The technology is fully intended to be used by a company like Youtube.
From what it sounds like from the press release:
The idea is that all content owners run a tool over their entire video archive. For each video it creates a sequence of 70 byte tags one per frame. This is stored in some sort of enormous database, which they provide to Youtube. When somebody uploads a new video, it takes each frame and computes the 70 byte signature for each frame. If a sequence of enough frames match or come close to matching between the uploaded video and something in the database.
If such a match is found, they want Youtube to delete the video, and frieze the user's account. Of course, the bigger the database the slow the matching will surely be. Not to mention surely the false positive rate is dependent on the size of the database you match against, so a bigger database will mean more false positives.
I think the sort of interface they show may be useful for some tasks, assuming it has access to traditional folders too.
Take the canonical example of taking photos of your low budget 9.0 megapixel (but 1.3 megapixel usable quality) point and click camera.
You dump all your photos out on your desktop, and begin sorting through them, separating into piles that correspond to the folders you plan on putting them in, or perhaps if you are enlightened, the piles correspond to tags you will put on them, and when fished sorting them, you could mass move or mass tag the whole pile.
Remember that fundamentally a pile is just a nameless folder. with an auto-generated icon. There is no reason why the interface could not offer the ability to name the piles, creating real folders.
Furthermore, there is also no reason why the "files" you see need to be the real files, rather than symlinks to the real files. Perhaps a search command results in a pile of items that respond to your query. The items are not the real files, just linked to them. You click the pile and chose the option to split into multiple piles based on file types. You then draw a lasso around the other piles to group them, and drag them to the filing cabinet, or something like that, which simply deletes the representation without deleting the file. To delete the actual file you would drag them to the trash can icon.
If you think like that you can see how the basic metaphor is serviceable without throwing everything out the window.
But the only so-called quantum cryptography I have heard of is hardly cryptography. It is a way to generate a shared secret between 2 computers that happen to be directly linked by optical fiber, while detecting any attempt at eavesdropping. It is worthless unless the two computers that want to communicate have a dedicated fiber line that connects them.[1]
Further while it prevents eavesdropping, it does not prevent full blown man-in-the-middle, where the fiber is severed, and converted into a pair of fibers with the adversary in the middle.
[1] Actually there are some tricks that allow for non-dedicated lines by using a chain of fully trusted relays with quantum secured channels between them, but that is still insane. A network of fully trusted relays simply cannot scale.
There is nothing binary-only about H.264. One of the best encoders for it is x264, which is available under the GPL. There are many decoder implementations, including several under GPL-compatible licenses. Basic decoding (as opposed to say hard-ware accelerated decoding) is generally a much easier task than encoding, although not without its challenges.
I think you are confusing this with the nonsense required by the Blue-ray consortium requiring kernel code to enforce secure communication pathways to the video card, which uses HDMI to ensure you never have access to the raw video frames. That is a completely different issue.
Ah, but there is a damn good chance your phone does have the hardware needed for hardwae support of VP8 and Theora. Most mobile devices supporting hardware acceleration are using a general purpose DSP, which should be able to accelerate those other codecs too. All that is lacking is software support for the acceleration.
For example every mobile device using a TI OMAP2 or OMAP3 system-on-a-chip has no h264-specific core but do have either a TMS320C55x DSP or an IVA2 or IVA2+ core. Those later cores can be use to accelerate any form of modern video decompression.
Thus, much of the market is a lot less invested in H.264 than you would think.
While that may be the legal theory, let us face it, being detained is a presumption that you are not innocent.
If you were really presumed innocent, they would just tell you that you are due to show up in court at some date and time, and hope you do, with failing to show up causing a rebuttable presumption of guilt.
But the system assumes you you are too dangerous to wander the streets until the Judge can get around to setting bail. Yeah, now that sounds like presuming innocence.
The real fact is that whole phrase has never had many real meaning. You are presumed guilty until cleared by the judge and/or jury. That just how the system works.
All the phrase really ever meant was that it was not permitted to hang a suspect until a court found them guilty.
Under the basic theory of copyright, this case would have been dismissed, and the court would have fined Omega several million dollars for wasting its time.
But unfortunately, nobody bothered to implement the abstract theory of copyright. They wrote some very confusing and poorly worded laws, that have special exceptions, and special exceptions to the special exceptions, and clauses that literally have no possible meaning that is consistent with the rest of the law, etc.
In the whole of this mess, the clauses regarding importation intended to keep out counterfeits were written incorrectly, such that they completely prohibit importation of a work without the permission of the copyright holder, even if the work was originally authorized by said copyright holder.
