I've been on a marrow donor list since '94, when a volunteer practically accosted me while I was walking to the school cafeteria and begged me to sign up since they were desperately short of Asian donors. They took a cheek swab and two blood samples; that was it.
A decade later, I got a call saying I was a preliminary match, and they needed my permission to unthaw one of the blood samples so they could run a more thorough compatibility test on it. I would receive a letter I'd need to sign consenting to further testing of my blood, plus some additional questions, which I filled out, signed, and returned. Sadly, the second test revealed my marrow wasn't a good enough match to warrant the risk of a transplant. But it did give me some insight into the process.
Donors and recipients are kept anonymous. You won't get to meet each other. There is no compensation, but as the donor you won't have to foot any of the medical bills. If you think about it, this is more like insurance than it is a donation. Because who knows, it could be you who needs the bone marrow transplant in the future. The procedure is low-risk, but they did say the area would be really sore for a week or two, like you'd run a marathon (they take the marrow from your hip). Which I thought was a funny analogy to use since I and I suspect most people have never run a marathon.
They're unlikely to execute their (indispensable) engineers and scientists for failure. They are however likely to imprison or execute their (expendable) extended family members. Often the people sent to North Korean prison camps have no idea they were even related to the person for whom they're being punished.
If the stuff you're seeing is sorted in reverse chronological order, you can browse through it up to the point where you last browsed through it. Then you're done becomes you know you've already seen everything that comes after.
If the stuff you're seeing is sorted in whatever way some mysterious algorithm thinks you'll find it interesting, you can't stop browsing. An interesting thing that you haven't seen before may be just one more page scroll down, buried among stuff you've already seen. This is like counting marbles by pulling one out of the bag, writing a number, throwing it back into the bag, and pulling out another marble. It's extraordinarily inefficient at finding all the unnumbered marbles, and guarantees you can never know if you've counted them all.
And what happens if you run across something you really do find interesting, but you have to put your phone down to do other stuff. And between then and the next time you're able to browse the algorithm has updated what it thinks your preferences are? You can't reproduce the previous sort order. So now it's like a magazine, where the pages can randomly rearrange their order while you're in the middle of browsing through it.
LOL. Take a 2013, 2014, 2015, and 2015 (again) Macbook Pro and put them side by side. See if you can tell them apart without looking up the serial number. A few years back, my cousin almost ended up paying Sandy Bridge prices for a Core 2 Duo MBP (new at his school store) because Apple makes it so hard to figure out what exactly you're buying.
They've been better about labeling their iPhones, but that seems predicated more on compelling people to upgrade because their phone's number is one lower than the current. Their iPad model naming started off sane, but has lately become a mess.
Putting the GPU in a separate box sitting between the computer and monitor and connected via a high-bandwidth cable like Thunderbolt, has already been done. This is just that idea, but combining the box and monitor. The only advantages I can think of over a separate GPU box are: You don't need a separate power cable because you can mooch off the monitor's power supply. And you could conceivably bypass any cable speed limits by running a direct channel from the GPU to the monitor (thinking ahead to when resolutions are higher than even Thunderbolt can support).
I can think of a lot of disadvantages though. Can't be repaired/upgraded separately. Destroys the thin profile of modern monitors. Overly complicated purchase choices (thinking ahead to a future when x different monitors and y different GPUs are available, you have to pick from x*y monitor/GPU combos, instead of just picking them separately for x+y choices). Hotspot created by GPU could damage the portion of the monitor it's adjacent to. Fan to cool the GPU is stuck in the monitor, so you can't shove it and the computer into a closet with only KVM cables leading to your desk, for some peace and quiet,
So then, what came before the Big Bang? What exists inside of a black hole?
Restricting yourself to purely scientific explanations doesn't make these sorts of questions go away. It just means you're willfully ignoring them - pretending stuff outside the reach of scientific inquiry doesn't exist, just like you're accusing others of pretending things outside the reach of scientific inquiry do exist.
Goedel proved nearly a century ago that any logical system is incomplete - there will always be things within the system which cannot be proven by the logic within that system. That is, the set of statements about the universe isn't divided into true and false things. It's divided into true, false, and cannot be determined. So any philosophy based on assuming things are false unless proven true is logically inconsistent. And believing nothing exists outside our current system is just as much a faith as believing something exists outside. The only logically sound stance is uncertainty about what if anything exists beyond our perception.
In that respect, Musk has the more logically consistent argument. He offers no proof but at least acknowledges the possibility that he may be wrong. You on the other hand offer no proof but seem certain that you are right (that he is wrong).
CPUs and GPUs are binned. They test each one which comes out of manufacturing. The ones which by pure luck (fewer impurities, cleaner etchings) can operate at a lower voltage (and thus draw less power) get binned as laptop parts. The rest of them become desktop parts.
It's not like they can manufacture these lower power consumption chips at will. The manufacturing process dictates that by pure change x% will be suitable for laptop use, leaving 100-x% for desktop use.
What about finding ways to protect people from the cosmic radiation during at least three years (x2) long journey to and from the planet?
A Hohmann transfer orbit to Mars only takes 9 months. 3 years is the time it would take to launch from Earth and travel to Mars via a Hohmann transfer orbit, wait for Mars to be in position for a Hohmann transfer back to Earth, then launch and travel to Earth.
By any objective measure, the second story is a bigger deal than the first one. The first story made the national news and was the headline on most news sites (883,000 articles on Google News). But the second was pretty much limited to local news (9340 articles on Google News - Fox was the only news service to carry it nationally). Drug overdoes deaths have more than doubled in the last decade, and at 47,000 per year have now become the leading cause of accidental death in the U.S., passing auto accidents at around 33,000.
