the energy is lost: a super-harmonic (e.g. 2x) will oscillate back and forth because the wave amplitude is about 0 at the end, so approximately 0 destructive interference occurs as the vibration reflects back. Basically your wave propagates and lifts the material in one direction, then snaps back in the other direction; at a harmonic, the reflecting wave is pulling along in the reverse direction as it went in, helping to accelerate the process in sync, causing amplification.
At subharmonics, this effect is diminished greatly, and the subharmonic eventually reverses. Likewise, because there is not a 0 amplitude at the end, some energy is lost (when you bend something upwards in the middle, it's pulled down by both ends until each end accelerates upwards to release the energy; but if there's nothing at the end, there's nothing there to anchor against, and you start generating odd-order harmonics and destructive interference). Subharmonics of a resonator are longer than the resonator itself, so the resonator is exposed to part of the energy: at 1nm, a 2^20 subharmonic (about 1mm) would expose 1/1000000 of the energy to the resonator, unless you can direct it into it (which is hard). It's why we need to actually coil 2 meters of wire into a 2/(2^n) meter antenna and then add a really strong amplifier to pick up 2 meter wave.
All of this becomes an interesting engineering problem until you realize you're dealing with DNA.
DNA is floppy and doesn't resonate well. It's the type of material we use to dampen resonant behavior. In other words: DNA isn't subject to resonance. Yes, DNA is one giant molecule and its individual components can't resonate independently because they will transfer the vibrational energy to the rest of the molecule.
Longer, lower-frequency waves have less energy, too.
So you're looking at THz band. Red light is 430THz. UVB is 1,000THz.
You might notice that red, green, and blue light don't cause skin cancer. UVA doesn't cause skin cancer. UVB causes DNA damage and skin cancer. The reasons for this are well-studied.
Part of the reason is just that there's so little energy. A big part is that the energy just goes into heating, and you're not microwaving your DNA; you're microwaving the entire mass in which your DNA sits.
Now that doesn't mean there's no risk. I did say energy. Think about what happens if you stick your face in a 4,000 watt microwave oven. One little photon of UVB can cause DNA damage, but one little microwave photon can't; yet if we pour enormous amounts of them together, we can kill you in seconds.
That's why we have wattage restrictions on transmitters used at ground level, in handheld devices, and so forth. Anything below a certain frequency is going to be harmless below a certain wattage; anything above that is going to start doing single-photon damage to your DNA.
Think about it for a minute. What would it be like if none of that were valid? Well, wind at 10mph doesn't kill you, and wind at 12mph doesn't kill you; but what about 11.7793mph? Only one way to find out!
In other words: we've already done testing. We even did testing on rats with cell phone radiation cranked up to ridiculous levels--which, on one hand, did cause some problems; but on the other, they used Sprague-Dawley rats for lifetime toxicity studies, which is bad methodology. I'm not surprised ridiculously-high-powered radiation at any frequency is harmful: you wouldn't stick your face in a microwave oven, unless you were completely stupid.
Looks like Luxottica had net units sold of 89 million in 2017, with $10,350 million of net revenue, an average $116 per unit sold.
Their cost of goods sold was $3,710 million, so $42 of cost of goods per unit sale.
Their selling and general administration expense was $5,170 million.
By comparison, Apple has $15.26 billion of general administration. Luxottica lists zero research and development, while Apple lists $11.58 billion.
Apple's unit sale is much, much larger than Luxottica's, so their general administration costs less of a percentage of general revenue per unit sold, of course. Apple also has higher net operating profit margin.
Zenni is using a lab in China, whereas Lensabl seems to use a lab in Los Angeles. Many of the manufacturers out there use New York, California, and other local labs. That doesn't mean Zenni can't make a quality product, although Zenni does have limited options compared to Zeiss or Lensabl's own labs--and the fixed costs for a more-complete lab are higher. Notably, Lensabl can provide a really high-end anti-reflective coating, while Zenni can provide a comparable but not-as-good (that means it's worth buying and is a real upgrade from basic, but it's not a Porsche) coating for about $20. Similarly, they can provide a greater range of Transitions lens coatings, and high-end digitally-modeled lenses shaped to remove visual aberration based on your eye shape as modeled from your prescription (this actually works, which is strange but well...I'm well-aware technology is frequently surprising).
Zenni has lower volume, less marketing, and a reduced need for complex logistics (logistics hickips are disastrous in a high-volume operation, so you can pretty much run out of your basement for sufficiently-low volume).
Higher automation or outsourcing to Mexico or China may be a good direction for the lens industry. Better business processes to scale the logistics up to huge and diverse supply chains and catalogs also could help Luxottica, Lensabl, and Zeiss get those costs down. I'll tell you one thing, though: those glasses do NOT costs $350 when they come out of Luxottica; they get there through the supply chain.
15.8% is still a rather high profit margin, and they are taking a significant cut. It's not much per unit, but sheer volume has a big economic effect.
Perhaps, but it's more like saying something that's 10C won't harm you, your body is about 36C, and something that's 70C will burn your skin; and someone says temperatures of 25C may be harmful.
Those mechanical effects are heat. The volume of particle exposure becomes the defining characteristic. Ionizing radiation can cause damage by individual particles (a collision can remove an electron from a molecule, thus breaking a molecular bond), while non-ionizing radiation can cause warming and thus can cause burns via a large volume of particles.
An extremely-bright beam of colored light, for example, can burn your arm off. Red, green, yellow, or blue, with enough wattage, it'll cut straight through.
The risk of microwave exposure is similar to the risk of getting near a hot object. We can build microwave generators to generate huge amounts of power in a narrow spectrum, so the effect is a bit more spectacular than looking inside your oven: we can basically divert the output of a blowtorch straight into your face. You'd notice.
it seems as if it'll only amount to a thought experiment while the political sphere is still flush with cash from corporate donations to super PACs.
Kind of irrelevant. The SuperPACs can affect who gets elected, but don't really affect their decisions. The politicians basically come set, and aren't easily bought; it's the media that leads to election results themselves which you can buy.
Current politicians have a will to solve these things, but not a will to move against their own fear of loss. We can't solve that by paying them; what we can do is dump money into funding a competing politician already interested in making such changes. Consider that the limit on what money can buy and you start to see opportunities you'd previously discounted.
Don't know what subscribestar is. I'm the sort that would crowdfund something on Kickstarter and not Indiegogo because the split between multiple platforms is irritating and shows different, non-composited numbers to backers (i.e. dishonest). That's also why I don't have a GoFundMe.