The courts should reject omega's claim, and rule that the importation was legal, but that would mean that importation of counterfeit works is also legal until congresses fixes the law. So that will never happen.
That case is just plain absurd. The argument in that case being that Importing and selling a branded item may violate the trademark rights of the brand holder? That argument could only possible be valid in the case of counterfeit goods. Otherwise trademark law just does not work like that. As long as I am selling genuine levi's jeans, I am using the trademark (the Levi's tag on the product) in a purely descriptive manner, which is part of what trademark law explicitly does not prohibit.
Unless I am missing something, I must conclude that the judge was clearly unfamiliar with the law, and depending on the lawyers to properly argue both law and fact, which is something judges do all too often despite the fact that that is forbidden. The results are verdicts that are inconsistent with the law itself.
For the English comment, even as an American I find the dropping of the second 'e' very odd.
It turns out that this is arguably a more correct spelling. Employee comes from employé, which is the past participle of the French verb Employer meaning, of course, "to employ". Being the past participle, according to french linguistic tradition it can also be used as a noun, having the expected meaning. For females, the word gets an extra unaccented 'e', becoming employée a feminine past participle, which if I remember correctly, is a form only ever used in noun form.
The only time credit cards get people into trouble is when they don't have the self control to pay it in full every time, or lack the ability to keep track of the amount charged, leading to a bill they just can't pay.
I for one would like to see a hybrid. I describe it as follows.
The way it would work is that in addition the the checking account the user has a special second account at the bank. When a charge authorization request is routed to the bank, if there is enough information in the checking account then a hold is placed on that amount just like with a debit card. However when the transaction completes, instead of the money being withdrawn, it is merely transered to the special second account.
Thus like a debit card, with only limited exceptions[0] by default it is not possible to spend more money than you have. However, unlike a debit card the money remains yours. The bank would not be allowed to touch the money in the special account without you authorizing it. When you get the credit card bill, you have all the money in that account to pay it in full. But it is still a credit card. If you don't pay it, the same consequences would occur. Notably this means that in the event of identity theft all the money is still in an account you own, so it is just transferred back to checking immediately upon you reporting the fraud.
Now this concept can go one of two ways. One is a soft option. Under this option they can transfer money back out of the special account at will, and when they get the bill, they can pay the minimum payment, pay in full, or anywhere in between. This option basically just prevents somebody from inadvertently charging more to the credit card then they have in the checking account.
The other is a hard option, where although the money in the special account remains yours, you cannot touch it, and it is only every transferred back in the case of reported fraud. Under this scenario, the bank would also not accept anything other than full payments on the credit card bill. This is the option for people who lack the self control to pay in full. That is their only option. Well they have to pay the charged amount in full. They may choose not to pay the late fees and interest when late, that is fine, but they won't be able to charge the card, even if they have the money in the checking account.
Now since these are credit cards, in addition to having authorization fail when there is not enough funds in the checking account, there would also be a credit limit. But some (many?) debit cards already have a monthy limit, which may prevent transaction even when there is enough money in the account.
[0] There is always the possibility of the completed transactions exceeding the amount listed in the initial authorization. That happens when you put a tip on the credit card at restaurants, among other occasions.
What I find interesting is that one of the biggest users of a BASE[0] non-relational database (a NoSQL database), namely Facebook, who uses Cassandra [1], has created an SQL style query interface named FBQL. The interface includes some rather advanced SQL features like embedded sub-queries in addition to the traditional selecting on joined tables.
Then again, that may be due in large part to the fact that they are using a database schema that is all but identical to a normalized schema used in relation ACID databases, and simply code with the expectation that the database may be inconsistent, so always expect broken references. That is not really the optimal way to use a non-relational database, but it works.
[0] Basically Available, Soft state, Eventual consistency. The somewhat the opposite of ACID.
[1] This can be pretty noticeable at peak hours, when you end up seeing an inconsistent database, one in which you are friends and are not friends with another user at the same time.
The problem is that the IPV6 process was done wrong.
The idea from the beginning should have been that except for edge-routers and edge-gateways no machine should ever speak both IPv4 and IPv6.
The process should have gone like this:
First define a packet format, and addressing scheme, with little regards for IPv4 compatibility. We did this step just fine.
Then define a way to encode IPv4 addresses into IPv6 address fields. I believe we also accomplished that just fine.
Design the mechanisms through which IPv6 packets are routed. (How are announcements done, and how to avoid insane sized routing tables). I believe we also accomplished this step.
Design a way to encapsulate and route IPv6 packets between two IPv6 networks separated by ipv4 networks. The IPv6 machines involved (except the edge routers) should see the entire IPv4 network as just one big hop between two IPv6 routers. I'm not very clear on how well we did here.