The media exerts considerable bias in the stories they choose to cover. They don't like guns, so they carry a disproportionate number of stories about gun violence. They like drugs, so they regularly ignore stories about the growing drug abuse problem. The stats I linked to above probably come as a complete surprise to most readers, because the media simply hasn't been giving it as much attention as it deserves, concentrating instead on publicizing their their other pet "issues." You can either believe whatever the media spoon-feeds you. Or you can educate yourself, do your own research, and figure out where the true problems are.
Yesterday there was a murder-suicide at UCLA which left two people dead and ended with no further incident, and a concert where 2 people died and an additional 57 were hospitalized after apparent drug overdoses. By any objective measure, the second is the bigger story. But the first made national news and was the biggest headline of the day. The second was only local news, with Fox the only national news outlet carrying the story according to Google News. Because a disproportionate number of people in the media think "guns bad, drugs good," and promote the gun death stories while suppressing the drug death stories.
So basically, even if the system were perfectly democratic and there were no manipulation by the party bosses, the Democrat nominee can be decided by just 17% of all eligible voters, and the Republican nominee by just 12%. In a close primary season such as this year's, our two choices for President are dictated by just 29% of the voting population. The way the parties are designed, this usually ends up the most politically extreme or bat-wing crazy 29%.
The problem is we use a plurality voting system. It's mathematically pretty much the worst possible system, scoring high only on clarity of who the winner is. Pretty much any other system (1) encourages more than two parties (you aren't "throwing away your vote" if you don't vote for the two major candidates), and (2) encourages selection of candidates more acceptable to the entire population, not just the party doing the nomination.
The Tesla S is approx 2 meters wide by 5 meters long, about half of which is windows. If you covered the rest with solar panels (only projected area matters), that's 5 m^2 of panels.
Figure you use the commercial 150 Watt/m^2 panels, and that's a peak generating capacity of 750 Watts. Capacity factor for solar in the U.S. is about 0.145 (this accounts for angle of the sun, weather, etc.). So (0.75 kW) * (0.145) * (24 hours) = 2.61 kWh. In other words, if you left your solar panel-covered Tesla S parked outside for a typical continental U.S. day, it would generate 2.61 kWh.
Charging efficiency of the Tesla battery is about 80%. So only about 2.09 kWh actually makes it into the battery (the rest heats up the battery and charger).
The best EPA-rated Tesla S uses 33 kWh/100 miles. So leaving your PV-encrusted Tesla parked out in the sun all day will charge the battery enough to move you 6.3 miles.
"The argument for lowercasing Internet is that is has become wholly generic, like electricity and the telephone. It never was trademarked and is not based on any proper noun," writes Tom Kent, AP Standards Editor. "The best reason for capitalizing it in the past may have been that the term was new.
Bozos like this is why general press coverage of technical and scientific stories still sucks. Rather than ask someone who knows for their informed opinion, they think they already know everything so can make decisions without having to ask.
An internet is any collection of interconnected networks. If a company connects its Los Angeles branch LAN with its New York branch LAN so they can share files, they now have an internet.
The Internet is the biggest grouping of such interconnected networks, which happens to span the globe (it didn't always). It is capitalized to distinguish it from other internets.
Lowercasing 'Internet' makes about as much sense as lowercasing Associated Press, because the AP used to be new, but now there are several other associations of press corps.
I upgraded my main PC to it a few weeks ago (after blacklisting a whole bunch of hosts and IPs on my router, and immediately installing Spybot Anti-Beacon after).
Pros:
UI makes more sense than Win 8.1. Less schizophrenia about whether it's a desktop OS or a tablet OS.
Games run better. A lot of the microstutters I attributed CPU load spikes or having to read stuff off the SSD are gone.
Icon/text scaling with DPI is much improved, though still not perfect.
I like the minimalist black and white icons in the notification bar, instead of the horrible color clash it used to be with different apps showing notifications with different colors.
They "fixed" the popup stealing focus problem. Now when you're typing a reply on slashdot and a system warning dialog pops up, focus stays with your browser. The dialog no longer disappears an instant after it pops up before you can read it because you happened to hit the space bar an instant after it popped up.
If you're used to Unix from the 1990s, Microsoft finally added multiple virtual desktop support.
The animated tiles in the Start menu are much less annoying that the full-screen animated tiles in the Win 8 Start menu.
Cons:
The animated tiles are still annoying.
Can't turn off updates. Not that big a deal for me since I run most of my apps in a VM running Windows 7 (I got tired of having to reinstall everything every time I upgraded laptops). But could be an issue for small businesses if you're running a mission-critical app, and a forced update breaks it.
Certain apps don't make the transition properly, and you may have to reinstall them. Others you can get working again with a few tweaks.
File explorer windows now default to quick access instead of library + This PC view. So it's now a two-click operation to actually browse your drives, instead of one-click.
It really, really pushes Cortana.
Network access is flakier. I'll try to open a network share or web page and sometimes it'll take a few seconds instead of opening instantly like on Win 7/8. Might be because I'm blocking certain hosts, and it's getting confused for a few seconds when it can't phone home to report which URL I'm visiting.
Task manager can't seem to remember the "hide when minimized" option even though I set it every time.
The popup stealing focus fix causes other problems. If I start a new app, it sometimes doesn't start with focus. I haven't quite figured out the pattern yet. e.g. I'll start a browser and immediately type ctr-l and the URL I wanted to go to, and nothing happens because the browser doesn't have focus. I have to click on it first before I can type ctrl-l and the URL.
