"What's that?" - thousands of journalists and their professors. I don't think this is a deliberate misuse of the word trojan - just non technical people getting confused between DarkComet and BlackHole. Similar to how some non-tech people call megabits per second a "bandwidth" or 9.8 meters per second the "force" of gravity.
I'm not disagreeing with anything you say; I'm just pointing out that Apple making single-button mice has had a disciplining effect on UI design. Control-clicking isn't discoverable (just the way keyboard command shortcuts without menu equivalents aren't) so isn't something that can relied on for prominent functionality. Say spell checking a document. If you had to control click to bring up a menu to spell check, and couldn't find it in the main menu, then a LOT of users would never find it and would assume spell checking didn't exist. If that's a prominent, advertised feature of the program then this would be problematic and a huge call generator. So the UI design needs to be adjusted accordingly - most obviously by putting all prominent features in the main menu, or a button on a toolbar, or equivalent, or all of the above. You frequently actually see Windows software that *doesn't* adhere to such obvious basic principles of design; the people who complain about the single-button mouse on the Mac are very frequently exactly the people who would create such seven-headed UI horrors in the first place.
And that explains the pathological hatred of buttons? Simplifying is good, but there's a reason why Apple is the only company to use a 1 button mouse, and why the early mouses all had 3 buttons. Single button mouses suck.
There are good reasons for a single-button mouse. The first is that non-technical users simply don't understand the concept of right or left clicking, and adding buttons that can be clicked just confuses. I should click the icon? The normal button or the menu button? How many times? Why do I sometimes use one button and at other times the other? There's also a menu up top already with all the same things and more... when should I use which? The second reason is that Apple selling a single-button mouse (not sure if they still do though) forces UI designers to make their products single-button navigable. This has a great simplifying effect and removes the ability to do things like different menus in different places, and puts all commands in a single unified menu - making the product more discoverable. Finally, people who want multibutton mice generally have a strong brand preference anyway; I like my Razers while you might order a Logitech the first thing you do. So the Apple mouse mainly ends up in the hands of a novice with no existing preference and likely limited experience with computers.
They could have my user names and logins, but not my passwords or any other credentials. Just asking would pretty much make at least me turn around and walk out, likely rather annoyed at having wasted my time on such losers. Why don't they ask for my address book and business card files while at it? Can I have theirs?
I think Terry Gilliam has come up with the best explanation:
It used to be that studio execs were entrepreneurs. Businessmen with vision willing to take risks.
They still take risks, it's just that they aren't going to invest in something that excludes half the 10-25 year old movie going audience before it's even made. That's not a risk, that's throwing money away. This of course is why we get so much formulaic crap with overacting, lack of all subtlety, predictable plots, the equivalent of poop and fart jokes, and characters intended to appeal to a simplistic childish world view. Adults don't go see movies because it's likely to be an uninteresting way to spend an evening - and it's uninteresting because movies are made to appeal to a pre-adult audience.
The punchline: Unless you are using an atypically paranoid browser config, there are a Lot of CAs and subordinate CAs(some of them known-slimy, others known-incompetent), whose certs your browser will silently trust.
Yeah, this just illustrates how worthless web certs are. Any government with a CA within its borders can get a cert to impersonate anyone or anything. So can any criminal organization in the same country as long as their government is sufficiently corrupt. The way it SHOULD work is that another party proves knowledge of a secret I gave them when I registered, so I can know when I contact them that it's the same organization I registered with and not an impostor. Plaintext can be avoided without certs; just a plain diffie-hellman handshake and a stream cipher works as well for that. (And in fact is how SSL works, other than the cert exchange. So just remove the latter.) But of course, then we wouldn't need an entire industry built around selling certs. We'd just need a few simple protocol changes...
HTTPS doesn't do much good if the country in question implements transparent proxies at the borders of their national network infrastructure that decrypt SSL traffic, inspect the contents, then re-encrypt it with an SSL certificate issued by one of the authorities registered for that country (which is certainly within the realm of possibility for most governments). Have you ever looked at (let alone modified) the list of SSL authorities that your web browser trusts by default?
