5 ms·

> Internet is a commodity for the average person, you can't make money in that market. This assumes that razor thin margins are bad, which is usually a safe as

> Internet is a commodity for the average person, you can't make money in that market. This assumes that razor thin margins are bad, which is usually a safe assumption but in the case of Starlink, I say it's not open minded enough. To boot Starlink online requires a minimum cost for service area (sats), but once you've got that area booted, onboarding new customers is trivial (vs laying fiber, cable, digging holes). The user terminal is for each individual - "Looks like a thin, flat, round UFO on a stick."[0] If you were offered $40 per Gb/s would you switch, 30? 20? Personally I would switch because it's simply not f'ing Comcast. [0] https://www.businessinsider.sg/spacex-starlink-user-terminal-flat-round-ufo-stick-elon-musk-2020-1?r=US&IR=T https://www.businessinsider.sg/spacex-starlink-user-terminal...

tialaramex 7y ago
How about if the offer was $1000 plus $10 per Mb/s and the maximum is 10Mbps ? You would switch because you've imagined this service is magically cheaper and better, but now I say it'll be expensive and worse are you still enthusiastic? Providing network service is a utility, and so yes razor thin margins are typical and that's too bad. It does make very high capital investment (such as launching huge numbers of satellites every few years) difficult to stomach though. These projects (whatever Musk says about them when hyping his company) deliver expensive network access to places where it's otherwise not practical to offer service at all. They aren't going head-to-head with Comcast, or even with AOL, they're competing with "Physically take a day's / week's work to the A site where they have network access".
- ghaff 7y ago
  >These projects (whatever Musk says about them when hyping his company) deliver expensive network access to places where it's otherwise not practical to offer service at all. They aren't going head-to-head with Comcast, or even with AOL, they're competing with "Physically take a day's / week's work to the A site where they have network access". Yeah, there are a lot of places in the US--some of which aren't even really all that rural--that just don't have good Internet options. Today, satellite basically falls into the slightly better than nothing category but it's neither price-competitive nor good if you have essentially any other option. While I know people in tech who make do with only a satellite connection, I'm sure many reading this wouldn't consider that viable and it will be increasingly less so over time. 5G will presumably help in some cases as it rolls out but it certainly won't reach everywhere. And, presumably, over time places that don't have a certain level of Internet access will be increasingly unsuitable at least for long-term residence.
  - ericd 7y ago
    I’d be careful about comparing current satellite internet with what Starlink is promising. The latency and total throughout differences are huge.
    - ghaff 7y ago
      That's why I specifically referred to current satellite. Of course, future satellite could be much better. (Although we'll have to see.)
      - ericd 7y ago
        Right right, I was guessing you were familiar, I was just pointing it out because I think a lot of people reading think about it in the context of HughesNet et al, and I think that leads them to write it off prematurely. I read a nice overview of the constellations a while back, but I realize now that it's actually pretty outdated (2018, and it talks about optical crosslinks between SpaceX's satellites, which has been punted for now): http://www.mit.edu/~portillo/files/Comparison-LEO-IAC-2018-slides.pdf http://www.mit.edu/~portillo/files/Comparison-LEO-IAC-2018-s... So I don't know if my comment about throughput is accurate anymore.
  - BenjiWiebe 7y ago
    5G? Around here in central Kansas 4G is spotty, and there's plenty of places that your phone will lose all data connection, and occasionally even lose the connection entirely. Also a max of 10Mbps would sure beat a max of 6Mbps from CenturyLink. Though at $1000 that would probably be too much...
martinald 7y ago
Doesn't work like that. Each starlink satellite has capacity of 20gig/sec. Assume that only 1 or 2 of the sats will be over a major metro area at any one time - that's 40gig/sec total capacity. That's only ~5,000 HD netflix streams at once (even less for 4K). It will soon become totally overloaded. A city like NYC probably has home/office bandwidth use of 40gig/sec per a few city blocks, nevermind the whole metro area. It's not going to work as a home internet replacement if you already have fibre or cable service. Furthermore, fibre and cable can be upgraded easily by putting new equipment on each side (even cable internet has went from 40mbit/sec to 10gig/sec - fibre has even more capacity). Elon himself has admitted that it won't work in cities for this reason.
- milankragujevic 7y ago
  Yes, but cities have FTTH, rural areas have.. often times nothing that approaches >5 Mbps.
  - reportingsjr 7y ago
    This will always be the case, unless rural subsidies are increased even more drastically (not really a good thing). Infrastructure in cities is far cheaper than rural areas, so this is kind of a cat and mouse game for rural areas where internet will be x times slower than in cities.
    - s_y_n_t_a_x 7y ago
      Rural internet is expensive because no PRIVATE company wants to put up a tower or run fiber for a couple dozen people. A satellite covers a much wider area. We'll see how it plays out, but as a person living out here, I'm excited.
      - leesalminen 7y ago
        It worked for our neighborhood. A retired engineer down the road did exactly this- started a private company to provide good wireless P2P internet to our rural, mountainous community. He’s about 7 years in and now serves 70% of our area. It ain’t cheap, but he’s been profitable enough now to be able to buy additional bandwidth from his upstream. We’re about to get 50/25 in the coming months. I offer my services to him for free just to help contribute and make sure he doesn’t go out of business. It feels like a community effort! I set up the online billing portal and recurring payments, another guy goes on people’s roof to install the ubiquiti receiver, etc. we all chip in.