Edge browser is extremely non-intuitive when changing the defaults (like homepage and search engine). You can't enter it manually. You have to browse to the page you want as your home page or your search engine, then go to the settings and the option to make that page your default shows up.
If you use IME to occasionally type in a foreign language, the desired setup is to make IME your default keyboard. That way you can use the right alt key to switch between typing in English and the other language. Unfortunately, they combined the keyboard preference option with the language preference option. If you make IME your default, now all your notifications and apps and even certain language-aware web pages default to the other language instead of English. If you leave the English keyboard as the default, any time you want to type in another language, you first need to click to switch from the English keyboard to the IME keyboard, then switch IME from English to foreign language typing mode. This is a major PITA for those of us who are multi-lingual but prefer everything be in English.
So yes it's worth upgrading, but no it's not quite ready yet. But you don't have to decide by July 29. You can upgrade to it, and r
I'm pretty sure YouTube knows this is futile too. But they have to do stuff like this to cover their butts legally. Otherwise Hollywood can go to a judge and say, "See? They're not trying hard enough to protect our copyrights. Strip them of their DMCA safe harbor provision!" (Never mind why it should be someone else's job to protect their copyrights.)
Especially since he explicitly says he cannot fathom why Trump is popular.
He probably has no experience with our two-party political system. The math shows that a single race, single vote, plurality wins voting system is pretty much the worst possible one you can come up with. It will eventually resolve into people having to split into two parties in order to maximize the chance of their favored candidate winning. Voting for any other candidate is equivalent to throwing away your vote.
Trumps' popularity (and also Hillary's) is because his unpopularity is mostly concentrated among people in the other party. There are sane segments of his party who see him for what he is, but they are outnumbered by the batshit crazy segment who likes him (same holds for Hillary in the Democratic party). I think that's the bigger story this election year that's completely being missed by the press - that our nomination process has somehow anointed the two most-disliked candidates as the nominees.
This is the end result of the polarization and extremism that's been controlling both parties - they've both been pulled so far from the country's political center that crazy stuff like this is possible. If you figure 20% in the center do not register with either party, and the remaining 80% split 40%/40% between the two parties, then to become the presidential nominee you only need to garner the support of 21% of the voters, whether they be an extremist 21% or crazy 21%.
We already went through this in the 1980s. The four color theorem - the idea that you can draw any map using just 4 colors and no neighboring regions with shared edges will have the same color - resolves down to 1476 possible simplified map configurations. Appel and Haken wrote a computer program to check each of those configurations, and none were found to violate the theorem. Thus the theorem must be true.
Mathematicians initially rejected the idea that this constituted a mathematical proof for the reasons that have been posted in the comments here. But in the 1980s they came around and begrudgingly accepted that the theorem had in fact been proven, thus legitimizing computer proofs by exhaustion of the solution space. It's really no different than proof by reduction, where you can reduce a problem down to just one or two cases, and simply show that those one or two cases don't violate the theorem being proven.
CDMA won the GSM vs. CDMA war. The HSPA/HSDPA 3G service in most GSM phones uses CDMA or wideband CDMA.
See, the original GSM spec used TDMA - basically each phone is assigned a timeslice and all phones take turns talking with the tower. This meant that even if the phone didn't have much data to transmit or didn't need to send any data at all, it still used a full timeslice. Couple that with time buffers to account for phones being different distances from the tower, and GSM ends up wasting a lot of bandwidth. CDMA allows all phones to transmit at the same time. The tower tells them apart by assigning an orthogonal code to each phone. Bandwidth scales automatically between phones because each phone sees the other phones' transmissions as noise, thus reducing the signal to noise ratio and reducing bandwidth. If a phone doesn't need data for a few seconds, the noise decreases, the SNR for the other phones increases, and that extra bandwidth is immediately available for all the other phones to use.
This is why CDMA carriers got 3G about a year before GSM carriers. Their towers could already provide 3G data speeds. GSM had to amend the GSM spec to specify CDMA and wideband CDMA data services, then wait for handset manufacturers and carriers to implement it. This is also why GSM phones could talk and use data at the same time. They had two separate radios - a TDMA radio for voice, a CDMA radio for data. CDMA phones only had a single radio which could be used for voice or data, but not both at the same time.
Most implementations of LTE use OFDMA. It does the same thing as CDMA, except using orthogonal frequencies instead of orthogonal codes. CDMA served as the proof of concept for widescale simultaneous orthogonal transmissions, so without it LTE probably would not exist or would not be as mature ias it is today. (OFDMA requires more processing power than CDMA, which is why it came later. WiMax used OFDMA, and my Galaxy S phone which used it would die after 3-4 hours on WiMax vs 8-12 hours on 3G CDMA.)
If the U.S. had followed the rest of the world in adopting GSM and had prohibited CDMA, our mobile data speeds today would probably be around 100-500 kbps. So be glad CDMA won, even if Qualcomm is evil.
Just to sidestep the PC vs Mac comparison, let's compare this to its predecessor the UX305CA.
UX305: $699, Core M, 8 MB RAM, 13.3" 1080p matte IPS screen (option for 3200x1800) covering 90% of sRGB, 256GB SSD, 3xUSB 3.0, mini HDMI, 12.3mm height, 1.192kg, 45 Wh battery giving 6+ hours. About the only thing it was missing was a backlit keyboard and a fan.
Zenbook 3: $999, Core i5, 4 MB RAM, 1080p IPS screen (implied touchscreen), 256GB PCIe SSD, 1xUSB 3.1/Thunderbolt, presumably HDMI via Thunderbolt, 11.9mm, 0.910kg, supposedly 9 hour battery though I usually reduce claims to 2/3 which would put it at 6 hours. Backlit keyboard, has a fan.
