5 ms·

Exactly. Fiber to the home is the only fiber than really matters. If there is fiber but my last mile is old copper, then that doesn’t do me a lot of good does i

Exactly. Fiber to the home is the only fiber than really matters. If there is fiber but my last mile is old copper, then that doesn’t do me a lot of good does it? France claiming 100% fiber by 2022? They are dreaming, or perhaps even delusional. No way they can get FTTH by 2022. There are some places that don’t have any connectivity and you have to use Satelite. They’re going to fix that in 4 years?

masklinn 9y ago
> Exactly. Fiber to the home is the only fiber than really matters. If there is fiber but my last mile is old copper, then that doesn’t do me a lot of good does it? I wouldn't go quite that far e.g. let's say you're my parents in a village of 300 (and as many cows) with the DSLAM >5km away, the best DSL you can get is 1024k on a good day (and on average closer to 512k with pings in the half-second range), my understanding is that for them the FTTN node would be roughly at the center of each village. All of a sudden, unless the village is huge there's <2km between any home and the node and you can expect 15~20Mbps over the entire village.
- kaennar 9y ago
  But how would you carry a model like that over to a place as big as the USA? Montana, Wyoming, Texas, Oklahoma, Kansas, and just about every state that isn't on the East/West coast would have an incredibly difficult time executing this kind of plan because of the distance between communities. I'm not saying that a similar model wouldn't be easy to implement in the boroughs of New York, Boston, SV, Austin, etc., but when you still have 20% of Americans living in not-urban environments. Indeed even the urban environments are far more sprawling and spaced out than anywhere in Europe/Asia (e.g. Dallas/FW, Houston, Phoenix, Austin, and even Atlanta). The kind of engineering to distribute such a large network seems to mirror the issues with the nationwide power grid that was proposed to allow for consistent solar energy. The USA is just a very, very big place with populations, even urban, that sprawl and are not condensed like in London, Paris, Munich or other countries in the fiber fight.
  - masklinn 9y ago
    > But how would you carry a model like that over to a place as big as the USA? Montana, Wyoming, Texas, Oklahoma, Kansas, and just about every state that isn't on the East/West coast would have an incredibly difficult time executing this kind of plan because of the distance between communities. https://en.wikipedia.org/wiki/Rural_Electrification_Act https://en.wikipedia.org/wiki/Rural_Electrification_Act > I'm not saying that a similar model wouldn't be easy to implement in the boroughs of New York, Boston, SV, Austin, etc. Oh please, the village I'm talking about is not a borough of a major metropolis, it's a small rural village, the closest thing which could be construed as a city has a population of under 50k and is 20km away. Boston and Austin have metro populations in the millions, there's barely 6 cities in all of France which reach that. > you still have 20% of Americans living in not-urban environments Oh good point, let's check the percentage of rural population in france. Hey look, it's 20%. Wow, probably an outlier, let's check Germany. Oh, 24%. Maybe Spain? 20% again. And these are highly urbanised countries mind, Ireland and Portugal are sitting at 36% rural population. > The USA is just a very, very big place with populations, even urban, that sprawl and are not condensed like in London, Paris, Munich or other countries in the fiber fight. None of London, Paris or Munich are countries, they're major/capital cities inside countries which have their own non-urban populations.
    - soundwave106 9y ago
      America as a whole does have a lot more remote areas. France only has one department with a population density of less than 10 per square kilometer according to Wikipedia (https://en.wikipedia.org/wiki/List_of_French_departments_by_population https://en.wikipedia.org/wiki/List_of_French_departments_by_...). Compare that to America and note all of the counties with a population density of less than 10 per square mile (https://www.census.gov/dmd/www/pdf/512popdn.pdf https://www.census.gov/dmd/www/pdf/512popdn.pdf) -- and a square mile is a larger unit than a square km. Fiber absolutely might be cost prohibitive in certain areas. I still agree with you, though, that America could do something if there was a strong enough will. Even if the cost of fiber is prohibitive, it seems like there are wireless technologies out there that would be better than the zilch broadband 23 million Americans currently have. (https://www.npr.org/2018/03/03/590546371/rural-communities-take-broadband-into-their-own-hands https://www.npr.org/2018/03/03/590546371/rural-communities-t...)
emilecantin 9y ago
It depends on what kind of copper. DSL suffers from the limits of twisted pairs over medium-long distances, with only about 1100 kHz available. DOCSIS, with its shielded coaxial cable, has much higher bandwidth available, up to 750 MHz. In practical terms, this means that DSL has already reached its practical limit with speeds around 30-50 Mbps, whereas Cable Internet still has room to grow around 2-4 Gbps, depending on how many TV channels the operator is ready to sacrifice to Internet traffic. Of course, this doesn't address congestion / oversubscription issues around cable, but assuming a competently managed network, FTTN with coax as the last mile is still a very good choice in 2018.
- amazon_not 9y ago
  Not quite. VDSL2 will nominally do 300 Mbps. Actual 100 Mbps service isn't uncommon. The successor, G.fast, will nominally do a gig, but about 500 Mbps over short loops.

