4 ms·
This article relies on Ookla speedtest results. I don't think those are representative for satellite internet, when speeds vary throughout the day based on sate
by diebeforei485 4y ago
This article relies on Ookla speedtest results. I don't think those are representative for satellite internet, when speeds vary throughout the day based on satellite positions relative to the user.
People usually go to test their speed when they feel their connection is slow, so it's not a uniform random sample.
- yencabulator 4y ago> I don't think those are representative for satellite internet, when speeds vary throughout the day based on satellite positions relative to the user. Huh? It's not Hughes, Starlink is not geostationary. Starlink satellites are in low orbit, orbiting the whole globe in about two hours. The dish switches satellites every few minutes. https://satellitemap.space/?constellation=starlink https://satellitemap.space/?constellation=starlink
- diebeforei485 4y agoYes, the satellites are moving, so the quality of service at any time varies depending on where the satellites are at that time.
- yencabulator 4y agoThey move quickly enough that your variation would be on the time scale of minutes, not over a day. (Earlier in the beta, I had a 20 ms hiccup every few minutes, during the hand-off to next satellite. I'm guessing the IP routing wasn't in sync with the satellite switch.)
- diebeforei485 4y agoCorrect, and it's when users feel those hiccups that they're going to speedtest.net - so the results are not representative. Terrestrial broadband doesn't have hiccups as frequently, so the speedtest.net results would have a better signal to noise ratio for terrestrial ISP's vs for Starlink.