5 ms·

We're talking about 3 different technologies here: 1. Fiber optics - this is growing, and companies which have a lot of it are continuing to lay it out (albeit

We're talking about 3 different technologies here: 1. Fiber optics - this is growing, and companies which have a lot of it are continuing to lay it out (albeit slowly). Also see, Altice's recent announcement. My hope is that post-election Google Fiber will ramp back up, simply for the legal defense of saying that they too are a cable company and the same laws should apply. In any case, consumers want it. High-end consumers will pay a premium, and the cost for urban footprint fill-out is not exorbitant. 2. Cable coaxial wire - this is a big thick wire that continues to upgrade nicely. DOCSIS 3.0 is basically the same old 1980s/1990s cable with better comms protocols. It's getting to the point to where they can compete on a reasonable basis with single strand fiber, at a lower speed and price point. So cable companies build out the back-end with fiber, but don't have to replant single homes. 3. Crappy copper telephone wire - this stuff is thin, it's painful to maintain and upgrade and always lags copper. Companies with a lot of this (Windstream, Verizon, AT&T, Frontier) are trying to run it for cash flow rather than spend a lot on upgrades. The telcos push fiber to the node, but it's a slow grind and involves very careful cost benefit analysis. 3 different technologies. 3 very different strategies. The election has further complicated things. Should VZ save up it's ammo and make a play for Sprint, Charter, Dish, or T-Mobile or invest in more broadband plants? I'd save the ammo given that Tom Wheeler is out on January 20th..

