Just how much of a chance is it to trust to a signature of someone on some paper, that says someone did something, with that something being specified elsewhere, and that elsewhere being paid to deliver that assessment, and evaluated by the number of signatures it delivers?
Most certifications(indeed a lot of diplomas) are not independant assessments, they don't all certify the same things to the same level, and accessing just what is certified is not always obvious. That means certifications remove a lot less risk than you seem to think.
I'm not saying they are bad though(I have some, some of them are good). But the idea that just holding the paper has any worth is laughable. EVERY ONE has to be cross-checked with the issuer before it should even be considered in the evaluation of a candidate. That means they can be used to tell candidates apart, but they cannot be used, like they are in most places, as a litmus test to accept a candidate, or to quickly sort a large number of candidates. They can do good, once you've whittled the less down to 5 or less.
RIAA: I own the content but you may use it User: If I pay you I own my copy, that's not negotiable. RIAA: Ownership is not something we're willing to give you. User: Well my money is not something I'm willing to give you, let's see how much content you can produce without an audience. RIAA: Government, User is using unfair negotiating tactics. User: Unfair? BAH! You're paid to encourage you to produce content. It's not a need, it's a want, but you need MY money. You will give me what I want, or you will get no money. Government: IANAL but I will ask counsel. Counsel: User is quoting straight from the history of copyright, the law says he's right, until we can change the law, no matter how much RIAA pays.
DRM negotiation in a perfect world, except if you're the RIAA
Do you seriously think a separatoin of State and Business is possible?
IANAL, but impossible necessary things don't stop being necessary, just because they are impossible. They're just impossible, and because they are necessary to the definition of a "Free" state, just like Microsoft produces non-Free software, they define western governments in general as non-Free. That in this particular case, it's quite legitimate for a French or a German not to be reminded every day of a lost loved one, or that a lost loved one may have been a war criminal, because that would border on harassment in many jurisdictions, means we have a hard(NP-Hard) problem.
How do we prevent an ideology who has been found guilty in court of crimes against humanity, from being hate and race-based, from promoting itself as hate and race-based? How many radical groups attract converts, not for their ideology or their ideas, but "who they are against"? When "who they are against" are a group of people that cannot be "discriminated against"(circa laws against discrimination in your own jurisdiction), you have two people who are arguing the position of a comma in a law. You have one person's face, and one person's fist, and the law becomes just how hard the fist has to hit before it's a crime...
As I said this is a hard problem, how do you determine the difference between the meaning of two relatively innocuous terms: National and Socialist, who basically mean "Us" and "Together", and their use as the label of a party who was identified as the perpetrator of (the numbers aren't all agreed on in all particulars) anywhere between 6 and 26 million dead. How do you tell the difference between "they had good words"(defensible, at least in the abstract, humans have defined "Us" vs "Them" since time immemorial) and "they had the right idea in killing those people"(which cannot be entertained lightly by a consciencious, sane human being even in numbers in single digits: I am trying to clarify the process in which law enforcement can express its reasoning here, because I believe that such a law enforcement process has to be transparent and understandable to someone without a law degree, not presenting my views). How do ideologies of damage and destruction can be defended, or find new converts, when the damage and the destruction itself are "not guaranteed rights" even when their espression in their abstract are inalienable rights? In this case, I'd redirect our readers to look a little closer to home, perhaps to Criminal Biker Gangs, who commit crimes, and where law enforcement officials demand methods to deal with such organised groups, and who are often denied, because "the right of association" is paramount; indeed it is paramount, yet its application, aka the details can certainly be argued for a more contextual interpretation: while the right to join into a group may certainly be a right, when this right is used to cover evidence of one or more crimes, we have to pick our poison... In light of the Enron and Worldcom scandal, we can also wonder just how many "legal associations" have been turned into a sham for using "rights" to hide evidence of crimes, or to protect those who think such schemes up from prosecution. How many people are hurt by such crimes? How many lives broken? What about the purchasing of political power through less-than-legal methods? Is a country where millions are affected by such crimes less or more free?
How can we tell? We usually can't, usually because the small, yet emotional cases can certainly be used as cover, for the large, unemotional cases that really affect our daily lives in a significant manner. How many policemen checking for "illegal content" do European governments in general have to task to finding paraphernalia vs locating child porn traders?
Europe is not more or less free than America or Canada, no, we're all non-free, to the last of us.
With Law enforcement demanding more and more tools, and tools that are more and more bigbrotherish each time, how long until we actually have laws that say people who mindlessly use tools without using their heads do jailtime equivalent to the judicial error they cause? (You accuse wrongfully someone of second degree murder because you didn't do a proper investigation, you do second degree murder time). Oh, and let's put it in the books that if a police union(or police department) tries to cover up something like this, not only do they commit a felony, but they share the punishment too.
Maybe this way everyone will be happy. We'll be giving law enforcement tools, but they'll actually be afraid of using them(and messing up). Fear of messing up seems to be underrepresented IMHO.
