The build out of the new internet starts this week....

XXVoid MainXX
Posts
7733
Joined
5/25/2012
Location
Schenectady, NY US
Edited Date/Time 5/23/2019 9:49pm
|
avidchimp
Posts
4550
Joined
7/9/2008
Location
Thousand Oaks, CA US
Fantasy
1182nd
5/12/2019 9:22am
That's nucking futs. That nose cone is massive!
Shawn142
Posts
2598
Joined
10/27/2008
Location
Burleson, TX US
5/12/2019 2:27pm Edited Date/Time 5/12/2019 2:30pm
Suppose to be a 4000+ satellite constellation in low Earth orbit which will allow for acceptable latency over the problematic stationary geosync Internet satellites of the past. Sats will also be able to communicate with eachother and share bandwidth load, providing high speed Internet to anywhere in the world.

SpaceX is only one company working on this though. OneWeb is also on the verge of launching a cluster of test satellites for their constellation, and Amazon has something similar planned once Jeff Bezos' other company Blue Origin brings their orbital launch vehicle online. SpaceX is definitely leading this race though. This has been a major goal for Elon and the company for about 10 years now. A lot was going on behind the scenes with this project. Everyone always asked how Elon planned on paying for his overly ambitious goals, this is the answer to that question. The entire world will be a potential customer.
XXVoid MainXX
Posts
7733
Joined
5/25/2012
Location
Schenectady, NY US
5/12/2019 3:36pm
Yep, this would be a way to generate HUGE funding for his mars plans. And, it would be great for Earth. It will actually be much faster than terrestrial based over long distances because laser light travels much slower through glass (fiber optics) than it does through the vacuum of space. So communications to other countries should be much faster.
plowboy
Posts
11628
Joined
1/3/2010
Location
Norwich, KS US
5/12/2019 5:30pm
I'm still waiting on this "high speed" we were promised. My phone, internet, tv, are still glitchy, and unreliable. The ads seem to be pretty sweet though.
3

The Shop

Falcon
Posts
10108
Joined
11/16/2011
Location
Menifee, CA US
Fantasy
798th
5/13/2019 10:35am Edited Date/Time 5/13/2019 10:37am
Yep, this would be a way to generate HUGE funding for his mars plans. And, it would be great for Earth. It will actually be much...
Yep, this would be a way to generate HUGE funding for his mars plans. And, it would be great for Earth. It will actually be much faster than terrestrial based over long distances because laser light travels much slower through glass (fiber optics) than it does through the vacuum of space. So communications to other countries should be much faster.
You are correct about the speed of electricity in a cable vs. light in a vacuum, but electricity in a fiber optic cable is still awesomely fast: about 69% the speed of light, or 128,000 miles per second (460 million MPH.)

My guess is that therefore, the difference in transit time between any two points on Earth is virtually null, or at least undetectable by humans. What we perceive as "speed" in internet terms has more to do with the processing speed of all the various devices in the signal chain, I think. Correct me if I'm wrong.




EDIT: Here I am, adding to the Vital brain trust by arguing about the speed of light instead of heralding the kick-ass new satellites. Way to go, Falcon!!! Hahaha.

XXVoid MainXX
Posts
7733
Joined
5/25/2012
Location
Schenectady, NY US
5/13/2019 10:41am
Latency is definitely huge in many circumstances. The more you can reduce it the better. Compared to geostationary satellites (like Hughsnet) an LEO solution is just as good or better than ground based. But the biggest advantage is now you'll have high speed connectivity anywhere in the world.
Falcon
Posts
10108
Joined
11/16/2011
Location
Menifee, CA US
Fantasy
798th
5/13/2019 1:30pm
Wouldn't the latency inherent in fiber optic only move the entire data packet back a few milliseconds? All the number crunching would still happen at the speed that the servers operate; it would simply begin and end some infinitesimally small amount of time later. Wouldn't it? IDK, I'm not an IT expert by any means.