We also need a similar mechanism to allow IPv4 packets to be routed between two IPv4 networks separated by an IPv6 network. That should be easy enough. It does mean the IPv6 networks would need to keep an IPv4 routing table, but that is not a terribly big deal, especially since the next step will need that anyway. I'm not really clear on the status of that.
Now the big one. 6to4 NAT should have been required on all edge routers that connect an IPv4 and IPv6 network. The problem was that this was not standardized and implemented anywhere near early enough in the process. It always should have been the case that an ipv6 machine could talk to ipv4 machines simply by creating an ipv6 packet where the destination is an ipv4 address encapsulated as an ipv6 address, and releasing it onto the IPv6 network.
The net effect would have been that ipv4 machines can talk to all ipv4 machines completely oblivious to the existence of IPv6. Similarly IPv6 machines can talk to all IPv6 machines while obvious of any IPv4 machines the packets were tunneled over. Finally the IPv6 machines would have the same level of access to IPv4 machines as many consumers currently have.
IPv6 had other promising things were the ball was dropped. For example, it almost made Anycast addresses useful. It requires that IPv6 anycast machines also have a unicast address, and prohibits an anycast address from being appearing in the source field of the ipv6 address. So the replies to a packet sent to an any-cast address have a unicast address unique to the machine who received the packet in them. It would not have been hard to create a specification for TCPv4onIPv6 that said that if a SYN was sent to an IPv6 anycast address, when the SYN/ACK is received backed, note the source address, and use that address for the ACK and all further communications until the connection is broken. But from what I have heard, that was not done.
What a shame. Anycast addresses that supported TCP properly could have been quite useful. It eliminate the need for abusing the DNS system for ensuring routing to the nearest datacenter. Further in the datacenter, it could be used for load balancing. Have all the servers for a domain accept the anycast address, and have them connected to a special router. The machines could periodically indicate their load level to the router, which could then decide which of the machines to route new connections. Telling the new connections from the established ones is stateless, since only new connections would use the anycast address. The only state the router would need to maintain beyond that all need to maintain, is the load level on a relatively small number of servers (a.k.a. a static amount of additional state amounting to a couple of kilobytes.)
Kotaku's video proves nothing. That is an example of a game for the regular DSi that happens to use that one method. Indeed latest speculation is that it uses parallax barriers which act much like lenticlar lenses in only working from a very specific and narrow viewpoint. To overcome this, dynamic forms that track a person have been developed, but those are suspected to be too expensive to use in a DS.
Fair enough, but Kerberos can be used as the basis of real SSO if done right. For example, Firefox, shh, and many other utilities can be configured to forward Kerberos tokens to avoid needing to sign in to the remote machine. Window's own AD based Domain architecutre allows access to other Windows machines without authenticating even if Kerberos was not configured, so that will not need multiple sign ons.
If done right, one would not need to sign in more than once unless doing something really unusual. If one is doing something really unusual, a second sign-in is not a very onerous requirement.
This actually has two parts. One is a requirement that the correct prefixes are used. Although slightly confusing to some computer users, this is still the correct thing to do. I do not however advocate the use of the fully spelled out names of the binary units, as those are just awful.
The policy notably requires that 'k' and not 'K' be used for Base 10 values since KB is actually Kelvin-Bytes, a unit that is virtually never useful. It also mandates the use of "B" for bytes, which means that "b" will mean bits. IT should make that explicit, but using "b" incorrectly for Bytes is much more common than using "B" incorrectly for bits.
But there is a second part that is more controversial. Applications must favor the base 10 units over the base 2 units. Thus when both are listed base 10 must come first. When only one is given, and is not configurable, it must be base 10. If only one is shown but it is configurable, the default must be base 10.
That part of the new policy is one I think I object to. For example, A program displaying RAM sizes while only showing one prefix should show MiB or GiB by default, since ram is sold in base 2 sizes. It would confuse everybody if the ram size was reported only in units of "MB" or "GB".
The 3D environment move when you look around the NDS screen. I think it's done by some image recognition with the front facing camera.
Actually that is something he already described:
A final kind of 3D is monocular 3D. there there is one image but you eye track or tilt track in a way that lets you update the image for the new point of view.
I think Kotaku's speculation may well be wrong. There is a technology to do 3d without special glasses. Unfortunately it sucks. The technology is based on lenticular lenses.
But that is the only known way (of which I am aware) to do stereographic 3d without glasses on something that is remotely like a standard display.
Most SSO in organizations I've ever seen seem to work by connecting the user directory to Kerberos, and use that for authentication to everything. Depending on how well IT department set up token forwarding, you may need to enter your credentials to access many systems, but everything in the company, from signing into the Windows Domain, to authenticating to the database uses Kerberos, so you have exactly one password to remember.