Now, IANAL but what I read on this topic seemed to indicate that only binaries would be signed, so even if you had the source, you'd be running untrusted binaries without any capacity to get them signed. This would of course be ok for IBM, but would cancel the benefit of having the source(you can't build a working binary from it). Maybe GPL4 can say that the source you get from a developer has to "be usable to generate a working binary equivalent to the binary you receive from vendor" next time...
I'd never thought I'd defend this kind of decision, but they have an (admittedly small) point:
Without some assurance of the identity of the originator of an email, making it a crime to send an email is an invitation to framing others. Also, it's not yet a crime to falsify headers or otherwise bypass the email identification process(there are laws that make it illegal to pretend you're someone else, even in email, but nothing says you can't pretend you're "nobody", nor does it mean you will necessarily be caught impersonating others(and a do-not-spam registry would be liable if it accused someone of spamming by going on what information is there, without some indication it's the whole of the true, that's called due process).
Let's hope the IETF finishes its sender-id mechanism soon, I can predict it will be illegal to circumvent it in short order, and the do not spam list will come into effect(but by then we won't need it anymore... We don't really need a do not spam law, as long as a person's email filters can gleefully butcher any email, with a 100% certainty that the email is legit, or not). We might find useful a clarification of the law that specifies explicitly that header-falsification IS indeed equivalent to the presenting of a false passport, in the eyes of the false document laws... But that could also be performed by jurisprudence. On second thought, let's get that clarification in the books right now, we don't want a judge to think that false id isn't false id if it's only 2.5 billion emails instead of just one passport...
Good thing ipv6 implies some ipsec, otherwise you could just anycast your machine to google's ip and watch em come, get a local multicast route from your uplink, and it'd be almost sniffing traffic. Just how easy is it to spoof pim packets anyways?
the net itself wasn't down, just happened that some centralized services for largish, geographically spread hosts were compromised all at the same time. That many of those are used mostly by end users makes it look like the net itself is damaged, but it's only edges of the network that are affected, not the core.
I have to wonder just how much of normal/not normal is due to the fact that we have a single inventor, someone whose revenue can be tracked pretty easily, and whose "development expenses" are similarly simpler, as opposed to a large corporation(say Microsoft). Would we be saying the same thing if a corporate juggernaut only got 3 million from a patent, and contrasted it with their investment?
I'm a lot more sympathetic with the single inventor case than with the corporate patent owner, and I'm sure I'm not the only one, but oddly enough, it seems the system rewards the corporate case a lot more(tax deductions presume you have a lot of money, for one).
Just how much of the problem with patents is because they are transferrable? I gotta wonder(licensing a company to make your product is one thing, selling them your patent is another).
You should rather say: they have to find something people are looking for, that only makes sense paying for as a service, that isn't offered by a cellphone. Inventing a new "thing" that gets charged monthly will only work if that "thing" doesn't make sense with an up-front fee.
While for manufacturers, pay-per-month is a great idea, for consumers, it's the worse idea, especially in a device you intend to keep(because it has your data in it). With the cellphone, we accept it partly because the alternatives to monthly billing can suck more, and partly because it makes sense to pay for access to the network(which doesn't belong to us, and whose value is created by other people: not the telco).
If this is true(and considering what I've seen of Aol's due process, I have no reason to doubt it) this just shows the actual damage from SPAM:
There is so much volume, it becomes uneconomical to have actual oversight of complaints, and with actual moving targets(spammers don't sit still) oversight is the only thing preventing collateral damage...
Hence, SPAM is causing a lot of damage, and a large portion of it is collateral(innocent bystanders affected by false positives).
It all comes down to execution, not only do you have a frigging large target(to optimize what usually happens to be 30 to 90% of the target organisation's "operations") but you have to optimize also the entire process surrounding that. You also have to co-opt practically the entire organisation to your cause, because the organisation's productivity is usually a chain, only as strong as the weakest link. ERP is an attempt to organize production around software, for efficiency purposes, so writing the software is not so hard, writing the software that matches the organisation, so as to minimize the disruptions, the expectation mismatches and to actually increase efficiency and functionality is non-trivial(and yes, most of that is not, technically, software engineering, it's closer to industrial engineering or process engineering).
When you bring in quickbooks, the software keeps track of the activities of the company, and yes, it does introduce some changes, but ERP is both more complex, and more tightly coupled to the activities of the company. The biological equivalent of a new ERP install for a corporation is replacement of a sizeable amount of the blood vessels and nerves in the body(aka replacing the command and control, and replacing the logistics channels to function with the new controllers). The proper title for ERP integrators should be "Corporation Heart and Brain Surgeon". Maybe seeing that on business cards, admin would finally get the message about just what they are inviting consultants to play with.
I believe you(got that from another thread about SCOX) have a minimum amount of shares to get issued paper, and there are fees to get them(s&h) and admin fees as well.
You think of it that way, and I might think of it that way, but until we explain it to Joe Sixpack and make it stick, we're a pair of irrelevant technophiles discussing on Slashdot.
Capitalism isn't simply to blame. The alternatives have been tried, and similarly failed, because men in power have too few incentives to remain honest. Aka that slashdot favorite, Power corrupts, absolute power corrupts absolutely. A friend's email sign actually contains part of a solution:
Diapers and politicians should be changed often, and for much the same reason.
