What the story on Netflix caching devices? Does Verizon use them?
My understanding is: No.
Verizon would rather sell more data, and having a local caching device on their network,
would decrease the amount of data they get to sell.
Even with local caching devices Verizon still has to do work to carry that data from the cache device across their global network, to the recipient.
As far as they're concerned --- it's more important that they get paid to carry that data,
than that the packets are are delivered with the lowest latency and greatest reliability.
I have been out of the Tier 1 ISP world for years - but at least historically Cogent was known as a bottom feeder of the industry. They charged dirt-cheap rates but ran a crappy network and skimped on their upstream connections.
Exactly. The Cogent rate for the same gigabits of data is probably about 70 to 80% lower than Verizon's rates.
In this case... if Cogent can get unlimited free peering to Verizon's network, then it would make no sense for Netflix to pay the extra premium Verizon charges per gigabit of committed demand, for access to their network.
Cogent may host much of Netflix, but they are by no means a Tier 1. This may no longer be the case - like I said
If you read Cogent's brochures and marketing material on cogentco.com, you will see them refer to their service as a Tier 1 opcal IP network.
I believe in the firm opinion of most of the influential industry players, Cogent is not really a Tier1; they are a wannabe, almost-Tier1.
It behooves Verizon to deliver the service their customers are PAYING for.
As far as Verizon is concerned, they are delivering that service. It's not Verizon's problem, that Netflix has chosen not to pay Verizon for a connection and transit, because Cogent is so much cheaper.
Meanwhile, Cogent wants to slurp off a free peering arrangement from Verizon, to assist Cogent in competing against Verizon in selling Netflix internet connectivity to Verizon's customers.
If customers were that concerned about it, they would be calling up Netflix to complain about the problem.
The thing is, private peering doesn't cost you anything.
No.... the data transferred still uses data capacity.
Lost revenue. Cogent is using Verizon's free peering to compete against Verizon, by selling the same bandwidth (sustained gigabits of data transfer) to Netflix, that Netflix would have to pay four to five times as much for if they bought an internet connection from Verizon.
In principle, to protect their revenues: the Tier1's need to ensure that they get paid by the other party connecting to their network for data coming into or leaving their network, that is using up some of their network capacity.
They either get paid via a transit fee from the party connecting, or by gaining an equivalent amount of usage of the peer's network.
Which means Netflix will have to pay their ISP more so they can pay to deliver the data.
Yep. And probably one reason all the Netflix traffic is going out Cogent, is Cogent has among the lowest fees per Gigabits of data transferred for data in the industry (But a reputation for being the internet's storm gutter due to unreliability in terms of quality metrics such as jitter, packet loss, latency). So instead of buying for $X a gigabit from Verizon, Netflix buys transit from Cogent instead, at a price that is most likely about 80% lower... which no doubt saves Netflix a lot of money.
The extremely low price is only possible by exploiting and abusing free peering relationships with other Tier1s.
If Cogent charged Netflix a higher, more reasonable price ---- that would be more in-line with the fees charged by Verizon, ATT, Level3, etc; it is likely, that Netflix would rationally engineer their traffic, among MULTIPLE transit providers, for best reliability across the board, so that they were more balanced among the Tier1 providers, and not pimping out Cogent links.
NETFLIX Could do this today, and they apparently choose not to ---- probably because the Netflix subscription prices would have to go up substantially.
Yes, they SELL connections to their customers, so the incoming flood of traffic from Cogent has already been paid for by Verizon's customers that are trying to watch their shows on Netflix. Verizon is trying to double-dip
NO. Verizon customers pay for a connection to Verizon's network --- you are paying for half of the equation, not the end-to-end transport; the fee paid by Verizon subscribers includes the customer's connection, but other Verizon peers and customers still have to pay in order to interconnect with Verizon's network and data sent to Verizon's other customers (unless a mutually beneficial peering arrangement can be reached that Verizon determines is either equally beneficial to both sides or benefits Verizon more).
This is not unusual, and all Large Tier1 providers behave in this manner, in order to protect their revenues and prevent loss-of-revenues due to providing free peering, when they could charge for service.
Double-dipping would be getting Cogent to pay extra, AND THEN..... limiting Netflix traffic, and reaching out to various contacts at Netflix to offer to prioritize their traffic, for an additional fee per terabyte, despite Cogent paying extra for all the paid transit of Netflix traffic through Verizon's network..