When I was in Vietnam recently, which blocks Facebook, they operated by intercepting DNS. They'd either make lookups fail or make them resolve to their own proxy. Before we realized this my wife uploaded a bunch of photos which then mysteriously disappeared overnight. We got around this by me firing up squid on my linode and using this as our web proxy, by IP address. (Authenticated obviously.) This way names are resolved in the good ole USA, geolocation says we're there (so get stuff in English), etc - AND the local government doesn't get to stick its grimy paws in my DNS lookups. To stops us they'd have to identify me personally, and spend resources on a single individual - and given we were foreign tourists they probably couldn't care less. After all, we'd leave in a few weeks and then we'd still post and say all the same things regardless. If we were locals we'd probably get on a watch list... They DID spend extra time on my exit processing at the airport, where the official wandered off with my passport and was gone 5-10 min.
At least they admit they don't have a clue as to what in hell is happening. In this case, they're asking others to duplicate their research, which is properly following the scientific method, so we can get more data and hopefully someone can come up with a hypothesis.
Actually, they're not even suggesting that - in fact, they seem perfectly happy being the only ones capable of building the devices. Which makes good sense if you want to make money selling them. Or scam investors. Your guess is as good as mine...
Yet they claim reliability in their own results, and commercial shipping of devices in the next year or so, so they either can reliably reproduce--and therefore accurately describe--working devices that others can build and test, or they are not telling the truth about something.
Just because they haven't published a paper detailing their setup and results doesn't mean it's automatically bogus. Ask yourself: if you have the choice between owning the rights to a revolutionary energy production system that could make you a multibillionaire overnight, or the choice of putting your name on a scientific paper outlining the details so others can get filthy rich while you get a pat on the back - what would you pick? So, yeah, it may be all a scam, but the absence of a paper isn't much of a reliable indicator. They're already stated themselves that they're far beyond publishing papers on this. Again, words are cheap and demos easily rigged.
I think 0.333... may only be a very close numerical approximation of 1/3
It's not. 0.333... isn't a number at all, it's a summation series of infinite length. That means it never ends. Ever. It's another exact way of stating 1/3. "0.3333" or any other finite series would be an approximation.
It's just like you can state cos(x) = 1-(x^2/2!)+(x^4/4!)+... This is an exact definition, not an approximation. If you want to say integrate cosines, or square the cosine, then you can derive another exact definition for the integral or square. In the case of the integral you might recognize that it can be refactored in terms of the series for the natural logarithm; the infinite series is a very useful analytical tool. And it's used to deal with exact definitions, not approximations.
I tried to look the problem up on Wikipedia, and all I got was inscrutable high-level math. From what I can gather, I seems like something that could be explained in layman's terms. Would someone be kind enough to do so?
3-sat: a series of 3-term clauses, of the form: _clause1_ AND _clause2_ AND... _clauseN_. Each clause[i] consists of (term1 OR term2 OR term3). Each term may be negated. The same term may appear in multiple clauses. The problem: find all solutions where this evaluates to true, if any. (Find the solution space.) An exhaustive search of the term space is NP-complete (this ought to be obvious).
Not a complicated problem to understand, at all. And a fairly useful one to solve. Since boolean logic isn't linear, gaussian reduction and elimination can't be used.
High ping, high jitter, low bandwidth once you factor in number of users and high cost, what could be better?
622 miles is really quite low and would only add about 10-12ms to the roundtrip. It's only a little more than the distance from San Francisco to Los Angeles. Jitter and bandwidth is a matter of pricing, and presumably there will be service tiers. Oversell it enough and it will be crap. Price it to manage demand and it could be excellent. If they can make this work anywhere in the world (why else 78 nodes) with an access device resembling a small book or hockey puck, then I predict monumental success.
Effectively what they're doing is turning sunlight into chemical energy. The process sounds complex at first glance, so can it be more efficient than other methods of capturing solar energy? From a technical POV the percentage of sunlight captured is interesting. But from a business POV the costs are interesting, and I think overall more important: real estate footprint, amortized capital costs, and operational costs. Where do these fall relative to other methods?
That's all fine and well for Google and their endless buckets of cash, but what about other companies, or importantly startups who want to get into the game.
