The "first" except for pretty much every single hyperbolic comet ever observed, right? The first as in not at all the first. The first in a meaning of the word first that does not, in fact, mean first.
The whole point of "hyperbolic orbit" is that it is an object that isn't bound to the sun, and objects that are not bound to the sun are by definition extrasolar. Yes, any given hyperbolic comet might have come out of the Oort cloud or some such from a previously bound orbit, but since we cannot tell what the origin of ANY such object really is by the time we see it at all, it is highly probable (at least) that many or most of these comets have been extrasolar objects and not just the results of collisions of bound objects making an unbound object.
It is going to be moderately difficult not to reinvent a wheel in/. commentary. This is big business, worth a lot of money, valuable as a research tool, critical to national defenses round the world, basis of modern commerce and banking, key (so to speak) to personal privacy in the internet era, and what the heck, essential to really excellent games.
One can buy hardware RNGs that plug into USB ports for a few tens of dollars. To efficiently whiten their output (if necessary -- a good one should already produce white/unbiased output) you can always xor them with a white PRNG -- it doesn't even have to be a terribly GOOD white PRNG, as long as it passes the bit distribution tests in dieharder (some from FIPS, some I made up myself) that rigorously check that all bit patterns occur according to the multinomial distribution as they should (a bit stronger than just counting 0's and 1's or 00 01 10 11 patterns, in other words).
That's why I'm surprised that supposedly professional companies selling VPNs did something as boneheaded as use a hard coded seed. Seriously? What, did their "security expert" sit at a table with the board and say "Hey, we need to make this encryption really, really random, based on a number nobody can possibly guess. Everybody give me your birthdays and we'll turn them into our secret key..."? You can do better -- you can do decently well, in fact, just reading pages on Wikipedia on the subject, e.g.: https://en.wikipedia.org/wiki/.... Cryptography has a long and colorful history, filled with illuminating examples such as the fact that World War II was to some extent won by the allies because they beat the shit out of the axis encryption schemes, inventing and building some of the earliest electronic computers to do so. Even now, civilization teeters on the brink because if somebody figures out solutions to certain "hard" computational problems (like prime number generation or factorization) SSL, SSH, and RSA-based methods in general are screwed, and with them pretty much all of modern e-commerce. To pull something as boneheaded as this requires real stupidity, almost wilful ignorance.
Or, if your application is really critical (read, it will cost you a mountain of money, human lives, etc if its encryption fails) you might throw in a well-designed commercial entropy source (a number are out there, including some that at least claim to be quantum based and hence "truly random" and not JUST unpredictable) and xor in that as well. Headless boxes also have a hard time keeping up with entropy demand, so adding a fast high entropy source is a good idea to avoid delays while the system accumulates more "random" bits.
There is nothing wrong with the FIPS tests, as far as they go, although they focus on bitlevel short-range (de)correlation and are relatively unsensitive to long range patterns. Marsaglia's tests are better for the latter. The problem with both is that they can't possibly protect you against this sort of thing. You can use the world's best (algorithmic) encryption-class RNG, one that passes every single test you throw at it with flying colors, and if you seed it with the number "42" every time you boot the system you will get the same random-looking sequence of bits every time you call the generator.
I do my best to educate people about this sort of thing (I'm the primary author of "dieharder", a testing tool that incorporates Marsaglia's diehard battery of tests, some of the FIPS tests, some tests of my own, and some tests contributed by others) but the fundamental difficulty is that many people just don't know what a random number is and don't understand that "random number generator" is an oxymoron. The XKCD joke posted above is actually remarkably apropos.
The problem is exhibited in the fact that many old books of tables had pages of "random numbers" in them. Even Marsaglia, when he first distributed his diehard test on CD, included files full of "high quality/certified random numbers". This sort of thing has led to people understandably having the misconception that it is numbers that are random-ish, not the PROCESS THAT GENERATES THE NUMBERS (sorry for shouting, but it is frustrating). If you have a "good" pseudorandom number generator (one that produces an acceptable spectrum of results when tested by a decent RNG tester and that has "algorithmic strength" on the basis of mathematics and not just empirical tests -- you really need both, especially for cryptography) and you seed it with a bitwise uniformly distributed and unpredictable seed, you are going to end up with a generator that returns a unique (to the seed) bitwise uniformly distributed, uncorrelated, unpredictable WITHOUT the seed bit sequence, that can be xor'd with a data stream to reversibly encrypt it. The problem is that understanding this much is only the tip of the iceberg as far as good encryption is concerned -- you can't reuse the same seed to send many messages, for example, especially messages that have the same bit patterns in predictable places such as a header (note that I'm NOT an expert on encryption per se, but know enough to appreciate the discipline and know a bit more than that in areas that overlap RNGs per se). Nor can you use a "good" algorithm (say AES or threefish) with a 32 bit seed -- too easy to exhaustively search the seed space. I'm sure there are other, more subtle, Don'ts out there.
But all encryption methods have the chicken/egg problem. To be secure and reproducible, you need a maximum entropy SOURCE to generate the seed bits, and there need to be enough seed bits to not be exhaustively searchable. You cannot get the seed from any sort of table. You cannot generate it with any RNG with its OWN -- 32 bit (don't laugh) -- seed. You cannot generate it with a process that has bias or internal correlation that effectively reduces the information content. It truly does need to be unpredictable, uniform, and not detectably correlated. Generating this isn't easy, and it isn't fast. Servers often lack sufficient entropy to keep up with a strong demand for random bits. Personal computers/laptops usually have more entropy sources and can do better. But ultimately, your entire encryption chain is always as weak as your primary seeding PROCESS.
Is the number 01010101 random? Maybe. If it was generated by a random process it is, even though it looks internally correlated. If you rolled an 8-bit perfect die and excluded all of the numbers that don't "look random", like this one and 00000000 and 11111111 etc, you'd end up with a a lot less than 8 bits of randomness. Numbers per se are never "random", but the process that generates them can be effectively unpredictable (and in the case of quantum processes, possibly "truly random" although this is more a metaphysical statement than an empirical one).
Without going into the impossibility of generating a "random" (as opposed to unpredictable and/or uncorrelated) number, the sad thing is that it is so easy to generate "good" seeds, as all of the primary operating systems have entropy gathering systems that are easy to call. The one in Windows is (sadly) closed source, but it is there, and M$ programmers are rarely overtly incompetent given a specific task. And for mission critical super-duper applications, there are always hardware sources (including many modern CPUs as well as over the counter ones). Attacks like this have been known and exploited for a long time, and it is honestly a bit difficult to imagine anybody who knows enough to write a VPN in the first place not knowing enough to avoid the problem, but hey, never attribute to malice and corruption that which can easily be explained by human stupidity.
Elon Musk has it all wrong. World War III won't be started by artificial intelligences run amok -- that's irresponsible science-fiction-fantasy-land talk designed to feed his personal narcissism with still more headlines. World War III will be started by purely human stupidity with a soupcon of narcissism on the side. FSM knows that there is plenty of both running rampant in the world today.
I also live in the southern US, and installed low E argon filled windows two decades ago. They dropped my electricity bill by maybe 10%. My original equipment furnace/AC units failed ten years ago, and I replaced them with ultra high efficiency HVAC (giant outside condensers etc). That cost a lot, but I had no choice about the replacement and the marginal cost of high efficiency was a few thousand each for the three floors of my house. That (plus finishing my attic and adding 16" of insulation under the roof in addition to the floored in attic insulation that was already there) dropped my bill by another 25%.
The problem with all of this is that now my house is down to an EPP payment of $145/month (big house, air conditioning, perimenopausal wife so we keep it pretty cold). This basically means that I have little marginal benefit left to receive if I install something like rooftop solar. The amortization time stretches out to over ten years, and while I've probably amortized the cost of the windows and the extra marginal cost (at least) of the HVAC, rooftop solar plus 30 KWH of battery (or more) would cost more than all three HVAC units combined. I keep looking into ways of doing it at zero out of pocket investment, and NC has just passed a bunch of laws that may make it possible -- basically finance it for the cost of my current electric bill over whatever, ten years. If I can work this out, I'll probably do it.
In other words, I agree that we are as efficient as the current economies dictate, although the rapidly changing CBA for both rooftop solar and commercial solar is going to be changing that over the next three years. Duke Power is planning to install 2.6 GW of solar in NC over the next three years. That is a bit more newsworthy and interesting than installing 30 MW of capacity in Scotland that (I suspect) will amortize "never" unless electricity is incredibly expensive in Scotland. At $0.11/KWH (which is pretty much what I pay in NC IIRC) it takes around 12 years for rooftop solar to pay for itself, assuming it lasts that long without any additional expense, and that is just too damn long.
I should have been clearer (but I was in a hurry to win the coveted first-post award:-). Hopefully:
a) No side effects, not because it might not be worth it even with side effects up to a direct mortality rate lower than the disease it cures, but because hey, side effects generally suck. If I'm going to expend valuable "hope", I might as well hope to get a real live pony as to get a tiny stuffed unicorn. (OK, not the best metaphor -- I really don't want a real live pony because I'm not that fond of horses and because of "side effects" on my yard, but you get what I mean). And sure, the side effects of making one immortal, restoring lost youth, functioning like permanent viagra (for either sex), causing lost hair to regrow, or making one devastatingly witty at parties all might be desirable, but what are the odds even with a dollop of "hope" thrown in? They are less likely than the pony.
b) And (omitted) one hope that this actually WORKS when one takes it out of the real of in vitro, through animal models, and into humans. In vitro we've cured lots and lots of cancer many time. In mice -- well damn, all I can say is that we could easily treble the life span of a typical mouse just by injecting them with this and that and adding a pinch of special mouse mix to their food supply (if you can believe the literature). But humans are not just cells, and we are not mice. Drugs that work in mice (or whose possibly fatal side effects don't show up in the lifetime of a typical mouse) don't always work in humans.