The UX305 was a worthy ultrabook that I've been recommending to a lot of people who otherwise would've settled for a low-end laptop. Usually their budget was around $500, while the UX305 frequently went on sale for $600 - the size, build, SSD, screen, and generous number of ports made it an easy up-sell for an extra $100. The Core M processor isn't a limitation for most people's computer use.
The new Zenbook 3 comes in at an extra $300 putting it out of reach of budget shoppers. It has a better CPU but lower base RAM, a faster SSD but only people doing video editing will notice the extra speed, loses all those ports (many people I know leave a nano receiver plugged in and use a wireless mouse), shaves a little off the weight and height, and has a backlit keyboard. Honestly, that doesn't seem worth an extra $300.
If you're AT&T and you notice one backbone is more saturated than another, that just means your funds allocation for bandwidth is wrong. Say (simplified) you're allocating 50% of your bandwidth funds to backbone 1 (B1), and 50% to backbone 2 (B2). This is based on the assumption that your customers will draw traffic equally between B1 and B2 (50% and 50% of traffic).
You notice that B1 is constantly congested. You take some measurements and find the correct bandwidth allocation would be 75% B1, 25% B2. So what you do is reduce your bandwidth to B2 from 50% to 25%, take the money you save by doing that and use it to pay for increased bandwidth on B1 from 50% to 75%.
Net Neutrality and Netflix have absolutely nothing to do with this. The only principle here is one of giving the customers what they want and have paid for, and doing so at minimum cost to yourself. The reason Net Neutrality had to be implemented was because ISPs were reducing bandwidth to B2 from 50% to 25%, pocketing the money they saved doing that, and trying to get Netflix to pay for increasing B1 from 50% to 75%.
Remember, the customers already paid the ISP for that bandwidth. It shouldn't matter to the ISP where the customers use the bandwidth - they've already been paid $x for y Mbps. If the customers decide to use more of that y Mbps on B1 instead of B2, then the ISP has to reduce bandwidth to B2 by the exact same amount that they increase bandwidth to B1, meaning there's virtually no difference in cost to them..
The fundamental problem here is that one group's wants and desires (the end-users') are being suppressed. This came about because nearly all local governments granted a monopoly to a single cable service provider. They were well intentioned - often making 99.9% coverage and bandwidth limits a condition of that monopoly. But at the same time they discounted or didn't believe in the market influence of competition, so didn't give much consideration to the harm that granting a monopoly - even a regulated monopoly - does.
Once the cable companies had the monopoly, they could basically ignore end-users' desires, thus depriving them of their voice in the ISP market. On top of that, they are now empowered because they are the only means of internet access to those customers. All the power of those customers' dollars, none of the drawbacks of having to listen to what those customers want! And they chose to leverage that power by trying to extort additional money from websites to deliver content that their customers had already paid them for.
Net neutrality is a band-aid to try to fix this problem. By prohibiting different pricing based on content source, it prevents this type of extortion. But like the original monopoly regulatory kludge, it kills off another aspect of the market - differential pricing based on the cost to actually deliver that content. If Netflix is streaming content to the ISP, that's a lot of bandwidth and so costs the ISP a lot of money. If Netflix installs content servers at the ISP, that eliminates the bandwidth consumption and so costs the ISP less money. But net neutrality essentially prohibits the ISP from passing that cost savings on to the customer. The ISP has to charge the same price for all content, regardless of source and the bandwidth cost to obtain data from that source. It's just one regulatory fix which mostly but not entirely works, trying to fix another regulatory fix which mostly but not entirely works.
The ultimate solution is to restore market power to the end-user. Let them vote with their dollars. This has the advantage of pitting dynamic human minds against any tricks the ISPs try to come up with to increase prices or degrade service. Right now we're trying to fight the ISPs' tricks using static laws, which take decades to implement in response to their previous tricks, giving them plenty of time to figure out new tricks.
Abolish the cable monopolies. Convert the monopoly into a tightly regulated service contact for the physical cable or fiber which runs to the homes, and only the physical cables. No content service allowed. The cable maintenance company then makes money by leasing bandwidth along that fiber to different ISPs at a fixed (regulated) rate. The ISPs then have to compete with each other based on quality of service and price. If an ISP tries to pull a Comcast and deliberately degrades Netflix, they will hemorrhage customers as they flee to a different ISP who isn't degrading Netflix. And they'll do it in a matter of weeks or months, not the years or decades it took to get Net Neutrality implemented. This is how we regulate utilities, and oddly enough this is how "socialist" Europe does it.
For a party's candidate to appear on a state's ballot, they need to jump through some hoops first (usually gather enough signatures of people saying they support that party - has to be completed far enough in advance of the election for the state to "verify" the signatures are authentic). The Libertarian party is the only 3rd party which has qualified in all 50 states, so is the only 3rd party whose candidates will show up on every state's ballot. Other 3rd party candidates start the election effectively conceding the electoral votes of some states (have no chance to win those states).
An unaffiliated candidate can still appear on a state's ballot if his/her supporters jump through the same hoops. That's what Ross Perot's supporters did in 1992. He did well enough in the polls that the media included him in all the Presidential debates.
Finally, most states allow each voter to write in a candidate of their choice. That's the reason Mickey Mouse and Charles Manson get a few votes each election.
All that said, our one-vote plurality-wins system means if you don't vote for a candidate from the two parties most likely to win, you're effectively throwing away your vote. Mathematically, it's the worst possible way to implement a voting system. There are much better systems out there.