You did mention the verbatim quote, but the general impression your post gave reinforced the idea that plagiarism is merely directly copying from a source. This impression is too wide spread even on Slashdot, mainly because of the crap idiot teachers spread. Therefore my posting was to help other readers understand what plagiarism really is, not to criticize your understanding, although I admit it did sound like the latter.
Of course, services like turnitin.com and other such tools don't distinguish between proper cited and quoted material, and material copied without attribution, considering both to be indicators of plagiarism, so if you have one of those idiot teachers that don't know what plagiarism really is, you would be best to avoid even properly cited quotations.
My guess is that since the optimal value for the cones varies between people anyway, they just split the difference between the average optimal. The result is that the image appears best to those closest to statistically average, but varies by person. That said, The brain is still rather good at smoothing over minor inconsistencies so the net impact may be rather small, but this may still be why some people have find Dolby3D the least pleasant to watch.
The "vertical" is relative to the eyeline as "horizontal". The images will always overlap horizontally relative to the floor of the theater. I'm pretty sure we are in agreement on that point. I was deliberate with the use of scare quotes around "vertically" in my original post.
The point I was trying to make: Consider standing on your head. Your right eye is now on the side of the body that would have had your left eye when upright. But with either linear or circular polarization when standing on your head your right eye, which is now on the left will be seeing the image meant for the eye on the right, and vice versa. That is just not going to work right. It would make anything that should appear recessed into the screen as sticking out of it.
But with your head turned only 90 degrees, the images will not overlap so as to form any coherent image. With circular technology the right eye is still seeing the same thing as when upright, as is the left eye. With linear technology the two are swapped. But for either they do not overlap along the plane of the eyes, while real world images would. Thus your mind would not be able to produce a coherent 3D image from the inputs.
It is also worth noting that despite that promise, GBP is still a fiat currency. The bank merely offers a currency exchange service where one of the currencies is GBP and the other is XAU.
Plagarism is not copying form a source. It is permissible to copy a definition from a source and use it as a direct quotation, as long as it is clear it is a quotation, and as long as you cite the source of the text.
Furthermore paraphrasing a unique idea without crediting the source of the idea is plagiarism, even though it is not an exact copy.
Plaragism is the use of an idea or wording that is not your own and not general knowledge without crediting the source. Specifically any relatively unique idea must be cited, and and borrowed wording must be clearly marked as such with the source given.
There is a general knowledge exception. Wording whose source is so well known that the source is implied may not need a specific citation. Traditional proverbs often fall into this category, although sometimes they fall into a category of general knowledge whose source it not known.
The general knowledge exception is more frequently found with ideas though.
Specifically, the concept of plagiarism (which is really only meaningful in an academic context, the rest of the world is only concerned about legal matters like Copyright infringement) is the use of specific ideas that are not your own without citing the source of the idea.
A paper that consists entirely of a long verbatim quote from some other source, but where the source is cited, and it has been made clear what was an exact quote is not plagarism. It would almost certainly fail to meet the requirements of the class.
On the other hand, if one paraphrases an idea they got from some other source, even without any exact quotes, if the source is not cited it is plagiarism.
However, plagiarism should not be the main concern of citing sources. The main concern should be to document where you got an idea, or a number, etc. One does not need to repeat all the evidence that supports an idea if one cites a source that provides said evidence, or equivalently one need not may a persuasive argument for some idea if one can site a source that provides said persuasive arguments. This allows you to focus on supporting your conclusion with evidence or arguments as the case may be, without having to justify every population statistic, or philosophical idea you use.
Citing sources can have other benefits too. It can help protect your reputation. If you used a source that should be reliable, and it is later determined the source had faulty data, by citing the source people will know that you were being mislead, and not that you were intentionally perpetuating bad data.
Actually, the law is quite clear that one cannot lose citizenship without intending to do so. So if one joins a terrorist organization, or a foreign military without the intent to relinquish citizenship, they remain US Citizens.
However, joining any group hostile to the United States does create a rebuttable presumption of the intention to relinquish citizenship.
Here is an abridgment of 8 USC 1481 (emphasis mine)
(a) A person who is a national of the United States whether by birth or naturalization, shall lose his nationality by voluntarily performing any of the following acts with the intention of relinquishing United States nationality—
[A list of actions including (among others) becoming a national of another country, treason, becoming a soldier in a hostile military, or a commissioned officer in any foreign military.]