ClassyJacket 10y ago
Point three makes me weep. In Australia the government started building a nationwide fibre to the home network. Wholesale only, so any ISP could provide services on it. They got about 10% of the way in, then a new conservative government got in power and stomped on the plan. Instead they're doing fibre to the node for everyone. Why have gigabit fibre when you can have slower, less reliable, more expensive copper connections that go a whopping 12Mbit!? Faster even, if the node happens to be sitting in your front lawn.
- PartTimeZombie 10y ago
  We started the same thing here in New Zealand. I was a bit outraged when $1.2 billion of taxpayers money was handed out to build a national fibre network, but it's turning out not too bad. Not perfect, but not bad either. It helps that the opposition bought in, despite disagreeing on details.
  - bitserf 10y ago
    NZ's UFB network has always been FTTH (meaning it terminates inside your house), there is no copper in this picture. This is why, for example, last month people who qualified for 200MB can now get gigabit plans.
  - yabatopia 10y ago
    Why were you outraged that your government invested $1.2 billion in a significant upgrade of an important piece of infrastructure, beneficial for citizens and companies, now and in the future? Was it because you've jumped on the anti-government train that's so popular right now, like the overused term "taxpayers money" suggests, or did you really believe that building a national fibre network was the dumbest idea ever?
    - CalRobert 10y ago
      Considering the population of New Zealand, that's a bill of around $255 per person. Perhaps they think the money could have been better spent?
      - theandrewbailey 10y ago
        $255 sounds extremely reasonable. That's much less than what most people pay annually for service.
        orclev 10y ago
        I would happily pay a one time cost of $255 to run fiber to my house. I'd even be willing to pay $500 to run fiber to my house. Sadly around here if I want fiber to my house, even assuming I could find an ISP willing to run it, it would probably cost me upwards of $2000 to have it run, and probably significantly more than that. That's also not counting the cost of actually getting service, that's just the "installation fee" to have it run.
        rubber_duck 10y ago
        He said per person, not per household - not the same unless you're living alone. Also this doesn't make your internet free you have to pay for the service on top of that. And believe it or not a lot of people don't care about having fiber internet available.
- rasz_pl 10y ago
  It doesnt make a bit of difference because whole project was doomed from the start (staffing committee with incumbent telco execs made sure of that). Whats the point of nation wide fiber network when you force data caps on it like it was some kind of undersea cable to mars?
  - ClassyJacket 10y ago
    It wasn't doomed from the start. The original plan was a good one. If we'd stuck to the fibre plan, nearly half the country would have fibre by now. The CVC prices are unreasonable but they can and will come down. Once the fibre is in place you can argue about the business details later. You can't, however, lay fibre overnight. That takes years of planning. The POI decision was dumb too but it can be worked around. As for the data caps, you can already buy an unlimited 100Mbps service right now. So it's just a matter of cost.
- gnopgnip 10y ago
  There is new technology like g.fast for 100+mbit speeds over a copper pair. Yes it is not as fast as either fiber or co-ax, and it is short sighted, but 100mbit is a huge improvement over legacy DSL speeds.
  - ClassyJacket 10y ago
    It already goes 100Mbit IF you're very lucky and happen to have the node right near your house (and then some poor sap at the other end gets 12). And no, they cannot upgrade FttN to G.fast, no matter what they tell you. That requires shorter copper lengths than all but the closest handful of houses to the node have. Plus you need everyone on the same technology so that the vectoring (like active noise cancellation) can work. Gfast drops off a cliff after a few hundred metres, and they have houses connected to nodes a kilometre away. Gfast and Xgfast are only appropriate for Fibre-to-the-curb. FttN can never be meaningfully upgraded.
  - scholia 10y ago
    You can get 80 meg broadband from BT Infinity with fibre to the street cabinet, so you might need to offer more than 100+. G.fast extends the fibre closer to the house but I suspect it's not very cost-effective unless the take-up is high, so perhaps it will mainly cater to rich people in suburbia. We've know that fibre (FTTH) was the answer since the 1980s but all British governments have been too short-sighted and stupid to finance it. Of course, they dug up the streets to install new pipes for the short-term benefits of North Sea gas, and they can find £50 billion for HS2, when they could have FTTH for less.
    - lostboys67 10y ago
      And realistically no consumer needs more than that.
      - razakel 10y ago
        >And realistically no consumer needs more than that. Yet. 640K ought to be enough for anybody.
- lostboys67 10y ago
  FTC (fiber To cab) makes economic sense as you can get decent speeds with the massive expense of doing FTP.
nickysielicki 10y ago
> DOCSIS 3.0 is basically the same old 1980s/1990s cable with better comms protocols. It's getting to the point to where they can compete on a reasonable basis with single strand fiber, at a lower speed and price point. It's not just that the communications protocols got better. I think that kind of implies that the medium is at it's end, and that the difference between DOCSIS 2 @ 38Mb/27Mb and DOCSIS 3.1 @ 10Gb/1Gb is some engineers squeezing more out of copper through ingenious protocol-level optimizations. And that's mostly not the case. The big difference is that 3.1 came out a decade later and can take advantage of amazing progress made with ADCs/DACs. I think it's reasonable to claim that the driving force in internet bandwidth gains have not been because of better mediums, like fiber instead of co-ax. The easiest place to look is wireless, where the medium hasn't changed and won't change, but the speeds have exploded. 802.11ax might push 10Gbit. LTE is an order of magnitude faster than EDGE. It's easier to push more data when there's less noise on the line. But if we can work around the noise instead, it makes little sense to spend millions laying new cables.