So what I should have said is "I hope it survives human trials and becomes a viable therapy that significantly reduces human mortality and morbidity from the disease because many therapies that make it to this stage do not." Including many "magic bullet" therapies (in vitro). Antibiotics were and still are largely a magic bullet therapy if you have any of a wide spectrum of infections and no allergies. Outside of that, magic bullet therapies -- take a dose of this and you are cured -- are comparatively rare, and we STILL have little that we can do for most viruses or cancers but stand by and watch them take their course (while providing metabolic support and controlling fevers etc that are the proximate cause of death), or poison the human ALMOST to death in the hope that the more rapidly dividing cancer cells die first. And of course these less than magic bullet therapies often fail.
Good points. For now, perhaps maybe, but in the long run we'll have to do better than that if we don't want to double the cost of solarizing a house over the lifetime of the cells. Then there is zero advantage -- hell, negative advantage -- compared to just buying power from a utility that makes more and more of its energy with cells that don't need the battery.
A big house needs around 30 KWH/day. When one isn't air conditioning or using an electric stove or oven or baseboard heating, a typical house probably draws less than a kilowatt, on average. Remember, at night and during the day usage usually drops -- it is mostly morning, evening, and AC. Two to three of the 14 KWH batteries would be plenty, especially with line voltage to back it up. But I foresee a time coming soon when 50 KWH will cost around $10000, and that would be enough to run your whole house for two or more days, depending.
I spend $145/month for a large house WITH three air central (high efficiency) ACs and a postmenopausal wife (house temperature kept around 68, in other words). At $0.10/KWH that is 1450 KWH/month or divide by 30 to get 50 KWH/day. If we kept the house at 72 instead of 68, we could probably shave 30% off of that. And this is a big, three floor house.
You are off by a factor of 2 to 3. The powerwall costs somewhere between $500 and $700 per KWH (I believe the 14 KWH powerwall retails for around $7000 -- the article is from last year) and competitors are clustered in the $500/KWH range. "No name" (to speak of) batteries from China are cheaper still, but not a LOT cheaper. Otherwise, you are correct that if the cost per KWH was $200 (ideally with a lot more than 1000 full cycles!) you are getting to where solarizing a house for full off-grid operation can be smoothly amortized and make you money over the lifetime of the hardware.
A lot depends on how the battery packs are designed -- in particular if they are designed for the actual power cells themselves to be replaced on a modular basis while preserving the inverter, the installation costs, the other electronics and chassis. If one can (at the end of 3, 5, or 10 years -- whatever one determines a "useful life" to be) just pop out the actual cells and trade them in for discounted new cells based on then current technology, one isn't really looking at a $700/KWH re-investment but rather more likely a $100/KWH or less re-investment, one that is likely to last twice as long and hold half again as much power as energy density and charge cycles may well have improved in the meantime.
Modular battery packs, in the meantime, will likely require congress to get in on the act to set standards. Otherwise Tesla will probably pull and Apple and make their battery packs non-serviceable so that they screw money out of the consumer at every turn, or arrange it so that only replacement cells from Tesla will work in their boxes. Ideally, congress would define form factors that ensure cross-compatibility from many manufacturers of the actual cells so that no matter what "box" one buys with the inverters etc in it, one can plug in any one of half a dozen competing replacement cells and be guaranteed that they will "just work". In our real non-ideal world, of course, that will never happen as long as corporate america pays for the political campaigns of all elected officials from both parties, but one can dream...
Straight up alkaline, abundant base materials, and the same technology can also be used to make other kinds of batteries. Bill Joy is Not An Idiot, although he did miss the chance to make Sun Microsystems' SunOS into Linux before Linux got off of the ground (and in the process, put the hurt on Microsoft) back when he had a perfectly good 386 Unix and nobody else did. Outside of that one mistake (and sure, it was a doozy) he isn't likely to be pulling a scam or anything like it. This is technology that is probably going to work, and will be commercialized (I'm guessing) in a lot less than five years if early mass-marketable prototypes work as well as the early demonstration cells do. Joy is talking about dropping the cost of house or car batteries from the current $500-ish/MWH to under $100/MWH, maybe as low as half that. And if you follow/. and other science sites, you must be aware that there are advances constantly being made in battery technologies that can, and almost certainly will, revolutionize energy storage within the next few years ASIDE from this one. Synthesis of different advances may yet make zinc oxide batteries work (which would be huge all by itself).
TESLA'S hype may not measure up to reality, but overall the reality is that battery technology is already cheap enough and robust enough to make houses that run well over 90% of the time fully off grid, houses that can run for one to two days on limited sun including AC, at an investment that amortizes in roughly 10 to 12 years at typical power prices. At PR prices of $0.20/KWH it would be more like 7 or 8 -- that is well above the national average. 8 year amortization is actually damn close to being a no-brainer, provided that you install hurricane-proof cells on hurricane-proof houses in a hurricane-prone part of the world, something that one really should do ANYWAY because it doesn't do anyone any good to build a house and have it blow away in a storm.
What Musk proposes is far from crazy, and could conceivably be one of the most cost-beneficial solutions -- but probably NOT if one buys it from Musk himself, as his prices for solar roofs and batteries are at least 50% higher than market average and kicks amortization back up to the 12 year plus level of not obviously worth it. If Joy's batteries come through with 1200+ recharge cycles at $60/KWH (and maybe fronted by supercapacitors to buffer daytime utilization and extend this still further) it will drop the amortization time for solarizing a house to five years at the up-front cost of a cheap car and with the loan repaid entirely by money saved on electricity. Well within five years all of this is going to happen no matter what happens with oil and coal and Paris accords or the like, driven by the simple fact that it will be the cheapest way to get power in 2/3 of the word. Power companies are already building solar as fast as they can afford to because for them, economies of scale already make it a no-brainer (they don't need batteries, for the most part, and can feed the energy straight into their grids to reduce fuel costs at their fuel based plants and avoid having to build expensive GW-scale new plants to keep up with demand).
It will be interesting to see whether or not centralized power generation and distribution survives my lifetime (at this point, likely to be somewhere between 10 and 20 years, small chances of 30 or more). I expect to put solar on both of my houses well within the decade not to save the Earth but to save money, sooner if batteries get cheaper faster and cell prices keep coming down or if meter prices for electricity go up. I'm waiting for
Sorry, used my mod points yesterday. Otherwise I'd mod you way up.
Even now, if somebody did steal my information from Equifax and uses it to borrow a zillion dollars, I have zero liability for that, as I did not do it and there is simply no way that they can prove that I did. So the loaning company will eat the debt, not me. It will be at most an annoyance to me, and probably not much of that.
Most responsible loaning agents already make it enormously difficult to borrow large amounts of money "anonymously", that is, without face time, notarized documents, and due diligence. The big exception is credit card companies, and all one can say to them is fuck them if they issue cards to an anonymous stranger just because they have my SSN and address in hand.
Yes, there have been people that have been royally screwed by all of this in the past, but it is not OUR RESPONSIBILITY to protect all of this. It is and always has been the responsibility of the loaner to verify that the loan is indeed a good risk, and hey, verifying that the loanee is who they claim to be is simply part of their due diligence, and accepting a single token like SSN as de facto evidence of this is just plain stupid, on THEIR part.
It would, of course, be lovely to implement a 3x penalty rule, but the top article is dead on the money -- SSN is a dated, stupid way of verifying identity. Ultimately, we're probably going to have to tie identity to something like DNA, so taking out a loan involves processing a cheek swab PLUS the usual due diligence. That would really put a stop to a lot of this even now -- if the loaning agency has a cheek swab, taken in front of witnesses, in a vault indexed by your loan, it is at the very least going to make prosecuting for fraud very, very easy.
Depending on the task being coded, sure. Operating systems are never complete. Programs written to count all of the letters in a standard text file and turn them into a simple table of frequencies, on the other hand, are, or would be if it weren't for the pesky operating systems and compilers and libraries and execution interfaces that keep changing.
But in context at the level of Alexa and other zwave smart device interfaces, hell yes, the software will never be complete because a lot of it IS operating system and the rest of it is interfaces to a dazzling and ever changing array of IoT devices, each with their own controls (and internal bugs!). I have a Smart Things (Samsung) controller running things in my house, and they update both the firmware and/or the phone apps every two to three months -- and some zwave devices I own STILL don't really work very well.
About the only way "5000 people" might make sense is if they had 2495 devices to interface, and assigned each device a two-person team. But then you'd still have a frightful collision at the OS level of the device itself where the extra 10 people could create a huge mess, and you'd have the further problem of 2495 teams reinventing the same wheels, differently, often badly.
A manager went to the master programmer and showed him the requirements document for a new application. The manager asked the master: ``How long will it take to design this system if I assign five programmers to it?''
``It will take one year,'' said the master promptly.
``But we need this system immediately or even sooner! How long will it take if I assign ten programmers to it?''
The master programmer frowned. ``In that case, it will take two years.''
``And what if I assign a hundred programmers to it?''
The master programmer shrugged. ``Then the design will never be completed,'' he said.
You'd think Amazon would know better... 5000 people, Sheesh. They'll actually be undesigning it -- "never completed" is nowhere near strong enough.