(b) Whenever the loss of United States nationality is put in issue in any action or proceeding commenced on or after September 26, 1961 under, or by virtue of, the provisions of this chapter or any other Act, the burden shall be upon the person or party claiming that such loss occurred, to establish such claim by a preponderance of the evidence. Any person who commits or performs, or who has committed or performed, any act of expatriation under the provisions of this chapter or any other Act shall be presumed to have done so voluntarily, but such presumption may be rebutted upon a showing, by a preponderance of the evidence, that the act or acts committed or performed were not done voluntarily.
[End of section]
If you read the whole thing[1] you will see that loss of citizenship for treason only occurs
Now, the law is more concerned with whether the actions are considered voluntary, as some of these acts, especially the ones related to treason are very strongly presumed to always be done with the intention of relinquishing citizenship. However, others such as becoming a national of another country carry only limited presumption of intent to relinquish citizenship. For example, one is allowed to become a national of another country while retaining US citizenship, the courts have been very clear about that.
Slashdot Mirror
The technology is fully intended to be used by a company like Youtube.
From what it sounds like from the press release:
The idea is that all content owners run a tool over their entire video archive. For each video it creates a sequence of 70 byte tags one per frame. This is stored in some sort of enormous database, which they provide to Youtube. When somebody uploads a new video, it takes each frame and computes the 70 byte signature for each frame. If a sequence of enough frames match or come close to matching between the uploaded video and something in the database.
If such a match is found, they want Youtube to delete the video, and frieze the user's account. Of course, the bigger the database the slow the matching will surely be. Not to mention surely the false positive rate is dependent on the size of the database you match against, so a bigger database will mean more false positives.
I think the sort of interface they show may be useful for some tasks, assuming it has access to traditional folders too.
Take the canonical example of taking photos of your low budget 9.0 megapixel (but 1.3 megapixel usable quality) point and click camera.
You dump all your photos out on your desktop, and begin sorting through them, separating into piles that correspond to the folders you plan on putting them in, or perhaps if you are enlightened, the piles correspond to tags you will put on them, and when fished sorting them, you could mass move or mass tag the whole pile.
Remember that fundamentally a pile is just a nameless folder. with an auto-generated icon. There is no reason why the interface could not offer the ability to name the piles, creating real folders.
Furthermore, there is also no reason why the "files" you see need to be the real files, rather than symlinks to the real files. Perhaps a search command results in a pile of items that respond to your query. The items are not the real files, just linked to them. You click the pile and chose the option to split into multiple piles based on file types. You then draw a lasso around the other piles to group them, and drag them to the filing cabinet, or something like that, which simply deletes the representation without deleting the file. To delete the actual file you would drag them to the trash can icon.
If you think like that you can see how the basic metaphor is serviceable without throwing everything out the window.
But the only so-called quantum cryptography I have heard of is hardly cryptography. It is a way to generate a shared secret between 2 computers that happen to be directly linked by optical fiber, while detecting any attempt at eavesdropping. It is worthless unless the two computers that want to communicate have a dedicated fiber line that connects them.[1]
Further while it prevents eavesdropping, it does not prevent full blown man-in-the-middle, where the fiber is severed, and converted into a pair of fibers with the adversary in the middle.
[1] Actually there are some tricks that allow for non-dedicated lines by using a chain of fully trusted relays with quantum secured channels between them, but that is still insane. A network of fully trusted relays simply cannot scale.
There is nothing binary-only about H.264. One of the best encoders for it is x264, which is available under the GPL. There are many decoder implementations, including several under GPL-compatible licenses. Basic decoding (as opposed to say hard-ware accelerated decoding) is generally a much easier task than encoding, although not without its challenges.
I think you are confusing this with the nonsense required by the Blue-ray consortium requiring kernel code to enforce secure communication pathways to the video card, which uses HDMI to ensure you never have access to the raw video frames. That is a completely different issue.
Ah, but there is a damn good chance your phone does have the hardware needed for hardwae support of VP8 and Theora. Most mobile devices supporting hardware acceleration are using a general purpose DSP, which should be able to accelerate those other codecs too. All that is lacking is software support for the acceleration.
For example every mobile device using a TI OMAP2 or OMAP3 system-on-a-chip has no h264-specific core but do have either a TMS320C55x DSP or an IVA2 or IVA2+ core. Those later cores can be use to accelerate any form of modern video decompression.
Thus, much of the market is a lot less invested in H.264 than you would think.
While that may be the legal theory, let us face it, being detained is a presumption that you are not innocent.
If you were really presumed innocent, they would just tell you that you are due to show up in court at some date and time, and hope you do, with failing to show up causing a rebuttable presumption of guilt.
But the system assumes you you are too dangerous to wander the streets until the Judge can get around to setting bail. Yeah, now that sounds like presuming innocence.