Why should Verizon expect to have an even in/out ratio? They sell the vast majority of their customers asymmetrical connections.
You only peer with another provider settlement-free, if the peering arrangement is mutually beneficial.
ISPs such as Verizon are in the business of selling connectivity, not giving it away for free.
If they are making settlement-free peerings where they don't benefit at least as much as the other party, then they are losing the revenue from selling connectivity, for less than the revenue would be worth in return.
Since Cogent is pushing much more data than they pull, then the peering relationship is disproportionately beneficial to Cogent.
Therefore, Cogent must pay more. It's really that simple.
A video as citation? That's a bit insulting if you expect me to take you seriously.
The medium doesn't matter. It's perfectly valid, notwithstanding your elitist high-horse act.
I think what you find insulting, is that the authorities don't share your highly narrow idea as to what the concept of peak oil refers to.
Note it's called peak oil and the likely peak of coal production is probably many decades away.
Yes. But 'oil' does not mean crude oil; it means all the oil products, which include gasoline ---- but not coal.
Peak oil looks like it happened in 2008 because it's the maximum point on the graph of crude oil extraction over time. Gas from shale, coal, whatever is something else.
NO. Peak oil refers to M. King Hubbert's theory about a bell-shaped pattern, and it is inclusive of all fossil fuels.
Google (or any hosting provider) is obliged, under the DMCA,
Not obliged. Expeditious removal is just a condition of the DMCA's special liability exclusions, AND not having those exclusions does not automatically make Google liable for any potential infringement ---- it reverts conditions to what they were before passage of the DMCA.
The DMCA's liability exclusions which are of questionable value anyways, since copyright holders have already sued and won against service providers and file sharing sites that MET the DMCA's safe harbor requirements.
That's nothing. Just wait until the ICF hits (internal combustion flu). Tesla will be laughing all the way to the bank.
It's called peak oil.... it just got delayed by 10 to 20 years, perhaps, due to the introduction of fracking.
When gasoline is no longer available due to global or local resource shortages, or prices --- the same could occur again.
It may be even worse, since the petroleum products are not merely used to fuel our vehicles, BUT they are also required to produce
fertilizers, so our farms can grow enough food for us to eat, AND required to produce plastics for new products that are critical to our daily lives.
Otherwise, you're out of luck, as claimant will simply reject your counter claim and copyright strike and takedown on your account and video will remain in force.
You may be able to sue the copyright holder for tortious interference in this case, and seek a court order for Google to disregard their malicious activity and reinstate the video.
No, it doesn't work that way. The same person can't issue a takedown for the same thing repeatedly. Claim, counter-claim, then it's up forever until sued for.
In reality.... they probably can, and Google is unlikely to notice that it's the same person that issued the second notice.
The same copyright holder may also have a second agent acting on their behalf to send the second takedown notice.
Nothing prevents them from sending the second letter.
Nothing prevents Google from making the mistake of failing to realize that the second letter was ultimately related to the same work claimed to be infringed, and the same alleged infringement.
If they are fair use, he can counter-claim and be done with it (until they sue him). He's effectively acknowledging that he's violating copyright by refusing to contest the assertions.
He does not have to make a formal counterclaim, until the matter goes before a judge.
He has an inalienable right to continue to post and publish his content, until ordered otherwise or enjoined from doing so by a court of law.
Otherwise all that leaked Windows source code would be fair use.
It is fair use to excerpt a video for educational purposes, and specifically --- to illustrate exactly what parts your criticism is about, and to crystalize in the minds of the viewers: portions relevant to the discussion in order to substantiate which claims are being debunked.
At some point the issue of what's reasonable has to kick in. If she lost a VHS tape 9 years ago, and the store went under since then, (1) there's no victim, and (2) the replacement cost for the videotape is probably only a few dollars
(1)Depends on how much revenue they lost due to that tape being tied up.
The victims are the shareholders --- probably they can still be repaid.
They need to be repaid in 2005 dollars adjusted for inflation, however; a ~20% increase on top reparations for what the value was in 2005... $100 worth of property and revenue loss in 2005 dollars, is a loss of $120.70 in 2014.