Early ideas tried around the birth of Democracy(I believe it was the league of Delos) actually had everyone having a turn being the head honcho, for a while, it was held that the duties and privileges of rulership were best served equally by everyone(although they had restrictive ideas about who could be "anyone"). Perhaps that is an idea, with the new technologies, we can impact our displeasure to politicians within 24 hours. We definitely need a form of control over our elected representatives, that keeps us more in control. Those that stray shouldn't need to wait 4 years before we show them the door... We need to tighten the reigns, technically they are working for us, and when they screw up, it's our fault, least we can do is not endure them and their failures their failures we will endure anyways, so at least, let's clean house when it happens...
Nice way of putting it, guess you're reduced to convince the 49 other states to fix the problem(the fcc and morality shouldn't mix) instead of moving. Its kinda scary that you think option 2 is more acceptable really...
I've never understood what Linux people are talking about when they say that Linux 'runs faster' than Windows.
I remember about hmm back before Linux 2, the speed difference was in the handling of interrupts(Windows back then also had ridiculously small memory space and virtual space limits). That's over 8-9 years ago WindowMaker/AfterStep were actually more in vogue than the KDE/Gnome offerings then, who were practically "upstart projects", Sun's OpenWindows ported to linux was also popular back then. Then Linux 2 came up, it was faster, stable, then Windows basically caught up, then Linux 2.2 came up, and added many features, and optimised some things, but the difference wasn't as noticeable, then 2.4 came up, and it was a speed demon, except for X(which to keep up with the windows improvements, needed video hardware acceleration support). Now with 2.6, and hardware accelerated graphics on a powerful machine, Linux is still a little faster, but to see the difference, you really need to do what most people only do with Linux: remove running programs you don't use. In some cases, the difference is pretty dramatic. Of course, it never really shows in competitive benchmarks(which usually use bare-metal machines, not pre-junked seven themes, iconbar/taskbar needs two rows just to fit installations). That Linux is less vulnerable to software accretion, because of better package managers, may also be a factor, but with lots of people reformatting every six months, in both camps, clueful people almost never see just how bad it gets...
Windows 2000 is probably still the fastest desktop for use(Windows XP is optimised more for boot time), provided you have an uncluttered system, and relatively recent/fast hardware(which is one of the reasons Microsoft was pushing manufacturers not to OEM 2000 with machines for a while when XP came out, it made XP look bad). As for linux desktops being slower than this, It's quite possible, depending on hardware(as an experiment, you might want to try windows 2000 and XP(in client mode) in a vmware windows, compare its graphics performance to linux clients) So far my testing shows Linux reacts better(speed wise) to the virtualized hardware, because the Windows speed boost come with directly hooking into the hardware, but when they go through the vmware shim, the fact that the linux kernel is smaller/leaner makes it edge out recent windows(Win98se is faster in the vm(smaller), but predictably, less stable). (Linux in a VM is actually faster in desktop performance than native kde-cygwin performance on that box, for that matter) This on an Athlon 1800+ with 756MB RAM host.
The fact that it's easier for Linux to switch to a lighter/less cluttered windows manager than for Windows(LiteStep is good though:) ) means that it's also easier for someone who finds his system slow to increase performance, while increasing ram almost universally helps, having less bytes to move around can make a system fly...
You might want to read up on this thread, and wonder with us how to improve the motion capture devices, just how sophisticated they have to be to capture full facial motions. IMHO They aren't quite there yet, most motion capture points capture perhaps a hundred points total for the head, when the face alone more than likely has half-again that number. Another factor to keep in mind is that these suits must be quite uncomfortable for the captured actor, who is then discouraged from making some of the most subtle motions. Perhaps putting trackable dots(RFID Tags?) sticking on the face, and tracking them externally, instead of using a heavy suit and mask, might be a better solution.
The difference is that paintings are static, therefore inherently non-threatening. Animations also have to stay both coherent, and realistic. This might actually be a result of the overuse of the motion-capture technologies(having a suit track the motions of a human, then take those motions, and reproduce them on a cgi).
Why is this important? well because with the technique, you track the motion of the bodymass of the actor, along with his skeletons, you don't track the motion of the texture of the human. Our minds are used to tracking the motions of even the blood inside a moving human body, to identify intent as well as capacity to threaten, so even seeing the sway of the body hairs of an opponent can contribute or detract from realism. That noone made a motion capture suit that can track that much detail(indeed most motion-capture suit obfuscate some of those, as they enclose the human in question, not even allowing sweat to escape) means that all motion-tracked(my word, you are free to trade me a better one) games will lack those telltales as muscles shifting, lipid flow, blood derivation or sweat traces, and those are all used by our instinctive mind as proof of "real human threat, approach with caution" or "woohoo matable member of the opposite sex, approach with caution if weapon is in view, otherwise, strut a bit" as opposed to "something fishy, alert alert alert". The last case has a bad effect in games because:
it prevents suspension of disbelief by engaging suspicion reflex
it leaves the primitive brain without a preprogrammed response, which makes the gamer somewhat uncomfortable (it's going improv without a script after all)
our higher brain functions may be unaffected, but they are pretty far from our pleasure centers, so pleasing the higher brain functions exclusively doesn't work as well as exciting the higher brain functions and eliciting survival/reproduction/lower brain reflexes or pleasure
As for the roomba, anyone notice how most cars also end up having super-deformed puppy faces on them? We thrive on the familiar, so using pet shapes, which are familiar and reassuring, works better than super-futuristic shapes, which is why the 60's fashion of "spaceclothes" never caught on since.