The real problem is that in order for the algorithm to work, there have to be enough people out there who are in fact buying the items together. Either those items are often bought together for some other non-bomb-making purpose -- people buying plastic buckets, nails, engine oil and high-nitrogen fertilzer because they are DIY gardners who are working on a shed to hold their lawnmower -- or there are enough people out there building bombs with ingredients purchased together on Amazon that their algorithm can identify the pattern and promote it over the alternative patterns of engine oil and lawnmower parts, fertilizer and garden trowels, nails and hammers and saws, plastic buckets and cleaning supplies.
This is the same problem the NSA's algorithms face. What is a true positive vs false positive signature of terrorist activity? The (Bayesian) answer depends on the PREVALANCE OF TERRORISTS as much as it does anything else. Even a tiny false positive probability (and in this kind of modeling, the false positive probability will not be small, I guarantee it) will swamp one with false positives if the prevalence in the community is low. If there are only ten terrorists in England who actually are trying to build a bomb, and the algorithm designed to pick out terrorists has 10% false positives when applied to a balanced population, then nearly all the positives one turns up will be false.
Note that this is the entire POINT of Amazon's algorithm. It isn't targeting terrorists. Either there is a large non-terrorist market out there for the items to be bought together for benign purposes, or else there are a small mountain of terrorists building bombs and buying the materials from Amazon. In the first case, this is much ado about nothing, and one can leave it to MI-whatever or Scotland Yard to work out some way of winnowing out true positives from false positives if Amazon ever does give them lists of people who actually buy the items together; in the latter case one has a real problem, but the problem isn't Amazon selling Joe Gardner a bunch of stuff that he suddenly decides to turn into a bomb just because he can instead of fertilizing his garden, lubricating his mower, and building his shed, which he keeps nice and clean with his bucket and ammonia.
Fair enough, and I apologize if I bristled a bit. Context is hard to see in nested replies:-)
I didn't think the kid I was replying to was trolling, and I didn't really care about the further upstream stuff. I was mostly trying to redirect him to ways he could learn enough to not make horrendous mistakes in his comment -- Debian part of Red Hat, jeeze. I'd think a real troll could manage to do better than that...:-)
To begin with, there are too many different flavors of RedHat. Browsing a list on Amazon, I saw they made variants under the codenames of Ubuntu, Debian and Slackware, just to name a few. I know that I'm very new to RedHat so maybe this is obvious but it seems like RedHat should just sell a few different flavors of its operating system. Perhaps one for the desktop and one for a server? Could someone explain why RedHat produces dozens of different versions of Linux?
Seriously? Dude, these aren't different flavors of Red Had Linux. Look, you need to read both:
to get some idea what you are talking about. Debian and Slackware are (for example) completely unrelated to Red Hat Linux. Slackware is very old, and used one of the original "packaging" systems -- installing everything from tar.gz images that slotted into a filesystem tree with some pre and post install scripting included. The Debian folks invented one of the original package management systems that was smarter than this (called "apt"). The purpose of apt was to allow a user to install packages from a common repository without entering what is called "dependency hell" - different versions of different programs built on different releases of common systems libraries so installing package B would break package A in unpredictable ways. Windows has a similar problem with "DLL (dynamic link library) hell -- you may have noticed lots of things like games have to install DLLs so that they have precisely the right version to work even if some other application on Windows has basically the same DLL installed (but maybe a different version). This is a nontrivial problem as some of the features a program uses go all the way into the operating systems itself as well as the many shared libraries that the system relies on in order to function. It doesn't have a really satisfactory solution even today, but apt is one that works pretty well. Ubuntu is arguably a variant of Debian that plays a bit better with proprietary tools -- Debian itself is religiously open source. Red Hat got its start by introducing the "Red Hat Package Manager" -- hence "RPM" -- packaging system. This too tried to avoid dependency hell, but sadly through most of the 90's it just didn't work very well -- it was way too easy to break and Red Hat installation because of package inconsistencies with stuff one "had" to have.
About that time we had left off using slackware for the Duke Physics department network and moved to Red Hat because SOME package management and automation was better than none. We had also just hired a young guy named Seth Vidal to be our primary systems administrator. He was a Very Smart Guy (tm) and started to look over package management alternatives that used RPMs but handled dependency hell better. He discovered Yellowdog Linux, which was Linux for the current generation of Apple hardware, and which actually layered a toplevel tool that maintained a local database on top of RPM itself to help track dependencies and block actions that would lead to an inconsistent installation while also automating updates. He (in the best of open source traditions) hacked Yellowdog Updater into "Yellowdog Updater, Modified" or "yum". Yum proved enormously popular and was ultimately adopted by Red Hat as it was the piece that made Red Hat really usable and maintainable on top of their otherwise adequate packaging specification.
Seth went to work for Red Hat after he left Duke and worked there until he was killed while riding his bike in Durham a few years ago. Since his death, yum has been taken over by a slightly different team and renamed "dnf", but dnf AFAICT maintains pretty much exactly the same functionality as yum while sadly not paying homage to the yellowdog folk
Eh, Fedora is free. Debian is free. Scientific Linux is free. Centos is free. And if you knew anything at all about Linux you wouldn't engage in open FUD like: "RedHat is $365 a year for security updates and access to the package repository. Which on *nix is essential." Centos is debranded RH. Fedora actually leads RH and trades a bit of stability for being closer to the cutting edge of development and available software (much like Debian). All of these give you access to the package repository. All of them get "security updates" whatever the hell that means for most of the flavors of linux, which typically get updates of all sorts on a nightly basis with "security" updates rolled transparently into all the rest.
None of them IMO having used both from IBM's PC-DOS through Windows 10 on the Microsoft side and from SLS and Slackware through to the present on the linux side (where I am typing this on Fedora 26 FWIW) are "comparable to Windows Pro". Nor is there any essential difference between workstation and server "editions" of Linux AFAIK -- they differ by at most the selection of packages out of the common repositories and how one configures them -- workstation leaves out most of the servers and daemons and a lot of the management tools, server includes them and focuses less on window managers and user applications, but one can turn a workstation into a server with a handful of dnf/yum/apt commands plus the necessary configuration and vice versa.
You also miss the point I was making to the person I was answering -- in case your "troll" comment was directed upstream somewhere. If you are learning to code and manage systems on a personal basis, linux isn't just free (assuming one is smart enough not to pay RH for a subscription when Centos is the exact same thing and where ANY of them can do "workstation" or "server" or any mix in between), it is an excellent learning platform because of its nearly total support for the most important coding languages, its superior IDE and debugging and programming tools, the fact that you can read the source of anything that interests you and learn from that (more or less impossible on Microsoft based systems), and yeah, the fact that it encourages the use of a CLI to learn what is actually happening and what is doing what from the OS itself up to userspace, with very nearly total control over everything in between, IF you get past using the GUI tools it provides for people that are sadly nearly illiterate about what the OS is actually doing so that the best they can do is pick from a list of menu entries that might or might not do what they would like to try to do.
As a single example, any Linux user can, if they wish, grab the kernel source and set it up to build. They can then, if they wish, configure a custom kernel, or write their own kernel driver, or tweak an existing driver. No, not everybody wants to do this, and most people never need to do this, but the ones who do have complete access. The ones who need to LEARN to do so have complete access. Now, which version of Windows do you need to buy, as a student, to get access to the operating system sources so that you can learn to actually hack the OS? Which version do you need to buy, as a systems manager, to get access at a level that would let you hack in new device drivers needed by your company or leave out the mountains of crap they put in because of its integration with GUI tools required to manage it?
When I teach people to code, one of the first things I ask them to do is install at least a Linux VM on whatever their favorite platform is (and to consider installing linux as a primary operating system and put WINDOWS inside a VM under Linux, where one can always boot the VM if there is something you really have to do under Windows). It is also a very good way to learn networking and how to manage and administer multiuser systems. Finally, yeah, compare batch jobs under Windows to shell scripts under Linux. O.M.G.
The cost is indeed an important factor, but the cost one should be comparing is "no cost at all" to all of the costs needed to install Windows Pro and a tiny fraction of the tools one gets for free with any of the major linux distributions.
I'm glad you have started to use Red Hat, although Fedora and Debian (at least) are free-er alternatives. You can learn to use a wide range of programming languages in a fully professional environment in Linux, and can do things like read source for everything from simple stuff to the operating system itself as you learn C (the "real programmer" language, as in the one people use to write operating systems and maybe 80 to 90% of all the code contributed to Linux and the other free operating systems).
However, you are mistaken (IMO) in your wish to see RH abandon the command line. And you are mistaken in your belief that Microsoft has abandoned the command line. Neither of these are correct. You can bring up a command line in Windows, for the excellent reason that in many circumstances a command line interface is the only way to accomplish things that lie outside of the range of functions supported by ANY GUI interface to a controlled system. It also provides an end run around many of the catch-22 situations you can find yourself in with GUI interfaces -- if the GUI interface itself has bugs or breaks, you can find yourself in a situation where you can't use the interface to fix the bugs or update the broken software and effectively have to reinstall the entire OS to regain control. With a CLI console, you can just login and fix it. CLI interfaces work as long as the kernel and a VERY minimal set of support works, which is why most operating systems have a primitive run level where this is pretty much all that works. Even Microsoft's, although in its "safe mode" you often can't fix what is wrong.