(2) Back in 2005, the replacement cost would have been much higher; the "special" videotapes that come with their licensing to rent, cost a pretty penny compared to the normal consumer VHS tapes....
Reasonable suspicion? Rather a low bar, considering that "probable cause" is the constitutional requirement for any search.
Since the patriot act was passed, probable cause was not required for wiretaps.
In fact, even reasonable suspicion is not required.
The intelligence agencies also never needed reasonable suspicion to make efforts to tap into a conversation crossing US borders or taking place outside the US.
Also, intelligence agencies' partner organizations such as GCHQ could do the tapping and cooperatively provide recordings to their buddies at the NSA through partner agency information sharing, if for some reason, the NSA themself was legally prevented from doing the tapping.
The really shocking thing is that nearly 10% of the respondents who comment self-identified as trolls. If your facts are accurate, that would mean that perhaps twice that many are trolls. So, 20%?
It doesn't follow..... there is another problem. It is not clear that the survey population is an example of a representative sample of the general population. In fact.... it seems their sample population / survey respondents may have been skewed towards higher participation by college-aged students.
Non-trolls identifying themselves as potential people who are trolls in the survey.
I.E. People who were lying, or mislead, causing them to answer the survey stating that they were trolls.
Also; it is not clear that people self-identifying as trolls are hard-core trolls, OR that they are acting as a troll 100% of the time.
100% of the survey respondents could have said there were trolls, AND their survey would still have said nothing about what percentage of the participants in your average slashdot discussion are posts by trolls, or what percentage of the comments are likely to be trollish comments.
Even Star Trek seems to often resorts to magical thinking and the "lone hero" narrative, devolving into a morality play or social drama with a futuristic backdrop and technobabble.
Yeah, but at least they get the math mostly right.. Spock famously calculates the right number of tribbles there would be after so many days, given the assumptions that he stated:)
Clearly, no one in the history of survey design has ever considered the idea of an unreliable respondent until you came along.
I am not claiming to have a novel point of view about what survey design should consider.
My argument is that regardless of how the survey is designed: people whom a majority of the internet community population would identify as trolls are likely to almost always lie on the survey in some manner and fail to be identified as a troll.
Meanwhile, their custom surveys are likely to flag people as trolls whom would not ordinarily be considered trolls.
EXAMPLE: for their 'GAIT' survey: "I have sent people to shock websites for the lulz."; "I enjoy griefing other players in multiplayer games."
The questions betray a narrow view of what constitutes a troll and what doesn't constitute one.
I'm not sure I would view a player in a competitive multi-player game slaying or sabotaging competitors as acts on the same level as trolling.
Is about material provided in confidence that cannot be used in court.
There's no reason the appropriate authorities can't listen in, if there is reasonable suspicion that the lawyer may be complicit in a future criminal act.
There have been cases where the lawyer became complicit in a later crime. case in point:
Lynne Stewart. Who perhaps should have been sentenced to death for her treasonous actions.
Attorney-client privilege protects information pertaining to their case and legal advise.
It doesn't protect against prosecution for conspiring with the lawyer, or using the lawyer as a channel to commit further crimes.
They should have included examples of "trolling" to illustrate their points.
Yes... I suppose it is also a finer point, whether the poster views their posts as trollish, whether
a plurality of the rest of the community views their posts as trollish, AND whether their assessments are in agreement.
They would have done well to determine 'what trolling is'... e.g. which criteria they will consider it,
and ask respondents to rank agreement/disagreement on a 5-point scale for various facts about their posts.
Such as, whether they enjoy playing practical jokes on other posters....
using sarcasm, personal insults, all caps, causing flamewars,
intentionally inciting or encouraging angry responses, manipulating other posters, etc
Furthermore --- a survey is not a good measure, they should have asked participants
to name the forums on which they most commonly participate, and a handle or URL
for extracting samples of their posts for review.
Only a review of their posts, some replies, and perhaps interviews of some others in
the community, will be a reliable indicator of whether the internet considers them to be
a troll or not.
The survey responses from potential trolls cannot really be trusted.... without resulting in
potentially unreliable data to base any findings on
Slashdot Mirror
What the story on Netflix caching devices? Does Verizon use them?
My understanding is: No.
Verizon would rather sell more data, and having a local caching device on their network, would decrease the amount of data they get to sell.