Since radio waves don't respect state borders, that makes broadcast waves a federal matter.
I can assure you that if broadcast was more subject to state regulations, there wouldn't be just three networks, you'd have 150 networks, organized in three cartels, just so they can take advantage of the loosest state policy when it comes to advertising, yet stay out of trouble at all the others. The Media wouldn't, however, be more free.
I've read somewhere, that there is a current of thought that the Ice Age was unleased by the global warming period, and that it's a sort of "correction" for global warming. Let me explain in more detail:
Global warming causes less ice at the poles(not really anyone notices), that turns up as coastal water, that water absorbs more heat from the sun, and keeps it in water(away from land), this in turns makes the land colder and colder, until bam you have an ice age. Exactly how long it lasts is until the sea freezes back enough that the landmass can regain its prior heat.
N.B. Global warming is a global phenomenon, nothing in it says humans have to be anywhere near where it's warmer... That's why we don't much like the idea...
You might also get away with interesting the uninformed by saying that Riemann's Function (not Conjecture, although they may be related, my math proof skill reading sucks) has some applications in cryptography.
I believe Stephenson mentions this in the Cryptonomicon(I believe he is quoting some early work done about the time of the Enigma cipher, but I can't find my copy at present).
Although I believe algorithms and implementations tend to stay away from using it, for a code size/in-place memory size/compute load perspective. (The Z function's Magnitude is quite large, at least without specialized instructions to deal with powers of complex functions).
While my postulate that this is related to cryptography may not be strictly true, at least, in currently-used implementations, it may yet serve to interest people to this applied science. Which wouldn't do that much harm, now would it?
On a related note, I keep reading what I can understand of the proof, and I can't help but noticing there's a LOT of hypotheses in there, perhaps he's proving the theorem for only a restricted set of conditions?
I wonder just how many people will react to a government body failing to do it's most important mission(as a principle, and not necessarily in the letter of the law) as a reason to abolish said government body. I wonder if we'd do the same thing with other fields. Insufficiently-tested medicine is still making it to the market and harming Americans, let's abolish the doctor's association.
Replacing the FCC with a body whose task is to monitor the Media and other for from undue politicial influence(yes that means both the parties AND the lobbies like the RIAA) might do a lot more good. Although, in this case, protecting said body from those influences would be the first hurdle. Maybe if the American people could do something without the Parties and the Lobbies actually having a say beforehand would be useful at that(the public doesn't have a tool that's untainted by the political movement at present, with which to act in its own unpartisan interest, while both parties are very partisan, but aren't exactly opposed, leaving a lot of people under-represented.
That the fact that the FCC is tasked with regulating communications, which communications ARE vital to both national-security interests, financial market regulatory compliance, integrity of the electory process and even the judicial system highlights just how important that this group be watched. The problem is, that right now, noone is watching the guardians... The political representation of the FCC should be just as important as the House of Representatives itself, as the power of communications, expressed through among other things, the Media, can have more far-reaching consequences than some laws... That noone is watching the FCC when it plays with that "field" is a sign that some interests are more represented than others.
That the FCC is trying to police content, and not even notice the effects of Media Consolidation should worry us a lot more than it does.
Perhaps the USA doesn't need an FCC, because its budget would be better applied at actually having anti-cartel laws that have teeth, and catch more than the occasional careless perpetrator, that I can believe. But not because the market can police itself, right now, it's more like because the police that's supposed to be watching the market is too busy watching nipples.
Actually, it's quite the opposite. OpenBSD's creed of "Security first" condemns it to a sort of "media obscurity" (nothing kills a story like "It doesn't do anything fancy, and just works") yet Theo's colourful disagreements with practically everyone under the sun keeps the mindshare of OpenBSD alive and well.
Couple that with their habit of doing things differently from everyone else just once in a while to keep track of who's watching, and you have a winner.
extra computing power is equivalent to about 90% of the computing power of the original machines, in terms of computing farms(if you're familiar with RAID technology, this is equivalent to RAID 0). Web serving requires a different kind of cluster, a high-availability cluster, where each machine is closer to RAID 5(or even RAID 1 in many cases), the use of the resources is completely different. Web service would also require shared storage, something which your pentiums aren't equipped to do(unless you have a very strange bunch of pentiums). Compute clusters like the top ten list basically are new machines, which calculate things. And only distributed programs(seti@home, distributed.net, etc...) match them currently in power, in fact, seti and the rest is mostly the application of grid technology, to an untrusted environment. Owning such a cluster makes it a trusted environment(you own all the machines and control which programs run, and can benchmark and fault detech to your heart's content). But coming back to your pentiums, if you aren't selling cpu power now, you aren't likely to benefit much from setting up a compute cluster.