However, there is another reason to use CLIs. They are much, much, much... much more powerful than GUIs for doing repetitive tasks efficiently and for doing almost ANYTHING quickly. An expert in Linux can go through an entire directory tree and patch every file inside to replace the word "frog" with the word "toad" with a single command line command -- probably two or three different ways using one of several tools that will do the job. With a GUI, you will die of old age before you finish even that simple a task in a large filesystem. Copying or moving files that match certain criteria, changing things involving regular expressions, working from one machine on a network of thirty machines to manage them or fix them -- CLI all the way.
CLIs are lightweight -- the Microsoft approach is to provide you with an entire virtual Windows interface to a remote machine simply because they assume that you can't do anything on the remote machine except through the GUI so they have to provide the whole damn thing in a virtual mode for you to make a single tiny change to a single file on the remote system. One system at a time, by hand.
Finally, CLIs are, to put it bluntly, faster for almost any task you EVER try to perform other than just editing text, assuming that you can type faster than 1 character per second or something absurd like that. Back in the days that Apple Macintoshi had pretty much "only" a GUI to the OS back end just as you describe, and when ALL applications were ONLY GUI apps -- a state of affairs that Apple abandoned, for good reason, when they transformed iOS into Unix -- there was a very apropos adage:
One can learn to use a Macintosh in a day, and then pay for that knowledge for the rest of your life.
This is the literal truth. Anybody could use a Mac, but they had to go through the GUI to (as illustrated above) move files, change things, do even simple stuff. Mousing is intuitive but SO SLOW! Even text editors that "only" let you move the cursor or select text or change things with the mouse are slow, slow, slow. The same thing is true about Android -- my biggest gripe with it is that it DOESN'T have a simple console interface built in to the OS any more (although one can install console-ish interfaces if one works at it or needs to root the system so you can actually do things with it). Ditto, only worse, for iphones and tablets. Consequently there is only on
Ah, but in the meantime, genetic research isolated the genes responsible for the condor's giant wing and wingspan. Researchers are in the process of selecting a species of very small pig to attempt for the first time to create a new species of pig, one with actual wings and hence in principle capable of flight according to the syllogism "If pigs had wings, they could fly".
Also, Pink Unicorn spotted trotting down US 70 near Dover, NC, halfheartedly pursued by gracefully dancing bears! More news at 11!
"Earth" slows asteroids down when they land on it. To rest, in the Earth's rest frame, in reasonable approximation. This is a completely inelastic collision, and given the disparity in their masses nearly all of the asteroid's relative kinetic energy is transformed into heat. Some of that heat heats the atmosphere as the atmosphere lands, sure, but seriously, look at the magnitudes involved. This is pretty much irrelevant, given that the impact is going to blow the friction-heated atmosphere near the impact point clean off of the planet.
As far as the second part of your question is concerned, yes, energy is conserved, so the minimum speed of an asteroid falling unimpeded from a place very far from Earth is also the minimum speed required to throw it back up to that place.
This isn't that radical an idea, after all. It works just fine for baseballs. If you drop a baseball in a vaccuum so that it falls some distance H, it will arrive with a speed of roughly \sqrt{2 g H} in near-Earth gravity g. If you want to throw it so that it rises to a maximum height H, you have to throw it up with speed \sqrt{2 g H}. This is simple algebra:
E_tot = U = mgH = 1/2 m v^2 = K
Solve for v. The solution doesn't care if v is directed up or down -- both are consistent with it having EITHER started at H and fallen to 0 (v negative or down) or started at 0 and risen to H (v positive or up). Or you can do the simple solution to Newton's second law and get there a bit harder from:
Falling: y(t) = H - 1/2 g t^2 and v(t) = - g t vs Rising: y(t) = v_0 t - 1/2 g t^2 and v(t) = v_0 - gt
If you solve the first for v when it hits the ground, you'll get \sqrt{2 g H}. If you use v_0 = \sqrt{2 g H} in the second one, it will rise precisely to height H and stop.
But seriously, all of this is in any introductory physics textbook, including mine. Look, here's a nice little lecture on this. Note especially slide 6. Yes, escape speed is also drop from infinity from rest speed when considering two bodies, and at the intro level that's all one teaches. But the energy concepts work just fine for whole solar systems, even when solving the dynamics problems involved becomes nearly impossible for long times:
Otherwise, find a physics textbook. Mine is online and free, and if you google up answers you are as likely as not to get directed to it just because there aren't many free competitors, but I promise, it is accurate enough and fairly complete (except where there are, no doubt, little errors or missing stuff -- an online textbook is never quite finished, sigh:-). But if you prefer Tipler and Mosca, or Halliday, Resnick and Walker or Serway and Jewett or Young and Freedman (old Sears and Zemansky) or Giancoli or Knight (I'm just reading off the authors of the stacks of the damn things in my office) IT DOESN'T MATTER. Look, I'm an expert on this. No kidding. Not expert at the level of a cosmologist maybe, but at the intro level it just isn't that difficult, and on a good day I can actually solve Newton's Law of Gravitation in Newton's Second Law and show that planets really DO move in elliptical orbits, which is a notch or two past intro. So just as a very elastic ball, dropped from a height, will bounce back up to almost the same height (difference lost to heat and sound during the bounce), so a comet that comes into the sun and passes some distance away from the sun at some speed will have the SAME speed as it departs from the sun at that distance on the far side.
So, you didn't bother to read any of the other two or three posts where I worked out the algebra (face-palm) before spouting crap. No, asteroids cannot hit "at any speed imaginable". They can hit at any speed that is GREATER THAN OR EQUAL TO Earth's escape speed, as I actually discussed and derived. Also, escape speed FROM EARTH is relative TO EARTH. So yes, I absolutely neglected solar potential energy because it just doesn't vary that much across the range where most of the actual acceleration of an object falling to Earth occurs -- inside lunar orbit. I'm not a "self-proclaimed" physics teacher, by the way. I just finished teaching university level calculus based introductory E&M to physics majors this morning, and will be teaching intro level mechanics to engineering students this afternoon.
You might also look at the post that I actually replied to, which suggested that they could land at 250 or 2500 kilometers PER HOUR. No, they can't. Yet your "knowledgeable" post implies that they can, to quote: "An asteroid can hit earth with any imaginable speed." Any imaginable speed faster than Earth's escape speed, 11.2 km/second, precisely as I replied. You can see the capitalized AT LEAST in the reply without even bothering to read the whole thing or read the wikipedia article on escape velocity:
where they do EXACTLY the same general algebra I do in my replies above, leading to the same general conclusion, before sure, going off to discuss escaping from the sun or wherever. "Escape energy" is zero total energy for forces that drop off like 1/r^2 (or technically, faster than 1/r), a definition that holds for Coulomb's Law and electrostatics as well. Orbits are categorized as elliptical (bound, E_tot less than 0), parabolic (E_tot = 0) or hyperbolic (E_tot great than 0) in all physics or astronomy textbooks, and yes, these are all defined in this way with respect to the specific two-body interaction involved in introductory treatments simply to avoid many body difficulties, just as they are almost invariably discussed in the limit where one body is much more massive than the other to avoid reduced mass coordinates and other difficulties.
As for adding relative velocities as you seem to have attempted to do -- it doesn't actually work that way, because the expressions that occur in the energy conservation equation are non-linear. Otherwise you would be right, one could somehow "drop" an asteroid in some way that lands at zero speed. What you mean to do is to compute the initial total mechanical energy of the dropped mass at any point in its free or bound trajectory, and compute its final mechanical energy as it is moving the the same velocity as the target object e.g. the Earth (or even pickier, the same velocity as the part of the Earth's surface it inelastically collides with, although this speed of roughly 1000 mph or less is pretty negligible relative to Earth escape speed). The difference has turned into "heat" during the collision. But since kinetic energy scales like velocity squared, and potential energy scales like 1/r (from each center being considered, most likely the Earth and the Sun if you want to talk about solar orbits and are willing to neglect other objects) it is absolutely not as simple as just adding relative velocities and is complicated further still if the object falling to Earth starts in a solar bound orbit. There is a nice equation for this in the Wikipedia article you are refusing to read although the discussion of the point is a bit abbreviated and isn't quite correct either, more of an approximation (note its "under simplified assumptions").
Anyway, if you want to (how was it you put it?:-) "make a fool of yourself" by looking a few things up and spouting nonsense, hey, bring it. Seriously, dude -- actual physics Ph.D. and everything, teaching intro physics since 1977, wrote my own textbooks, don't even bother with lecture notes because I just plain know this stuff and can lecture on any topic at the intro level at least cold. So bring it.
You are mistaken. Escape velocity has nothing to do with launching something into orbit. Well, yes it does. One has to add half of escape energy to an object as kinetic energy to establish it in a low, circular orbit, but that is more or less an interesting algebraic coincidence (related to the virial theorem). Furthermore, as I work out algebraically above, escape speed IS EXACTLY the speed of an object with escape energy, and is in turn BY DEFINITION the speed of an object dropped from infinitely far away, initially at rest. You can learn all of this from literally any decent introductory physics textbook, from any teacher of physics (including me:-), and probably from wikipedia. I already cited the intro physics textbook I wrote above, so you can learn it there, but since you are arguing with me you might understandably refuse to accept that is is authoritative -- I might be a nut instead of somebody who has been teaching intro physics for just about 40 years. So go find one of your own, or google "escape velocity" or "escape speed".
Slashdot Mirror
The "first" except for pretty much every single hyperbolic comet ever observed, right? The first as in not at all the first. The first in a meaning of the word first that does not, in fact, mean first.