Above a certain frequency, the waves are small enough to penetrate and hit DNA.
Below a certain frequency, those waves have macro-effects. They cause molecules to wobble and heat up. Infra-red is the highest-energy known for this in particular, as it's just below visible light in terms of energy (visible light is shorter wavelength).
What? You didn't know? Light and EM are the same thing. They're both radio waves. Your wifi antenna does the same thing as a light bulb; so does an X-ray machine.
why would anyone assume there is a relationship between cost to produce something and the cost we have to pay?
Because there is.
I'm curious about if the prices above are gross or net costs. Basically, what's the profit margin?
Luxottica's net profit margin is about 15.8%, meaning that for every $1 of revenue they keep about 16 cents of profit. This is similar to Apple, Microsoft, and Google (20%-22%), and dissimilar from Comcast (11%-13%, depending on year), Adidas (8%), and Walmart (3%). Large pharmaceutical companies pull fluctuating profits generally holding around 12% for a 5-year average.
I generally describe 8% as a fair net profit margin. It's somewhat arbitrary, but viable, and is derived from common numbers. Farmers believe their correct profit margin is 20%, and agricultural business literature often states that farmers should draw a 20% profit margin but tend to struggle to hold 10%--a similarly-arbitrary statement, and one that forms the foundation for agricultural industry welfare (not that welfare under our food source is a bad thing). Generally, modern regulations actually hold the profit margin of farmers down, rather than subsidizing it up (the farm subsidies basically set a price limit and offset it with a government payment).
We can surmise, then, that the 15.8% net profit margin is not as egregious as considered--a $350 pair of glasses would be roughly $302 of net costs and $50 of revenues if sold direct from a single company, although the reality is a lot of small profit margins along the way--yet it is quite egregious compared to standard 8%-10% profit margins in durable goods industries, and even exceeds the 12% margins of some industries with more control of the market consumer.
In short: the situation is not as bad as presented; but it does demonstrate market power abuse.
The main time I see this happening is when some game releases DLC. Rather than selectively install the DLC they pack it into their data files and inflict the download and footprint cost on everyone whether they want it or not.
In a similar sense, I've frequently suggested they should profile or self-profile games and stream content.
Think about something like Breath of the Wild, 13 gigabytes. Do you need 13 gigabytes to play the opening scene?
When you start the game, the very first assets you need are identifiable. You can profile the loading screens and such, or you can speculatively identify assets by predicting what the next screen will load based on where menu entries go etc. and what assets (and code!) they call up.
So as soon as you turn it on, you have a list of things you need to get to a new game.
The same is true of starting a new game: you know what scene it calls into, and can download that. You can inspect the scene and see what assets it calls, and download those. You can speculatively-render: don't rasterize, but call out what assets would be used in the render, and identify what is visible and what is in the local scene but not visible. You can look for sector changes (doors) and scene changes (transitioning doors). You can look for events and movies.
You can project ahead and identify what you're going to need. Then, if you encounter something not loaded, you can pause and download it.
In development, you can profile this: you can speculatively load (with all assets available) and then have the profiler catch anything that was loaded but not used (load last) or used but not loaded (add to the forced speculation at this point). Developers can tweak the speculation to improve its base functionality.
Much of this already exists: the game loads up everything it needs into memory as you enter a scene (preloading), rather than streaming it off disk as it comes into the render view. We're mainly talking about leveraging that, but with a little look-ahead as to where you could end up immediately (what's the next room?).
You're coming within range of several shrines and dungeons. Grab their base assets.
You're getting closer to a particular shrine entrance. Prioritize its assets. Move those to the front of the queue.
You passed it and are now closer to some other entrance. Change the queue, download those assets instead.
Imagine: you buy the game and you're playing it 12 seconds later. It's going to take 18 hours to download, but you're getting 21% through it in the next 10 hours.
How do you diff, say, a megatexture atlas which you've tweaked some of the dimensions to remove an unused image and repack the rest?
You use the original file as a dictionary and designate the locations of each texture, with procedural instructions to reconstruct the output.
(i.e. changing the compression on the textures to improve performance or avoid a licensing cost, which means changing the code, plus all of the texture atlases, plus re-optimising/recompressing everything)
Now that one you need to reissue the files. You could theoretically write a deterministic decompress-compress procedure, though.
If it's encrypted and the user at the end can't decrypt it, it can't be used, and the user at the end doesn't have to have it.
For example: every DVD player has the capacity to decrypt a DVD. If you decap the chips or otherwise get into the player, you can retrieve the key. The software decrypts the video as it plays.
So you can sign all the data (encrypted or decrypted) with your private key, then use the session key to decrypt, patch, re-encrypt, and verify that the result is the same as the signature presented.
But getting there is extremely difficult -- those two parties have a vested interest in maintaining this system, and will resist any change that undermines it.
No they won't. Well, the Republicans will.
The Democrats in power in Baltimore and in Maryland seem to resist change of which they are uncertain. Undermining their party power base is not a large concern; political fall-out--that the change has unintended consequences which make their voters angry--has heavier weight. They're all loss-averse and vulnerable to other typical human cognitive flaws, so they hyperfocus on such things.
Where I live we have switched to ranked choice voting at the municipal level, but we're also a town controlled by Democrats, so switching didn't imperil party control of the elected offices, and only boiled down to making party insiders' choices slightly more at risk.
The Instant Runoff Voting switch actually leaves the election vulnerable to some manipulation. Tideman's Alternative isn't.
Generally, if your party has more than 50% of the voting base, it has control. That doesn't mean the same candidate has control.
I often use an example for Unified Majority with 40% Hard Right, 29% Moderate Conservative, and 31% Liberal because the 69% always vote in a republican, but each group impacts the outcome. Under UM and Tideman's Alternative, they elect a Moderate Republican; under IRV, the Moderate Republican is eliminated and the Hard Right Republican is elected. Likewise, removing the Hard Right under UM gets you a Moderate Liberal; removing the Moderate Conservatives gets you a Hard Right Republican; and removing the Liberals gets you a Hard Right Republican.
As you can see, every vote is factored into the final outcome of Tideman's Alternative. Unified Majority uses an electoral structure to ensure the final candidate set represents the span of voters so that Tideman's Alternative has something with which to work.
Even so, if more than half your base is behind a single candidate, that candidate wins. If more than half your base is behind a party, that party wins--albeit the particular candidate may be softer than if you had a party primary.
Switching was aided by a fairly long-term trend of rising progressives who both supported the change and were wresting control.