The real fact is that whole phrase has never had many real meaning. You are presumed guilty until cleared by the judge and/or jury. That just how the system works.
All the phrase really ever meant was that it was not permitted to hang a suspect until a court found them guilty.
Under the basic theory of copyright, this case would have been dismissed, and the court would have fined Omega several million dollars for wasting its time.
But unfortunately, nobody bothered to implement the abstract theory of copyright. They wrote some very confusing and poorly worded laws, that have special exceptions, and special exceptions to the special exceptions, and clauses that literally have no possible meaning that is consistent with the rest of the law, etc.
In the whole of this mess, the clauses regarding importation intended to keep out counterfeits were written incorrectly, such that they completely prohibit importation of a work without the permission of the copyright holder, even if the work was originally authorized by said copyright holder.
The courts should reject omega's claim, and rule that the importation was legal, but that would mean that importation of counterfeit works is also legal until congresses fixes the law. So that will never happen.
That case is just plain absurd. The argument in that case being that Importing and selling a branded item may violate the trademark rights of the brand holder? That argument could only possible be valid in the case of counterfeit goods. Otherwise trademark law just does not work like that. As long as I am selling genuine levi's jeans, I am using the trademark (the Levi's tag on the product) in a purely descriptive manner, which is part of what trademark law explicitly does not prohibit.
Unless I am missing something, I must conclude that the judge was clearly unfamiliar with the law, and depending on the lawyers to properly argue both law and fact, which is something judges do all too often despite the fact that that is forbidden. The results are verdicts that are inconsistent with the law itself.
For the English comment, even as an American I find the dropping of the second 'e' very odd.
It turns out that this is arguably a more correct spelling. Employee comes from employé, which is the past participle of the French verb Employer meaning, of course, "to employ". Being the past participle, according to french linguistic tradition it can also be used as a noun, having the expected meaning. For females, the word gets an extra unaccented 'e', becoming employée a feminine past participle, which if I remember correctly, is a form only ever used in noun form.
Go figure.
The only time credit cards get people into trouble is when they don't have the self control to pay it in full every time, or lack the ability to keep track of the amount charged, leading to a bill they just can't pay.
I for one would like to see a hybrid. I describe it as follows.
The way it would work is that in addition the the checking account the user has a special second account at the bank. When a charge authorization request is routed to the bank, if there is enough information in the checking account then a hold is placed on that amount just like with a debit card. However when the transaction completes, instead of the money being withdrawn, it is merely transered to the special second account.
Thus like a debit card, with only limited exceptions[0] by default it is not possible to spend more money than you have. However, unlike a debit card the money remains yours. The bank would not be allowed to touch the money in the special account without you authorizing it. When you get the credit card bill, you have all the money in that account to pay it in full. But it is still a credit card. If you don't pay it, the same consequences would occur. Notably this means that in the event of identity theft all the money is still in an account you own, so it is just transferred back to checking immediately upon you reporting the fraud.
Now this concept can go one of two ways. One is a soft option. Under this option they can transfer money back out of the special account at will, and when they get the bill, they can pay the minimum payment, pay in full, or anywhere in between. This option basically just prevents somebody from inadvertently charging more to the credit card then they have in the checking account.
The other is a hard option, where although the money in the special account remains yours, you cannot touch it, and it is only every transferred back in the case of reported fraud. Under this scenario, the bank would also not accept anything other than full payments on the credit card bill. This is the option for people who lack the self control to pay in full. That is their only option. Well they have to pay the charged amount in full. They may choose not to pay the late fees and interest when late, that is fine, but they won't be able to charge the card, even if they have the money in the checking account.
Now since these are credit cards, in addition to having authorization fail when there is not enough funds in the checking account, there would also be a credit limit. But some (many?) debit cards already have a monthy limit, which may prevent transaction even when there is enough money in the account.
[0] There is always the possibility of the completed transactions exceeding the amount listed in the initial authorization. That happens when you put a tip on the credit card at restaurants, among other occasions.
What I find interesting is that one of the biggest users of a BASE[0] non-relational database (a NoSQL database), namely Facebook, who uses Cassandra [1], has created an SQL style query interface named FBQL. The interface includes some rather advanced SQL features like embedded sub-queries in addition to the traditional selecting on joined tables.
Then again, that may be due in large part to the fact that they are using a database schema that is all but identical to a normalized schema used in relation ACID databases, and simply code with the expectation that the database may be inconsistent, so always expect broken references. That is not really the optimal way to use a non-relational database, but it works.
[0] Basically Available, Soft state, Eventual consistency. The somewhat the opposite of ACID.