The whole point of "hyperbolic orbit" is that it is an object that isn't bound to the sun, and objects that are not bound to the sun are by definition extrasolar. Yes, any given hyperbolic comet might have come out of the Oort cloud or some such from a previously bound orbit, but since we cannot tell what the origin of ANY such object really is by the time we see it at all, it is highly probable (at least) that many or most of these comets have been extrasolar objects and not just the results of collisions of bound objects making an unbound object.
https://en.wikipedia.org/wiki/...
It is going to be moderately difficult not to reinvent a wheel in /. commentary. This is big business, worth a lot of money, valuable as a research tool, critical to national defenses round the world, basis of modern commerce and banking, key (so to speak) to personal privacy in the internet era, and what the heck, essential to really excellent games.
One can buy hardware RNGs that plug into USB ports for a few tens of dollars. To efficiently whiten their output (if necessary -- a good one should already produce white/unbiased output) you can always xor them with a white PRNG -- it doesn't even have to be a terribly GOOD white PRNG, as long as it passes the bit distribution tests in dieharder (some from FIPS, some I made up myself) that rigorously check that all bit patterns occur according to the multinomial distribution as they should (a bit stronger than just counting 0's and 1's or 00 01 10 11 patterns, in other words).
That's why I'm surprised that supposedly professional companies selling VPNs did something as boneheaded as use a hard coded seed. Seriously? What, did their "security expert" sit at a table with the board and say "Hey, we need to make this encryption really, really random, based on a number nobody can possibly guess. Everybody give me your birthdays and we'll turn them into our secret key..."? You can do better -- you can do decently well, in fact, just reading pages on Wikipedia on the subject, e.g.: https://en.wikipedia.org/wiki/.... Cryptography has a long and colorful history, filled with illuminating examples such as the fact that World War II was to some extent won by the allies because they beat the shit out of the axis encryption schemes, inventing and building some of the earliest electronic computers to do so. Even now, civilization teeters on the brink because if somebody figures out solutions to certain "hard" computational problems (like prime number generation or factorization) SSL, SSH, and RSA-based methods in general are screwed, and with them pretty much all of modern e-commerce. To pull something as boneheaded as this requires real stupidity, almost wilful ignorance.
Or, if your application is really critical (read, it will cost you a mountain of money, human lives, etc if its encryption fails) you might throw in a well-designed commercial entropy source (a number are out there, including some that at least claim to be quantum based and hence "truly random" and not JUST unpredictable) and xor in that as well. Headless boxes also have a hard time keeping up with entropy demand, so adding a fast high entropy source is a good idea to avoid delays while the system accumulates more "random" bits.
There is nothing wrong with the FIPS tests, as far as they go, although they focus on bitlevel short-range (de)correlation and are relatively unsensitive to long range patterns. Marsaglia's tests are better for the latter. The problem with both is that they can't possibly protect you against this sort of thing. You can use the world's best (algorithmic) encryption-class RNG, one that passes every single test you throw at it with flying colors, and if you seed it with the number "42" every time you boot the system you will get the same random-looking sequence of bits every time you call the generator.
I do my best to educate people about this sort of thing (I'm the primary author of "dieharder", a testing tool that incorporates Marsaglia's diehard battery of tests, some of the FIPS tests, some tests of my own, and some tests contributed by others) but the fundamental difficulty is that many people just don't know what a random number is and don't understand that "random number generator" is an oxymoron. The XKCD joke posted above is actually remarkably apropos.
The problem is exhibited in the fact that many old books of tables had pages of "random numbers" in them. Even Marsaglia, when he first distributed his diehard test on CD, included files full of "high quality/certified random numbers". This sort of thing has led to people understandably having the misconception that it is numbers that are random-ish, not the PROCESS THAT GENERATES THE NUMBERS (sorry for shouting, but it is frustrating). If you have a "good" pseudorandom number generator (one that produces an acceptable spectrum of results when tested by a decent RNG tester and that has "algorithmic strength" on the basis of mathematics and not just empirical tests -- you really need both, especially for cryptography) and you seed it with a bitwise uniformly distributed and unpredictable seed, you are going to end up with a generator that returns a unique (to the seed) bitwise uniformly distributed, uncorrelated, unpredictable WITHOUT the seed bit sequence, that can be xor'd with a data stream to reversibly encrypt it. The problem is that understanding this much is only the tip of the iceberg as far as good encryption is concerned -- you can't reuse the same seed to send many messages, for example, especially messages that have the same bit patterns in predictable places such as a header (note that I'm NOT an expert on encryption per se, but know enough to appreciate the discipline and know a bit more than that in areas that overlap RNGs per se). Nor can you use a "good" algorithm (say AES or threefish) with a 32 bit seed -- too easy to exhaustively search the seed space. I'm sure there are other, more subtle, Don'ts out there.
But all encryption methods have the chicken/egg problem. To be secure and reproducible, you need a maximum entropy SOURCE to generate the seed bits, and there need to be enough seed bits to not be exhaustively searchable. You cannot get the seed from any sort of table. You cannot generate it with any RNG with its OWN -- 32 bit (don't laugh) -- seed. You cannot generate it with a process that has bias or internal correlation that effectively reduces the information content. It truly does need to be unpredictable, uniform, and not detectably correlated. Generating this isn't easy, and it isn't fast. Servers often lack sufficient entropy to keep up with a strong demand for random bits. Personal computers/laptops usually have more entropy sources and can do better. But ultimately, your entire encryption chain is always as weak as your primary seeding PROCESS.
Is the number 01010101 random? Maybe. If it was generated by a random process it is, even though it looks internally correlated. If you rolled an 8-bit perfect die and excluded all of the numbers that don't "look random", like this one and 00000000 and 11111111 etc, you'd end up with a a lot less than 8 bits of randomness. Numbers per se are never "random", but the process that generates them can be effectively unpredictable (and in the case of quantum processes, possibly "truly random" although this is more a metaphysical statement than an empirical one).
xkcd mod informative? Not funny? Seriously?
Without going into the impossibility of generating a "random" (as opposed to unpredictable and/or uncorrelated) number, the sad thing is that it is so easy to generate "good" seeds, as all of the primary operating systems have entropy gathering systems that are easy to call. The one in Windows is (sadly) closed source, but it is there, and M$ programmers are rarely overtly incompetent given a specific task. And for mission critical super-duper applications, there are always hardware sources (including many modern CPUs as well as over the counter ones). Attacks like this have been known and exploited for a long time, and it is honestly a bit difficult to imagine anybody who knows enough to write a VPN in the first place not knowing enough to avoid the problem, but hey, never attribute to malice and corruption that which can easily be explained by human stupidity.
Elon Musk has it all wrong. World War III won't be started by artificial intelligences run amok -- that's irresponsible science-fiction-fantasy-land talk designed to feed his personal narcissism with still more headlines. World War III will be started by purely human stupidity with a soupcon of narcissism on the side. FSM knows that there is plenty of both running rampant in the world today.
I also live in the southern US, and installed low E argon filled windows two decades ago. They dropped my electricity bill by maybe 10%. My original equipment furnace/AC units failed ten years ago, and I replaced them with ultra high efficiency HVAC (giant outside condensers etc). That cost a lot, but I had no choice about the replacement and the marginal cost of high efficiency was a few thousand each for the three floors of my house. That (plus finishing my attic and adding 16" of insulation under the roof in addition to the floored in attic insulation that was already there) dropped my bill by another 25%.
The problem with all of this is that now my house is down to an EPP payment of $145/month (big house, air conditioning, perimenopausal wife so we keep it pretty cold). This basically means that I have little marginal benefit left to receive if I install something like rooftop solar. The amortization time stretches out to over ten years, and while I've probably amortized the cost of the windows and the extra marginal cost (at least) of the HVAC, rooftop solar plus 30 KWH of battery (or more) would cost more than all three HVAC units combined. I keep looking into ways of doing it at zero out of pocket investment, and NC has just passed a bunch of laws that may make it possible -- basically finance it for the cost of my current electric bill over whatever, ten years. If I can work this out, I'll probably do it.
In other words, I agree that we are as efficient as the current economies dictate, although the rapidly changing CBA for both rooftop solar and commercial solar is going to be changing that over the next three years. Duke Power is planning to install 2.6 GW of solar in NC over the next three years. That is a bit more newsworthy and interesting than installing 30 MW of capacity in Scotland that (I suspect) will amortize "never" unless electricity is incredibly expensive in Scotland. At $0.11/KWH (which is pretty much what I pay in NC IIRC) it takes around 12 years for rooftop solar to pay for itself, assuming it lasts that long without any additional expense, and that is just too damn long.
I should have been clearer (but I was in a hurry to win the coveted first-post award:-). Hopefully:
a) No side effects, not because it might not be worth it even with side effects up to a direct mortality rate lower than the disease it cures, but because hey, side effects generally suck. If I'm going to expend valuable "hope", I might as well hope to get a real live pony as to get a tiny stuffed unicorn. (OK, not the best metaphor -- I really don't want a real live pony because I'm not that fond of horses and because of "side effects" on my yard, but you get what I mean). And sure, the side effects of making one immortal, restoring lost youth, functioning like permanent viagra (for either sex), causing lost hair to regrow, or making one devastatingly witty at parties all might be desirable, but what are the odds even with a dollop of "hope" thrown in? They are less likely than the pony.
b) And (omitted) one hope that this actually WORKS when one takes it out of the real of in vitro, through animal models, and into humans. In vitro we've cured lots and lots of cancer many time. In mice -- well damn, all I can say is that we could easily treble the life span of a typical mouse just by injecting them with this and that and adding a pinch of special mouse mix to their food supply (if you can believe the literature). But humans are not just cells, and we are not mice. Drugs that work in mice (or whose possibly fatal side effects don't show up in the lifetime of a typical mouse) don't always work in humans.
