6 ms·

Thing about QOS is that if I buy a level of service from my ISP and you buy an equivalent or greater level of service from your ISP, you and I should be able to

Thing about QOS is that if I buy a level of service from my ISP and you buy an equivalent or greater level of service from your ISP, you and I should be able to communicate at that level of service. The which bits we decide to send to each other should be completely up to us. QOS should happen at the edges of the network, certainly not in the middle because the middle has no clue what bits count to me. Common Carrier law is exactly what Title II is, and what currently applies to trucks driving down the roads (can't buy a faster speed limit), electricity, shipping, oil pipelines, landline telephony, postage, etc.

gumby 12y ago
Oh I completely agree in regards to QoS! That was not my point. My point is that the technology is still immature. We haven't really considered what our true criteria are, what layer they live at etc. There are still many innovations ahead. And don't think the carriers aren't well versed in how to game the system! You say, correctly, that Title II is common carriage. But consider what came along with that. Note that I think CC has mostly worked with busses (you can't discriminate by race, at least) and the roads. It's worked well with trains. In other areas it hasn't got such a good record (and my politics are what is called in America "liberal"). In the case of telephony, title II unarguably held back technological development. All the development was done at the center. AT&T and Bell Labs did a really astounding job of development in the core (FDM in the 1930's for chrissake, and direct dialing with routing tables on punched metal cards in the 1940s). Not to mention SS7, the transistor and fibre optics. But all that came at a huge cost: phone calls were expensive, long distance was a limited service, and all innovation at the edge was explicitly, and by legal force, suppressed. You probably don't remember how horrible it was to use a Bell 103 modem, or the crappy approved phones. Mobile telephony was expensive and essentially useless. The communication infrastructure of today would be unrecognizable to a telephone engineer of 1990, while the communication infrastructure of 1990 would have been be easily comprehensible to a telephone engineer of 1925. My point is: that it worked at all was because we had a useful metric: 99% of all calls had to be completed, even if it meant overbuilding to guarantee that on Mother's day. On the other hand we don't even have an equivalent of an Erlang [1] in the internet domain. I don't believe NN can legitimately be legislated in 2014, and any attempt will screw us at LEAST as much as ATT/comcast/verizon etc al are screwing us today. [1] http://en.wikipedia.org/wiki/Erlang_(unit) http://en.wikipedia.org/wiki/Erlang_(unit)

gumby 12y ago

Oh I completely agree in regards to QoS! That was not my point. My point is that the technology is still immature. We haven't really considered what our true criteria are, what layer they live at etc. There are still many innovations ahead. And don't think the carriers aren't well versed in how to game the system! You say, correctly, that Title II is common carriage. But consider what came along with that. Note that I think CC has mostly worked with busses (you can't discriminate by race, at least) and the roads. It's worked well with trains. In other areas it hasn't got such a good record (and my politics are what is called in America "liberal"). In the case of telephony, title II unarguably held back technological development. All the development was done at the center. AT&T and Bell Labs did a really astounding job of development in the core (FDM in the 1930's for chrissake, and direct dialing with routing tables on punched metal cards in the 1940s). Not to mention SS7, the transistor and fibre optics. But all that came at a huge cost: phone calls were expensive, long distance was a limited service, and all innovation at the edge was explicitly, and by legal force, suppressed. You probably don't remember how horrible it was to use a Bell 103 modem, or the crappy approved phones. Mobile telephony was expensive and essentially useless. The communication infrastructure of today would be unrecognizable to a telephone engineer of 1990, while the communication infrastructure of 1990 would have been be easily comprehensible to a telephone engineer of 1925. My point is: that it worked at all was because we had a useful metric: 99% of all calls had to be completed, even if it meant overbuilding to guarantee that on Mother's day. On the other hand we don't even have an equivalent of an Erlang [1] in the internet domain. I don't believe NN can legitimately be legislated in 2014, and any attempt will screw us at LEAST as much as ATT/comcast/verizon etc al are screwing us today. [1] http://en.wikipedia.org/wiki/Erlang_(unit) http://en.wikipedia.org/wiki/Erlang_(unit)