H.264 is a standard; not a de-facto, or "industry" standard, but one adopted by an international standards body with wide representation. It publishes specs. If you build a part to do something with H.264 video, as long as it conforms to spec, it will work with others' products. You know, like the way any unlocked GSM phone works on any GSM network that operates on the same frequency band. It's ideal for startups, because you only need expertise in your own narrow product field, not in the entire much broader space. To build say an innovative silicon decoder you don't need to know how to build an encoder, because the elementary stream conforms to the standard. You don't need to know whether it came off a disc or ethernet. And while you occasionally run into interop issues this is positively nothing compared to the alternative of having inhouse expertise for *everything*. Not to mention the cost of dealing with some hacker who thinks they're doing something smart in the encoder, blowing up your taped-out decoder you've sent off to fab!
Compared to other costs, licensing fees are fairly trivial. $100k doesn't even buy a competent engineer for a year.
I take it you're not from the UK. Over here, "my client is likely to be executed" is the standard first line of defence against an extradition request. If you can make it sound halfway believable it's practically guaranteed to work. Also high up on the list are other standard human rights, e.g. "my client will be denied a fair trial". Our judges are quite astoundingly liberal and concerned by the rights of those before them in comparison to many other countries, and certainly in comparison to our politicians...
You'd think that "my client isn't suspected of ANY crime, and is only wanted for interviewing in an investigation of a crime that can only result in a fine comparable to a traffic violation" ought to be enough for the judge to tell Sweden to stop wasting resources and to go stuff it. Would the U.K. extradite someone who MAY have acted disorderly?
As for Sweden, by LAW they're not supposed to extradite anyone suspected of a political crime, and espionage is a political crime. Nobody anywhere extradites spies, friendly or not, unless they're wanted for non-political crimes as well. And then, the presence of political crimes may itself be a barrier to extradition. However, Sweden doesn't have a strong constitutional separation of its branches of government; parliament can pretty much hire and fire among the judiciary as they see fit. Politicians and bureaucrats are immune to lawsuits and criminal repercussions, and the political class generally protects itself. It's a monarchy where parliament has taken over the role of the King, not a republic. Basically, legal protections are very weak and they could just make something up - like they have before when dealing with inconvenient individuals. People have no constitutional right not to incriminate themselves, there is no system for warrants, and much other due process simply isn't there.
The problem I have with magazines (not so much e-books) is that I pay to subscribe to read an issue, but then have to continue to subscribe indefinitely to read that issue or a story in it again. But I already paid for it! Then there's the problem with family and friends, and fair use. If my wife buys a magazine (she in fact subscribes to Nature, Science, and about five others for work) I can read it, if for no other reason than CA is a community property state so it belongs to me as well. If she got it on the Kindle or iPad (she has both) I'd have to have her device on hand to read it, there's no way to manage community property between multiple devices. While I have her Kindle she couldn't access anything. And then, in the future, if she no longer needed a particular journal for work she'd lose all OLD ones she already paid for, and she might still need those (because she reported on something and needs to retain the original references). And on top of that, she can't just photocopy an article to show someone for reference - fair use, but blocked by DRM. She could print them and file, but now we're adding work and hassles AND still pay more for less. If, in fact, there is a way to print without running afoul of some obscure IP law.
If publishers could sort these things out I think digital would be vastly superior to paper; it certainly requires a whole lot less physical space to store it!
Wait, wait, wait. The whole point of publication is to open up your results so that other scientists can poke holes in it and the science can be redone and improved upon. Isn't it kind of a bogus statement say something like "this paper shouldn't have been published"? And with outrage, no less. Could the science really have been that bad and still be approved for publication to begin with? It must have been subject to at least a bit of peer review prior to its release.
How come no one was outraged about the guy who reinvented integration (http://science.slashdot.org/story/10/12/06/0416250/Medical-Researcher-Rediscovers-Integration)?!
The paper spurs justified criticism of methodology; that's perfectly reasonable. What ruffles people's feathers is using the resources of NASA to peddle their results in highly hyped press conferences. The lesson here is that if you're going to do that your research better be airtight. And that would include correlating the research by others using different methods. What they have in no way correlates with the presentation, which makes them look like used car salesmen.
Not to mention using a typedef so you can type PFOO instead of *FOO is like '#define ZERO 0" so you can type ZERO instead of 0. There's just no point to it.