It's more than RCV has become a meme and has politically less risk, so the perceived losses for not switching are large. That's how politics operates these days. It's what Trump has refined to its absolute conclusion.
My complex analysis is not something you see in the political sphere. The political sphere is about a brand. It's about $15/hr, whenever that gets here, and to hell with inflation and productivity gains and a fair wage. $15 in 2005 or $15 in 2025, as long as it's $15.
Mostly I don't think there's any defense against this.
I designed a defense specifically against this.
Let's say you have two political philosophies: Liberal and Conservative. They're diametrically-opposed, and adherents carry them to various degrees.
In a one-vote system or an approval system, you get two parties. Ranked systems can also raise two parties, but there's less need. The two-plus system appears because of damage in these systems: the candidate with the strongest favor--not the majority favor--wins.
With majority-runoff and instant runoff voting, this occurs in early rounds and runs down to a later round, so candidates with the strongest non-majority favor advance while strong consensus candidates are eliminated early. I've actually published crude methods to manipulate IRV (the most common ranked choice voting method) to precision-select a candidate by adding a third candidate and campaigning to eliminate the original winner in the penultimate round, selecting the original loser (the third candidate can't beat the loser, but can make the winner the lowest vote-getter in the three-way round).
Okay, that sounds like we could change to a Condorcet system; but that's not quite enough.
A Condorcet system eliminates the IRV/Plurality problem. Condorcet systems are frequently manipulable; Tideman's Alternative is almost-certainly non-manipulable (its resistance in practice is near 100% because you need more votes to manipulate it under any failure mode than you do to legitimately win).
Party primaries, however, provide a huge problem.
Because of the aforementioned liberal-conservative divide into two parties, we have what amounts to a top-two system with severe distortions. Top-two is itself horribly broken (it has the plurality flaw with a nonpartisan blanket primary on plurality, and by IRV/STV it nominates two wildly-biased candidates not fitting the overall consensus).
The most-severe distortion is the primary itself. Base voters--people who vote straight party--don't have any marginal utility for primary voting. That means activist voters--people well away from the consensus and moderate positions--drive the primary.
With 40% hard-right, 29% moderate-right, and 31% liberal voters, you'd think that the 40% hard-right would drive the Republican primary. The problem is with 20% hard-right, 49% moderate-right, and 31% liberal, you'll probably actually see 15% hard-right and 10% moderate-right voters get out in the primary. Obviously, the 69% of Republican voters carry the General Election.
If that sounds fanciful, think about this: a propaganda campaign can excite your activists and move more people to the activist camp, getting a bigger turn-out for your polarized candidate in the primary.
There's your exploit.
Use the defect of the party primary to select an extreme.
In the general election, you just have to lift base voters and tilt a minority of swing voters. That's all. You do that, you've hand-selected a candidate.
So what's the weakness of this exploit?
Primary Weakness: it requires manipulation of a small subset of the population.
It's a weak attack. It's fragile. It can't succeed if it needs a major shift.
First, let's add another consideration about Ranked Choice Voting: people reliably rank to six candidates; the number of ballots ranking seven or more candidates is much smaller than the number ranking six.
Unified Majority replaces the Party Primary with a nonpartisan blanket primary run via STV. For single-seat elections, Unified Majority no
Fifth amendment stuff should probably apply here, although I don't see what law they can possibly use to compel a witness in an investigation not related to any law which that witness may have broken.
As for the two parties, one is the Liberal philosophy and one is the Conservative philosophy. They're polar opposites, with degrees of how strongly people push those philosophies. That's the real difference, and it's important. Legislators and administrators will take the country in different directions depending on which philosophical ideal underlies their party, even if you have things like unmitigated corruption.
To a degree, government structures can mitigate corruption; not much else can. Municipal governments, in particular, can use the council-manager structure to ensure a highly-responsive government with minimal corruption. More-representative bicameral government structures with highly-representative and manipulation-resistant electoral systems can also mitigate corruption as a secondary effect.
Under the Model City Charter, the City Council appoints and may dismiss at any time a City Manager to perform all administrative functions; the Council has no direct administrative power. This contrasts with a Council-Strong Mayor form, wherein the Mayor is elected, powerful, and can freely run the city while hindering Council severely.
The Mayor can easily appoint all their friends and donors to high-power, lucrative positions; whereas a City Manager is not an elected official, has little reason to do that, and can be thrown out by City Council. City Council needs to appoint the City Manager, so they sort of share influence and, thus, will have severe conflicts about which of their friends and donors should be appointed where, creating a stopgap for such corruption.
Unicameral City Councils can operate with a mixed election, such as 14 Districts each electing one Council member, plus 7 more at-large by proportional vote. This creates a modal difference between members and tends to hinder corruption as well.
Bicameral structures, on the other hand, can operate by districts of Condorcet single representatives (Senate) and proportional multi-representatives (House). This ensures that Senators represent the overall consensus, which severely hinders power-grab politics focusing on small groups and majority power; while House Representatives are fit to the largest cohesive subgroups in the voting population, creating more-extreme voices while essentially making competition between candidates unappealing to their own base moot (i.e. a Republican and a Democrat in a mixed district have basically no reason to campaign against each other, and need to focus heavily on beating the candidates similar to themselves).
By using a system of Unified Majority, we eliminate tampering by strategic nomination, party primary manipulation, and propaganda attack. Unified Majority uses nonpartisan blanket primaries held by STV. For single-seats, it nominates 5 (or 7 or 9) Candidates from which we elect via the non-manipulable Condorcet system, Tideman's Alternative. For multi-seats, it nominates between 2n and 5n, from which we elect (n) candidates via a second round of Single Transferable Vote. STV itself is highly tamper-resistant.
Unified Majority ensures your nominated candidates represent the span of the voters. Baltimore City is 7% Republican and like 0.5% Green; you're not going to elect a Green or Republican mayor, and Unified Majority gives you a span of various Democrats--unless, of course, a Green or Republican runs such a strong campaign that they have a real shot at winning the general election (that's happened, and they won by being more than 50% of voters's first choice, so they would have been nominated under Unified Majority's nonpartisan blanket primary).
Party Primary gives Baltimore City one Democrat for 92% of the voters and one Republican for 7% of the voters, which is not choice. Under Unified Majority, you'd need
Thing is a self-driving car might get confused, swerve off the road, and cause some minor property damage. A human in the same situation might get confused, lock up the brakes, and either fail to steer entirely or steer into opposing traffic.
