4 ms·

Great in theory, but I just don't see this happening with the telcos (AT&T and TWC) controlling the entire broadband footprint in SoCal. It's called collusion,

Great in theory, but I just don't see this happening with the telcos (AT&T and TWC) controlling the entire broadband footprint in SoCal. It's called collusion, look it up. These two telcos (the worst in the world most likely) base their entire business and offerings on providing high-speed internet of yesteryear. AT&T UVerse service is a step up from traditional ADSL with speeds up to 25Mbps (I get 19Mbps regularly and live about 2 miles from the CO). Telco execs have been quoted as saying "consumer just don't need higher speeds." And most probably do not, the niche consumers (probably the HNers) would love 1Gbps speeds for reasons unknown (perhaps to host their own in-house EC2/S3!) but the niche consumers are the 1% and telcos don't care about us. The biggest problem with this concept of Gigabit internet in Los Angeles is who will pay the bill, which will most certainly exceed the estimate. AT&T already has plenty of fiber running to the pole with copper to the household. There are tens of thousands of old buildings in LA that make it impossible or extremely difficult to run fiber to the home. Not to mention, a ton of landlords will simply not care to tear up their infrastructure even if it costs them nothing. There is a VERY serious disease people in LA have which is "We just don't care." Additionally, the "free" access of 2Mbps - 5Mbps is redundant since these are quite frankly ancient speeds regardless of their price economics. So with that said, if you do a simple analysis of how many customers would actually pay for high-speed (>100Mbps) access vs. the "free hoarders" it would be dismal. And would never cover the costs of the roll out. Finally, EVEN if the economics made sense, the telcos would simply fight it, create problems for potential bidders, collude enough to force bidders out and continue their shitty business.

josh2600 13y ago
This is correct, but you do need to understand that there are technical limitations to twisted pair. Even if AT&T or $TELCO runs fiber to the node everywhere, twisted pair has gigantic line loss issues over long distances. We really need fiber to the prem or frankly we're just messing around. There's no financial model that will support building fiber networks except the obvious societal benefit. Having fast fiber everywhere is incredibly valuable from both a social and economic perspective. To put it bluntly, GDP goes up when internet gets faster. Now, the costs will always be wrong for a fiber network. I question why LA would do this instead of building a WISP (wireless ISP) that would deliver speed over radiowave. Forget tearing up all the streets, just put radio towers on the buildings! The costs are lower, the bandwidth is fast and the rollout is asynchronous (you roll the radio towers as you sign people up). Source: I come from the Internet and stuff. My credentials: Comcast fiber rollout, AT&T sales, Avaya consulting and 2600hz (open-source telco infrastructure).
- dakrisht 13y ago
  I completely agree with you - twisted pair has go to go and fiber needs to terminate in the home. It's actually a embarrassing that this country can neither build a somewhat simple healthcare website and has some of the slowest access speeds in the world. I remember going to Bulgaria over the years for business and you can get a 100Mbps symmetrical connection for - wait for it - $40 USD. WISP is an interesting concept and has a better economic model. If you look at proposed 5G wireless speeds, you're looking at 10Gbps+ Even LTE today at 30-50Mbps is better than any non-Gig Fiber to the home (aside from the FIOS service available to maybe 10,000 people in LA) The biggest problem is and will be AT&T and TWC with their collusion. Government regulation needs to intervene here otherwise private sector simply stands no chance. Trust me - I want even 100Mbps connection. It's frustrating. Going back to the concept of a WISP - now that's some innovation if you could make it happen. There was a company on HN a few months back that made long-range WiFi domed Tx/Rx units that could be mounted on poles, etc. Anyone have that link?
  - jauer 13y ago
    The problem with WISP deployments is without exclusive use of spectrum (which you have to win at auction from the FCC vs Verizon/at&t/etc) you can't do LTE-type deployments where you can count on a low noise etc. Beyond that, keep in mind that LTE 30-50Mbps is dependent on the spectrum set on a sector. If everyone ditched the local MSO & Telco for LTE it would get so overloaded that everyone would be saying it sucks and switching back to the MSO. Generally speaking you are left to deal with licensed microwave in the ~$10k per link for ~260Mbps or unlicensed microwave in 2.4/5.8Ghz where you have to deal with noise from every WiFi router and other random junk. Depending on how many other devices are out there you get less than 1 Mbps to 200 Mbps depending on spectrum availability. The technology for WISP deployments is fairly mature, to the point that if you are using straight-up WiFi you are doing it wrong. If you want to learn more, the current market leaders in the US for unlicensed microwave are Cambium (former Motorola)'s PMP series and Ubiquiti's AirMax series. Check them out. Anyway, there just isn't enough spectrum available to do a robust city-wide WISP network with significant uptake.
kalleboo 13y ago
> There are tens of thousands of old buildings in LA that make it impossible or extremely difficult to run fiber to the home Really? Here in Sweden, I haven't seen any problems with running fiber to at least ~80 year old apartment buildings (I recently went apartment hunting, and literally zero of the apartments I looked at didn't have fiber internet)