Definitely a positive to have high-speed from any location.
1
XXVoid MainXX
Posts
7733
Joined
5/25/2012
Location
Schenectady, NY US
5/13/2019 5:03pm Edited Date/Time 5/13/2019 6:24pm
Latency is cumulative. That is, every device in your path (number of "hops") including the fiber optic cables will add to the total latency . You will notice latency issues on apps that send a lot of smaller packets back and forth (tcp request/reply). Both to and from trips are also added together. Also, one transaction might take several packets back and forth from source to destintation and back. Higher latency becomes very noticeable in gaming apps, voice over ip, web based communications, etc, etc. Latency isn't as important if you are just transferring large files. With LEO satellite based communication I would think the number of hops should be lower between source and destination, and combined with low latency lasers in a vacuum to communicate between the sats I would think it should be really low latency compared to most other end to end network paths.
XXVoid MainXX
Posts
7733
Joined
5/25/2012
Location
Schenectady, NY US
5/13/2019 5:23pm Edited Date/Time 5/13/2019 5:50pm
So for example, here's a traceroute that shows all the hops between me and a random host I picked in Germany:

[void@laplinux ~]$ traceroute ping.eu
traceroute to ping.eu (88.198.46.60), 30 hops max, 60 byte packets
1 buffalo.void.home (192.168.0.1) 1.360 ms 1.409 ms 1.358 ms
2 modem.void.home (192.168.1.254) 2.004 ms 2.157 ms 2.130 ms
3 107-136-16-1.lightspeed.stlsmo.sbcglobal.net (107.136.16.1) 20.791 ms 20.897 ms 21.541 ms
4 64.148.121.156 (64.148.121.156) 22.447 ms 23.530 ms 23.503 ms
5 71.144.225.140 (71.144.225.140) 26.260 ms 26.494 ms 23.120 ms
6 12.83.40.177 (12.83.40.177) 24.376 ms 24.252 ms 12.83.40.173 (12.83.40.173) 25.237 ms
7 cgcil403igs.ip.att.net (12.122.132.121) 32.923 ms 30.367 ms 32.726 ms
8 chi-b21-link.telia.net (213.248.87.253) 30.221 ms 30.256 ms 30.113 ms
9 nyk-bb4-link.telia.net (62.115.137.58) 46.191 ms nyk-bb3-link.telia.net (80.91.246.163) 48.501 ms 48.392 ms
10 ldn-bb4-link.telia.net (62.115.112.245) 112.488 ms ldn-bb3-link.telia.net (62.115.113.21) 118.639 ms ldn-bb4-link.telia.net (62.115.112.245) 112.542 ms
11 prs-bb4-link.telia.net (62.115.114.229) 120.522 ms prs-bb3-link.telia.net (62.115.134.92) 136.514 ms prs-bb4-link.telia.net (62.115.114.229) 120.507 ms
12 ffm-bb3-link.telia.net (62.115.123.12) 141.716 ms 142.078 ms 145.041 ms
13 ffm-b4-link.telia.net (62.115.120.0) 136.437 ms ffm-b4-link.telia.net (62.115.120.6) 144.900 ms 144.864 ms
14 hetzner-ic-326013-ffm-b4.c.telia.net (213.248.70.3) 137.384 ms 133.025 ms 168.116 ms
15 core22.fsn1.hetzner.com (213.239.224.245) 143.029 ms 139.144 ms core21.fsn1.hetzner.com (213.239.224.241) 137.720 ms
16 ex9k1.dc13.fsn1.hetzner.com (213.239.245.238) 139.793 ms 141.255 ms 139.658 ms
17 faces.eu (88.198.46.60) 140.086 ms 140.417 ms *


And ping checks to that host:

[void@laplinux ~]$ ping ping.eu
PING ping.eu (88.198.46.60) 56(84) bytes of data.
64 bytes from faces.eu (88.198.46.60): icmp_seq=1 ttl=45 time=234 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=2 ttl=45 time=153 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=3 ttl=45 time=176 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=4 ttl=45 time=199 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=5 ttl=45 time=223 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=6 ttl=45 time=512 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=7 ttl=45 time=165 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=8 ttl=45 time=191 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=9 ttl=45 time=210 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=10 ttl=45 time=233 ms
^C
--- ping.eu ping statistics ---
10 packets transmitted, 10 received, 0% packet loss, time 20ms
rtt min/avg/max/mdev = 153.437/229.610/511.886/97.656 ms