[1] This can be pretty noticeable at peak hours, when you end up seeing an inconsistent database, one in which you are friends and are not friends with another user at the same time.
The problem is that the IPV6 process was done wrong.
The idea from the beginning should have been that except for edge-routers and edge-gateways no machine should ever speak both IPv4 and IPv6.
The process should have gone like this:
The net effect would have been that ipv4 machines can talk to all ipv4 machines completely oblivious to the existence of IPv6. Similarly IPv6 machines can talk to all IPv6 machines while obvious of any IPv4 machines the packets were tunneled over. Finally the IPv6 machines would have the same level of access to IPv4 machines as many consumers currently have.
IPv6 had other promising things were the ball was dropped. For example, it almost made Anycast addresses useful. It requires that IPv6 anycast machines also have a unicast address, and prohibits an anycast address from being appearing in the source field of the ipv6 address. So the replies to a packet sent to an any-cast address have a unicast address unique to the machine who received the packet in them. It would not have been hard to create a specification for TCPv4onIPv6 that said that if a SYN was sent to an IPv6 anycast address, when the SYN/ACK is received backed, note the source address, and use that address for the ACK and all further communications until the connection is broken. But from what I have heard, that was not done.
What a shame. Anycast addresses that supported TCP properly could have been quite useful. It eliminate the need for abusing the DNS system for ensuring routing to the nearest datacenter. Further in the datacenter, it could be used for load balancing. Have all the servers for a domain accept the anycast address, and have them connected to a special router. The machines could periodically indicate their load level to the router, which could then decide which of the machines to route new connections. Telling the new connections from the established ones is stateless, since only new connections would use the anycast address. The only state the router would need to maintain beyond that all need to maintain, is the load level on a relatively small number of servers (a.k.a. a static amount of additional state amounting to a couple of kilobytes.)
Kotaku's video proves nothing. That is an example of a game for the regular DSi that happens to use that one method. Indeed latest speculation is that it uses parallax barriers which act much like lenticlar lenses in only working from a very specific and narrow viewpoint. To overcome this, dynamic forms that track a person have been developed, but those are suspected to be too expensive to use in a DS.
Fair enough, but Kerberos can be used as the basis of real SSO if done right. For example, Firefox, shh, and many other utilities can be configured to forward Kerberos tokens to avoid needing to sign in to the remote machine. Window's own AD based Domain architecutre allows access to other Windows machines without authenticating even if Kerberos was not configured, so that will not need multiple sign ons.
If done right, one would not need to sign in more than once unless doing something really unusual. If one is doing something really unusual, a second sign-in is not a very onerous requirement.
This actually has two parts.
One is a requirement that the correct prefixes are used. Although slightly confusing to some computer users, this is still the correct thing to do. I do not however advocate the use of the fully spelled out names of the binary units, as those are just awful.
The policy notably requires that 'k' and not 'K' be used for Base 10 values since KB is actually Kelvin-Bytes, a unit that is virtually never useful. It also mandates the use of "B" for bytes, which means that "b" will mean bits. IT should make that explicit, but using "b" incorrectly for Bytes is much more common than using "B" incorrectly for bits.
But there is a second part that is more controversial. Applications must favor the base 10 units over the base 2 units. Thus when both are listed base 10 must come first. When only one is given, and is not configurable, it must be base 10. If only one is shown but it is configurable, the default must be base 10.
That part of the new policy is one I think I object to. For example, A program displaying RAM sizes while only showing one prefix should show MiB or GiB by default, since ram is sold in base 2 sizes. It would confuse everybody if the ram size was reported only in units of "MB" or "GB".
that is if you move you'd expect to be able to see around one object blocking your view of another. but that does not happen with this kind of 3D.
Apparently, Nintendo has done this...and it's available today in NDS!
Please see this game: http://www.nintendo.co.jp/ds/dsiware/krgj/index.html
The 3D environment move when you look around the NDS screen. I think it's done by some image recognition with the front facing camera.
Actually that is something he already described:
A final kind of 3D is monocular 3D. there there is one image but you eye track or tilt track in a way that lets you update the image for the new point of view.
I think Kotaku's speculation may well be wrong. There is a technology to do 3d without special glasses. Unfortunately it sucks. The technology is based on lenticular lenses.
But that is the only known way (of which I am aware) to do stereographic 3d without glasses on something that is remotely like a standard display.
Most SSO in organizations I've ever seen seem to work by connecting the user directory to Kerberos, and use that for authentication to everything. Depending on how well IT department set up token forwarding, you may need to enter your credentials to access many systems, but everything in the company, from signing into the Windows Domain, to authenticating to the database uses Kerberos, so you have exactly one password to remember.