masklinn 9y ago

> Exactly. Fiber to the home is the only fiber than really matters. If there is fiber but my last mile is old copper, then that doesn’t do me a lot of good does it? I wouldn't go quite that far e.g. let's say you're my parents in a village of 300 (and as many cows) with the DSLAM >5km away, the best DSL you can get is 1024k on a good day (and on average closer to 512k with pings in the half-second range), my understanding is that for them the FTTN node would be roughly at the center of each village. All of a sudden, unless the village is huge there's <2km between any home and the node and you can expect 15~20Mbps over the entire village.

kaennar 9y ago
But how would you carry a model like that over to a place as big as the USA? Montana, Wyoming, Texas, Oklahoma, Kansas, and just about every state that isn't on the East/West coast would have an incredibly difficult time executing this kind of plan because of the distance between communities. I'm not saying that a similar model wouldn't be easy to implement in the boroughs of New York, Boston, SV, Austin, etc., but when you still have 20% of Americans living in not-urban environments. Indeed even the urban environments are far more sprawling and spaced out than anywhere in Europe/Asia (e.g. Dallas/FW, Houston, Phoenix, Austin, and even Atlanta). The kind of engineering to distribute such a large network seems to mirror the issues with the nationwide power grid that was proposed to allow for consistent solar energy. The USA is just a very, very big place with populations, even urban, that sprawl and are not condensed like in London, Paris, Munich or other countries in the fiber fight.
- masklinn 9y ago
  > But how would you carry a model like that over to a place as big as the USA? Montana, Wyoming, Texas, Oklahoma, Kansas, and just about every state that isn't on the East/West coast would have an incredibly difficult time executing this kind of plan because of the distance between communities. https://en.wikipedia.org/wiki/Rural_Electrification_Act https://en.wikipedia.org/wiki/Rural_Electrification_Act > I'm not saying that a similar model wouldn't be easy to implement in the boroughs of New York, Boston, SV, Austin, etc. Oh please, the village I'm talking about is not a borough of a major metropolis, it's a small rural village, the closest thing which could be construed as a city has a population of under 50k and is 20km away. Boston and Austin have metro populations in the millions, there's barely 6 cities in all of France which reach that. > you still have 20% of Americans living in not-urban environments Oh good point, let's check the percentage of rural population in france. Hey look, it's 20%. Wow, probably an outlier, let's check Germany. Oh, 24%. Maybe Spain? 20% again. And these are highly urbanised countries mind, Ireland and Portugal are sitting at 36% rural population. > The USA is just a very, very big place with populations, even urban, that sprawl and are not condensed like in London, Paris, Munich or other countries in the fiber fight. None of London, Paris or Munich are countries, they're major/capital cities inside countries which have their own non-urban populations.
  - soundwave106 9y ago
    America as a whole does have a lot more remote areas. France only has one department with a population density of less than 10 per square kilometer according to Wikipedia (https://en.wikipedia.org/wiki/List_of_French_departments_by_population https://en.wikipedia.org/wiki/List_of_French_departments_by_...). Compare that to America and note all of the counties with a population density of less than 10 per square mile (https://www.census.gov/dmd/www/pdf/512popdn.pdf https://www.census.gov/dmd/www/pdf/512popdn.pdf) -- and a square mile is a larger unit than a square km. Fiber absolutely might be cost prohibitive in certain areas. I still agree with you, though, that America could do something if there was a strong enough will. Even if the cost of fiber is prohibitive, it seems like there are wireless technologies out there that would be better than the zilch broadband 23 million Americans currently have. (https://www.npr.org/2018/03/03/590546371/rural-communities-take-broadband-into-their-own-hands https://www.npr.org/2018/03/03/590546371/rural-communities-t...)

emilecantin 9y ago

It depends on what kind of copper. DSL suffers from the limits of twisted pairs over medium-long distances, with only about 1100 kHz available. DOCSIS, with its shielded coaxial cable, has much higher bandwidth available, up to 750 MHz. In practical terms, this means that DSL has already reached its practical limit with speeds around 30-50 Mbps, whereas Cable Internet still has room to grow around 2-4 Gbps, depending on how many TV channels the operator is ready to sacrifice to Internet traffic. Of course, this doesn't address congestion / oversubscription issues around cable, but assuming a competently managed network, FTTN with coax as the last mile is still a very good choice in 2018.

amazon_not 9y ago
Not quite. VDSL2 will nominally do 300 Mbps. Actual 100 Mbps service isn't uncommon. The successor, G.fast, will nominally do a gig, but about 500 Mbps over short loops.