Slashdot Mirror
Just how much of a chance is it to trust to a signature of someone on some paper, that says someone did something, with that something being specified elsewhere, and that elsewhere being paid to deliver that assessment, and evaluated by the number of signatures it delivers?
Most certifications(indeed a lot of diplomas) are not independant assessments, they don't all certify the same things to the same level, and accessing just what is certified is not always obvious. That means certifications remove a lot less risk than you seem to think.
I'm not saying they are bad though(I have some, some of them are good). But the idea that just holding the paper has any worth is laughable. EVERY ONE has to be cross-checked with the issuer before it should even be considered in the evaluation of a candidate. That means they can be used to tell candidates apart, but they cannot be used, like they are in most places, as a litmus test to accept a candidate, or to quickly sort a large number of candidates. They can do good, once you've whittled the less down to 5 or less.
RIAA: I own the content but you may use it
User: If I pay you I own my copy, that's not negotiable.
RIAA: Ownership is not something we're willing to give you.
User: Well my money is not something I'm willing to give you, let's see how much content you can produce without an audience.
RIAA: Government, User is using unfair negotiating tactics.
User: Unfair? BAH! You're paid to encourage you to produce content. It's not a need, it's a want, but you need MY money. You will give me what I want, or you will get no money.
Government: IANAL but I will ask counsel.
Counsel: User is quoting straight from the history of copyright, the law says he's right, until we can change the law, no matter how much RIAA pays.
DRM negotiation in a perfect world, except if you're the RIAA
IANAL, but impossible necessary things don't stop being necessary, just because they are impossible. They're just impossible, and because they are necessary to the definition of a "Free" state, just like Microsoft produces non-Free software, they define western governments in general as non-Free. That in this particular case, it's quite legitimate for a French or a German not to be reminded every day of a lost loved one, or that a lost loved one may have been a war criminal, because that would border on harassment in many jurisdictions, means we have a hard(NP-Hard) problem.
How do we prevent an ideology who has been found guilty in court of crimes against humanity, from being hate and race-based, from promoting itself as hate and race-based? How many radical groups attract converts, not for their ideology or their ideas, but "who they are against"? When "who they are against" are a group of people that cannot be "discriminated against"(circa laws against discrimination in your own jurisdiction), you have two people who are arguing the position of a comma in a law. You have one person's face, and one person's fist, and the law becomes just how hard the fist has to hit before it's a crime...
As I said this is a hard problem, how do you determine the difference between the meaning of two relatively innocuous terms: National and Socialist, who basically mean "Us" and "Together", and their use as the label of a party who was identified as the perpetrator of (the numbers aren't all agreed on in all particulars) anywhere between 6 and 26 million dead. How do you tell the difference between "they had good words"(defensible, at least in the abstract, humans have defined "Us" vs "Them" since time immemorial) and "they had the right idea in killing those people"(which cannot be entertained lightly by a consciencious, sane human being even in numbers in single digits: I am trying to clarify the process in which law enforcement can express its reasoning here, because I believe that such a law enforcement process has to be transparent and understandable to someone without a law degree, not presenting my views). How do ideologies of damage and destruction can be defended, or find new converts, when the damage and the destruction itself are "not guaranteed rights" even when their espression in their abstract are inalienable rights? In this case, I'd redirect our readers to look a little closer to home, perhaps to Criminal Biker Gangs, who commit crimes, and where law enforcement officials demand methods to deal with such organised groups, and who are often denied, because "the right of association" is paramount; indeed it is paramount, yet its application, aka the details can certainly be argued for a more contextual interpretation: while the right to join into a group may certainly be a right, when this right is used to cover evidence of one or more crimes, we have to pick our poison... In light of the Enron and Worldcom scandal, we can also wonder just how many "legal associations" have been turned into a sham for using "rights" to hide evidence of crimes, or to protect those who think such schemes up from prosecution. How many people are hurt by such crimes? How many lives broken? What about the purchasing of political power through less-than-legal methods? Is a country where millions are affected by such crimes less or more free?
How can we tell? We usually can't, usually because the small, yet emotional cases can certainly be used as cover, for the large, unemotional cases that really affect our daily lives in a significant manner. How many policemen checking for "illegal content" do European governments in general have to task to finding paraphernalia vs locating child porn traders?
Europe is not more or less free than America or Canada, no, we're all non-free, to the last of us.
With Law enforcement demanding more and more tools, and tools that are more and more bigbrotherish each time, how long until we actually have laws that say people who mindlessly use tools without using their heads do jailtime equivalent to the judicial error they cause? (You accuse wrongfully someone of second degree murder because you didn't do a proper investigation, you do second degree murder time). Oh, and let's put it in the books that if a police union(or police department) tries to cover up something like this, not only do they commit a felony, but they share the punishment too.