- azernik 10y ago
  From the infrastructure-investment perspective, only changing the medium is a big-ticket item (digging up cables, installing new ones, etc.). Swapping out newer transceivers on the ends is a tiny investment.
rayiner 10y ago
> My hope is that post-election Google Fiber will ramp back up, simply for the legal defense of saying that they too are a cable company and the same laws should apply. In any case, consumers want it. High-end consumers will pay a premium, and the cost for urban footprint fill-out is not exorbitant. I'm not sure I see why the election will make things more favorable for Google Fiber. Nor do I necessarily agree with the statement that "consumers want it." FiOS is awesome in the D.C. metro area. I get 150 mbps symmetric all times of day with 8 ms pings to my hosted server. In rural Maryland. Yet, Verizon has struggled to hit even 33% market penetration. It's incomprehensible to me, but most people seem to care more about television viewing options than broadband. My parents wanted to ditch FiOS because of all the Indian channels Cox Cable had!! > 2. Cable coaxial wire - this is a big thick wire that continues to upgrade nicely. DOCSIS 3.0 is basically the same old 1980s/1990s cable with better comms protocols. It's getting to the point to where they can compete on a reasonable basis with single strand fiber, at a lower speed and price point. So cable companies build out the back-end with fiber, but don't have to replant single homes. Modern HFC networks are nothing like the 1980s/1990s networks even when you're talking about the wires in the ground. The only thing that hasn't been upgraded is the coaxial cable into the house.
- dforrestwilson1 10y ago
  I live in Kansas City, and people here are pretty bonkers for it. Prior to Google Fiber we had AT&T and Time Warner Cable. I can't even begin to list all of the amazing features Google Fiber provides, and the service experience is far superior to the typical ISP. We recently house-hunted and Google Fiber is one of the must-have items. I would guess that Google Fiber would give Alphabet wiggle room when it comes to net neutrality etc. They will have to buy content from Comcast and AT&T. They could end up competing on search and advertising against Verizon. I could see advantages to being able to argue legally that they also are an ISP as well as a media company and search provider. Just my opinion/hope though.
- massysett 10y ago
  > It's incomprehensible to me, but most people seem to care more about television viewing options than broadband. I'm also in DC area and live in a house where the previous owners had Fios. We have Comcast and we are happy with it. I get over 200 megabits down, only about 12 megabits up. I know Fios can get us faster uplink speed. But I don't care. This uplink is fast enough. I seldom upload much of anything. I remember when my uplink was 1.5 megabits years ago, and I was happy with that, so I'm not going to get Fios just to get more speed I don't need. The Comcast X1 box is excellent and it's the first time I've had a cable box that is even decent. The voice command is great. One night I hit the mic button and said "Saturday Night Fever" and seconds later we were watching the movie. Maybe Fios has a box that is as good. I have no idea. But their marketing gives me no reason to believe they have a box that is as good, so I have no reason to switch. When both have Internet that is more than good enough for me, things like a box are going to make all the difference. The simple fact is that cable Internet is more than good enough for most people, so they are not going to switch to get some uplink speed or some faster ping. My neighbor has Fios and asked me about how happy I am with Comcast and said he is considering getting Comcast. Clearly Fios is not offering something so amazing that people would not consider leaving.
scholia 10y ago
> We're talking about 3 different technologies here I'd be interested to know why we aren't talking about four different technologies, with the fourth one being wireless. You can roll out fast broadband to large areas just by building a few towers to broadcast WiMAX. There's no "backhoe time" and you avoid the cost of digging up the streets. IEEE 802.16m will now give you a gigabit to a fixed point You could do the same with LTE but you need a lot more towers and I don't know about spectrum availability. However, WiMAX seems to have failed in the US, with Clearwire shutting down in March...
- sgt101 10y ago
  The bearer is contended. There are 100s of users sharing the cell throughput, which works for the normal burst mobile cases but when you get to streaming cases like updates and video not so much.
- dforrestwilson1 10y ago
  Everything I've read about wireless is that we're hitting a wall when it comes to wireless ranges. 5GHz for example is much faster, but also a much shorter range (and bad wall/tree penetration) vs 2.4GHz. Maybe I'm wrong though?
- jaredklewis 10y ago
  > building a few towers to broadcast WiMAX Easier said than done. It might be workable in dense urban areas where there are already lots of tall buildings, but getting approval to build towers in many residential areas is like pulling teeth. To support peak usage hours when everyone is at home HD streaming hulu and netflix, even with WiMax, you will need way more (i.e. smaller) cells than wireless carriers have today, which means lots more towers. Plus you'll need fiber going to all of these small cell towers, which means you're half way to FTTN anyway. I think wireless internet might fit some niche situations well, but I'm skeptical there is significant capital or long terms savings in the general case. Not to mention that it will never be able to compete with fiber on latency or bandwidth.
- darpa_escapee 10y ago
  > However, WiMAX seems to have failed in the US, with Clearwire shutting down in March... I played the WiMAX game in the US as a consumer. If you didn't have clear line of site with the tower, you got terrible speeds if you still had a connection. Driving + WiMAX was a no go unless it was through an empty field (never got to try this one). Using WiMAX near a window and getting up for some coffee was a no go. Hell, being outside that pane of glass made a huge difference. It was really bad for mobile. > You can roll out fast broadband to large areas just by building a few towers to broadcast WiMAX. I can see this working if customers had outdoor antennas installed.