So what I should have said is "I hope it survives human trials and becomes a viable therapy that significantly reduces human mortality and morbidity from the disease because many therapies that make it to this stage do not." Including many "magic bullet" therapies (in vitro). Antibiotics were and still are largely a magic bullet therapy if you have any of a wide spectrum of infections and no allergies. Outside of that, magic bullet therapies -- take a dose of this and you are cured -- are comparatively rare, and we STILL have little that we can do for most viruses or cancers but stand by and watch them take their course (while providing metabolic support and controlling fevers etc that are the proximate cause of death), or poison the human ALMOST to death in the hope that the more rapidly dividing cancer cells die first. And of course these less than magic bullet therapies often fail.
... otherwise, good news.
Good points. For now, perhaps maybe, but in the long run we'll have to do better than that if we don't want to double the cost of solarizing a house over the lifetime of the cells. Then there is zero advantage -- hell, negative advantage -- compared to just buying power from a utility that makes more and more of its energy with cells that don't need the battery.
A big house needs around 30 KWH/day. When one isn't air conditioning or using an electric stove or oven or baseboard heating, a typical house probably draws less than a kilowatt, on average. Remember, at night and during the day usage usually drops -- it is mostly morning, evening, and AC. Two to three of the 14 KWH batteries would be plenty, especially with line voltage to back it up. But I foresee a time coming soon when 50 KWH will cost around $10000, and that would be enough to run your whole house for two or more days, depending.
I spend $145/month for a large house WITH three air central (high efficiency) ACs and a postmenopausal wife (house temperature kept around 68, in other words). At $0.10/KWH that is 1450 KWH/month or divide by 30 to get 50 KWH/day. If we kept the house at 72 instead of 68, we could probably shave 30% off of that. And this is a big, three floor house.
Tesla's batteries cost about $200 per kWh capacity
https://www.forbes.com/sites/p...
You are off by a factor of 2 to 3. The powerwall costs somewhere between $500 and $700 per KWH (I believe the 14 KWH powerwall retails for around $7000 -- the article is from last year) and competitors are clustered in the $500/KWH range. "No name" (to speak of) batteries from China are cheaper still, but not a LOT cheaper. Otherwise, you are correct that if the cost per KWH was $200 (ideally with a lot more than 1000 full cycles!) you are getting to where solarizing a house for full off-grid operation can be smoothly amortized and make you money over the lifetime of the hardware.
A lot depends on how the battery packs are designed -- in particular if they are designed for the actual power cells themselves to be replaced on a modular basis while preserving the inverter, the installation costs, the other electronics and chassis. If one can (at the end of 3, 5, or 10 years -- whatever one determines a "useful life" to be) just pop out the actual cells and trade them in for discounted new cells based on then current technology, one isn't really looking at a $700/KWH re-investment but rather more likely a $100/KWH or less re-investment, one that is likely to last twice as long and hold half again as much power as energy density and charge cycles may well have improved in the meantime.
Modular battery packs, in the meantime, will likely require congress to get in on the act to set standards. Otherwise Tesla will probably pull and Apple and make their battery packs non-serviceable so that they screw money out of the consumer at every turn, or arrange it so that only replacement cells from Tesla will work in their boxes. Ideally, congress would define form factors that ensure cross-compatibility from many manufacturers of the actual cells so that no matter what "box" one buys with the inverters etc in it, one can plug in any one of half a dozen competing replacement cells and be guaranteed that they will "just work". In our real non-ideal world, of course, that will never happen as long as corporate america pays for the political campaigns of all elected officials from both parties, but one can dream...
I'd take issue, AC, with "other technologies". For example:
https://www.bloomberg.com/news...
https://www.greentechmedia.com...
Straight up alkaline, abundant base materials, and the same technology can also be used to make other kinds of batteries. Bill Joy is Not An Idiot, although he did miss the chance to make Sun Microsystems' SunOS into Linux before Linux got off of the ground (and in the process, put the hurt on Microsoft) back when he had a perfectly good 386 Unix and nobody else did. Outside of that one mistake (and sure, it was a doozy) he isn't likely to be pulling a scam or anything like it. This is technology that is probably going to work, and will be commercialized (I'm guessing) in a lot less than five years if early mass-marketable prototypes work as well as the early demonstration cells do. Joy is talking about dropping the cost of house or car batteries from the current $500-ish/MWH to under $100/MWH, maybe as low as half that. And if you follow /. and other science sites, you must be aware that there are advances constantly being made in battery technologies that can, and almost certainly will, revolutionize energy storage within the next few years ASIDE from this one. Synthesis of different advances may yet make zinc oxide batteries work (which would be huge all by itself).
TESLA'S hype may not measure up to reality, but overall the reality is that battery technology is already cheap enough and robust enough to make houses that run well over 90% of the time fully off grid, houses that can run for one to two days on limited sun including AC, at an investment that amortizes in roughly 10 to 12 years at typical power prices. At PR prices of $0.20/KWH it would be more like 7 or 8 -- that is well above the national average. 8 year amortization is actually damn close to being a no-brainer, provided that you install hurricane-proof cells on hurricane-proof houses in a hurricane-prone part of the world, something that one really should do ANYWAY because it doesn't do anyone any good to build a house and have it blow away in a storm.
What Musk proposes is far from crazy, and could conceivably be one of the most cost-beneficial solutions -- but probably NOT if one buys it from Musk himself, as his prices for solar roofs and batteries are at least 50% higher than market average and kicks amortization back up to the 12 year plus level of not obviously worth it. If Joy's batteries come through with 1200+ recharge cycles at $60/KWH (and maybe fronted by supercapacitors to buffer daytime utilization and extend this still further) it will drop the amortization time for solarizing a house to five years at the up-front cost of a cheap car and with the loan repaid entirely by money saved on electricity. Well within five years all of this is going to happen no matter what happens with oil and coal and Paris accords or the like, driven by the simple fact that it will be the cheapest way to get power in 2/3 of the word. Power companies are already building solar as fast as they can afford to because for them, economies of scale already make it a no-brainer (they don't need batteries, for the most part, and can feed the energy straight into their grids to reduce fuel costs at their fuel based plants and avoid having to build expensive GW-scale new plants to keep up with demand).
It will be interesting to see whether or not centralized power generation and distribution survives my lifetime (at this point, likely to be somewhere between 10 and 20 years, small chances of 30 or more). I expect to put solar on both of my houses well within the decade not to save the Earth but to save money, sooner if batteries get cheaper faster and cell prices keep coming down or if meter prices for electricity go up. I'm waiting for
Sorry, used my mod points yesterday. Otherwise I'd mod you way up.
Even now, if somebody did steal my information from Equifax and uses it to borrow a zillion dollars, I have zero liability for that, as I did not do it and there is simply no way that they can prove that I did. So the loaning company will eat the debt, not me. It will be at most an annoyance to me, and probably not much of that.
Most responsible loaning agents already make it enormously difficult to borrow large amounts of money "anonymously", that is, without face time, notarized documents, and due diligence. The big exception is credit card companies, and all one can say to them is fuck them if they issue cards to an anonymous stranger just because they have my SSN and address in hand.
Yes, there have been people that have been royally screwed by all of this in the past, but it is not OUR RESPONSIBILITY to protect all of this. It is and always has been the responsibility of the loaner to verify that the loan is indeed a good risk, and hey, verifying that the loanee is who they claim to be is simply part of their due diligence, and accepting a single token like SSN as de facto evidence of this is just plain stupid, on THEIR part.
It would, of course, be lovely to implement a 3x penalty rule, but the top article is dead on the money -- SSN is a dated, stupid way of verifying identity. Ultimately, we're probably going to have to tie identity to something like DNA, so taking out a loan involves processing a cheek swab PLUS the usual due diligence. That would really put a stop to a lot of this even now -- if the loaning agency has a cheek swab, taken in front of witnesses, in a vault indexed by your loan, it is at the very least going to make prosecuting for fraud very, very easy.
Depending on the task being coded, sure. Operating systems are never complete. Programs written to count all of the letters in a standard text file and turn them into a simple table of frequencies, on the other hand, are, or would be if it weren't for the pesky operating systems and compilers and libraries and execution interfaces that keep changing.
But in context at the level of Alexa and other zwave smart device interfaces, hell yes, the software will never be complete because a lot of it IS operating system and the rest of it is interfaces to a dazzling and ever changing array of IoT devices, each with their own controls (and internal bugs!). I have a Smart Things (Samsung) controller running things in my house, and they update both the firmware and/or the phone apps every two to three months -- and some zwave devices I own STILL don't really work very well.
About the only way "5000 people" might make sense is if they had 2495 devices to interface, and assigned each device a two-person team. But then you'd still have a frightful collision at the OS level of the device itself where the extra 10 people could create a huge mess, and you'd have the further problem of 2495 teams reinventing the same wheels, differently, often badly.
3.4
A manager went to the master programmer and showed him the requirements document for a new application. The manager asked the master: ``How long will it take to design this system if I assign five programmers to it?''
``It will take one year,'' said the master promptly.
``But we need this system immediately or even sooner! How long will it take if I assign ten programmers to it?''
The master programmer frowned. ``In that case, it will take two years.''
``And what if I assign a hundred programmers to it?''
The master programmer shrugged. ``Then the design will never be completed,'' he said.
You'd think Amazon would know better... 5000 people, Sheesh. They'll actually be undesigning it -- "never completed" is nowhere near strong enough.