Maybe this way everyone will be happy. We'll be giving law enforcement tools, but they'll actually be afraid of using them(and messing up). Fear of messing up seems to be underrepresented IMHO.
Now, IANAL but what I read on this topic seemed to indicate that only binaries would be signed, so even if you had the source, you'd be running untrusted binaries without any capacity to get them signed. This would of course be ok for IBM, but would cancel the benefit of having the source(you can't build a working binary from it). Maybe GPL4 can say that the source you get from a developer has to "be usable to generate a working binary equivalent to the binary you receive from vendor" next time...
I'd never thought I'd defend this kind of decision, but they have an (admittedly small) point:
Without some assurance of the identity of the originator of an email, making it a crime to send an email is an invitation to framing others. Also, it's not yet a crime to falsify headers or otherwise bypass the email identification process(there are laws that make it illegal to pretend you're someone else, even in email, but nothing says you can't pretend you're "nobody", nor does it mean you will necessarily be caught impersonating others(and a do-not-spam registry would be liable if it accused someone of spamming by going on what information is there, without some indication it's the whole of the true, that's called due process).
Let's hope the IETF finishes its sender-id mechanism soon, I can predict it will be illegal to circumvent it in short order, and the do not spam list will come into effect(but by then we won't need it anymore... We don't really need a do not spam law, as long as a person's email filters can gleefully butcher any email, with a 100% certainty that the email is legit, or not). We might find useful a clarification of the law that specifies explicitly that header-falsification IS indeed equivalent to the presenting of a false passport, in the eyes of the false document laws... But that could also be performed by jurisprudence. On second thought, let's get that clarification in the books right now, we don't want a judge to think that false id isn't false id if it's only 2.5 billion emails instead of just one passport...
Good thing ipv6 implies some ipsec, otherwise you could just anycast your machine to google's ip and watch em come, get a local multicast route from your uplink, and it'd be almost sniffing traffic. Just how easy is it to spoof pim packets anyways?
the net itself wasn't down, just happened that some centralized services for largish, geographically spread hosts were compromised all at the same time. That many of those are used mostly by end users makes it look like the net itself is damaged, but it's only edges of the network that are affected, not the core.
I have to wonder just how much of normal/not normal is due to the fact that we have a single inventor, someone whose revenue can be tracked pretty easily, and whose "development expenses" are similarly simpler, as opposed to a large corporation(say Microsoft). Would we be saying the same thing if a corporate juggernaut only got 3 million from a patent, and contrasted it with their investment?
I'm a lot more sympathetic with the single inventor case than with the corporate patent owner, and I'm sure I'm not the only one, but oddly enough, it seems the system rewards the corporate case a lot more(tax deductions presume you have a lot of money, for one).
Just how much of the problem with patents is because they are transferrable? I gotta wonder(licensing a company to make your product is one thing, selling them your patent is another).
You should rather say: they have to find something people are looking for, that only makes sense paying for as a service, that isn't offered by a cellphone. Inventing a new "thing" that gets charged monthly will only work if that "thing" doesn't make sense with an up-front fee.
While for manufacturers, pay-per-month is a great idea, for consumers, it's the worse idea, especially in a device you intend to keep(because it has your data in it). With the cellphone, we accept it partly because the alternatives to monthly billing can suck more, and partly because it makes sense to pay for access to the network(which doesn't belong to us, and whose value is created by other people: not the telco).
If this is true(and considering what I've seen of Aol's due process, I have no reason to doubt it) this just shows the actual damage from SPAM:
There is so much volume, it becomes uneconomical to have actual oversight of complaints, and with actual moving targets(spammers don't sit still) oversight is the only thing preventing collateral damage...
Hence, SPAM is causing a lot of damage, and a large portion of it is collateral(innocent bystanders affected by false positives).
It all comes down to execution, not only do you have a frigging large target(to optimize what usually happens to be 30 to 90% of the target organisation's "operations") but you have to optimize also the entire process surrounding that. You also have to co-opt practically the entire organisation to your cause, because the organisation's productivity is usually a chain, only as strong as the weakest link. ERP is an attempt to organize production around software, for efficiency purposes, so writing the software is not so hard, writing the software that matches the organisation, so as to minimize the disruptions, the expectation mismatches and to actually increase efficiency and functionality is non-trivial(and yes, most of that is not, technically, software engineering, it's closer to industrial engineering or process engineering).
When you bring in quickbooks, the software keeps track of the activities of the company, and yes, it does introduce some changes, but ERP is both more complex, and more tightly coupled to the activities of the company. The biological equivalent of a new ERP install for a corporation is replacement of a sizeable amount of the blood vessels and nerves in the body(aka replacing the command and control, and replacing the logistics channels to function with the new controllers). The proper title for ERP integrators should be "Corporation Heart and Brain Surgeon". Maybe seeing that on business cards, admin would finally get the message about just what they are inviting consultants to play with.
I believe you(got that from another thread about SCOX) have a minimum amount of shares to get issued paper, and there are fees to get them(s&h) and admin fees as well.
You think of it that way, and I might think of it that way, but until we explain it to Joe Sixpack and make it stick, we're a pair of irrelevant technophiles discussing on Slashdot.