It looks like they broke the test rats into groups with 1.5, 3, and 6 W/kg exposure, CDMA and GSM, male and female. That's 3*2*2 = 12 groups. For the brain section, they looked for two types of tumors. So now they've got 24 groupings that they're searching for possible correlations.
The statistical significance of the one correlation they found (male, CDMA, 6 W/kg, malignant glioma) was p < 0.05. In other words, due to their limited sample size, just by random chance alone you'd expect such a blip to occur about 1 in 20 times even when there is no real correlation. Well they tried 24 times and got one blip.
Same thing with the heart results. 24 groupings, one blip with p < 0.05, one blip with p = 0.052. Again, almost exactly what you'd expect by pure chance alone.
What's even more scary is that Facebook is now tracking and advertising to you when they see you outside of Facebook
They've been doing that for close to a decade. The non-technical press has just finally figured it out. It's the main reason I used extensions like noscript, cookiesafe, and ghostery, and now browse in incognito mode all the time. Every 'f' icon on a web page is a little eyeball tracking what pages you're visiting.
I'm sorry if you're just learning this now. Facebook probably already knows your name, your address, where you work, who your family, friends, and co-workers are, what you look like (courtesy of FB members who "helpfully" tag you in their photos with your name), what websites you like to visit, what products you like to buy (or window shop for), and thus can make a pretty good guess what your hobbies are, how much money you make, and what your family life is like. Even though you've never created a Facebook account, courtesy of millions of those little 'f' eyeballs.
Slashdot Mirror
I've been on a marrow donor list since '94, when a volunteer practically accosted me while I was walking to the school cafeteria and begged me to sign up since they were desperately short of Asian donors. They took a cheek swab and two blood samples; that was it.
A decade later, I got a call saying I was a preliminary match, and they needed my permission to unthaw one of the blood samples so they could run a more thorough compatibility test on it. I would receive a letter I'd need to sign consenting to further testing of my blood, plus some additional questions, which I filled out, signed, and returned. Sadly, the second test revealed my marrow wasn't a good enough match to warrant the risk of a transplant. But it did give me some insight into the process.
Donors and recipients are kept anonymous. You won't get to meet each other. There is no compensation, but as the donor you won't have to foot any of the medical bills. If you think about it, this is more like insurance than it is a donation. Because who knows, it could be you who needs the bone marrow transplant in the future. The procedure is low-risk, but they did say the area would be really sore for a week or two, like you'd run a marathon (they take the marrow from your hip). Which I thought was a funny analogy to use since I and I suspect most people have never run a marathon.
Anyway, it's a small price to pay for potentially saving someone's life. Go do it if you haven't yet.
https://bethematch.org/support-the-cause/
They're unlikely to execute their (indispensable) engineers and scientists for failure. They are however likely to imprison or execute their (expendable) extended family members. Often the people sent to North Korean prison camps have no idea they were even related to the person for whom they're being punished.
If the stuff you're seeing is sorted in reverse chronological order, you can browse through it up to the point where you last browsed through it. Then you're done becomes you know you've already seen everything that comes after.
If the stuff you're seeing is sorted in whatever way some mysterious algorithm thinks you'll find it interesting, you can't stop browsing. An interesting thing that you haven't seen before may be just one more page scroll down, buried among stuff you've already seen. This is like counting marbles by pulling one out of the bag, writing a number, throwing it back into the bag, and pulling out another marble. It's extraordinarily inefficient at finding all the unnumbered marbles, and guarantees you can never know if you've counted them all.
And what happens if you run across something you really do find interesting, but you have to put your phone down to do other stuff. And between then and the next time you're able to browse the algorithm has updated what it thinks your preferences are? You can't reproduce the previous sort order. So now it's like a magazine, where the pages can randomly rearrange their order while you're in the middle of browsing through it.
LOL. Take a 2013, 2014, 2015, and 2015 (again) Macbook Pro and put them side by side. See if you can tell them apart without looking up the serial number. A few years back, my cousin almost ended up paying Sandy Bridge prices for a Core 2 Duo MBP (new at his school store) because Apple makes it so hard to figure out what exactly you're buying.
They've been better about labeling their iPhones, but that seems predicated more on compelling people to upgrade because their phone's number is one lower than the current. Their iPad model naming started off sane, but has lately become a mess.
Putting the GPU in a separate box sitting between the computer and monitor and connected via a high-bandwidth cable like Thunderbolt, has already been done. This is just that idea, but combining the box and monitor. The only advantages I can think of over a separate GPU box are: You don't need a separate power cable because you can mooch off the monitor's power supply. And you could conceivably bypass any cable speed limits by running a direct channel from the GPU to the monitor (thinking ahead to when resolutions are higher than even Thunderbolt can support).
I can think of a lot of disadvantages though. Can't be repaired/upgraded separately. Destroys the thin profile of modern monitors. Overly complicated purchase choices (thinking ahead to a future when x different monitors and y different GPUs are available, you have to pick from x*y monitor/GPU combos, instead of just picking them separately for x+y choices). Hotspot created by GPU could damage the portion of the monitor it's adjacent to. Fan to cool the GPU is stuck in the monitor, so you can't shove it and the computer into a closet with only KVM cables leading to your desk, for some peace and quiet,
So then, what came before the Big Bang? What exists inside of a black hole?
Restricting yourself to purely scientific explanations doesn't make these sorts of questions go away. It just means you're willfully ignoring them - pretending stuff outside the reach of scientific inquiry doesn't exist, just like you're accusing others of pretending things outside the reach of scientific inquiry do exist.