Even with local caching devices Verizon still has to do work to carry that data from the cache device across their global network, to the recipient.
As far as they're concerned --- it's more important that they get paid to carry that data, than that the packets are are delivered with the lowest latency and greatest reliability.
I have been out of the Tier 1 ISP world for years - but at least historically Cogent was known as a bottom feeder of the industry. They charged dirt-cheap rates but ran a crappy network and skimped on their upstream connections.
Exactly. The Cogent rate for the same gigabits of data is probably about 70 to 80% lower than Verizon's rates. In this case... if Cogent can get unlimited free peering to Verizon's network, then it would make no sense for Netflix to pay the extra premium Verizon charges per gigabit of committed demand, for access to their network.
Cogent may host much of Netflix, but they are by no means a Tier 1. This may no longer be the case - like I said
If you read Cogent's brochures and marketing material on cogentco.com, you will see them refer to their service as a Tier 1 opcal IP network.
I believe in the firm opinion of most of the influential industry players, Cogent is not really a Tier1; they are a wannabe, almost-Tier1.
It behooves Verizon to deliver the service their customers are PAYING for.
As far as Verizon is concerned, they are delivering that service. It's not Verizon's problem, that Netflix has chosen not to pay Verizon for a connection and transit, because Cogent is so much cheaper.
Meanwhile, Cogent wants to slurp off a free peering arrangement from Verizon, to assist Cogent in competing against Verizon in selling Netflix internet connectivity to Verizon's customers.
If customers were that concerned about it, they would be calling up Netflix to complain about the problem.
The thing is, private peering doesn't cost you anything.
No.... the data transferred still uses data capacity.
Lost revenue. Cogent is using Verizon's free peering to compete against Verizon, by selling the same bandwidth (sustained gigabits of data transfer) to Netflix, that Netflix would have to pay four to five times as much for if they bought an internet connection from Verizon.
In principle, to protect their revenues: the Tier1's need to ensure that they get paid by the other party connecting to their network for data coming into or leaving their network, that is using up some of their network capacity.
They either get paid via a transit fee from the party connecting, or by gaining an equivalent amount of usage of the peer's network.
Which means Netflix will have to pay their ISP more so they can pay to deliver the data.
Yep. And probably one reason all the Netflix traffic is going out Cogent, is Cogent has among the lowest fees per Gigabits of data transferred for data in the industry (But a reputation for being the internet's storm gutter due to unreliability in terms of quality metrics such as jitter, packet loss, latency). So instead of buying for $X a gigabit from Verizon, Netflix buys transit from Cogent instead, at a price that is most likely about 80% lower... which no doubt saves Netflix a lot of money.
The extremely low price is only possible by exploiting and abusing free peering relationships with other Tier1s. If Cogent charged Netflix a higher, more reasonable price ---- that would be more in-line with the fees charged by Verizon, ATT, Level3, etc; it is likely, that Netflix would rationally engineer their traffic, among MULTIPLE transit providers, for best reliability across the board, so that they were more balanced among the Tier1 providers, and not pimping out Cogent links.
NETFLIX Could do this today, and they apparently choose not to ---- probably because the Netflix subscription prices would have to go up substantially.
Yes, they SELL connections to their customers, so the incoming flood of traffic from Cogent has already been paid for by Verizon's customers that are trying to watch their shows on Netflix. Verizon is trying to double-dip
NO. Verizon customers pay for a connection to Verizon's network --- you are paying for half of the equation, not the end-to-end transport; the fee paid by Verizon subscribers includes the customer's connection, but other Verizon peers and customers still have to pay in order to interconnect with Verizon's network and data sent to Verizon's other customers (unless a mutually beneficial peering arrangement can be reached that Verizon determines is either equally beneficial to both sides or benefits Verizon more).
This is not unusual, and all Large Tier1 providers behave in this manner, in order to protect their revenues and prevent loss-of-revenues due to providing free peering, when they could charge for service.
Double-dipping would be getting Cogent to pay extra, AND THEN..... limiting Netflix traffic, and reaching out to various contacts at Netflix to offer to prioritize their traffic, for an additional fee per terabyte, despite Cogent paying extra for all the paid transit of Netflix traffic through Verizon's network..
Why should Verizon expect to have an even in/out ratio? They sell the vast majority of their customers asymmetrical connections.
