3 ms·

This net neutrality is far from it when Google, Amazon, etc have their own private networks to which net neutrality rules don't apply. If carriers are prevented

This net neutrality is far from it when Google, Amazon, etc have their own private networks to which net neutrality rules don't apply. If carriers are prevented from offering differentiated services, which are required for next Gen products like self driving cars, video games in the cloud, Tele presence, and even just reliable video conferencing, then the only companies that will be able to offer these services will be the ones with existing multi billion private networks. Or you have to run your business on aws or gce and pay them 10X and they get all the profit. The only way a true competitor could get started is if they can buy the network service they need from an independent carrier. The fcc is right on this and Google is full of bs.

AnthonyMouse 8y ago
Network neutrality applies to last mile providers like Comcast and Verizon, and only to the extent that one of the endpoints is one of their last mile customers. You're talking about transit providers like Cogent and Level 3, which it doesn't apply to any more than it does to Google. It also applies to Google to the extent that they have Google Fiber last mile customers. It's true that GCE and AWS have expensive bandwidth prices, but then buy from someone else. It's not as if they have no effective competition, unlike Comcast.
- carnagii 8y ago
  Correct, but suppose one company has hundreds of edge's to their private network very close to the last mile and they run 10-20% of all internet traffic going across that last mile. If you are Google or Amazon and push a lot of low priority streaming bandwidth, a neutral last mile dedicates a lot of capacity to you, so when you prioritize high value data on your own network that priority ends up applying to the last mile as well. The only way a new company without a massive nationwide network could compete with Google or Amazon is if they could pay a carrier for prioritized service. If you prevent the carriers from offering prioritized service then you are reducing competition because Google and Amazon can and do offer this. Google and Amazon are far more dangerous monopolists than the traditional telcos. They are already vertically integrated hardware/software/media/everything companies. If some innovative startup builds a killer product that requires their network to function then they will either buy it, or create their own competitive product and kill it. AT&T, Verizon, and Comcast are relatively small companies now and they are in direct competition with Google and Amazon on many fronts, including this one. The Google version of net-neutrality is nothing more than an attempt to stifle what little competition they have left.
  - AnthonyMouse 8y ago
    > The only way a new company without a massive nationwide network could compete with Google or Amazon is if they could pay a carrier for prioritized service. If you prevent the carriers from offering prioritized service then you are reducing competition because Google and Amazon can and do offer this. You're confusing two different things. The reason companies build networks with many edges is the speed of light. If the server is in California and the customer is in New York, you're looking at about 100ms of round trip latency, because physics. Putting a server in the Northeast can get the RTT to that customer down to around 20ms. There is nothing the ISP can do about this; if the server is thousands of miles away, you can't change the speed of light. (Another reason to do that is then you don't need as much long-distance fiber capacity, but that's a cost savings and doesn't really affect performance one way or the other.) Moreover, small customers with latency-sensitive applications can do the same thing as Google and Amazon -- use a CDN. There are multiple competitive CDNs with edges all over the place that will cache your content closer to your users. By contrast, if the ISP's network isn't exceeding capacity, ISP prioritization does nothing. A well-provisioned network forwards 100% of packets immediately. There is nothing to prioritize when nothing has to be dropped. Now suppose you want paid prioritization. The first thing you've done is encourage the ISPs to underprovision their networks sufficiently to cause significant congestion, because they can't charge to relieve congestion if there isn't any. So yay, now the ISP purposely saturates their uplink and the default type of network connection you get is one with 90ms of bufferbloat and significant packet loss, that way they can charge extra to make it the way it should have been to begin with. Now you can pay the ISP to put it back the way it was. But that isn't any advantage over Google or Amazon because they can do the same thing, only they can negotiate a better price than you because they're bigger. (They still don't like it because they'd rather not pay monopoly rents to Comcast at all than have to pay $X even if smaller competitors have to pay $2X.) The only thing paid prioritization gets the little guy is a bill from the ISP -- from every ISP -- and a correspondingly even more competitive disadvantage against larger competitors with more leverage. Along with more transaction costs, because good luck negotiating as a small business with every ISP everywhere. > Google and Amazon are far more dangerous monopolists than the traditional telcos. Google and Amazon are not monopolists in this context at all. Google has a dominant position in search and Amazon has a dominant position in online retail, but that has very little to do with networks or data centers. They have no monopolies there -- they compete with each other, along with Microsoft, IBM, Oracle, the ISPs themselves and a zillion different smaller providers.