Often we see the severity of each incident lower than a likely comparable scenario with a human. Think more property damage, less severe property damage and loss of life and limb. Gives you a huge dilemma.
we are marching towards no accountability with e-voting machines that have proven to be insecure as hell.
Current systems are broken. It's possible to establish something more-secure than paper (a lot more secure); and the primary attacker isn't hackers (we can trivially exclude hackers), but the electoral authority (extremely difficult to exclude). It's a big effort, though.
Sadly, I worry more about the selection of idiots we're putting on the ballot than the voting system itself
It's up to the people to decide; and that gets manipulated by the broken electoral systems we use, notably party primaries and single-vote systems. IRV is also easy to manipulate (it's a lot more effort and strategy than plurality, but I've developed the exploit to a fair degree).
Single Transferable Vote (multiple-winner) and Tideman's Alternative (single-winner) are near-impossible to manipulate, even with social media propaganda attacks. I've used those to assemble a highly-representative, manipulation-resistant electoral system called Unified Majority.
I'm going to write two books, one about sabotaging elections and one about securing elections. Maybe then, after it becomes clear any group of half-stoned college kids can hijack an election just by attacking the voting rule, people will get it through their heads.
question Bruce's logic here, since he's not exactly ignorant of either technology or law.
That happens once in a while. He's one of the foremost information security gurus in the world, but he still backs paper ballots as if they're magic. He won't even have a conversation about systemic problems in proving paper ballot integrity, in the lack of auditability of paper ballots, and so forth.
When such experts speak, people follow. That gets us HR1, the Black Box Voting Act, which threatens to cripple our democracy by expanding vote-by-mail not just to those with disabilities and other accessibility issues, but to everybody.
Imagine you go to vote. They use paper ballots.
You walk in, they have a box that was put there the previous night. It's black, sealed, and nobody can look inside. The box is also hidden behind a curtain.
You fill out a ballot, go behind the curtain, cast a vote.
Then they take the box out of the room and come back three weeks later with a different box filled with ballots that definitely came out of that box and haven't been altered. They count those.
That's vote-by-mail.
If 20% (a lot!) of votes by mail are tampered and 0.1% of votes are cast by mail, then 1 in 5,000 votes are tampered. If 50% of votes are cast by mail, that's 1 in 10. It matters.
Meanwhile, the paper ballots are the record, and they're transferred by a chain-of-custody, held in secure trust, etc.. If you open the ballots up to recount, whatever they say is the correct record.
That's not an audit trail. You don't have record of the contents of the ballot box; only that a ballot box with contents was handed off. We can duplicate or reverse seals readily enough.
To really provide an audit trail, you need data. Records that prove a set of ballots came from a particular polling center. Those records need to be produced in a manner in which the public can verify the election is not tamperable right up to the point of producing those records, and then can record those records themselves.
For plurality, you just use vote counts, as you can sum them; but plurality is hackable by adding candidates (strategic nomination).
For ranked ballots, the amount of data grows factorially with the number of candidates. It's impossible to prove ranked ballots without a computer involved. That means you need universal verifiability for electronic voting--this is a harder problem, but one I've explored; nothing today achieves this, although it's trivial to achieve.
HR1 of course outright prohibits the use of electronic records as an official count against the paper record. It doesn't make an exception when using systems of established universal verifiability.
So we've got a bill for establishing the tyranny of black box voting and insecure elections. Fucking great.
Firstly the EMI envelope within a PC is controlled and far lower than what you compare it to. Short lengths of very low current very low voltage signals at high frequencies radiate but do so poorly.
The PC envelope is specifically designed to be an EMI shield due to EMI generated by the PC. It's an FCC compliance point.
And that is not remotely true. The data lines are twisted to prevent radiation and have been for a long time.
Twisting the data lines causes them to self-shield against near-end cross-talk (for round-trip pairs e.g. Ethernet, they'll have opposing magnetic fields which self-cancel), and also causes LVDS pairs to remain at the same base voltage when acting as antenna (these pairs don't self-shield against NEXT). It doesn't prevent them from radiating outward in an LVDS setup.
and in fact USB 3.0 all things being equal would be less likely to cause external interference than USB 2.0 based on signalling alone.
With the HDD connected, the noise floor in the 2.4 GHz band is raised by nearly 20 dB. This could impact wireless device sensitivity significantly.
With a wireless mouse, performance is fine at 2, 3, and 5 feet. Attach a USB 3.0 hard drive (no writing to it) and the mouse is fine at 2 feet, but lags at 3 feet and 5 feet. Modifying the USB 3.0 connector at the host device itself improves performance of the wireless mouse.
Here's the thing: wifi signals don't cause autism; they just cause other wifi signals to fail. That's true when the signal isn't even a wifi signal, but is in the same band. USB 3.0 emits EMI in that band.
SATA and eSATA are the same interface, but not the same cable. SATA is inside a high-EMI envelope within a computer casing; eSATA is exposed to less electromagnetic interference.
SATA doesn't have shielding; eSATA requires shielding.
eSATA will work without shielding; so will USB 3.0. Both will also emit large amounts of EMI outside the shielded envelope, interfering with other electronic devices and violating FCC regulations.
USB 3.0 cables aren't shielded from outside electronics; outside electronics are shielded from USB 3.0 cables.
Slashdot Mirror
both the previous day crew that survived and the ones that died diagnosed properly the problem: something is fucking moving the control surfaces. What neither did, was understanding the underlying cause was MCAS command.
Hey I diagnosed the problem: the plane fell out of the air. Future crews should keep the plane in the air. I am very smart.
and have likely been flying around on aircraft without this feature for most of your life.
Correct: the MCAS is a new feature in the 747 MAX, and is misbehaving.
The Lion Air pilots couldn't figure out why the plane was fucking around on them, and a pilot riding along diagnosed the issue.
The next day, a Lion Air plane crashed due to the same issue.
An indicator that suggests a particular issue which looks like other issues but isn't would tell the pilots immediately that this particular issue is happening, thus they could take immediate action instead of flailing around confused.
Glycerin and propylene glycol are irritants. They're bigger irritants than tobacco smoke.
They're also well-known low-toxicity. They don't cause the damage that smoke causes. These compounds can absorb through the lungs into the blood without producing cytotoxicity in the doses consumed.
PG is actually less-toxic than VG, but VG is less-irritating. VG is natural, and PG is an organic compound.
I think the health effects are pretty well expected to be less-damaging than the health effects of huffing charred plant aerosol.