You only peer with another provider settlement-free, if the peering arrangement is mutually beneficial. ISPs such as Verizon are in the business of selling connectivity, not giving it away for free. If they are making settlement-free peerings where they don't benefit at least as much as the other party, then they are losing the revenue from selling connectivity, for less than the revenue would be worth in return.
Since Cogent is pushing much more data than they pull, then the peering relationship is disproportionately beneficial to Cogent.
Therefore, Cogent must pay more. It's really that simple.
A video as citation? That's a bit insulting if you expect me to take you seriously.
The medium doesn't matter. It's perfectly valid, notwithstanding your elitist high-horse act. I think what you find insulting, is that the authorities don't share your highly narrow idea as to what the concept of peak oil refers to.
Note it's called peak oil and the likely peak of coal production is probably many decades away.
Yes. But 'oil' does not mean crude oil; it means all the oil products, which include gasoline ---- but not coal.
Except that without the burger flippers, I don't see how McDonalds' business model could work.
We'll find out most likely next time the minimum wage goes up. The cooks can be replaced with robots.
I think you're better off building a big fence around your property, with a secured gate, and supplying UPS/Fedex with the "gate code".
Security gates are common.
Bins are not.
Peak oil looks like it happened in 2008 because it's the maximum point on the graph of crude oil extraction over time. Gas from shale, coal, whatever is something else.
NO. Peak oil refers to M. King Hubbert's theory about a bell-shaped pattern, and it is inclusive of all fossil fuels.
no one will be given enough alcohol to put them over the legal driving limit.
Real pubs aren't supposed to give anyone enough alcohol to get them drunk, either.
Google (or any hosting provider) is obliged, under the DMCA,
Not obliged. Expeditious removal is just a condition of the DMCA's special liability exclusions, AND not having those exclusions does not automatically make Google liable for any potential infringement ---- it reverts conditions to what they were before passage of the DMCA.
The DMCA's liability exclusions which are of questionable value anyways, since copyright holders have already sued and won against service providers and file sharing sites that MET the DMCA's safe harbor requirements.
That's nothing. Just wait until the ICF hits (internal combustion flu). Tesla will be laughing all the way to the bank.
It's called peak oil.... it just got delayed by 10 to 20 years, perhaps, due to the introduction of fracking.
When gasoline is no longer available due to global or local resource shortages, or prices --- the same could occur again.
It may be even worse, since the petroleum products are not merely used to fuel our vehicles, BUT they are also required to produce fertilizers, so our farms can grow enough food for us to eat, AND required to produce plastics for new products that are critical to our daily lives.
Otherwise, you're out of luck, as claimant will simply reject your counter claim and copyright strike and takedown on your account and video will remain in force.
You may be able to sue the copyright holder for tortious interference in this case, and seek a court order for Google to disregard their malicious activity and reinstate the video.
No, it doesn't work that way. The same person can't issue a takedown for the same thing repeatedly. Claim, counter-claim, then it's up forever until sued for.
In reality.... they probably can, and Google is unlikely to notice that it's the same person that issued the second notice. The same copyright holder may also have a second agent acting on their behalf to send the second takedown notice.
Nothing prevents them from sending the second letter. Nothing prevents Google from making the mistake of failing to realize that the second letter was ultimately related to the same work claimed to be infringed, and the same alleged infringement.
If they are fair use, he can counter-claim and be done with it (until they sue him). He's effectively acknowledging that he's violating copyright by refusing to contest the assertions.
He does not have to make a formal counterclaim, until the matter goes before a judge.
He has an inalienable right to continue to post and publish his content, until ordered otherwise or enjoined from doing so by a court of law.
Otherwise all that leaked Windows source code would be fair use.
It is fair use to excerpt a video for educational purposes, and specifically --- to illustrate exactly what parts your criticism is about, and to crystalize in the minds of the viewers: portions relevant to the discussion in order to substantiate which claims are being debunked.
At some point the issue of what's reasonable has to kick in. If she lost a VHS tape 9 years ago, and the store went under since then, (1) there's no victim, and (2) the replacement cost for the videotape is probably only a few dollars
(1)Depends on how much revenue they lost due to that tape being tied up. The victims are the shareholders --- probably they can still be repaid. They need to be repaid in 2005 dollars adjusted for inflation, however; a ~20% increase on top reparations for what the value was in 2005... $100 worth of property and revenue loss in 2005 dollars, is a loss of $120.70 in 2014.