ClassyJacket 10y ago

Point three makes me weep. In Australia the government started building a nationwide fibre to the home network. Wholesale only, so any ISP could provide services on it. They got about 10% of the way in, then a new conservative government got in power and stomped on the plan. Instead they're doing fibre to the node for everyone. Why have gigabit fibre when you can have slower, less reliable, more expensive copper connections that go a whopping 12Mbit!? Faster even, if the node happens to be sitting in your front lawn.

PartTimeZombie 10y ago
We started the same thing here in New Zealand. I was a bit outraged when $1.2 billion of taxpayers money was handed out to build a national fibre network, but it's turning out not too bad. Not perfect, but not bad either. It helps that the opposition bought in, despite disagreeing on details.
- bitserf 10y ago
  NZ's UFB network has always been FTTH (meaning it terminates inside your house), there is no copper in this picture. This is why, for example, last month people who qualified for 200MB can now get gigabit plans.
- yabatopia 10y ago
  Why were you outraged that your government invested $1.2 billion in a significant upgrade of an important piece of infrastructure, beneficial for citizens and companies, now and in the future? Was it because you've jumped on the anti-government train that's so popular right now, like the overused term "taxpayers money" suggests, or did you really believe that building a national fibre network was the dumbest idea ever?
  - CalRobert 10y ago
    Considering the population of New Zealand, that's a bill of around $255 per person. Perhaps they think the money could have been better spent?
    - theandrewbailey 10y ago
      $255 sounds extremely reasonable. That's much less than what most people pay annually for service.
      - orclev 10y ago
        I would happily pay a one time cost of $255 to run fiber to my house. I'd even be willing to pay $500 to run fiber to my house. Sadly around here if I want fiber to my house, even assuming I could find an ISP willing to run it, it would probably cost me upwards of $2000 to have it run, and probably significantly more than that. That's also not counting the cost of actually getting service, that's just the "installation fee" to have it run.
      - rubber_duck 10y ago
        He said per person, not per household - not the same unless you're living alone. Also this doesn't make your internet free you have to pay for the service on top of that. And believe it or not a lot of people don't care about having fiber internet available.
rasz_pl 10y ago
It doesnt make a bit of difference because whole project was doomed from the start (staffing committee with incumbent telco execs made sure of that). Whats the point of nation wide fiber network when you force data caps on it like it was some kind of undersea cable to mars?
- ClassyJacket 10y ago
  It wasn't doomed from the start. The original plan was a good one. If we'd stuck to the fibre plan, nearly half the country would have fibre by now. The CVC prices are unreasonable but they can and will come down. Once the fibre is in place you can argue about the business details later. You can't, however, lay fibre overnight. That takes years of planning. The POI decision was dumb too but it can be worked around. As for the data caps, you can already buy an unlimited 100Mbps service right now. So it's just a matter of cost.
gnopgnip 10y ago
There is new technology like g.fast for 100+mbit speeds over a copper pair. Yes it is not as fast as either fiber or co-ax, and it is short sighted, but 100mbit is a huge improvement over legacy DSL speeds.
- ClassyJacket 10y ago
  It already goes 100Mbit IF you're very lucky and happen to have the node right near your house (and then some poor sap at the other end gets 12). And no, they cannot upgrade FttN to G.fast, no matter what they tell you. That requires shorter copper lengths than all but the closest handful of houses to the node have. Plus you need everyone on the same technology so that the vectoring (like active noise cancellation) can work. Gfast drops off a cliff after a few hundred metres, and they have houses connected to nodes a kilometre away. Gfast and Xgfast are only appropriate for Fibre-to-the-curb. FttN can never be meaningfully upgraded.
- scholia 10y ago
  You can get 80 meg broadband from BT Infinity with fibre to the street cabinet, so you might need to offer more than 100+. G.fast extends the fibre closer to the house but I suspect it's not very cost-effective unless the take-up is high, so perhaps it will mainly cater to rich people in suburbia. We've know that fibre (FTTH) was the answer since the 1980s but all British governments have been too short-sighted and stupid to finance it. Of course, they dug up the streets to install new pipes for the short-term benefits of North Sea gas, and they can find £50 billion for HS2, when they could have FTTH for less.
  - lostboys67 10y ago
    And realistically no consumer needs more than that.
    - razakel 10y ago
      >And realistically no consumer needs more than that. Yet. 640K ought to be enough for anybody.
lostboys67 10y ago
FTC (fiber To cab) makes economic sense as you can get decent speeds with the massive expense of doing FTP.