That's a quarter to a half of a second latency per packet. Compared with a google address:

[void@laplinux ~]$ ping www.google.com
PING www.google.com (172.217.4.196) 56(84) bytes of data.
64 bytes from lga15s48-in-f196.1e100.net (172.217.4.196): icmp_seq=1 ttl=51 time=29.9 ms
64 bytes from lga15s48-in-f196.1e100.net (172.217.4.196): icmp_seq=2 ttl=51 time=31.8 ms
64 bytes from lga15s48-in-f196.1e100.net (172.217.4.196): icmp_seq=3 ttl=51 time=33.3 ms
64 bytes from lga15s48-in-f196.1e100.net (172.217.4.196): icmp_seq=4 ttl=51 time=33.9 ms
64 bytes from lga15s48-in-f196.1e100.net (172.217.4.196): icmp_seq=5 ttl=51 time=33.8 ms
64 bytes from lga15s48-in-f196.1e100.net (172.217.4.196): icmp_seq=6 ttl=51 time=30.4 ms
64 bytes from lga15s48-in-f196.1e100.net (172.217.4.196): icmp_seq=7 ttl=51 time=30.3 ms
64 bytes from lga15s48-in-f196.1e100.net (172.217.4.196): icmp_seq=8 ttl=51 time=30.9 ms
64 bytes from lga15s48-in-f196.1e100.net (172.217.4.196): icmp_seq=9 ttl=51 time=33.7 ms
64 bytes from lga15s48-in-f196.1e100.net (172.217.4.196): icmp_seq=10 ttl=51 time=30.4 ms
^C
--- www.google.com ping statistics ---
10 packets transmitted, 10 received, 0% packet loss, time 24ms
rtt min/avg/max/mdev = 29.915/31.846/33.883/1.578 ms

And a traceroute for comparison:

[void@laplinux ~]$ traceroute www.google.com
traceroute to www.google.com (172.217.4.68), 30 hops max, 60 byte packets
1 buffalo.void.home (192.168.0.1) 1.683 ms 1.637 ms 1.625 ms
2 modem.void.home (192.168.1.254) 2.486 ms 2.589 ms 2.587 ms
3 107-136-16-1.lightspeed.stlsmo.sbcglobal.net (107.136.16.1) 26.180 ms 27.520 ms 28.169 ms
4 64.148.121.156 (64.148.121.156) 21.562 ms 22.565 ms 22.638 ms
5 71.144.225.140 (71.144.225.140) 22.713 ms 22.909 ms 22.606 ms
6 12.83.40.177 (12.83.40.177) 22.454 ms 12.83.40.173 (12.83.40.173) 24.767 ms 24.516 ms
7 12.122.99.101 (12.122.99.101) 30.050 ms 30.423 ms 30.657 ms
8 12.247.108.6 (12.247.108.6) 30.575 ms 31.490 ms 12.247.108.10 (12.247.108.10) 30.230 ms
9 * * *
10 108.170.238.236 (108.170.238.236) 33.868 ms 33.826 ms 216.239.42.32 (216.239.42.32) 30.857 ms
11 108.170.243.197 (108.170.243.197) 30.570 ms 30.372 ms 216.239.51.145 (216.239.51.145) 32.260 ms
12 209.85.241.124 (209.85.241.124) 30.509 ms 30.532 ms 209.85.250.144 (209.85.250.144) 31.059 ms
13 209.85.241.122 (209.85.241.122) 32.583 ms 216.239.63.32 (216.239.63.32) 30.435 ms ord37s18-in-f4.1e100.net (172.217.4.68) 29.245 ms

Some apps would be much more noticeable than others.
XXVoid MainXX
Posts
7733
Joined
5/25/2012
Location
Schenectady, NY US
5/13/2019 5:33pm Edited Date/Time 5/13/2019 5:44pm
I actually have 2 different internet providers here at the house. The above traces were over my ATT Uverse connection. Here's the same traces from my Charter/Spectrum connection:

Germany:

[void@laplinux ~]$ traceroute ping.eu
traceroute to ping.eu (88.198.46.60), 30 hops max, 60 byte packets
1 buffalo.void.home (192.168.0.1) 1.432 ms 1.484 ms 1.557 ms
2 * * *
3 acr08ovldmo-gbe-1-35.ovld.mo.charter.com (96.34.48.0) 36.775 ms 36.648 ms 33.233 ms
4 crr02blvlil-bue-10.blvl.il.charter.com (96.34.50.97) 36.689 ms 36.661 ms 36.635 ms
5 bbr01blvlil-bue-110.blvl.il.charter.com (96.34.2.170) 49.934 ms 44.039 ms 36.743 ms
6 bbr01olvemo-bue-3.olve.mo.charter.com (96.34.0.14) 37.692 ms 42.969 ms 30.752 ms
7 bbr02chcgil-bue-2.chcg.il.charter.com (96.34.0.12) 42.801 ms 22.192 ms 23.821 ms
8 prr01chcgil-bue-4.chcg.il.charter.com (96.34.3.11) 21.006 ms 19.951 ms 21.053 ms
9 * * *
10 * * *
11 ae0-3356.lon10.core-backbone.com (212.113.8.42) 135.323 ms 134.968 ms 129.184 ms
12 ae2-2077.fra20.core-backbone.com (5.56.18.1) 126.008 ms 125.811 ms 124.862 ms
13 core-backbone.hetzner.com (80.255.15.122) 126.727 ms 127.428 ms 128.369 ms
14 core21.fsn1.hetzner.com (213.239.224.241) 142.715 ms core22.fsn1.hetzner.com (213.239.224.245) 130.239 ms core21.fsn1.hetzner.com (213.239.224.241) 141.399 ms
15 ex9k1.dc13.fsn1.hetzner.com (213.239.245.242) 146.221 ms ex9k1.dc13.fsn1.hetzner.com (213.239.245.238) 133.678 ms ex9k1.dc13.fsn1.hetzner.com (213.239.245.242) 138.072 ms
16 faces.eu (88.198.46.60) 127.608 ms 124.594 ms 125.006 ms

[void@laplinux ~]$ ping ping.eu
PING ping.eu (88.198.46.60) 56(84) bytes of data.
64 bytes from faces.eu (88.198.46.60): icmp_seq=1 ttl=45 time=174 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=2 ttl=45 time=197 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=3 ttl=45 time=219 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=4 ttl=45 time=139 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=5 ttl=45 time=162 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=6 ttl=45 time=184 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=7 ttl=45 time=207 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=8 ttl=45 time=230 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=9 ttl=45 time=150 ms
64 bytes from faces.eu (88.198.46.60): icmp_seq=10 ttl=45 time=172 ms
^C
--- ping.eu ping statistics ---
10 packets transmitted, 10 received, 0% packet loss, time 23ms
rtt min/avg/max/mdev = 139.494/183.520/230.337/28.192 ms

Google:

[void@laplinux ~]$ traceroute www.google.com
traceroute to www.google.com (172.217.6.4), 30 hops max, 60 byte packets
1 buffalo.void.home (192.168.0.1) 1.475 ms 1.504 ms 1.526 ms
2 * * *
3 acr08ovldmo-gbe-1-35.ovld.mo.charter.com (96.34.48.0) 14.163 ms 23.568 ms 23.435 ms
4 crr02blvlil-bue-10.blvl.il.charter.com (96.34.50.97) 23.586 ms 23.619 ms 23.513 ms
5 bbr01blvlil-bue-110.blvl.il.charter.com (96.34.2.170) 25.786 ms 25.230 ms 26.713 ms
6 bbr01olvemo-bue-3.olve.mo.charter.com (96.34.0.14) 28.864 ms 22.563 ms 27.715 ms
7 bbr02chcgil-bue-2.chcg.il.charter.com (96.34.0.12) 31.953 ms 21.833 ms 33.719 ms
8 prr01chcgil-bue-4.chcg.il.charter.com (96.34.3.11) 26.806 ms 26.287 ms 26.727 ms
9 prr01chcgil-gbe-0-7-0-3.chcg.il.charter.com (96.34.152.117) 26.207 ms 23.851 ms 24.179 ms
10 108.170.243.193 (108.170.243.193) 23.557 ms 21.601 ms 24.177 ms
11 72.14.232.152 (72.14.232.152) 35.240 ms 216.239.42.150 (216.239.42.150) 33.286 ms 216.239.41.160 (216.239.41.160) 30.323 ms
12 108.170.238.137 (108.170.238.137) 26.320 ms 108.170.243.254 (108.170.243.254) 25.452 ms ord38s01-in-f4.1e100.net (172.217.6.4) 23.610 ms