Goedel proved nearly a century ago that any logical system is incomplete - there will always be things within the system which cannot be proven by the logic within that system. That is, the set of statements about the universe isn't divided into true and false things. It's divided into true, false, and cannot be determined. So any philosophy based on assuming things are false unless proven true is logically inconsistent. And believing nothing exists outside our current system is just as much a faith as believing something exists outside. The only logically sound stance is uncertainty about what if anything exists beyond our perception.
In that respect, Musk has the more logically consistent argument. He offers no proof but at least acknowledges the possibility that he may be wrong. You on the other hand offer no proof but seem certain that you are right (that he is wrong).
CPUs and GPUs are binned. They test each one which comes out of manufacturing. The ones which by pure luck (fewer impurities, cleaner etchings) can operate at a lower voltage (and thus draw less power) get binned as laptop parts. The rest of them become desktop parts.
It's not like they can manufacture these lower power consumption chips at will. The manufacturing process dictates that by pure change x% will be suitable for laptop use, leaving 100-x% for desktop use.
A Hohmann transfer orbit to Mars only takes 9 months. 3 years is the time it would take to launch from Earth and travel to Mars via a Hohmann transfer orbit, wait for Mars to be in position for a Hohmann transfer back to Earth, then launch and travel to Earth.
Yesterday, June 1, there was a gun-related murder-suicide at UCLA. Two people were killed. There were no other incidental injuries.
Yesterday, June 1, there was also a concert where 2 people died and 57 were hospitalized. Right now they think it was drug-related - either overdoses, or bad synthetic drugs.
By any objective measure, the second story is a bigger deal than the first one. The first story made the national news and was the headline on most news sites (883,000 articles on Google News). But the second was pretty much limited to local news (9340 articles on Google News - Fox was the only news service to carry it nationally). Drug overdoes deaths have more than doubled in the last decade, and at 47,000 per year have now become the leading cause of accidental death in the U.S., passing auto accidents at around 33,000.
The media exerts considerable bias in the stories they choose to cover. They don't like guns, so they carry a disproportionate number of stories about gun violence. They like drugs, so they regularly ignore stories about the growing drug abuse problem. The stats I linked to above probably come as a complete surprise to most readers, because the media simply hasn't been giving it as much attention as it deserves, concentrating instead on publicizing their their other pet "issues." You can either believe whatever the media spoon-feeds you. Or you can educate yourself, do your own research, and figure out where the true problems are.
Drug abuse deaths have more than doubled in the last decade. They've now passed car accidents as the leading cause of accidental death. This is probably the biggest scourge in the U.S. this century, and most people are clueless it's even happening because the media isn't reporting it.
Yesterday there was a murder-suicide at UCLA which left two people dead and ended with no further incident, and a concert where 2 people died and an additional 57 were hospitalized after apparent drug overdoses. By any objective measure, the second is the bigger story. But the first made national news and was the biggest headline of the day. The second was only local news, with Fox the only national news outlet carrying the story according to Google News. Because a disproportionate number of people in the media think "guns bad, drugs good," and promote the gun death stories while suppressing the drug death stories.
32% of voters are Democrats, 23% Republicans. Each has primaries where the winner is chosen by a majority of votes.
So basically, even if the system were perfectly democratic and there were no manipulation by the party bosses, the Democrat nominee can be decided by just 17% of all eligible voters, and the Republican nominee by just 12%. In a close primary season such as this year's, our two choices for President are dictated by just 29% of the voting population. The way the parties are designed, this usually ends up the most politically extreme or bat-wing crazy 29%.
The problem is we use a plurality voting system. It's mathematically pretty much the worst possible system, scoring high only on clarity of who the winner is. Pretty much any other system (1) encourages more than two parties (you aren't "throwing away your vote" if you don't vote for the two major candidates), and (2) encourages selection of candidates more acceptable to the entire population, not just the party doing the nomination.
The Tesla S is approx 2 meters wide by 5 meters long, about half of which is windows. If you covered the rest with solar panels (only projected area matters), that's 5 m^2 of panels.
Figure you use the commercial 150 Watt/m^2 panels, and that's a peak generating capacity of 750 Watts. Capacity factor for solar in the U.S. is about 0.145 (this accounts for angle of the sun, weather, etc.). So (0.75 kW) * (0.145) * (24 hours) = 2.61 kWh. In other words, if you left your solar panel-covered Tesla S parked outside for a typical continental U.S. day, it would generate 2.61 kWh.
Charging efficiency of the Tesla battery is about 80%. So only about 2.09 kWh actually makes it into the battery (the rest heats up the battery and charger).
The best EPA-rated Tesla S uses 33 kWh/100 miles. So leaving your PV-encrusted Tesla parked out in the sun all day will charge the battery enough to move you 6.3 miles.
Bozos like this is why general press coverage of technical and scientific stories still sucks. Rather than ask someone who knows for their informed opinion, they think they already know everything so can make decisions without having to ask.
Lowercasing 'Internet' makes about as much sense as lowercasing Associated Press, because the AP used to be new, but now there are several other associations of press corps.
Pros:
Cons:
So yes it's worth upgrading, but no it's not quite ready yet. But you don't have to decide by July 29. You can upgrade to it, and r
I'm pretty sure YouTube knows this is futile too. But they have to do stuff like this to cover their butts legally. Otherwise Hollywood can go to a judge and say, "See? They're not trying hard enough to protect our copyrights. Strip them of their DMCA safe harbor provision!" (Never mind why it should be someone else's job to protect their copyrights.)
He probably has no experience with our two-party political system. The math shows that a single race, single vote, plurality wins voting system is pretty much the worst possible one you can come up with. It will eventually resolve into people having to split into two parties in order to maximize the chance of their favored candidate winning. Voting for any other candidate is equivalent to throwing away your vote.