The real problem is that in order for the algorithm to work, there have to be enough people out there who are in fact buying the items together. Either those items are often bought together for some other non-bomb-making purpose -- people buying plastic buckets, nails, engine oil and high-nitrogen fertilzer because they are DIY gardners who are working on a shed to hold their lawnmower -- or there are enough people out there building bombs with ingredients purchased together on Amazon that their algorithm can identify the pattern and promote it over the alternative patterns of engine oil and lawnmower parts, fertilizer and garden trowels, nails and hammers and saws, plastic buckets and cleaning supplies.
This is the same problem the NSA's algorithms face. What is a true positive vs false positive signature of terrorist activity? The (Bayesian) answer depends on the PREVALANCE OF TERRORISTS as much as it does anything else. Even a tiny false positive probability (and in this kind of modeling, the false positive probability will not be small, I guarantee it) will swamp one with false positives if the prevalence in the community is low. If there are only ten terrorists in England who actually are trying to build a bomb, and the algorithm designed to pick out terrorists has 10% false positives when applied to a balanced population, then nearly all the positives one turns up will be false.
Note that this is the entire POINT of Amazon's algorithm. It isn't targeting terrorists. Either there is a large non-terrorist market out there for the items to be bought together for benign purposes, or else there are a small mountain of terrorists building bombs and buying the materials from Amazon. In the first case, this is much ado about nothing, and one can leave it to MI-whatever or Scotland Yard to work out some way of winnowing out true positives from false positives if Amazon ever does give them lists of people who actually buy the items together; in the latter case one has a real problem, but the problem isn't Amazon selling Joe Gardner a bunch of stuff that he suddenly decides to turn into a bomb just because he can instead of fertilizing his garden, lubricating his mower, and building his shed, which he keeps nice and clean with his bucket and ammonia.
Fair enough, and I apologize if I bristled a bit. Context is hard to see in nested replies:-)
I didn't think the kid I was replying to was trolling, and I didn't really care about the further upstream stuff. I was mostly trying to redirect him to ways he could learn enough to not make horrendous mistakes in his comment -- Debian part of Red Hat, jeeze. I'd think a real troll could manage to do better than that...:-)
I missed this one in my previous reply to you:
To begin with, there are too many different flavors of RedHat. Browsing a list on Amazon, I saw they made variants under the codenames of Ubuntu, Debian and Slackware, just to name a few. I know that I'm very new to RedHat so maybe this is obvious but it seems like RedHat should just sell a few different flavors of its operating system. Perhaps one for the desktop and one for a server? Could someone explain why RedHat produces dozens of different versions of Linux?
Seriously? Dude, these aren't different flavors of Red Had Linux. Look, you need to read both:
https://en.wikipedia.org/wiki/...
and
https://en.wikipedia.org/wiki/...
to get some idea what you are talking about. Debian and Slackware are (for example) completely unrelated to Red Hat Linux. Slackware is very old, and used one of the original "packaging" systems -- installing everything from tar.gz images that slotted into a filesystem tree with some pre and post install scripting included. The Debian folks invented one of the original package management systems that was smarter than this (called "apt"). The purpose of apt was to allow a user to install packages from a common repository without entering what is called "dependency hell" - different versions of different programs built on different releases of common systems libraries so installing package B would break package A in unpredictable ways. Windows has a similar problem with "DLL (dynamic link library) hell -- you may have noticed lots of things like games have to install DLLs so that they have precisely the right version to work even if some other application on Windows has basically the same DLL installed (but maybe a different version). This is a nontrivial problem as some of the features a program uses go all the way into the operating systems itself as well as the many shared libraries that the system relies on in order to function. It doesn't have a really satisfactory solution even today, but apt is one that works pretty well. Ubuntu is arguably a variant of Debian that plays a bit better with proprietary tools -- Debian itself is religiously open source. Red Hat got its start by introducing the "Red Hat Package Manager" -- hence "RPM" -- packaging system. This too tried to avoid dependency hell, but sadly through most of the 90's it just didn't work very well -- it was way too easy to break and Red Hat installation because of package inconsistencies with stuff one "had" to have.
About that time we had left off using slackware for the Duke Physics department network and moved to Red Hat because SOME package management and automation was better than none. We had also just hired a young guy named Seth Vidal to be our primary systems administrator. He was a Very Smart Guy (tm) and started to look over package management alternatives that used RPMs but handled dependency hell better. He discovered Yellowdog Linux, which was Linux for the current generation of Apple hardware, and which actually layered a toplevel tool that maintained a local database on top of RPM itself to help track dependencies and block actions that would lead to an inconsistent installation while also automating updates. He (in the best of open source traditions) hacked Yellowdog Updater into "Yellowdog Updater, Modified" or "yum". Yum proved enormously popular and was ultimately adopted by Red Hat as it was the piece that made Red Hat really usable and maintainable on top of their otherwise adequate packaging specification.
Seth went to work for Red Hat after he left Duke and worked there until he was killed while riding his bike in Durham a few years ago. Since his death, yum has been taken over by a slightly different team and renamed "dnf", but dnf AFAICT maintains pretty much exactly the same functionality as yum while sadly not paying homage to the yellowdog folk
Eh, Fedora is free. Debian is free. Scientific Linux is free. Centos is free. And if you knew anything at all about Linux you wouldn't engage in open FUD like: "RedHat is $365 a year for security updates and access to the package repository. Which on *nix is essential." Centos is debranded RH. Fedora actually leads RH and trades a bit of stability for being closer to the cutting edge of development and available software (much like Debian). All of these give you access to the package repository. All of them get "security updates" whatever the hell that means for most of the flavors of linux, which typically get updates of all sorts on a nightly basis with "security" updates rolled transparently into all the rest.
None of them IMO having used both from IBM's PC-DOS through Windows 10 on the Microsoft side and from SLS and Slackware through to the present on the linux side (where I am typing this on Fedora 26 FWIW) are "comparable to Windows Pro". Nor is there any essential difference between workstation and server "editions" of Linux AFAIK -- they differ by at most the selection of packages out of the common repositories and how one configures them -- workstation leaves out most of the servers and daemons and a lot of the management tools, server includes them and focuses less on window managers and user applications, but one can turn a workstation into a server with a handful of dnf/yum/apt commands plus the necessary configuration and vice versa.
You also miss the point I was making to the person I was answering -- in case your "troll" comment was directed upstream somewhere. If you are learning to code and manage systems on a personal basis, linux isn't just free (assuming one is smart enough not to pay RH for a subscription when Centos is the exact same thing and where ANY of them can do "workstation" or "server" or any mix in between), it is an excellent learning platform because of its nearly total support for the most important coding languages, its superior IDE and debugging and programming tools, the fact that you can read the source of anything that interests you and learn from that (more or less impossible on Microsoft based systems), and yeah, the fact that it encourages the use of a CLI to learn what is actually happening and what is doing what from the OS itself up to userspace, with very nearly total control over everything in between, IF you get past using the GUI tools it provides for people that are sadly nearly illiterate about what the OS is actually doing so that the best they can do is pick from a list of menu entries that might or might not do what they would like to try to do.
As a single example, any Linux user can, if they wish, grab the kernel source and set it up to build. They can then, if they wish, configure a custom kernel, or write their own kernel driver, or tweak an existing driver. No, not everybody wants to do this, and most people never need to do this, but the ones who do have complete access. The ones who need to LEARN to do so have complete access. Now, which version of Windows do you need to buy, as a student, to get access to the operating system sources so that you can learn to actually hack the OS? Which version do you need to buy, as a systems manager, to get access at a level that would let you hack in new device drivers needed by your company or leave out the mountains of crap they put in because of its integration with GUI tools required to manage it?
When I teach people to code, one of the first things I ask them to do is install at least a Linux VM on whatever their favorite platform is (and to consider installing linux as a primary operating system and put WINDOWS inside a VM under Linux, where one can always boot the VM if there is something you really have to do under Windows). It is also a very good way to learn networking and how to manage and administer multiuser systems. Finally, yeah, compare batch jobs under Windows to shell scripts under Linux. O.M.G.
The cost is indeed an important factor, but the cost one should be comparing is "no cost at all" to all of the costs needed to install Windows Pro and a tiny fraction of the tools one gets for free with any of the major linux distributions.
I'm glad you have started to use Red Hat, although Fedora and Debian (at least) are free-er alternatives. You can learn to use a wide range of programming languages in a fully professional environment in Linux, and can do things like read source for everything from simple stuff to the operating system itself as you learn C (the "real programmer" language, as in the one people use to write operating systems and maybe 80 to 90% of all the code contributed to Linux and the other free operating systems).
However, you are mistaken (IMO) in your wish to see RH abandon the command line. And you are mistaken in your belief that Microsoft has abandoned the command line. Neither of these are correct. You can bring up a command line in Windows, for the excellent reason that in many circumstances a command line interface is the only way to accomplish things that lie outside of the range of functions supported by ANY GUI interface to a controlled system. It also provides an end run around many of the catch-22 situations you can find yourself in with GUI interfaces -- if the GUI interface itself has bugs or breaks, you can find yourself in a situation where you can't use the interface to fix the bugs or update the broken software and effectively have to reinstall the entire OS to regain control. With a CLI console, you can just login and fix it. CLI interfaces work as long as the kernel and a VERY minimal set of support works, which is why most operating systems have a primitive run level where this is pretty much all that works. Even Microsoft's, although in its "safe mode" you often can't fix what is wrong.