josh2600 13y ago

This is correct, but you do need to understand that there are technical limitations to twisted pair. Even if AT&T or $TELCO runs fiber to the node everywhere, twisted pair has gigantic line loss issues over long distances. We really need fiber to the prem or frankly we're just messing around. There's no financial model that will support building fiber networks except the obvious societal benefit. Having fast fiber everywhere is incredibly valuable from both a social and economic perspective. To put it bluntly, GDP goes up when internet gets faster. Now, the costs will always be wrong for a fiber network. I question why LA would do this instead of building a WISP (wireless ISP) that would deliver speed over radiowave. Forget tearing up all the streets, just put radio towers on the buildings! The costs are lower, the bandwidth is fast and the rollout is asynchronous (you roll the radio towers as you sign people up). Source: I come from the Internet and stuff. My credentials: Comcast fiber rollout, AT&T sales, Avaya consulting and 2600hz (open-source telco infrastructure).

dakrisht 13y ago
I completely agree with you - twisted pair has go to go and fiber needs to terminate in the home. It's actually a embarrassing that this country can neither build a somewhat simple healthcare website and has some of the slowest access speeds in the world. I remember going to Bulgaria over the years for business and you can get a 100Mbps symmetrical connection for - wait for it - $40 USD. WISP is an interesting concept and has a better economic model. If you look at proposed 5G wireless speeds, you're looking at 10Gbps+ Even LTE today at 30-50Mbps is better than any non-Gig Fiber to the home (aside from the FIOS service available to maybe 10,000 people in LA) The biggest problem is and will be AT&T and TWC with their collusion. Government regulation needs to intervene here otherwise private sector simply stands no chance. Trust me - I want even 100Mbps connection. It's frustrating. Going back to the concept of a WISP - now that's some innovation if you could make it happen. There was a company on HN a few months back that made long-range WiFi domed Tx/Rx units that could be mounted on poles, etc. Anyone have that link?
- jauer 13y ago
  The problem with WISP deployments is without exclusive use of spectrum (which you have to win at auction from the FCC vs Verizon/at&t/etc) you can't do LTE-type deployments where you can count on a low noise etc. Beyond that, keep in mind that LTE 30-50Mbps is dependent on the spectrum set on a sector. If everyone ditched the local MSO & Telco for LTE it would get so overloaded that everyone would be saying it sucks and switching back to the MSO. Generally speaking you are left to deal with licensed microwave in the ~$10k per link for ~260Mbps or unlicensed microwave in 2.4/5.8Ghz where you have to deal with noise from every WiFi router and other random junk. Depending on how many other devices are out there you get less than 1 Mbps to 200 Mbps depending on spectrum availability. The technology for WISP deployments is fairly mature, to the point that if you are using straight-up WiFi you are doing it wrong. If you want to learn more, the current market leaders in the US for unlicensed microwave are Cambium (former Motorola)'s PMP series and Ubiquiti's AirMax series. Check them out. Anyway, there just isn't enough spectrum available to do a robust city-wide WISP network with significant uptake.

kalleboo 13y ago

> There are tens of thousands of old buildings in LA that make it impossible or extremely difficult to run fiber to the home Really? Here in Sweden, I haven't seen any problems with running fiber to at least ~80 year old apartment buildings (I recently went apartment hunting, and literally zero of the apartments I looked at didn't have fiber internet)