Have a look at the info about LTE frequency assignments. OK, all you hams out there, how many MHz of the frequency band to carry a data rate of 21MHz at the various assigned frequencies? How much frequency spectrum is available? Divide X by Y and you get the number of simultaneous full-speed downloads.
Cell towers have phased antennas. This means there's a bunch of dipoles, usually in a circular arrangement. When a signal is received it can be angularly located by the relative phases of arrival at the dipoles. With a protocol incorporating chirping (a unit impulse convolved with a chirp system, then on receipt deconvolved to retrieve the pulse) the angle can be very precisely located. Once a transmitter is located the phased receiver can very accurately reject noise and other transmitters on the same frequency, based on their angle - in other words, if the phase(s) of the signal is wrong it's a reflection, echo, noise, artifact, or a different transmitter. This works so well that it can listen to a whole bunch of transmitters on the same frequency simultaneously by separation based on the transmission angle. The converse works as well; by aligning phases of the different antennas the transmission (virtual lobe) can be made highly directional (some will be out of phase to cancel, others in-phase for gain). In fact, the phase array can transmit to a whole bunch of receivers (stations) at the same time by simply adding up the signals for each antenna. So by using a bunch of phased antennas you create a large number of independent virtual narrow lobes of communication. As a mobile station moves around it's tracked and the lobe used to communicate with it changed as needed.
Finally, modern communications in the GHz band aren't based on FM modulated channels. They're time multiplexed. When a station transmits, it bursts for a few milliseconds. The rest of the time it's silent and other stations can use the frequency. In addition, there may be a channel arrangement, but in general collision domains can be self-managed by the stations while channel assignment requires centralized logic. But the biggest benefit is that the transmitter can output relatively high power (a few watts) during a brief period, vs relatively low power for longer periods. This improves range.
As a result of this, you can never get more bandwidth than that sustainable with a single virtual lobe. Your station will have a maximum power output (e.g. 3 mW/s, or in reality more like 3 W for 1ms); the better it can communicate with the tower the more packets it can send without exceeding its power limit. The same holds in the other direction, though it's limited by its single dipole receiver.
I can count the overages we've had on the 5GB plan over the last two years on one hand.
Now most of them are using the aircards supplementary to a regular connection. They typically have cable/dsl available at home, and are also occasionally in branch offices.
Yeah, I have Verizon Mobile Broadband (3G) and get around 1.5-3Mbit depending on location, in urban areas. Sometimes a lot less inside offices, hotel lobbies, etc, but rarely less than 300k. I use it all the time for everything from Skype and VoIP to mailing PDFs, downloading data sheets, occasional emergency software downloads, etc. And of course, ssh, git, svn, rsync. Usual engineering work stuff. I've shared it with coworkers at conferences. I've never gotten close to the 5GB cap, not once. I think I've maxed it out at about 1.5GB, which included sharing it with coworkers for an entire week during CES. I haven't looked into upgrading to 4G yet, but if it can fall back on the existing 3G net I think it'll be easy to sell me on it.
Now go build a radix tree for a routing table of 128 bit IPv6 addresses - let's see how well that works.
No problem. IPv6 uses, canonically, 64 bits for network routing and 64 bits for host identification. Routers have a forwarding tree and a separate local delivery tree for locally attached hosts. Contrary to your statement, a 64+64 bit address scheme makes for efficient router implementation.
Why is IPv6 not based on MAC adresses? I've never understood this. Every piece of electronics capable of connecting to a network has at least one unique hardware id already. Why do we need a new one?
MAC addresses aren't useful for routing since they end up scattered all over.
However, IPv6 in its most basic form actually uses the MAC address combined with a routing prefix. But it still needs DHCP for things like DNS and default gateways. IIRC it uses an all-zero prefix until someone (like a router) sends it something back to its actual assigned prefix, at which point it remembers. One of the early IPv6 headaches was that this could change (aka renumbering); as someone who in the late 90s went through large server codebases to make them IPv6 compatible this was a huge PITA. Servers configured to listen to different traffic on different interfaces really don't like to have their addresses suddenly changed. IPv6 at the time was cute but a useless academic exercise. Fortunately it has been fixed since...
Slashdot Mirror
Soon you'll have to have AV and actually pay attention to what you are doing same as the Windows guys
I doubt AV helps against people who install remote management tools. It's easier to make sure they can't install software in the first place.