Trumps' popularity (and also Hillary's) is because his unpopularity is mostly concentrated among people in the other party. There are sane segments of his party who see him for what he is, but they are outnumbered by the batshit crazy segment who likes him (same holds for Hillary in the Democratic party). I think that's the bigger story this election year that's completely being missed by the press - that our nomination process has somehow anointed the two most-disliked candidates as the nominees.
This is the end result of the polarization and extremism that's been controlling both parties - they've both been pulled so far from the country's political center that crazy stuff like this is possible. If you figure 20% in the center do not register with either party, and the remaining 80% split 40%/40% between the two parties, then to become the presidential nominee you only need to garner the support of 21% of the voters, whether they be an extremist 21% or crazy 21%.
We already went through this in the 1980s. The four color theorem - the idea that you can draw any map using just 4 colors and no neighboring regions with shared edges will have the same color - resolves down to 1476 possible simplified map configurations. Appel and Haken wrote a computer program to check each of those configurations, and none were found to violate the theorem. Thus the theorem must be true.
Mathematicians initially rejected the idea that this constituted a mathematical proof for the reasons that have been posted in the comments here. But in the 1980s they came around and begrudgingly accepted that the theorem had in fact been proven, thus legitimizing computer proofs by exhaustion of the solution space. It's really no different than proof by reduction, where you can reduce a problem down to just one or two cases, and simply show that those one or two cases don't violate the theorem being proven.
CDMA won the GSM vs. CDMA war. The HSPA/HSDPA 3G service in most GSM phones uses CDMA or wideband CDMA.
See, the original GSM spec used TDMA - basically each phone is assigned a timeslice and all phones take turns talking with the tower. This meant that even if the phone didn't have much data to transmit or didn't need to send any data at all, it still used a full timeslice. Couple that with time buffers to account for phones being different distances from the tower, and GSM ends up wasting a lot of bandwidth. CDMA allows all phones to transmit at the same time. The tower tells them apart by assigning an orthogonal code to each phone. Bandwidth scales automatically between phones because each phone sees the other phones' transmissions as noise, thus reducing the signal to noise ratio and reducing bandwidth. If a phone doesn't need data for a few seconds, the noise decreases, the SNR for the other phones increases, and that extra bandwidth is immediately available for all the other phones to use.
This is why CDMA carriers got 3G about a year before GSM carriers. Their towers could already provide 3G data speeds. GSM had to amend the GSM spec to specify CDMA and wideband CDMA data services, then wait for handset manufacturers and carriers to implement it. This is also why GSM phones could talk and use data at the same time. They had two separate radios - a TDMA radio for voice, a CDMA radio for data. CDMA phones only had a single radio which could be used for voice or data, but not both at the same time.
Most implementations of LTE use OFDMA. It does the same thing as CDMA, except using orthogonal frequencies instead of orthogonal codes. CDMA served as the proof of concept for widescale simultaneous orthogonal transmissions, so without it LTE probably would not exist or would not be as mature ias it is today. (OFDMA requires more processing power than CDMA, which is why it came later. WiMax used OFDMA, and my Galaxy S phone which used it would die after 3-4 hours on WiMax vs 8-12 hours on 3G CDMA.)
If the U.S. had followed the rest of the world in adopting GSM and had prohibited CDMA, our mobile data speeds today would probably be around 100-500 kbps. So be glad CDMA won, even if Qualcomm is evil.
Just to sidestep the PC vs Mac comparison, let's compare this to its predecessor the UX305CA.
UX305: $699, Core M, 8 MB RAM, 13.3" 1080p matte IPS screen (option for 3200x1800) covering 90% of sRGB, 256GB SSD, 3xUSB 3.0, mini HDMI, 12.3mm height, 1.192kg, 45 Wh battery giving 6+ hours. About the only thing it was missing was a backlit keyboard and a fan.
Zenbook 3: $999, Core i5, 4 MB RAM, 1080p IPS screen (implied touchscreen), 256GB PCIe SSD, 1xUSB 3.1/Thunderbolt, presumably HDMI via Thunderbolt, 11.9mm, 0.910kg, supposedly 9 hour battery though I usually reduce claims to 2/3 which would put it at 6 hours. Backlit keyboard, has a fan.
The UX305 was a worthy ultrabook that I've been recommending to a lot of people who otherwise would've settled for a low-end laptop. Usually their budget was around $500, while the UX305 frequently went on sale for $600 - the size, build, SSD, screen, and generous number of ports made it an easy up-sell for an extra $100. The Core M processor isn't a limitation for most people's computer use.
The new Zenbook 3 comes in at an extra $300 putting it out of reach of budget shoppers. It has a better CPU but lower base RAM, a faster SSD but only people doing video editing will notice the extra speed, loses all those ports (many people I know leave a nano receiver plugged in and use a wireless mouse), shaves a little off the weight and height, and has a backlit keyboard. Honestly, that doesn't seem worth an extra $300.
If you're AT&T and you notice one backbone is more saturated than another, that just means your funds allocation for bandwidth is wrong. Say (simplified) you're allocating 50% of your bandwidth funds to backbone 1 (B1), and 50% to backbone 2 (B2). This is based on the assumption that your customers will draw traffic equally between B1 and B2 (50% and 50% of traffic).
You notice that B1 is constantly congested. You take some measurements and find the correct bandwidth allocation would be 75% B1, 25% B2. So what you do is reduce your bandwidth to B2 from 50% to 25%, take the money you save by doing that and use it to pay for increased bandwidth on B1 from 50% to 75%.