[void@laplinux ~]$ ping www.google.com
PING www.google.com (172.217.6.4) 56(84) bytes of data.
64 bytes from ord38s01-in-f4.1e100.net (172.217.6.4): icmp_seq=1 ttl=51 time=19.10 ms
64 bytes from ord38s01-in-f4.1e100.net (172.217.6.4): icmp_seq=2 ttl=51 time=24.4 ms
64 bytes from ord38s01-in-f4.1e100.net (172.217.6.4): icmp_seq=3 ttl=51 time=20.5 ms
64 bytes from ord38s01-in-f4.1e100.net (172.217.6.4): icmp_seq=4 ttl=51 time=18.6 ms
64 bytes from ord38s01-in-f4.1e100.net (172.217.6.4): icmp_seq=5 ttl=51 time=20.3 ms
64 bytes from ord38s01-in-f4.1e100.net (172.217.6.4): icmp_seq=6 ttl=51 time=27.7 ms
64 bytes from ord38s01-in-f4.1e100.net (172.217.6.4): icmp_seq=7 ttl=51 time=17.4 ms
64 bytes from ord38s01-in-f4.1e100.net (172.217.6.4): icmp_seq=8 ttl=51 time=25.2 ms
64 bytes from ord38s01-in-f4.1e100.net (172.217.6.4): icmp_seq=9 ttl=51 time=30.5 ms
64 bytes from ord38s01-in-f4.1e100.net (172.217.6.4): icmp_seq=10 ttl=51 time=38.4 ms
^C
--- www.google.com ping statistics ---
10 packets transmitted, 10 received, 0% packet loss, time 9ms
rtt min/avg/max/mdev = 17.435/24.291/38.352/6.139 ms
imoto34
Posts
3781
Joined
1/28/2010
Location
TN US
Fantasy
3174th
5/14/2019 10:00am
Just curious, why the two isps?
hard2kill
Posts
369
Joined
9/8/2010
Location
Flag Pond, TN US
5/14/2019 11:24am
Didn't google have a similar project they scraped a while back? I can see this kind of internet distribution allowing blockchain technology to take hold in a big way. Possibly even lead to a global economic solution similar to bitcoin?
Falcon
Posts
10108
Joined
11/16/2011
Location
Menifee, CA US
Fantasy
798th
5/14/2019 11:28am Edited Date/Time 5/14/2019 11:30am
Latency is cumulative. That is, every device in your path (number of "hops") including the fiber optic cables will add to the total latency . You...
Latency is cumulative. That is, every device in your path (number of "hops") including the fiber optic cables will add to the total latency . You will notice latency issues on apps that send a lot of smaller packets back and forth (tcp request/reply). Both to and from trips are also added together. Also, one transaction might take several packets back and forth from source to destintation and back. Higher latency becomes very noticeable in gaming apps, voice over ip, web based communications, etc, etc. Latency isn't as important if you are just transferring large files. With LEO satellite based communication I would think the number of hops should be lower between source and destination, and combined with low latency lasers in a vacuum to communicate between the sats I would think it should be really low latency compared to most other end to end network paths.
Ah, now that makes more sense. I wasn't considering back/forth transmissions.
XXVoid MainXX
Posts
7733
Joined
5/25/2012
Location
Schenectady, NY US
5/14/2019 5:35pm
imoto34 wrote:
Just curious, why the two isps?
I work from home and need reliability. If one goes down the other takes over.
Shawn142
Posts
2598
Joined
10/27/2008
Location
Burleson, TX US
5/14/2019 6:29pm
hard2kill wrote:
Didn't google have a similar project they scraped a while back? I can see this kind of internet distribution allowing blockchain technology to take hold in...
Didn't google have a similar project they scraped a while back? I can see this kind of internet distribution allowing blockchain technology to take hold in a big way. Possibly even lead to a global economic solution similar to bitcoin?
I'm not sure if the project was similar or not but Google has put significant investment into SpaceX's Starlink program rather than try to compete against it.
early
Posts
8260
Joined
2/13/2013
Location
University Heights, OH US
Fantasy
2212th
5/15/2019 6:07am
What is the probability that existing isp's are successful in delaying the widespread adoption of this technology through regulation or exclusive rights deals?
Falcon
Posts
10108
Joined
11/16/2011
Location
Menifee, CA US
Fantasy
798th
5/15/2019 11:55am
That's great info, thanks. It made me think of the articles below, however. Hmmmm.....