tialaramex 7y ago

How about if the offer was $1000 plus $10 per Mb/s and the maximum is 10Mbps ? You would switch because you've imagined this service is magically cheaper and better, but now I say it'll be expensive and worse are you still enthusiastic? Providing network service is a utility, and so yes razor thin margins are typical and that's too bad. It does make very high capital investment (such as launching huge numbers of satellites every few years) difficult to stomach though. These projects (whatever Musk says about them when hyping his company) deliver expensive network access to places where it's otherwise not practical to offer service at all. They aren't going head-to-head with Comcast, or even with AOL, they're competing with "Physically take a day's / week's work to the A site where they have network access".

ghaff 7y ago
>These projects (whatever Musk says about them when hyping his company) deliver expensive network access to places where it's otherwise not practical to offer service at all. They aren't going head-to-head with Comcast, or even with AOL, they're competing with "Physically take a day's / week's work to the A site where they have network access". Yeah, there are a lot of places in the US--some of which aren't even really all that rural--that just don't have good Internet options. Today, satellite basically falls into the slightly better than nothing category but it's neither price-competitive nor good if you have essentially any other option. While I know people in tech who make do with only a satellite connection, I'm sure many reading this wouldn't consider that viable and it will be increasingly less so over time. 5G will presumably help in some cases as it rolls out but it certainly won't reach everywhere. And, presumably, over time places that don't have a certain level of Internet access will be increasingly unsuitable at least for long-term residence.
- ericd 7y ago
  I’d be careful about comparing current satellite internet with what Starlink is promising. The latency and total throughout differences are huge.
  - ghaff 7y ago
    That's why I specifically referred to current satellite. Of course, future satellite could be much better. (Although we'll have to see.)
    - ericd 7y ago
      Right right, I was guessing you were familiar, I was just pointing it out because I think a lot of people reading think about it in the context of HughesNet et al, and I think that leads them to write it off prematurely. I read a nice overview of the constellations a while back, but I realize now that it's actually pretty outdated (2018, and it talks about optical crosslinks between SpaceX's satellites, which has been punted for now): http://www.mit.edu/~portillo/files/Comparison-LEO-IAC-2018-slides.pdf http://www.mit.edu/~portillo/files/Comparison-LEO-IAC-2018-s... So I don't know if my comment about throughput is accurate anymore.
- BenjiWiebe 7y ago
  5G? Around here in central Kansas 4G is spotty, and there's plenty of places that your phone will lose all data connection, and occasionally even lose the connection entirely. Also a max of 10Mbps would sure beat a max of 6Mbps from CenturyLink. Though at $1000 that would probably be too much...

martinald 7y ago

Doesn't work like that. Each starlink satellite has capacity of 20gig/sec. Assume that only 1 or 2 of the sats will be over a major metro area at any one time - that's 40gig/sec total capacity. That's only ~5,000 HD netflix streams at once (even less for 4K). It will soon become totally overloaded. A city like NYC probably has home/office bandwidth use of 40gig/sec per a few city blocks, nevermind the whole metro area. It's not going to work as a home internet replacement if you already have fibre or cable service. Furthermore, fibre and cable can be upgraded easily by putting new equipment on each side (even cable internet has went from 40mbit/sec to 10gig/sec - fibre has even more capacity). Elon himself has admitted that it won't work in cities for this reason.

milankragujevic 7y ago
Yes, but cities have FTTH, rural areas have.. often times nothing that approaches >5 Mbps.
- reportingsjr 7y ago
  This will always be the case, unless rural subsidies are increased even more drastically (not really a good thing). Infrastructure in cities is far cheaper than rural areas, so this is kind of a cat and mouse game for rural areas where internet will be x times slower than in cities.
  - s_y_n_t_a_x 7y ago
    Rural internet is expensive because no PRIVATE company wants to put up a tower or run fiber for a couple dozen people. A satellite covers a much wider area. We'll see how it plays out, but as a person living out here, I'm excited.
    - leesalminen 7y ago
      It worked for our neighborhood. A retired engineer down the road did exactly this- started a private company to provide good wireless P2P internet to our rural, mountainous community. He’s about 7 years in and now serves 70% of our area. It ain’t cheap, but he’s been profitable enough now to be able to buy additional bandwidth from his upstream. We’re about to get 50/25 in the coming months. I offer my services to him for free just to help contribute and make sure he doesn’t go out of business. It feels like a community effort! I set up the online billing portal and recurring payments, another guy goes on people’s roof to install the ubiquiti receiver, etc. we all chip in.