Be that as it may, why the fuck would you put the plane and crew back in the air the next day without an investigation? As soon as this irregularity occurred, the plane and crew should have been grounded. Determining the cause, the response, and the obvious (not likely, but observable without projecting other possibilities) outcome without the irregular situation which corrected the issue, an immediate action would become visible.
That action may be preventative (fix the problem) or contingency (ensure all pilots have training for familiarity with the situation and its appropriate response).
When an irregularity occurs as such, you immediately don't know that the individual plane is unsafe. You may discover by investigation that the entire fleet is unsafe; but you don't know if the plane or its pilots are unique to the failure. You ground that whole damned set of factors, examine the situation, quickly identify the contributing parameters, and decide if you need to ground all planes, issue new training, or just deal with the specific set involved.
That's just basic risk management. It has nothing to do with airplanes.
Those studies were on cell phones in the sub-2GHz band.
Humans are routinely exposed to over half a kilowatt in the 300GHz-430,000GHz band.
And at the resonant frequencies of the various physical structures, or at sufficiently close subharmonics thereof, at a microscopic level, non-ionizing radiation can also cause enough flexing and stretching to break the covalent bonds that hold together strands of DNA. That was what the study on THz EM concluded.
Oh for...really?
At subharmonics, much of the energy is lost: a super-harmonic (e.g. 2x) will oscillate back and forth because the wave amplitude is about 0 at the end, so approximately 0 destructive interference occurs as the vibration reflects back. Basically your wave propagates and lifts the material in one direction, then snaps back in the other direction; at a harmonic, the reflecting wave is pulling along in the reverse direction as it went in, helping to accelerate the process in sync, causing amplification.
At subharmonics, this effect is diminished greatly, and the subharmonic eventually reverses. Likewise, because there is not a 0 amplitude at the end, some energy is lost (when you bend something upwards in the middle, it's pulled down by both ends until each end accelerates upwards to release the energy; but if there's nothing at the end, there's nothing there to anchor against, and you start generating odd-order harmonics and destructive interference). Subharmonics of a resonator are longer than the resonator itself, so the resonator is exposed to part of the energy: at 1nm, a 2^20 subharmonic (about 1mm) would expose 1/1000000 of the energy to the resonator, unless you can direct it into it (which is hard). It's why we need to actually coil 2 meters of wire into a 2/(2^n) meter antenna and then add a really strong amplifier to pick up 2 meter wave.
All of this becomes an interesting engineering problem until you realize you're dealing with DNA.
DNA is floppy and doesn't resonate well. It's the type of material we use to dampen resonant behavior. In other words: DNA isn't subject to resonance. Yes, DNA is one giant molecule and its individual components can't resonate independently because they will transfer the vibrational energy to the rest of the molecule.
Longer, lower-frequency waves have less energy, too.
So you're looking at THz band. Red light is 430THz. UVB is 1,000THz.
You might notice that red, green, and blue light don't cause skin cancer. UVA doesn't cause skin cancer. UVB causes DNA damage and skin cancer. The reasons for this are well-studied.
Part of the reason is just that there's so little energy. A big part is that the energy just goes into heating, and you're not microwaving your DNA; you're microwaving the entire mass in which your DNA sits.
Now that doesn't mean there's no risk. I did say energy. Think about what happens if you stick your face in a 4,000 watt microwave oven. One little photon of UVB can cause DNA damage, but one little microwave photon can't; yet if we pour enormous amounts of them together, we can kill you in seconds.
That's why we have wattage restrictions on transmitters used at ground level, in handheld devices, and so forth. Anything below a certain frequency is going to be harmless below a certain wattage; anything above that is going to start doing single-photon damage to your DNA.
Think about it for a minute. What would it be like if none of that were valid? Well, wind at 10mph doesn't kill you, and wind at 12mph doesn't kill you; but what about 11.7793mph? Only one way to find out!
In other words: we've already done testing. We even did testing on rats with cell phone radiation cranked up to ridiculous levels--which, on one hand, did cause some problems; but on the other, they used Sprague-Dawley rats for lifetime toxicity studies, which is bad methodology. I'm not surprised ridiculously-high-powered radiation at any frequency is harmful: you wouldn't stick your face in a microwave oven, unless you were completely stupid.
Looks like Luxottica had net units sold of 89 million in 2017, with $10,350 million of net revenue, an average $116 per unit sold.
Their cost of goods sold was $3,710 million, so $42 of cost of goods per unit sale.
Their selling and general administration expense was $5,170 million.
By comparison, Apple has $15.26 billion of general administration. Luxottica lists zero research and development, while Apple lists $11.58 billion.
Apple's unit sale is much, much larger than Luxottica's, so their general administration costs less of a percentage of general revenue per unit sold, of course. Apple also has higher net operating profit margin.
Zenni is using a lab in China, whereas Lensabl seems to use a lab in Los Angeles. Many of the manufacturers out there use New York, California, and other local labs. That doesn't mean Zenni can't make a quality product, although Zenni does have limited options compared to Zeiss or Lensabl's own labs--and the fixed costs for a more-complete lab are higher. Notably, Lensabl can provide a really high-end anti-reflective coating, while Zenni can provide a comparable but not-as-good (that means it's worth buying and is a real upgrade from basic, but it's not a Porsche) coating for about $20. Similarly, they can provide a greater range of Transitions lens coatings, and high-end digitally-modeled lenses shaped to remove visual aberration based on your eye shape as modeled from your prescription (this actually works, which is strange but well...I'm well-aware technology is frequently surprising).
Zenni has lower volume, less marketing, and a reduced need for complex logistics (logistics hickips are disastrous in a high-volume operation, so you can pretty much run out of your basement for sufficiently-low volume).
Higher automation or outsourcing to Mexico or China may be a good direction for the lens industry. Better business processes to scale the logistics up to huge and diverse supply chains and catalogs also could help Luxottica, Lensabl, and Zeiss get those costs down. I'll tell you one thing, though: those glasses do NOT costs $350 when they come out of Luxottica; they get there through the supply chain.
15.8% is still a rather high profit margin, and they are taking a significant cut. It's not much per unit, but sheer volume has a big economic effect.
Perhaps, but it's more like saying something that's 10C won't harm you, your body is about 36C, and something that's 70C will burn your skin; and someone says temperatures of 25C may be harmful.
Those mechanical effects are heat. The volume of particle exposure becomes the defining characteristic. Ionizing radiation can cause damage by individual particles (a collision can remove an electron from a molecule, thus breaking a molecular bond), while non-ionizing radiation can cause warming and thus can cause burns via a large volume of particles.