nickysielicki 10y ago

> DOCSIS 3.0 is basically the same old 1980s/1990s cable with better comms protocols. It's getting to the point to where they can compete on a reasonable basis with single strand fiber, at a lower speed and price point. It's not just that the communications protocols got better. I think that kind of implies that the medium is at it's end, and that the difference between DOCSIS 2 @ 38Mb/27Mb and DOCSIS 3.1 @ 10Gb/1Gb is some engineers squeezing more out of copper through ingenious protocol-level optimizations. And that's mostly not the case. The big difference is that 3.1 came out a decade later and can take advantage of amazing progress made with ADCs/DACs. I think it's reasonable to claim that the driving force in internet bandwidth gains have not been because of better mediums, like fiber instead of co-ax. The easiest place to look is wireless, where the medium hasn't changed and won't change, but the speeds have exploded. 802.11ax might push 10Gbit. LTE is an order of magnitude faster than EDGE. It's easier to push more data when there's less noise on the line. But if we can work around the noise instead, it makes little sense to spend millions laying new cables.

azernik 10y ago
From the infrastructure-investment perspective, only changing the medium is a big-ticket item (digging up cables, installing new ones, etc.). Swapping out newer transceivers on the ends is a tiny investment.

rayiner 10y ago

> My hope is that post-election Google Fiber will ramp back up, simply for the legal defense of saying that they too are a cable company and the same laws should apply. In any case, consumers want it. High-end consumers will pay a premium, and the cost for urban footprint fill-out is not exorbitant. I'm not sure I see why the election will make things more favorable for Google Fiber. Nor do I necessarily agree with the statement that "consumers want it." FiOS is awesome in the D.C. metro area. I get 150 mbps symmetric all times of day with 8 ms pings to my hosted server. In rural Maryland. Yet, Verizon has struggled to hit even 33% market penetration. It's incomprehensible to me, but most people seem to care more about television viewing options than broadband. My parents wanted to ditch FiOS because of all the Indian channels Cox Cable had!! > 2. Cable coaxial wire - this is a big thick wire that continues to upgrade nicely. DOCSIS 3.0 is basically the same old 1980s/1990s cable with better comms protocols. It's getting to the point to where they can compete on a reasonable basis with single strand fiber, at a lower speed and price point. So cable companies build out the back-end with fiber, but don't have to replant single homes. Modern HFC networks are nothing like the 1980s/1990s networks even when you're talking about the wires in the ground. The only thing that hasn't been upgraded is the coaxial cable into the house.