>>>do some damn fact checking
"What's that?" - thousands of journalists and their professors. I don't think this is a deliberate misuse of the word trojan - just non technical people getting confused between DarkComet and BlackHole. Similar to how some non-tech people call megabits per second a "bandwidth" or 9.8 meters per second the "force" of gravity.
Lets not forget the ever popular "rate of speed"!
I'm not disagreeing with anything you say; I'm just pointing out that Apple making single-button mice has had a disciplining effect on UI design. Control-clicking isn't discoverable (just the way keyboard command shortcuts without menu equivalents aren't) so isn't something that can relied on for prominent functionality. Say spell checking a document. If you had to control click to bring up a menu to spell check, and couldn't find it in the main menu, then a LOT of users would never find it and would assume spell checking didn't exist. If that's a prominent, advertised feature of the program then this would be problematic and a huge call generator. So the UI design needs to be adjusted accordingly - most obviously by putting all prominent features in the main menu, or a button on a toolbar, or equivalent, or all of the above. You frequently actually see Windows software that *doesn't* adhere to such obvious basic principles of design; the people who complain about the single-button mouse on the Mac are very frequently exactly the people who would create such seven-headed UI horrors in the first place.
And that explains the pathological hatred of buttons? Simplifying is good, but there's a reason why Apple is the only company to use a 1 button mouse, and why the early mouses all had 3 buttons. Single button mouses suck.
There are good reasons for a single-button mouse. The first is that non-technical users simply don't understand the concept of right or left clicking, and adding buttons that can be clicked just confuses. I should click the icon? The normal button or the menu button? How many times? Why do I sometimes use one button and at other times the other? There's also a menu up top already with all the same things and more... when should I use which? The second reason is that Apple selling a single-button mouse (not sure if they still do though) forces UI designers to make their products single-button navigable. This has a great simplifying effect and removes the ability to do things like different menus in different places, and puts all commands in a single unified menu - making the product more discoverable. Finally, people who want multibutton mice generally have a strong brand preference anyway; I like my Razers while you might order a Logitech the first thing you do. So the Apple mouse mainly ends up in the hands of a novice with no existing preference and likely limited experience with computers.
They could have my user names and logins, but not my passwords or any other credentials. Just asking would pretty much make at least me turn around and walk out, likely rather annoyed at having wasted my time on such losers. Why don't they ask for my address book and business card files while at it? Can I have theirs?
I think Terry Gilliam has come up with the best explanation:
It used to be that studio execs were entrepreneurs. Businessmen with vision willing to take risks.
They still take risks, it's just that they aren't going to invest in something that excludes half the 10-25 year old movie going audience before it's even made. That's not a risk, that's throwing money away. This of course is why we get so much formulaic crap with overacting, lack of all subtlety, predictable plots, the equivalent of poop and fart jokes, and characters intended to appeal to a simplistic childish world view. Adults don't go see movies because it's likely to be an uninteresting way to spend an evening - and it's uninteresting because movies are made to appeal to a pre-adult audience.
The punchline: Unless you are using an atypically paranoid browser config, there are a Lot of CAs and subordinate CAs(some of them known-slimy, others known-incompetent), whose certs your browser will silently trust.
Yeah, this just illustrates how worthless web certs are. Any government with a CA within its borders can get a cert to impersonate anyone or anything. So can any criminal organization in the same country as long as their government is sufficiently corrupt. The way it SHOULD work is that another party proves knowledge of a secret I gave them when I registered, so I can know when I contact them that it's the same organization I registered with and not an impostor. Plaintext can be avoided without certs; just a plain diffie-hellman handshake and a stream cipher works as well for that. (And in fact is how SSL works, other than the cert exchange. So just remove the latter.) But of course, then we wouldn't need an entire industry built around selling certs. We'd just need a few simple protocol changes...
HTTPS doesn't do much good if the country in question implements transparent proxies at the borders of their national network infrastructure that decrypt SSL traffic, inspect the contents, then re-encrypt it with an SSL certificate issued by one of the authorities registered for that country (which is certainly within the realm of possibility for most governments). Have you ever looked at (let alone modified) the list of SSL authorities that your web browser trusts by default?