An extremely-bright beam of colored light, for example, can burn your arm off. Red, green, yellow, or blue, with enough wattage, it'll cut straight through.
The risk of microwave exposure is similar to the risk of getting near a hot object. We can build microwave generators to generate huge amounts of power in a narrow spectrum, so the effect is a bit more spectacular than looking inside your oven: we can basically divert the output of a blowtorch straight into your face. You'd notice.
it seems as if it'll only amount to a thought experiment while the political sphere is still flush with cash from corporate donations to super PACs.
Kind of irrelevant. The SuperPACs can affect who gets elected, but don't really affect their decisions. The politicians basically come set, and aren't easily bought; it's the media that leads to election results themselves which you can buy.
Current politicians have a will to solve these things, but not a will to move against their own fear of loss. We can't solve that by paying them; what we can do is dump money into funding a competing politician already interested in making such changes. Consider that the limit on what money can buy and you start to see opportunities you'd previously discounted.
Don't know what subscribestar is. I'm the sort that would crowdfund something on Kickstarter and not Indiegogo because the split between multiple platforms is irritating and shows different, non-composited numbers to backers (i.e. dishonest). That's also why I don't have a GoFundMe.
Above a certain frequency, the waves are small enough to penetrate and hit DNA.
Below a certain frequency, those waves have macro-effects. They cause molecules to wobble and heat up. Infra-red is the highest-energy known for this in particular, as it's just below visible light in terms of energy (visible light is shorter wavelength).
What? You didn't know? Light and EM are the same thing. They're both radio waves. Your wifi antenna does the same thing as a light bulb; so does an X-ray machine.
Mine was better.
why would anyone assume there is a relationship between cost to produce something and the cost we have to pay?
Because there is.
I'm curious about if the prices above are gross or net costs. Basically, what's the profit margin?
Luxottica's net profit margin is about 15.8%, meaning that for every $1 of revenue they keep about 16 cents of profit. This is similar to Apple, Microsoft, and Google (20%-22%), and dissimilar from Comcast (11%-13%, depending on year), Adidas (8%), and Walmart (3%). Large pharmaceutical companies pull fluctuating profits generally holding around 12% for a 5-year average.
I generally describe 8% as a fair net profit margin. It's somewhat arbitrary, but viable, and is derived from common numbers. Farmers believe their correct profit margin is 20%, and agricultural business literature often states that farmers should draw a 20% profit margin but tend to struggle to hold 10%--a similarly-arbitrary statement, and one that forms the foundation for agricultural industry welfare (not that welfare under our food source is a bad thing). Generally, modern regulations actually hold the profit margin of farmers down, rather than subsidizing it up (the farm subsidies basically set a price limit and offset it with a government payment).
We can surmise, then, that the 15.8% net profit margin is not as egregious as considered--a $350 pair of glasses would be roughly $302 of net costs and $50 of revenues if sold direct from a single company, although the reality is a lot of small profit margins along the way--yet it is quite egregious compared to standard 8%-10% profit margins in durable goods industries, and even exceeds the 12% margins of some industries with more control of the market consumer.
In short: the situation is not as bad as presented; but it does demonstrate market power abuse.
The main time I see this happening is when some game releases DLC. Rather than selectively install the DLC they pack it into their data files and inflict the download and footprint cost on everyone whether they want it or not.
In a similar sense, I've frequently suggested they should profile or self-profile games and stream content.
Think about something like Breath of the Wild, 13 gigabytes. Do you need 13 gigabytes to play the opening scene?
When you start the game, the very first assets you need are identifiable. You can profile the loading screens and such, or you can speculatively identify assets by predicting what the next screen will load based on where menu entries go etc. and what assets (and code!) they call up.
So as soon as you turn it on, you have a list of things you need to get to a new game.
The same is true of starting a new game: you know what scene it calls into, and can download that. You can inspect the scene and see what assets it calls, and download those. You can speculatively-render: don't rasterize, but call out what assets would be used in the render, and identify what is visible and what is in the local scene but not visible. You can look for sector changes (doors) and scene changes (transitioning doors). You can look for events and movies.
You can project ahead and identify what you're going to need. Then, if you encounter something not loaded, you can pause and download it.
In development, you can profile this: you can speculatively load (with all assets available) and then have the profiler catch anything that was loaded but not used (load last) or used but not loaded (add to the forced speculation at this point). Developers can tweak the speculation to improve its base functionality.
Much of this already exists: the game loads up everything it needs into memory as you enter a scene (preloading), rather than streaming it off disk as it comes into the render view. We're mainly talking about leveraging that, but with a little look-ahead as to where you could end up immediately (what's the next room?).
You're coming within range of several shrines and dungeons. Grab their base assets.
You're getting closer to a particular shrine entrance. Prioritize its assets. Move those to the front of the queue.
You passed it and are now closer to some other entrance. Change the queue, download those assets instead.
Imagine: you buy the game and you're playing it 12 seconds later. It's going to take 18 hours to download, but you're getting 21% through it in the next 10 hours.
How do you diff, say, a megatexture atlas which you've tweaked some of the dimensions to remove an unused image and repack the rest?
You use the original file as a dictionary and designate the locations of each texture, with procedural instructions to reconstruct the output.
(i.e. changing the compression on the textures to improve performance or avoid a licensing cost, which means changing the code, plus all of the texture atlases, plus re-optimising/recompressing everything)
Now that one you need to reissue the files. You could theoretically write a deterministic decompress-compress procedure, though.
If it's encrypted and the user at the end can't decrypt it, it can't be used, and the user at the end doesn't have to have it.
For example: every DVD player has the capacity to decrypt a DVD. If you decap the chips or otherwise get into the player, you can retrieve the key. The software decrypts the video as it plays.
So you can sign all the data (encrypted or decrypted) with your private key, then use the session key to decrypt, patch, re-encrypt, and verify that the result is the same as the signature presented.
But getting there is extremely difficult -- those two parties have a vested interest in maintaining this system, and will resist any change that undermines it.
No they won't. Well, the Republicans will.
The Democrats in power in Baltimore and in Maryland seem to resist change of which they are uncertain. Undermining their party power base is not a large concern; political fall-out--that the change has unintended consequences which make their voters angry--has heavier weight. They're all loss-averse and vulnerable to other typical human cognitive flaws, so they hyperfocus on such things.
Where I live we have switched to ranked choice voting at the municipal level, but we're also a town controlled by Democrats, so switching didn't imperil party control of the elected offices, and only boiled down to making party insiders' choices slightly more at risk.