AnthonyMouse 8y ago

Network neutrality applies to last mile providers like Comcast and Verizon, and only to the extent that one of the endpoints is one of their last mile customers. You're talking about transit providers like Cogent and Level 3, which it doesn't apply to any more than it does to Google. It also applies to Google to the extent that they have Google Fiber last mile customers. It's true that GCE and AWS have expensive bandwidth prices, but then buy from someone else. It's not as if they have no effective competition, unlike Comcast.

carnagii 8y ago
Correct, but suppose one company has hundreds of edge's to their private network very close to the last mile and they run 10-20% of all internet traffic going across that last mile. If you are Google or Amazon and push a lot of low priority streaming bandwidth, a neutral last mile dedicates a lot of capacity to you, so when you prioritize high value data on your own network that priority ends up applying to the last mile as well. The only way a new company without a massive nationwide network could compete with Google or Amazon is if they could pay a carrier for prioritized service. If you prevent the carriers from offering prioritized service then you are reducing competition because Google and Amazon can and do offer this. Google and Amazon are far more dangerous monopolists than the traditional telcos. They are already vertically integrated hardware/software/media/everything companies. If some innovative startup builds a killer product that requires their network to function then they will either buy it, or create their own competitive product and kill it. AT&T, Verizon, and Comcast are relatively small companies now and they are in direct competition with Google and Amazon on many fronts, including this one. The Google version of net-neutrality is nothing more than an attempt to stifle what little competition they have left.
- AnthonyMouse 8y ago
  > The only way a new company without a massive nationwide network could compete with Google or Amazon is if they could pay a carrier for prioritized service. If you prevent the carriers from offering prioritized service then you are reducing competition because Google and Amazon can and do offer this. You're confusing two different things. The reason companies build networks with many edges is the speed of light. If the server is in California and the customer is in New York, you're looking at about 100ms of round trip latency, because physics. Putting a server in the Northeast can get the RTT to that customer down to around 20ms. There is nothing the ISP can do about this; if the server is thousands of miles away, you can't change the speed of light. (Another reason to do that is then you don't need as much long-distance fiber capacity, but that's a cost savings and doesn't really affect performance one way or the other.) Moreover, small customers with latency-sensitive applications can do the same thing as Google and Amazon -- use a CDN. There are multiple competitive CDNs with edges all over the place that will cache your content closer to your users. By contrast, if the ISP's network isn't exceeding capacity, ISP prioritization does nothing. A well-provisioned network forwards 100% of packets immediately. There is nothing to prioritize when nothing has to be dropped. Now suppose you want paid prioritization. The first thing you've done is encourage the ISPs to underprovision their networks sufficiently to cause significant congestion, because they can't charge to relieve congestion if there isn't any. So yay, now the ISP purposely saturates their uplink and the default type of network connection you get is one with 90ms of bufferbloat and significant packet loss, that way they can charge extra to make it the way it should have been to begin with. Now you can pay the ISP to put it back the way it was. But that isn't any advantage over Google or Amazon because they can do the same thing, only they can negotiate a better price than you because they're bigger. (They still don't like it because they'd rather not pay monopoly rents to Comcast at all than have to pay $X even if smaller competitors have to pay $2X.) The only thing paid prioritization gets the little guy is a bill from the ISP -- from every ISP -- and a correspondingly even more competitive disadvantage against larger competitors with more leverage. Along with more transaction costs, because good luck negotiating as a small business with every ISP everywhere. > Google and Amazon are far more dangerous monopolists than the traditional telcos. Google and Amazon are not monopolists in this context at all. Google has a dominant position in search and Amazon has a dominant position in online retail, but that has very little to do with networks or data centers. They have no monopolies there -- they compete with each other, along with Microsoft, IBM, Oracle, the ISPs themselves and a zillion different smaller providers.