Net Neutrality and Netflix have absolutely nothing to do with this. The only principle here is one of giving the customers what they want and have paid for, and doing so at minimum cost to yourself. The reason Net Neutrality had to be implemented was because ISPs were reducing bandwidth to B2 from 50% to 25%, pocketing the money they saved doing that, and trying to get Netflix to pay for increasing B1 from 50% to 75%.
Remember, the customers already paid the ISP for that bandwidth. It shouldn't matter to the ISP where the customers use the bandwidth - they've already been paid $x for y Mbps. If the customers decide to use more of that y Mbps on B1 instead of B2, then the ISP has to reduce bandwidth to B2 by the exact same amount that they increase bandwidth to B1, meaning there's virtually no difference in cost to them..
The fundamental problem here is that one group's wants and desires (the end-users') are being suppressed. This came about because nearly all local governments granted a monopoly to a single cable service provider. They were well intentioned - often making 99.9% coverage and bandwidth limits a condition of that monopoly. But at the same time they discounted or didn't believe in the market influence of competition, so didn't give much consideration to the harm that granting a monopoly - even a regulated monopoly - does.
Once the cable companies had the monopoly, they could basically ignore end-users' desires, thus depriving them of their voice in the ISP market. On top of that, they are now empowered because they are the only means of internet access to those customers. All the power of those customers' dollars, none of the drawbacks of having to listen to what those customers want! And they chose to leverage that power by trying to extort additional money from websites to deliver content that their customers had already paid them for.
Net neutrality is a band-aid to try to fix this problem. By prohibiting different pricing based on content source, it prevents this type of extortion. But like the original monopoly regulatory kludge, it kills off another aspect of the market - differential pricing based on the cost to actually deliver that content. If Netflix is streaming content to the ISP, that's a lot of bandwidth and so costs the ISP a lot of money. If Netflix installs content servers at the ISP, that eliminates the bandwidth consumption and so costs the ISP less money. But net neutrality essentially prohibits the ISP from passing that cost savings on to the customer. The ISP has to charge the same price for all content, regardless of source and the bandwidth cost to obtain data from that source. It's just one regulatory fix which mostly but not entirely works, trying to fix another regulatory fix which mostly but not entirely works.
The ultimate solution is to restore market power to the end-user. Let them vote with their dollars. This has the advantage of pitting dynamic human minds against any tricks the ISPs try to come up with to increase prices or degrade service. Right now we're trying to fight the ISPs' tricks using static laws, which take decades to implement in response to their previous tricks, giving them plenty of time to figure out new tricks.
Abolish the cable monopolies. Convert the monopoly into a tightly regulated service contact for the physical cable or fiber which runs to the homes, and only the physical cables. No content service allowed. The cable maintenance company then makes money by leasing bandwidth along that fiber to different ISPs at a fixed (regulated) rate. The ISPs then have to compete with each other based on quality of service and price. If an ISP tries to pull a Comcast and deliberately degrades Netflix, they will hemorrhage customers as they flee to a different ISP who isn't degrading Netflix. And they'll do it in a matter of weeks or months, not the years or decades it took to get Net Neutrality implemented. This is how we regulate utilities, and oddly enough this is how "socialist" Europe does it.
For a party's candidate to appear on a state's ballot, they need to jump through some hoops first (usually gather enough signatures of people saying they support that party - has to be completed far enough in advance of the election for the state to "verify" the signatures are authentic). The Libertarian party is the only 3rd party which has qualified in all 50 states, so is the only 3rd party whose candidates will show up on every state's ballot. Other 3rd party candidates start the election effectively conceding the electoral votes of some states (have no chance to win those states).
An unaffiliated candidate can still appear on a state's ballot if his/her supporters jump through the same hoops. That's what Ross Perot's supporters did in 1992. He did well enough in the polls that the media included him in all the Presidential debates.
Finally, most states allow each voter to write in a candidate of their choice. That's the reason Mickey Mouse and Charles Manson get a few votes each election.
All that said, our one-vote plurality-wins system means if you don't vote for a candidate from the two parties most likely to win, you're effectively throwing away your vote. Mathematically, it's the worst possible way to implement a voting system. There are much better systems out there.
Forgot to link the relevant XKCD cartoon.
I didn't. :) But just reading the first dozen pages...
It looks like they broke the test rats into groups with 1.5, 3, and 6 W/kg exposure, CDMA and GSM, male and female. That's 3*2*2 = 12 groups. For the brain section, they looked for two types of tumors. So now they've got 24 groupings that they're searching for possible correlations.
The statistical significance of the one correlation they found (male, CDMA, 6 W/kg, malignant glioma) was p < 0.05. In other words, due to their limited sample size, just by random chance alone you'd expect such a blip to occur about 1 in 20 times even when there is no real correlation. Well they tried 24 times and got one blip.
Same thing with the heart results. 24 groupings, one blip with p < 0.05, one blip with p = 0.052. Again, almost exactly what you'd expect by pure chance alone.
They've been doing that for close to a decade. The non-technical press has just finally figured it out. It's the main reason I used extensions like noscript, cookiesafe, and ghostery, and now browse in incognito mode all the time. Every 'f' icon on a web page is a little eyeball tracking what pages you're visiting.
I'm sorry if you're just learning this now. Facebook probably already knows your name, your address, where you work, who your family, friends, and co-workers are, what you look like (courtesy of FB members who "helpfully" tag you in their photos with your name), what websites you like to visit, what products you like to buy (or window shop for), and thus can make a pretty good guess what your hobbies are, how much money you make, and what your family life is like. Even though you've never created a Facebook account, courtesy of millions of those little 'f' eyeballs.