Re: [EToSat] New Version Notification for draft-kuhn-quic-4-sat-04.txt

Yes. By the time traffic leaves the aggregation level, I'd expect most 
packets to travel over a well-provisioned core. Sadly, not true for all.

However, "Yourself" is an interesting term: If one tests an aggressive 
app, I'd have thought the pain was more likely caused to others in your 
house/company/cafe or whatever that share the service. Measuring a 
method to see whether it performs well, is unlikely to show how 
disruptive the tested application is to other possibly quite different, 
traffic.

Gorry

On 24/04/2020 11:58, Lars Eggert wrote:
> Hi,
>
> On 2020-4-24, at 12:02, Jeremy Harris <jgh@wizmail.org> wrote:
>> Those distinguished colleagues have forgotten, or weren't around,
>> for the congestion-collapse phase of the Internet (around 1990?)
>> which drove the development of congestion control.
>>
>> We only have robustness because of such consideration.  Selfishness
>> is not ok.
> the Internet is different in many places now. Selfishness is still not OK, but usually your are (only) hurting yourself vs. impacting others. Back then, the core was under-provisioned compared to the access networks, these days it's usually the opposite.
>
> Lars
>
>
> _______________________________________________
> EToSat mailing list
> EToSat@ietf.org
> https://www.ietf.org/mailman/listinfo/etosat