Capitalism isn't simply to blame. The alternatives have been tried, and similarly failed, because men in power have too few incentives to remain honest. Aka that slashdot favorite, Power corrupts, absolute power corrupts absolutely. A friend's email sign actually contains part of a solution:
Diapers and politicians should be changed often, and for much the same reason.
Early ideas tried around the birth of Democracy(I believe it was the league of Delos) actually had everyone having a turn being the head honcho, for a while, it was held that the duties and privileges of rulership were best served equally by everyone(although they had restrictive ideas about who could be "anyone"). Perhaps that is an idea, with the new technologies, we can impact our displeasure to politicians within 24 hours. We definitely need a form of control over our elected representatives, that keeps us more in control. Those that stray shouldn't need to wait 4 years before we show them the door... We need to tighten the reigns, technically they are working for us, and when they screw up, it's our fault, least we can do is not endure them and their failures their failures we will endure anyways, so at least, let's clean house when it happens...
Nice way of putting it, guess you're reduced to convince the 49 other states to fix the problem(the fcc and morality shouldn't mix) instead of moving. Its kinda scary that you think option 2 is more acceptable really...
I remember about hmm back before Linux 2, the speed difference was in the handling of interrupts(Windows back then also had ridiculously small memory space and virtual space limits). That's over 8-9 years ago WindowMaker/AfterStep were actually more in vogue than the KDE/Gnome offerings then, who were practically "upstart projects", Sun's OpenWindows ported to linux was also popular back then. Then Linux 2 came up, it was faster, stable, then Windows basically caught up, then Linux 2.2 came up, and added many features, and optimised some things, but the difference wasn't as noticeable, then 2.4 came up, and it was a speed demon, except for X(which to keep up with the windows improvements, needed video hardware acceleration support). Now with 2.6, and hardware accelerated graphics on a powerful machine, Linux is still a little faster, but to see the difference, you really need to do what most people only do with Linux: remove running programs you don't use. In some cases, the difference is pretty dramatic. Of course, it never really shows in competitive benchmarks(which usually use bare-metal machines, not pre-junked seven themes, iconbar/taskbar needs two rows just to fit installations). That Linux is less vulnerable to software accretion, because of better package managers, may also be a factor, but with lots of people reformatting every six months, in both camps, clueful people almost never see just how bad it gets...
Windows 2000 is probably still the fastest desktop for use(Windows XP is optimised more for boot time), provided you have an uncluttered system, and relatively recent/fast hardware(which is one of the reasons Microsoft was pushing manufacturers not to OEM 2000 with machines for a while when XP came out, it made XP look bad). As for linux desktops being slower than this, It's quite possible, depending on hardware(as an experiment, you might want to try windows 2000 and XP(in client mode) in a vmware windows, compare its graphics performance to linux clients) So far my testing shows Linux reacts better(speed wise) to the virtualized hardware, because the Windows speed boost come with directly hooking into the hardware, but when they go through the vmware shim, the fact that the linux kernel is smaller/leaner makes it edge out recent windows(Win98se is faster in the vm(smaller), but predictably, less stable). (Linux in a VM is actually faster in desktop performance than native kde-cygwin performance on that box, for that matter) This on an Athlon 1800+ with 756MB RAM host.
The fact that it's easier for Linux to switch to a lighter/less cluttered windows manager than for Windows(LiteStep is good though
You might want to read up on this thread, and wonder with us how to improve the motion capture devices, just how sophisticated they have to be to capture full facial motions. IMHO They aren't quite there yet, most motion capture points capture perhaps a hundred points total for the head, when the face alone more than likely has half-again that number. Another factor to keep in mind is that these suits must be quite uncomfortable for the captured actor, who is then discouraged from making some of the most subtle motions. Perhaps putting trackable dots(RFID Tags?) sticking on the face, and tracking them externally, instead of using a heavy suit and mask, might be a better solution.
Why is this important? well because with the technique, you track the motion of the bodymass of the actor, along with his skeletons, you don't track the motion of the texture of the human. Our minds are used to tracking the motions of even the blood inside a moving human body, to identify intent as well as capacity to threaten, so even seeing the sway of the body hairs of an opponent can contribute or detract from realism. That noone made a motion capture suit that can track that much detail(indeed most motion-capture suit obfuscate some of those, as they enclose the human in question, not even allowing sweat to escape) means that all motion-tracked(my word, you are free to trade me a better one) games will lack those telltales as muscles shifting, lipid flow, blood derivation or sweat traces, and those are all used by our instinctive mind as proof of "real human threat, approach with caution" or "woohoo matable member of the opposite sex, approach with caution if weapon is in view, otherwise, strut a bit" as opposed to "something fishy, alert alert alert". The last case has a bad effect in games because:
it prevents suspension of disbelief by engaging suspicion reflex
it leaves the primitive brain without a preprogrammed response, which makes the gamer somewhat uncomfortable (it's going improv without a script after all)
our higher brain functions may be unaffected, but they are pretty far from our pleasure centers, so pleasing the higher brain functions exclusively doesn't work as well as exciting the higher brain functions and eliciting survival/reproduction/lower brain reflexes or pleasure
As for the roomba, anyone notice how most cars also end up having super-deformed puppy faces on them? We thrive on the familiar, so using pet shapes, which are familiar and reassuring, works better than super-futuristic shapes, which is why the 60's fashion of "spaceclothes" never caught on since.