However, there is another reason to use CLIs. They are much, much, much... much more powerful than GUIs for doing repetitive tasks efficiently and for doing almost ANYTHING quickly. An expert in Linux can go through an entire directory tree and patch every file inside to replace the word "frog" with the word "toad" with a single command line command -- probably two or three different ways using one of several tools that will do the job. With a GUI, you will die of old age before you finish even that simple a task in a large filesystem. Copying or moving files that match certain criteria, changing things involving regular expressions, working from one machine on a network of thirty machines to manage them or fix them -- CLI all the way.
CLIs are lightweight -- the Microsoft approach is to provide you with an entire virtual Windows interface to a remote machine simply because they assume that you can't do anything on the remote machine except through the GUI so they have to provide the whole damn thing in a virtual mode for you to make a single tiny change to a single file on the remote system. One system at a time, by hand.
Finally, CLIs are, to put it bluntly, faster for almost any task you EVER try to perform other than just editing text, assuming that you can type faster than 1 character per second or something absurd like that. Back in the days that Apple Macintoshi had pretty much "only" a GUI to the OS back end just as you describe, and when ALL applications were ONLY GUI apps -- a state of affairs that Apple abandoned, for good reason, when they transformed iOS into Unix -- there was a very apropos adage:
One can learn to use a Macintosh in a day, and then pay for that knowledge for the rest of your life.
This is the literal truth. Anybody could use a Mac, but they had to go through the GUI to (as illustrated above) move files, change things, do even simple stuff. Mousing is intuitive but SO SLOW! Even text editors that "only" let you move the cursor or select text or change things with the mouse are slow, slow, slow. The same thing is true about Android -- my biggest gripe with it is that it DOESN'T have a simple console interface built in to the OS any more (although one can install console-ish interfaces if one works at it or needs to root the system so you can actually do things with it). Ditto, only worse, for iphones and tablets. Consequently there is only on
As Old Ben Kanobi once said, "It's been a long time since I heard that name..."
Lids, heh, lids...
Ah, but in the meantime, genetic research isolated the genes responsible for the condor's giant wing and wingspan. Researchers are in the process of selecting a species of very small pig to attempt for the first time to create a new species of pig, one with actual wings and hence in principle capable of flight according to the syllogism "If pigs had wings, they could fly".
Also, Pink Unicorn spotted trotting down US 70 near Dover, NC, halfheartedly pursued by gracefully dancing bears! More news at 11!
"Earth" slows asteroids down when they land on it. To rest, in the Earth's rest frame, in reasonable approximation. This is a completely inelastic collision, and given the disparity in their masses nearly all of the asteroid's relative kinetic energy is transformed into heat. Some of that heat heats the atmosphere as the atmosphere lands, sure, but seriously, look at the magnitudes involved. This is pretty much irrelevant, given that the impact is going to blow the friction-heated atmosphere near the impact point clean off of the planet.
As far as the second part of your question is concerned, yes, energy is conserved, so the minimum speed of an asteroid falling unimpeded from a place very far from Earth is also the minimum speed required to throw it back up to that place.
This isn't that radical an idea, after all. It works just fine for baseballs. If you drop a baseball in a vaccuum so that it falls some distance H, it will arrive with a speed of roughly \sqrt{2 g H} in near-Earth gravity g. If you want to throw it so that it rises to a maximum height H, you have to throw it up with speed \sqrt{2 g H}. This is simple algebra:
E_tot = U = mgH = 1/2 m v^2 = K
Solve for v. The solution doesn't care if v is directed up or down -- both are consistent with it having EITHER started at H and fallen to 0 (v negative or down) or started at 0 and risen to H (v positive or up). Or you can do the simple solution to Newton's second law and get there a bit harder from:
Falling: y(t) = H - 1/2 g t^2 and v(t) = - g t
vs
Rising: y(t) = v_0 t - 1/2 g t^2 and v(t) = v_0 - gt
If you solve the first for v when it hits the ground, you'll get \sqrt{2 g H}. If you use v_0 = \sqrt{2 g H} in the second one, it will rise precisely to height H and stop.
But seriously, all of this is in any introductory physics textbook, including mine. Look, here's a nice little lecture on this. Note especially slide 6. Yes, escape speed is also drop from infinity from rest speed when considering two bodies, and at the intro level that's all one teaches. But the energy concepts work just fine for whole solar systems, even when solving the dynamics problems involved becomes nearly impossible for long times:
http://www.phys.hawaii.edu/~mo...
Otherwise, find a physics textbook. Mine is online and free, and if you google up answers you are as likely as not to get directed to it just because there aren't many free competitors, but I promise, it is accurate enough and fairly complete (except where there are, no doubt, little errors or missing stuff -- an online textbook is never quite finished, sigh:-). But if you prefer Tipler and Mosca, or Halliday, Resnick and Walker or Serway and Jewett or Young and Freedman (old Sears and Zemansky) or Giancoli or Knight (I'm just reading off the authors of the stacks of the damn things in my office) IT DOESN'T MATTER. Look, I'm an expert on this. No kidding. Not expert at the level of a cosmologist maybe, but at the intro level it just isn't that difficult, and on a good day I can actually solve Newton's Law of Gravitation in Newton's Second Law and show that planets really DO move in elliptical orbits, which is a notch or two past intro. So just as a very elastic ball, dropped from a height, will bounce back up to almost the same height (difference lost to heat and sound during the bounce), so a comet that comes into the sun and passes some distance away from the sun at some speed will have the SAME speed as it departs from the sun at that distance on the far side.
So, you didn't bother to read any of the other two or three posts where I worked out the algebra (face-palm) before spouting crap. No, asteroids cannot hit "at any speed imaginable". They can hit at any speed that is GREATER THAN OR EQUAL TO Earth's escape speed, as I actually discussed and derived. Also, escape speed FROM EARTH is relative TO EARTH. So yes, I absolutely neglected solar potential energy because it just doesn't vary that much across the range where most of the actual acceleration of an object falling to Earth occurs -- inside lunar orbit. I'm not a "self-proclaimed" physics teacher, by the way. I just finished teaching university level calculus based introductory E&M to physics majors this morning, and will be teaching intro level mechanics to engineering students this afternoon.
You might also look at the post that I actually replied to, which suggested that they could land at 250 or 2500 kilometers PER HOUR. No, they can't. Yet your "knowledgeable" post implies that they can, to quote: "An asteroid can hit earth with any imaginable speed." Any imaginable speed faster than Earth's escape speed, 11.2 km/second, precisely as I replied. You can see the capitalized AT LEAST in the reply without even bothering to read the whole thing or read the wikipedia article on escape velocity:
https://en.wikipedia.org/wiki/...
where they do EXACTLY the same general algebra I do in my replies above, leading to the same general conclusion, before sure, going off to discuss escaping from the sun or wherever. "Escape energy" is zero total energy for forces that drop off like 1/r^2 (or technically, faster than 1/r), a definition that holds for Coulomb's Law and electrostatics as well. Orbits are categorized as elliptical (bound, E_tot less than 0), parabolic (E_tot = 0) or hyperbolic (E_tot great than 0) in all physics or astronomy textbooks, and yes, these are all defined in this way with respect to the specific two-body interaction involved in introductory treatments simply to avoid many body difficulties, just as they are almost invariably discussed in the limit where one body is much more massive than the other to avoid reduced mass coordinates and other difficulties.
As for adding relative velocities as you seem to have attempted to do -- it doesn't actually work that way, because the expressions that occur in the energy conservation equation are non-linear. Otherwise you would be right, one could somehow "drop" an asteroid in some way that lands at zero speed. What you mean to do is to compute the initial total mechanical energy of the dropped mass at any point in its free or bound trajectory, and compute its final mechanical energy as it is moving the the same velocity as the target object e.g. the Earth (or even pickier, the same velocity as the part of the Earth's surface it inelastically collides with, although this speed of roughly 1000 mph or less is pretty negligible relative to Earth escape speed). The difference has turned into "heat" during the collision. But since kinetic energy scales like velocity squared, and potential energy scales like 1/r (from each center being considered, most likely the Earth and the Sun if you want to talk about solar orbits and are willing to neglect other objects) it is absolutely not as simple as just adding relative velocities and is complicated further still if the object falling to Earth starts in a solar bound orbit. There is a nice equation for this in the Wikipedia article you are refusing to read although the discussion of the point is a bit abbreviated and isn't quite correct either, more of an approximation (note its "under simplified assumptions").
Anyway, if you want to (how was it you put it?:-) "make a fool of yourself" by looking a few things up and spouting nonsense, hey, bring it. Seriously, dude -- actual physics Ph.D. and everything, teaching intro physics since 1977, wrote my own textbooks, don't even bother with lecture notes because I just plain know this stuff and can lecture on any topic at the intro level at least cold. So bring it.
You are mistaken. Escape velocity has nothing to do with launching something into orbit. Well, yes it does. One has to add half of escape energy to an object as kinetic energy to establish it in a low, circular orbit, but that is more or less an interesting algebraic coincidence (related to the virial theorem). Furthermore, as I work out algebraically above, escape speed IS EXACTLY the speed of an object with escape energy, and is in turn BY DEFINITION the speed of an object dropped from infinitely far away, initially at rest. You can learn all of this from literally any decent introductory physics textbook, from any teacher of physics (including me:-), and probably from wikipedia. I already cited the intro physics textbook I wrote above, so you can learn it there, but since you are arguing with me you might understandably refuse to accept that is is authoritative -- I might be a nut instead of somebody who has been teaching intro physics for just about 40 years. So go find one of your own, or google "escape velocity" or "escape speed".