The Instant Runoff Voting switch actually leaves the election vulnerable to some manipulation. Tideman's Alternative isn't.
Generally, if your party has more than 50% of the voting base, it has control. That doesn't mean the same candidate has control.
I often use an example for Unified Majority with 40% Hard Right, 29% Moderate Conservative, and 31% Liberal because the 69% always vote in a republican, but each group impacts the outcome. Under UM and Tideman's Alternative, they elect a Moderate Republican; under IRV, the Moderate Republican is eliminated and the Hard Right Republican is elected. Likewise, removing the Hard Right under UM gets you a Moderate Liberal; removing the Moderate Conservatives gets you a Hard Right Republican; and removing the Liberals gets you a Hard Right Republican.
As you can see, every vote is factored into the final outcome of Tideman's Alternative. Unified Majority uses an electoral structure to ensure the final candidate set represents the span of voters so that Tideman's Alternative has something with which to work.
Even so, if more than half your base is behind a single candidate, that candidate wins. If more than half your base is behind a party, that party wins--albeit the particular candidate may be softer than if you had a party primary.
Switching was aided by a fairly long-term trend of rising progressives who both supported the change and were wresting control.
It's more than RCV has become a meme and has politically less risk, so the perceived losses for not switching are large. That's how politics operates these days. It's what Trump has refined to its absolute conclusion.
My complex analysis is not something you see in the political sphere. The political sphere is about a brand. It's about $15/hr, whenever that gets here, and to hell with inflation and productivity gains and a fair wage. $15 in 2005 or $15 in 2025, as long as it's $15.
There's no defense against that "exploit"
Mostly I don't think there's any defense against this.
I designed a defense specifically against this.
Let's say you have two political philosophies: Liberal and Conservative. They're diametrically-opposed, and adherents carry them to various degrees.
In a one-vote system or an approval system, you get two parties. Ranked systems can also raise two parties, but there's less need. The two-plus system appears because of damage in these systems: the candidate with the strongest favor--not the majority favor--wins.
With majority-runoff and instant runoff voting, this occurs in early rounds and runs down to a later round, so candidates with the strongest non-majority favor advance while strong consensus candidates are eliminated early. I've actually published crude methods to manipulate IRV (the most common ranked choice voting method) to precision-select a candidate by adding a third candidate and campaigning to eliminate the original winner in the penultimate round, selecting the original loser (the third candidate can't beat the loser, but can make the winner the lowest vote-getter in the three-way round).
Okay, that sounds like we could change to a Condorcet system; but that's not quite enough.
A Condorcet system eliminates the IRV/Plurality problem. Condorcet systems are frequently manipulable; Tideman's Alternative is almost-certainly non-manipulable (its resistance in practice is near 100% because you need more votes to manipulate it under any failure mode than you do to legitimately win).
Party primaries, however, provide a huge problem.
Because of the aforementioned liberal-conservative divide into two parties, we have what amounts to a top-two system with severe distortions. Top-two is itself horribly broken (it has the plurality flaw with a nonpartisan blanket primary on plurality, and by IRV/STV it nominates two wildly-biased candidates not fitting the overall consensus).
The most-severe distortion is the primary itself. Base voters--people who vote straight party--don't have any marginal utility for primary voting. That means activist voters--people well away from the consensus and moderate positions--drive the primary.
With 40% hard-right, 29% moderate-right, and 31% liberal voters, you'd think that the 40% hard-right would drive the Republican primary. The problem is with 20% hard-right, 49% moderate-right, and 31% liberal, you'll probably actually see 15% hard-right and 10% moderate-right voters get out in the primary. Obviously, the 69% of Republican voters carry the General Election.
If that sounds fanciful, think about this: a propaganda campaign can excite your activists and move more people to the activist camp, getting a bigger turn-out for your polarized candidate in the primary.
There's your exploit.
Use the defect of the party primary to select an extreme.
In the general election, you just have to lift base voters and tilt a minority of swing voters. That's all. You do that, you've hand-selected a candidate.
So what's the weakness of this exploit?
Primary Weakness: it requires manipulation of a small subset of the population.
It's a weak attack. It's fragile. It can't succeed if it needs a major shift.
The fix is called Unified Majority.
First, let's add another consideration about Ranked Choice Voting: people reliably rank to six candidates; the number of ballots ranking seven or more candidates is much smaller than the number ranking six.
Unified Majority replaces the Party Primary with a nonpartisan blanket primary run via STV. For single-seat elections, Unified Majority no
How do you know that your knowledge of a crime isn't a crime itself, or doesn't reveal that you have likely committed crimes?
Fifth amendment stuff should probably apply here, although I don't see what law they can possibly use to compel a witness in an investigation not related to any law which that witness may have broken.
As for the two parties, one is the Liberal philosophy and one is the Conservative philosophy. They're polar opposites, with degrees of how strongly people push those philosophies. That's the real difference, and it's important. Legislators and administrators will take the country in different directions depending on which philosophical ideal underlies their party, even if you have things like unmitigated corruption.
To a degree, government structures can mitigate corruption; not much else can. Municipal governments, in particular, can use the council-manager structure to ensure a highly-responsive government with minimal corruption. More-representative bicameral government structures with highly-representative and manipulation-resistant electoral systems can also mitigate corruption as a secondary effect.
Under the Model City Charter, the City Council appoints and may dismiss at any time a City Manager to perform all administrative functions; the Council has no direct administrative power. This contrasts with a Council-Strong Mayor form, wherein the Mayor is elected, powerful, and can freely run the city while hindering Council severely.
The Mayor can easily appoint all their friends and donors to high-power, lucrative positions; whereas a City Manager is not an elected official, has little reason to do that, and can be thrown out by City Council. City Council needs to appoint the City Manager, so they sort of share influence and, thus, will have severe conflicts about which of their friends and donors should be appointed where, creating a stopgap for such corruption.
Unicameral City Councils can operate with a mixed election, such as 14 Districts each electing one Council member, plus 7 more at-large by proportional vote. This creates a modal difference between members and tends to hinder corruption as well.
Bicameral structures, on the other hand, can operate by districts of Condorcet single representatives (Senate) and proportional multi-representatives (House). This ensures that Senators represent the overall consensus, which severely hinders power-grab politics focusing on small groups and majority power; while House Representatives are fit to the largest cohesive subgroups in the voting population, creating more-extreme voices while essentially making competition between candidates unappealing to their own base moot (i.e. a Republican and a Democrat in a mixed district have basically no reason to campaign against each other, and need to focus heavily on beating the candidates similar to themselves).