You did mention the verbatim quote, but the general impression your post gave reinforced the idea that plagiarism is merely directly copying from a source. This impression is too wide spread even on Slashdot, mainly because of the crap idiot teachers spread. Therefore my posting was to help other readers understand what plagiarism really is, not to criticize your understanding, although I admit it did sound like the latter.
Of course, services like turnitin.com and other such tools don't distinguish between proper cited and quoted material, and material copied without attribution, considering both to be indicators of plagiarism, so if you have one of those idiot teachers that don't know what plagiarism really is, you would be best to avoid even properly cited quotations.
My guess is that since the optimal value for the cones varies between people anyway, they just split the difference between the average optimal. The result is that the image appears best to those closest to statistically average, but varies by person. That said, The brain is still rather good at smoothing over minor inconsistencies so the net impact may be rather small, but this may still be why some people have find Dolby3D the least pleasant to watch.
The "vertical" is relative to the eyeline as "horizontal". The images will always overlap horizontally relative to the floor of the theater. I'm pretty sure we are in agreement on that point. I was deliberate with the use of scare quotes around "vertically" in my original post.
The point I was trying to make: Consider standing on your head. Your right eye is now on the side of the body that would have had your left eye when upright. But with either linear or circular polarization when standing on your head your right eye, which is now on the left will be seeing the image meant for the eye on the right, and vice versa. That is just not going to work right. It would make anything that should appear recessed into the screen as sticking out of it.
But with your head turned only 90 degrees, the images will not overlap so as to form any coherent image. With circular technology the right eye is still seeing the same thing as when upright, as is the left eye. With linear technology the two are swapped. But for either they do not overlap along the plane of the eyes, while real world images would. Thus your mind would not be able to produce a coherent 3D image from the inputs.
It is also worth noting that despite that promise, GBP is still a fiat currency. The bank merely offers a currency exchange service where one of the currencies is GBP and the other is XAU.
Plagarism is not copying form a source. It is permissible to copy a definition from a source and use it as a direct quotation, as long as it is clear it is a quotation, and as long as you cite the source of the text.
Furthermore paraphrasing a unique idea without crediting the source of the idea is plagiarism, even though it is not an exact copy.
Plaragism is the use of an idea or wording that is not your own and not general knowledge without crediting the source. Specifically any relatively unique idea must be cited, and and borrowed wording must be clearly marked as such with the source given.
There is a general knowledge exception. Wording whose source is so well known that the source is implied may not need a specific citation. Traditional proverbs often fall into this category, although sometimes they fall into a category of general knowledge whose source it not known.
The general knowledge exception is more frequently found with ideas though.
Specifically, the concept of plagiarism (which is really only meaningful in an academic context, the rest of the world is only concerned about legal matters like Copyright infringement) is the use of specific ideas that are not your own without citing the source of the idea.
A paper that consists entirely of a long verbatim quote from some other source, but where the source is cited, and it has been made clear what was an exact quote is not plagarism. It would almost certainly fail to meet the requirements of the class.
On the other hand, if one paraphrases an idea they got from some other source, even without any exact quotes, if the source is not cited it is plagiarism.
However, plagiarism should not be the main concern of citing sources. The main concern should be to document where you got an idea, or a number, etc. One does not need to repeat all the evidence that supports an idea if one cites a source that provides said evidence, or equivalently one need not may a persuasive argument for some idea if one can site a source that provides said persuasive arguments. This allows you to focus on supporting your conclusion with evidence or arguments as the case may be, without having to justify every population statistic, or philosophical idea you use.
Citing sources can have other benefits too. It can help protect your reputation. If you used a source that should be reliable, and it is later determined the source had faulty data, by citing the source people will know that you were being mislead, and not that you were intentionally perpetuating bad data.
Actually, the law is quite clear that one cannot lose citizenship without intending to do so. So if one joins a terrorist organization, or a foreign military without the intent to relinquish citizenship, they remain US Citizens.
However, joining any group hostile to the United States does create a rebuttable presumption of the intention to relinquish citizenship.
Here is an abridgment of 8 USC 1481 (emphasis mine)
If you read the whole thing[1] you will see that loss of citizenship for treason only occurs
Now, the law is more concerned with whether the actions are considered voluntary, as some of these acts, especially the ones related to treason are very strongly presumed to always be done with the intention of relinquishing citizenship. However, others such as becoming a national of another country carry only limited presumption of intent to relinquish citizenship. For example, one is allowed to become a national of another country while retaining US citizenship, the courts have been very clear about that.
[1] http://www.law.cornell.edu/uscode/uscode08/usc_sec_08_00001481----000-.html#a