(2) Back in 2005, the replacement cost would have been much higher; the "special" videotapes that come with their licensing to rent, cost a pretty penny compared to the normal consumer VHS tapes....
Reasonable suspicion? Rather a low bar, considering that "probable cause" is the constitutional requirement for any search.
Since the patriot act was passed, probable cause was not required for wiretaps. In fact, even reasonable suspicion is not required.
The intelligence agencies also never needed reasonable suspicion to make efforts to tap into a conversation crossing US borders or taking place outside the US.
Also, intelligence agencies' partner organizations such as GCHQ could do the tapping and cooperatively provide recordings to their buddies at the NSA through partner agency information sharing, if for some reason, the NSA themself was legally prevented from doing the tapping.
The really shocking thing is that nearly 10% of the respondents who comment self-identified as trolls. If your facts are accurate, that would mean that perhaps twice that many are trolls. So, 20%?
It doesn't follow..... there is another problem. It is not clear that the survey population is an example of a representative sample of the general population. In fact.... it seems their sample population / survey respondents may have been skewed towards higher participation by college-aged students.
Non-trolls identifying themselves as potential people who are trolls in the survey.
I.E. People who were lying, or mislead, causing them to answer the survey stating that they were trolls.
Also; it is not clear that people self-identifying as trolls are hard-core trolls, OR that they are acting as a troll 100% of the time.
100% of the survey respondents could have said there were trolls, AND their survey would still have said nothing about what percentage of the participants in your average slashdot discussion are posts by trolls, or what percentage of the comments are likely to be trollish comments.
Even Star Trek seems to often resorts to magical thinking and the "lone hero" narrative, devolving into a morality play or social drama with a futuristic backdrop and technobabble.
Yeah, but at least they get the math mostly right.. Spock famously calculates the right number of tribbles there would be after so many days, given the assumptions that he stated :)
Clearly, no one in the history of survey design has ever considered the idea of an unreliable respondent until you came along.
I am not claiming to have a novel point of view about what survey design should consider.
My argument is that regardless of how the survey is designed: people whom a majority of the internet community population would identify as trolls are likely to almost always lie on the survey in some manner and fail to be identified as a troll.
Meanwhile, their custom surveys are likely to flag people as trolls whom would not ordinarily be considered trolls.
EXAMPLE: for their 'GAIT' survey: "I have sent people to shock websites for the lulz."; "I enjoy griefing other players in multiplayer games."
The questions betray a narrow view of what constitutes a troll and what doesn't constitute one.
I'm not sure I would view a player in a competitive multi-player game slaying or sabotaging competitors as acts on the same level as trolling.
Is about material provided in confidence that cannot be used in court.
There's no reason the appropriate authorities can't listen in, if there is reasonable suspicion that the lawyer may be complicit in a future criminal act.
There have been cases where the lawyer became complicit in a later crime. case in point: Lynne Stewart. Who perhaps should have been sentenced to death for her treasonous actions.
Attorney-client privilege protects information pertaining to their case and legal advise. It doesn't protect against prosecution for conspiring with the lawyer, or using the lawyer as a channel to commit further crimes.
They should have included examples of "trolling" to illustrate their points.
Yes... I suppose it is also a finer point, whether the poster views their posts as trollish, whether a plurality of the rest of the community views their posts as trollish, AND whether their assessments are in agreement.
They would have done well to determine 'what trolling is' ... e.g. which criteria they will consider it,
and ask respondents to rank agreement/disagreement on a 5-point scale for various facts about their posts.
Such as, whether they enjoy playing practical jokes on other posters.... using sarcasm, personal insults, all caps, causing flamewars, intentionally inciting or encouraging angry responses, manipulating other posters, etc
Furthermore --- a survey is not a good measure, they should have asked participants to name the forums on which they most commonly participate, and a handle or URL for extracting samples of their posts for review.
Only a review of their posts, some replies, and perhaps interviews of some others in the community, will be a reliable indicator of whether the internet considers them to be a troll or not.
The survey responses from potential trolls cannot really be trusted.... without resulting in potentially unreliable data to base any findings on