dforrestwilson1 10y ago
I live in Kansas City, and people here are pretty bonkers for it. Prior to Google Fiber we had AT&T and Time Warner Cable. I can't even begin to list all of the amazing features Google Fiber provides, and the service experience is far superior to the typical ISP. We recently house-hunted and Google Fiber is one of the must-have items. I would guess that Google Fiber would give Alphabet wiggle room when it comes to net neutrality etc. They will have to buy content from Comcast and AT&T. They could end up competing on search and advertising against Verizon. I could see advantages to being able to argue legally that they also are an ISP as well as a media company and search provider. Just my opinion/hope though.
massysett 10y ago
> It's incomprehensible to me, but most people seem to care more about television viewing options than broadband. I'm also in DC area and live in a house where the previous owners had Fios. We have Comcast and we are happy with it. I get over 200 megabits down, only about 12 megabits up. I know Fios can get us faster uplink speed. But I don't care. This uplink is fast enough. I seldom upload much of anything. I remember when my uplink was 1.5 megabits years ago, and I was happy with that, so I'm not going to get Fios just to get more speed I don't need. The Comcast X1 box is excellent and it's the first time I've had a cable box that is even decent. The voice command is great. One night I hit the mic button and said "Saturday Night Fever" and seconds later we were watching the movie. Maybe Fios has a box that is as good. I have no idea. But their marketing gives me no reason to believe they have a box that is as good, so I have no reason to switch. When both have Internet that is more than good enough for me, things like a box are going to make all the difference. The simple fact is that cable Internet is more than good enough for most people, so they are not going to switch to get some uplink speed or some faster ping. My neighbor has Fios and asked me about how happy I am with Comcast and said he is considering getting Comcast. Clearly Fios is not offering something so amazing that people would not consider leaving.

scholia 10y ago

> We're talking about 3 different technologies here I'd be interested to know why we aren't talking about four different technologies, with the fourth one being wireless. You can roll out fast broadband to large areas just by building a few towers to broadcast WiMAX. There's no "backhoe time" and you avoid the cost of digging up the streets. IEEE 802.16m will now give you a gigabit to a fixed point You could do the same with LTE but you need a lot more towers and I don't know about spectrum availability. However, WiMAX seems to have failed in the US, with Clearwire shutting down in March...

sgt101 10y ago
The bearer is contended. There are 100s of users sharing the cell throughput, which works for the normal burst mobile cases but when you get to streaming cases like updates and video not so much.
dforrestwilson1 10y ago
Everything I've read about wireless is that we're hitting a wall when it comes to wireless ranges. 5GHz for example is much faster, but also a much shorter range (and bad wall/tree penetration) vs 2.4GHz. Maybe I'm wrong though?
jaredklewis 10y ago
> building a few towers to broadcast WiMAX Easier said than done. It might be workable in dense urban areas where there are already lots of tall buildings, but getting approval to build towers in many residential areas is like pulling teeth. To support peak usage hours when everyone is at home HD streaming hulu and netflix, even with WiMax, you will need way more (i.e. smaller) cells than wireless carriers have today, which means lots more towers. Plus you'll need fiber going to all of these small cell towers, which means you're half way to FTTN anyway. I think wireless internet might fit some niche situations well, but I'm skeptical there is significant capital or long terms savings in the general case. Not to mention that it will never be able to compete with fiber on latency or bandwidth.
darpa_escapee 10y ago
> However, WiMAX seems to have failed in the US, with Clearwire shutting down in March... I played the WiMAX game in the US as a consumer. If you didn't have clear line of site with the tower, you got terrible speeds if you still had a connection. Driving + WiMAX was a no go unless it was through an empty field (never got to try this one). Using WiMAX near a window and getting up for some coffee was a no go. Hell, being outside that pane of glass made a huge difference. It was really bad for mobile. > You can roll out fast broadband to large areas just by building a few towers to broadcast WiMAX. I can see this working if customers had outdoor antennas installed.