When I was in Vietnam recently, which blocks Facebook, they operated by intercepting DNS. They'd either make lookups fail or make them resolve to their own proxy. Before we realized this my wife uploaded a bunch of photos which then mysteriously disappeared overnight. We got around this by me firing up squid on my linode and using this as our web proxy, by IP address. (Authenticated obviously.) This way names are resolved in the good ole USA, geolocation says we're there (so get stuff in English), etc - AND the local government doesn't get to stick its grimy paws in my DNS lookups. To stops us they'd have to identify me personally, and spend resources on a single individual - and given we were foreign tourists they probably couldn't care less. After all, we'd leave in a few weeks and then we'd still post and say all the same things regardless. If we were locals we'd probably get on a watch list... They DID spend extra time on my exit processing at the airport, where the official wandered off with my passport and was gone 5-10 min.
At least they admit they don't have a clue as to what in hell is happening. In this case, they're asking others to duplicate their research, which is properly following the scientific method, so we can get more data and hopefully someone can come up with a hypothesis.
Actually, they're not even suggesting that - in fact, they seem perfectly happy being the only ones capable of building the devices. Which makes good sense if you want to make money selling them. Or scam investors. Your guess is as good as mine...
Yet they claim reliability in their own results, and commercial shipping of devices in the next year or so, so they either can reliably reproduce--and therefore accurately describe--working devices that others can build and test, or they are not telling the truth about something.
Just because they haven't published a paper detailing their setup and results doesn't mean it's automatically bogus. Ask yourself: if you have the choice between owning the rights to a revolutionary energy production system that could make you a multibillionaire overnight, or the choice of putting your name on a scientific paper outlining the details so others can get filthy rich while you get a pat on the back - what would you pick? So, yeah, it may be all a scam, but the absence of a paper isn't much of a reliable indicator. They're already stated themselves that they're far beyond publishing papers on this. Again, words are cheap and demos easily rigged.
I think 0.333... may only be a very close numerical approximation of 1/3
It's not. 0.333... isn't a number at all, it's a summation series of infinite length. That means it never ends. Ever. It's another exact way of stating 1/3. "0.3333" or any other finite series would be an approximation.
It's just like you can state cos(x) = 1-(x^2/2!)+(x^4/4!)+... This is an exact definition, not an approximation. If you want to say integrate cosines, or square the cosine, then you can derive another exact definition for the integral or square. In the case of the integral you might recognize that it can be refactored in terms of the series for the natural logarithm; the infinite series is a very useful analytical tool. And it's used to deal with exact definitions, not approximations.
I tried to look the problem up on Wikipedia, and all I got was inscrutable high-level math. From what I can gather, I seems like something that could be explained in layman's terms. Would someone be kind enough to do so?
3-sat: a series of 3-term clauses, of the form: _clause1_ AND _clause2_ AND ... _clauseN_. Each clause[i] consists of (term1 OR term2 OR term3). Each term may be negated. The same term may appear in multiple clauses. The problem: find all solutions where this evaluates to true, if any. (Find the solution space.) An exhaustive search of the term space is NP-complete (this ought to be obvious).
Not a complicated problem to understand, at all. And a fairly useful one to solve. Since boolean logic isn't linear, gaussian reduction and elimination can't be used.
High ping, high jitter, low bandwidth once you factor in number of users and high cost, what could be better?
622 miles is really quite low and would only add about 10-12ms to the roundtrip. It's only a little more than the distance from San Francisco to Los Angeles. Jitter and bandwidth is a matter of pricing, and presumably there will be service tiers. Oversell it enough and it will be crap. Price it to manage demand and it could be excellent. If they can make this work anywhere in the world (why else 78 nodes) with an access device resembling a small book or hockey puck, then I predict monumental success.
Effectively what they're doing is turning sunlight into chemical energy. The process sounds complex at first glance, so can it be more efficient than other methods of capturing solar energy? From a technical POV the percentage of sunlight captured is interesting. But from a business POV the costs are interesting, and I think overall more important: real estate footprint, amortized capital costs, and operational costs. Where do these fall relative to other methods?
It gets hot, steamy, wet and wild!!!
That's all fine and well for Google and their endless buckets of cash, but what about other companies, or importantly startups who want to get into the game.