https://www.popularmechanics.com/space/a26885/space-junk-cleanup/

(Here's a pretty unabashed anti-starlink article. Not sure how accurate it all is - the dude sounds like he's getting ready for a tin foil capSmile
https://www.3blmedia.com/News/Are-We-Locking-Earth-Space-Prison




https://youtu.be/yS1ibDImAYU
XXVoid MainXX
Posts
7733
Joined
5/25/2012
Location
Schenectady, NY US
5/15/2019 7:08pm
The orbit that these satellites are going in is basically self cleaning within 5 years. The drag from the atmosphere slows them down and they drop back to earth and burn up. Looks like launch time has been moved up to about 53 minutes from now.
XXVoid MainXX
Posts
7733
Joined
5/25/2012
Location
Schenectady, NY US
5/15/2019 7:30pm
I wonder if Hughsnet and Iridium and others are crapping in their boots?
yz133rider
Posts
4467
Joined
8/1/2013
Location
Avondale, PA US
5/15/2019 7:32pm
What in the hell are you guys talking about? Way way too technical for me!
Shawn142
Posts
2598
Joined
10/27/2008
Location
Burleson, TX US
5/15/2019 7:42pm
Falcon wrote:
That's great info, thanks. It made me think of the articles below, however. Hmmmm..... https://www.popularmechanics.com/space/a26885/space-junk-cleanup/ (Here's a pretty unabashed anti-starlink article. Not sure how accurate it...
That's great info, thanks. It made me think of the articles below, however. Hmmmm.....

https://www.popularmechanics.com/space/a26885/space-junk-cleanup/

(Here's a pretty unabashed anti-starlink article. Not sure how accurate it all is - the dude sounds like he's getting ready for a tin foil capSmile
https://www.3blmedia.com/News/Are-We-Locking-Earth-Space-Prison




https://youtu.be/yS1ibDImAYU
I don't think people who make this argument really understand just how much space is up there in orbit. A lot of what we track in these situations is natural debris caught as the Earth travels. Despite what the movie Gravity had to show you that's not how orbital debris actually works.

While we are certainly in a golden age of being able to launch satellites there are only a few thousand up there old and new. Things like spent rocket stages are often not at orbital velocity and will slowly decay back into the atmosphere. It is of course worth the time figuring out how to de-orbit old tech that is just floating up there. That's an orbital slot someone else could use, and there are programs working on how to deal with it.
imoto34
Posts
3781
Joined
1/28/2010
Location
TN US
Fantasy
3174th
5/15/2019 7:42pm
imoto34 wrote:
Just curious, why the two isps?
I work from home and need reliability. If one goes down the other takes over.
Understandable.
IWreckALot
Posts
8677
Joined
3/12/2011
Location
Fort Worth, TX US
5/16/2019 7:23am
imoto34 wrote:
Just curious, why the two isps?
He also likes to listen to the radio while he watches TV.
1

Post a reply to: The build out of the new internet starts this week....

The Latest