Beta Starlink with a slight tree obstruction vs degraded DOCSIS3 last mile

Wed Jun 30 20:55:47 UTC 2021

On Tue, Jun 29, 2021 at 9:04 PM Eric Kuhnke <eric.kuhnke at gmail.com> wrote:
>
> I'd much rather have 0.11 to 0.25% average packet loss over a 3 to 24 hour period than 2-3%.
>
> The starlink experience is overall superior at present from a subjective user point of view.

... and I think that this last sentence is the most important one.

My current Internet is a 4.3 mile WiFi link to a tower, which has a
wifi link to another tower, which has a wifi link to the POP with
fiber....[0]
There is some correlation between what smokeping shows and what my
subjective experience is[1],  but it's not absolute. Hearing "this
feels better/faster" is sometimes more useful than "this graph shows
X".
Yup, the Starlink graph showed some spikes of 280+ms, but the general
latency looked low. Goodput might be great, or not -- but the
important bit remains "overall superior at present from a subjective
user point of view."

W
[0]: Yes, I did the Starlink preorder the second it was available. No,
sadly I don't have it yet...
[1]: Usually poor. Especially when anyone else is using it...

>
> On Mon, Jun 28, 2021, 10:45 AM Matt Hoppes <mattlists at rivervalleyinternet.net> wrote:
>>
>> I don't know how you can be embarrassed when you have a pretty solid
>> 30ms ping constantly, and Starlink has jitter all over the place and
>> spikes as high as 280ms.
>>
>> I'll take the DOCSIS3 system....
>>
>> On 6/25/21 8:49 PM, Eric Kuhnke wrote:
>> > I thought I would post an interesting comparison between a degraded
>> > DOCSIS3 link, of a carrier that shall remain nameless to avoid
>> > embarrassing anybody, and a starlink CPE with a slight 1/12th tree
>> > obstruction in a portion of its view.
>> >
>> > First two screenshots are the docsis3, to its gateway and to a very
>> > reliable hosted asterisk system in Seattle.
>> >
>> > Second two screenshots are starlink, also to its gateway and to the same
>> > destination.
>> >
>> > https://imgur.com/a/OQ5wyDr
>> >

-- 
The computing scientist’s main challenge is not to get confused by the
complexities of his own making.
  -- E. W. Dijkstra