H.264 is a standard; not a de-facto, or "industry" standard, but one adopted by an international standards body with wide representation. It publishes specs. If you build a part to do something with H.264 video, as long as it conforms to spec, it will work with others' products. You know, like the way any unlocked GSM phone works on any GSM network that operates on the same frequency band. It's ideal for startups, because you only need expertise in your own narrow product field, not in the entire much broader space. To build say an innovative silicon decoder you don't need to know how to build an encoder, because the elementary stream conforms to the standard. You don't need to know whether it came off a disc or ethernet. And while you occasionally run into interop issues this is positively nothing compared to the alternative of having inhouse expertise for *everything*. Not to mention the cost of dealing with some hacker who thinks they're doing something smart in the encoder, blowing up your taped-out decoder you've sent off to fab!
Compared to other costs, licensing fees are fairly trivial. $100k doesn't even buy a competent engineer for a year.
I take it you're not from the UK. Over here, "my client is likely to be executed" is the standard first line of defence against an extradition request. If you can make it sound halfway believable it's practically guaranteed to work. Also high up on the list are other standard human rights, e.g. "my client will be denied a fair trial". Our judges are quite astoundingly liberal and concerned by the rights of those before them in comparison to many other countries, and certainly in comparison to our politicians...
You'd think that "my client isn't suspected of ANY crime, and is only wanted for interviewing in an investigation of a crime that can only result in a fine comparable to a traffic violation" ought to be enough for the judge to tell Sweden to stop wasting resources and to go stuff it. Would the U.K. extradite someone who MAY have acted disorderly?
As for Sweden, by LAW they're not supposed to extradite anyone suspected of a political crime, and espionage is a political crime. Nobody anywhere extradites spies, friendly or not, unless they're wanted for non-political crimes as well. And then, the presence of political crimes may itself be a barrier to extradition. However, Sweden doesn't have a strong constitutional separation of its branches of government; parliament can pretty much hire and fire among the judiciary as they see fit. Politicians and bureaucrats are immune to lawsuits and criminal repercussions, and the political class generally protects itself. It's a monarchy where parliament has taken over the role of the King, not a republic. Basically, legal protections are very weak and they could just make something up - like they have before when dealing with inconvenient individuals. People have no constitutional right not to incriminate themselves, there is no system for warrants, and much other due process simply isn't there.
The problem I have with magazines (not so much e-books) is that I pay to subscribe to read an issue, but then have to continue to subscribe indefinitely to read that issue or a story in it again. But I already paid for it! Then there's the problem with family and friends, and fair use. If my wife buys a magazine (she in fact subscribes to Nature, Science, and about five others for work) I can read it, if for no other reason than CA is a community property state so it belongs to me as well. If she got it on the Kindle or iPad (she has both) I'd have to have her device on hand to read it, there's no way to manage community property between multiple devices. While I have her Kindle she couldn't access anything. And then, in the future, if she no longer needed a particular journal for work she'd lose all OLD ones she already paid for, and she might still need those (because she reported on something and needs to retain the original references). And on top of that, she can't just photocopy an article to show someone for reference - fair use, but blocked by DRM. She could print them and file, but now we're adding work and hassles AND still pay more for less. If, in fact, there is a way to print without running afoul of some obscure IP law.
If publishers could sort these things out I think digital would be vastly superior to paper; it certainly requires a whole lot less physical space to store it!
Wait, wait, wait. The whole point of publication is to open up your results so that other scientists can poke holes in it and the science can be redone and improved upon. Isn't it kind of a bogus statement say something like "this paper shouldn't have been published"? And with outrage, no less. Could the science really have been that bad and still be approved for publication to begin with? It must have been subject to at least a bit of peer review prior to its release. How come no one was outraged about the guy who reinvented integration (http://science.slashdot.org/story/10/12/06/0416250/Medical-Researcher-Rediscovers-Integration)?!
The paper spurs justified criticism of methodology; that's perfectly reasonable. What ruffles people's feathers is using the resources of NASA to peddle their results in highly hyped press conferences. The lesson here is that if you're going to do that your research better be airtight. And that would include correlating the research by others using different methods. What they have in no way correlates with the presentation, which makes them look like used car salesmen.
Not to mention using a typedef so you can type PFOO instead of *FOO is like '#define ZERO 0" so you can type ZERO instead of 0. There's just no point to it.