By using a system of Unified Majority, we eliminate tampering by strategic nomination, party primary manipulation, and propaganda attack. Unified Majority uses nonpartisan blanket primaries held by STV. For single-seats, it nominates 5 (or 7 or 9) Candidates from which we elect via the non-manipulable Condorcet system, Tideman's Alternative. For multi-seats, it nominates between 2n and 5n, from which we elect (n) candidates via a second round of Single Transferable Vote. STV itself is highly tamper-resistant.
Unified Majority ensures your nominated candidates represent the span of the voters. Baltimore City is 7% Republican and like 0.5% Green; you're not going to elect a Green or Republican mayor, and Unified Majority gives you a span of various Democrats--unless, of course, a Green or Republican runs such a strong campaign that they have a real shot at winning the general election (that's happened, and they won by being more than 50% of voters's first choice, so they would have been nominated under Unified Majority's nonpartisan blanket primary).
Party Primary gives Baltimore City one Democrat for 92% of the voters and one Republican for 7% of the voters, which is not choice. Under Unified Majority, you'd need
Thing is a self-driving car might get confused, swerve off the road, and cause some minor property damage. A human in the same situation might get confused, lock up the brakes, and either fail to steer entirely or steer into opposing traffic.
Often we see the severity of each incident lower than a likely comparable scenario with a human. Think more property damage, less severe property damage and loss of life and limb. Gives you a huge dilemma.
we are marching towards no accountability with e-voting machines that have proven to be insecure as hell.
Current systems are broken. It's possible to establish something more-secure than paper (a lot more secure); and the primary attacker isn't hackers (we can trivially exclude hackers), but the electoral authority (extremely difficult to exclude). It's a big effort, though.
Sadly, I worry more about the selection of idiots we're putting on the ballot than the voting system itself
It's up to the people to decide; and that gets manipulated by the broken electoral systems we use, notably party primaries and single-vote systems. IRV is also easy to manipulate (it's a lot more effort and strategy than plurality, but I've developed the exploit to a fair degree).
Single Transferable Vote (multiple-winner) and Tideman's Alternative (single-winner) are near-impossible to manipulate, even with social media propaganda attacks. I've used those to assemble a highly-representative, manipulation-resistant electoral system called Unified Majority.
I'm going to write two books, one about sabotaging elections and one about securing elections. Maybe then, after it becomes clear any group of half-stoned college kids can hijack an election just by attacking the voting rule, people will get it through their heads.
question Bruce's logic here, since he's not exactly ignorant of either technology or law.
That happens once in a while. He's one of the foremost information security gurus in the world, but he still backs paper ballots as if they're magic. He won't even have a conversation about systemic problems in proving paper ballot integrity, in the lack of auditability of paper ballots, and so forth.
When such experts speak, people follow. That gets us HR1, the Black Box Voting Act, which threatens to cripple our democracy by expanding vote-by-mail not just to those with disabilities and other accessibility issues, but to everybody.
Imagine you go to vote. They use paper ballots.
You walk in, they have a box that was put there the previous night. It's black, sealed, and nobody can look inside. The box is also hidden behind a curtain.
You fill out a ballot, go behind the curtain, cast a vote.
Then they take the box out of the room and come back three weeks later with a different box filled with ballots that definitely came out of that box and haven't been altered. They count those.
That's vote-by-mail.
If 20% (a lot!) of votes by mail are tampered and 0.1% of votes are cast by mail, then 1 in 5,000 votes are tampered. If 50% of votes are cast by mail, that's 1 in 10. It matters.
Meanwhile, the paper ballots are the record, and they're transferred by a chain-of-custody, held in secure trust, etc.. If you open the ballots up to recount, whatever they say is the correct record.
That's not an audit trail. You don't have record of the contents of the ballot box; only that a ballot box with contents was handed off. We can duplicate or reverse seals readily enough.
To really provide an audit trail, you need data. Records that prove a set of ballots came from a particular polling center. Those records need to be produced in a manner in which the public can verify the election is not tamperable right up to the point of producing those records, and then can record those records themselves.
For plurality, you just use vote counts, as you can sum them; but plurality is hackable by adding candidates (strategic nomination).
For ranked ballots, the amount of data grows factorially with the number of candidates. It's impossible to prove ranked ballots without a computer involved. That means you need universal verifiability for electronic voting--this is a harder problem, but one I've explored; nothing today achieves this, although it's trivial to achieve.
HR1 of course outright prohibits the use of electronic records as an official count against the paper record. It doesn't make an exception when using systems of established universal verifiability.
So we've got a bill for establishing the tyranny of black box voting and insecure elections. Fucking great.
Firstly the EMI envelope within a PC is controlled and far lower than what you compare it to. Short lengths of very low current very low voltage signals at high frequencies radiate but do so poorly.
The PC envelope is specifically designed to be an EMI shield due to EMI generated by the PC. It's an FCC compliance point.
And that is not remotely true. The data lines are twisted to prevent radiation and have been for a long time.
Twisting the data lines causes them to self-shield against near-end cross-talk (for round-trip pairs e.g. Ethernet, they'll have opposing magnetic fields which self-cancel), and also causes LVDS pairs to remain at the same base voltage when acting as antenna (these pairs don't self-shield against NEXT). It doesn't prevent them from radiating outward in an LVDS setup.
and in fact USB 3.0 all things being equal would be less likely to cause external interference than USB 2.0 based on signalling alone.
Oh really?
With the HDD connected, the noise floor in the 2.4 GHz band is raised by nearly 20 dB. This could impact wireless device sensitivity significantly.
With a wireless mouse, performance is fine at 2, 3, and 5 feet. Attach a USB 3.0 hard drive (no writing to it) and the mouse is fine at 2 feet, but lags at 3 feet and 5 feet. Modifying the USB 3.0 connector at the host device itself improves performance of the wireless mouse.
Here's the thing: wifi signals don't cause autism; they just cause other wifi signals to fail. That's true when the signal isn't even a wifi signal, but is in the same band. USB 3.0 emits EMI in that band.
SATA and eSATA are the same interface, but not the same cable. SATA is inside a high-EMI envelope within a computer casing; eSATA is exposed to less electromagnetic interference.
SATA doesn't have shielding; eSATA requires shielding.
eSATA will work without shielding; so will USB 3.0. Both will also emit large amounts of EMI outside the shielded envelope, interfering with other electronic devices and violating FCC regulations.
USB 3.0 cables aren't shielded from outside electronics; outside electronics are shielded from USB 3.0 cables.