Since radio waves don't respect state borders, that makes broadcast waves a federal matter.
I can assure you that if broadcast was more subject to state regulations, there wouldn't be just three networks, you'd have 150 networks, organized in three cartels, just so they can take advantage of the loosest state policy when it comes to advertising, yet stay out of trouble at all the others. The Media wouldn't, however, be more free.
I've read somewhere, that there is a current of thought that the Ice Age was unleased by the global warming period, and that it's a sort of "correction" for global warming. Let me explain in more detail:
Global warming causes less ice at the poles(not really anyone notices), that turns up as coastal water, that water absorbs more heat from the sun, and keeps it in water(away from land), this in turns makes the land colder and colder, until bam you have an ice age. Exactly how long it lasts is until the sea freezes back enough that the landmass can regain its prior heat.
N.B. Global warming is a global phenomenon, nothing in it says humans have to be anywhere near where it's warmer... That's why we don't much like the idea...
You might also get away with interesting the uninformed by saying that Riemann's Function (not Conjecture, although they may be related, my math proof skill reading sucks) has some applications in cryptography.
I believe Stephenson mentions this in the Cryptonomicon(I believe he is quoting some early work done about the time of the Enigma cipher, but I can't find my copy at present).
Although I believe algorithms and implementations tend to stay away from using it, for a code size/in-place memory size/compute load perspective. (The Z function's Magnitude is quite large, at least without specialized instructions to deal with powers of complex functions).
While my postulate that this is related to cryptography may not be strictly true, at least, in currently-used implementations, it may yet serve to interest people to this applied science. Which wouldn't do that much harm, now would it?
On a related note, I keep reading what I can understand of the proof, and I can't help but noticing there's a LOT of hypotheses in there, perhaps he's proving the theorem for only a restricted set of conditions?
I wonder just how many people will react to a government body failing to do it's most important mission(as a principle, and not necessarily in the letter of the law) as a reason to abolish said government body. I wonder if we'd do the same thing with other fields. Insufficiently-tested medicine is still making it to the market and harming Americans, let's abolish the doctor's association.
Replacing the FCC with a body whose task is to monitor the Media and other for from undue politicial influence(yes that means both the parties AND the lobbies like the RIAA) might do a lot more good. Although, in this case, protecting said body from those influences would be the first hurdle. Maybe if the American people could do something without the Parties and the Lobbies actually having a say beforehand would be useful at that(the public doesn't have a tool that's untainted by the political movement at present, with which to act in its own unpartisan interest, while both parties are very partisan, but aren't exactly opposed, leaving a lot of people under-represented.
That the fact that the FCC is tasked with regulating communications, which communications ARE vital to both national-security interests, financial market regulatory compliance, integrity of the electory process and even the judicial system highlights just how important that this group be watched. The problem is, that right now, noone is watching the guardians... The political representation of the FCC should be just as important as the House of Representatives itself, as the power of communications, expressed through among other things, the Media, can have more far-reaching consequences than some laws... That noone is watching the FCC when it plays with that "field" is a sign that some interests are more represented than others.
That the FCC is trying to police content, and not even notice the effects of Media Consolidation should worry us a lot more than it does.
Perhaps the USA doesn't need an FCC, because its budget would be better applied at actually having anti-cartel laws that have teeth, and catch more than the occasional careless perpetrator, that I can believe. But not because the market can police itself, right now, it's more like because the police that's supposed to be watching the market is too busy watching nipples.
Actually, it's quite the opposite. OpenBSD's creed of "Security first" condemns it to a sort of "media obscurity" (nothing kills a story like "It doesn't do anything fancy, and just works") yet Theo's colourful disagreements with practically everyone under the sun keeps the mindshare of OpenBSD alive and well.
Couple that with their habit of doing things differently from everyone else just once in a while to keep track of who's watching, and you have a winner.
extra computing power is equivalent to about 90% of the computing power of the original machines, in terms of computing farms(if you're familiar with RAID technology, this is equivalent to RAID 0). Web serving requires a different kind of cluster, a high-availability cluster, where each machine is closer to RAID 5(or even RAID 1 in many cases), the use of the resources is completely different. Web service would also require shared storage, something which your pentiums aren't equipped to do(unless you have a very strange bunch of pentiums). Compute clusters like the top ten list basically are new machines, which calculate things. And only distributed programs(seti@home, distributed.net, etc...) match them currently in power, in fact, seti and the rest is mostly the application of grid technology, to an untrusted environment. Owning such a cluster makes it a trusted environment(you own all the machines and control which programs run, and can benchmark and fault detech to your heart's content).
But coming back to your pentiums, if you aren't selling cpu power now, you aren't likely to benefit much from setting up a compute cluster.