[tsvwg] The implication of L4S increased RTT bias on T1-ISPs and transit

[Sebastian Moeller <[e-mail]>]

Dear list,

while responding to a related L4S thread, I realized, that we typically mostly argue about L4S merits or lack thereof from the perspective of end-users/leaf nodes and ISPs connecting these to rest of the internet, but so far I do not think that the transit side of things has actually been asked for their opinion. So....

QUESTION: has anybody discussed the current design with its noticeable increase in RTT bias with T1-ISPs that mainly deal in (long-range) transit? I might be worhtwile getting their opinion on whether we should prioritize even more bandwidth for short range connections than TCP in FIFOs already does? 
IMHO this is an interesting question to answer, after all the internet is about the potential to connect to all end-points, so deciding whether we want to give near hosts an even greater boost and are happy with L4S weak guarantees in that regard (will not starve, with not starvation being defined as >= ~1/16 of capacity)?



Best Regards
	Sebastian