Ever see Windows code declare anything const?
Have a look at the info about LTE frequency assignments. OK, all you hams out there, how many MHz of the frequency band to carry a data rate of 21MHz at the various assigned frequencies? How much frequency spectrum is available? Divide X by Y and you get the number of simultaneous full-speed downloads.
Cell towers have phased antennas. This means there's a bunch of dipoles, usually in a circular arrangement. When a signal is received it can be angularly located by the relative phases of arrival at the dipoles. With a protocol incorporating chirping (a unit impulse convolved with a chirp system, then on receipt deconvolved to retrieve the pulse) the angle can be very precisely located. Once a transmitter is located the phased receiver can very accurately reject noise and other transmitters on the same frequency, based on their angle - in other words, if the phase(s) of the signal is wrong it's a reflection, echo, noise, artifact, or a different transmitter. This works so well that it can listen to a whole bunch of transmitters on the same frequency simultaneously by separation based on the transmission angle. The converse works as well; by aligning phases of the different antennas the transmission (virtual lobe) can be made highly directional (some will be out of phase to cancel, others in-phase for gain). In fact, the phase array can transmit to a whole bunch of receivers (stations) at the same time by simply adding up the signals for each antenna. So by using a bunch of phased antennas you create a large number of independent virtual narrow lobes of communication. As a mobile station moves around it's tracked and the lobe used to communicate with it changed as needed.
Finally, modern communications in the GHz band aren't based on FM modulated channels. They're time multiplexed. When a station transmits, it bursts for a few milliseconds. The rest of the time it's silent and other stations can use the frequency. In addition, there may be a channel arrangement, but in general collision domains can be self-managed by the stations while channel assignment requires centralized logic. But the biggest benefit is that the transmitter can output relatively high power (a few watts) during a brief period, vs relatively low power for longer periods. This improves range.
As a result of this, you can never get more bandwidth than that sustainable with a single virtual lobe. Your station will have a maximum power output (e.g. 3 mW/s, or in reality more like 3 W for 1ms); the better it can communicate with the tower the more packets it can send without exceeding its power limit. The same holds in the other direction, though it's limited by its single dipole receiver.
I can count the overages we've had on the 5GB plan over the last two years on one hand.
Now most of them are using the aircards supplementary to a regular connection. They typically have cable/dsl available at home, and are also occasionally in branch offices.
Yeah, I have Verizon Mobile Broadband (3G) and get around 1.5-3Mbit depending on location, in urban areas. Sometimes a lot less inside offices, hotel lobbies, etc, but rarely less than 300k. I use it all the time for everything from Skype and VoIP to mailing PDFs, downloading data sheets, occasional emergency software downloads, etc. And of course, ssh, git, svn, rsync. Usual engineering work stuff. I've shared it with coworkers at conferences. I've never gotten close to the 5GB cap, not once. I think I've maxed it out at about 1.5GB, which included sharing it with coworkers for an entire week during CES. I haven't looked into upgrading to 4G yet, but if it can fall back on the existing 3G net I think it'll be easy to sell me on it.
Now go build a radix tree for a routing table of 128 bit IPv6 addresses - let's see how well that works.
No problem. IPv6 uses, canonically, 64 bits for network routing and 64 bits for host identification. Routers have a forwarding tree and a separate local delivery tree for locally attached hosts. Contrary to your statement, a 64+64 bit address scheme makes for efficient router implementation.
Why is IPv6 not based on MAC adresses? I've never understood this. Every piece of electronics capable of connecting to a network has at least one unique hardware id already. Why do we need a new one?
MAC addresses aren't useful for routing since they end up scattered all over.
However, IPv6 in its most basic form actually uses the MAC address combined with a routing prefix. But it still needs DHCP for things like DNS and default gateways. IIRC it uses an all-zero prefix until someone (like a router) sends it something back to its actual assigned prefix, at which point it remembers. One of the early IPv6 headaches was that this could change (aka renumbering); as someone who in the late 90s went through large server codebases to make them IPv6 compatible this was a huge PITA. Servers configured to listen to different traffic on different interfaces really don't like to have their addresses suddenly changed. IPv6 at the time was cute but a useless academic exercise. Fortunately it has been fixed since...