Re: [tsvwg] L4S vs SCE

[Wesley Eddy <[e-mail]>]

On 11/19/2019 7:12 AM, Ingemar Johansson S wrote:
> Hi Chairs + others
>
> I am not convinced that all these L4S vs SCE mail threads will make us any wiser.
>   
> The first question is around how safe L4S is to deploy on the internet and the proof points for either option differ widely.
>
> The other question is if we should have another ECN experiment (SCE) at this stage, also here I see that we have vastly differing opinions.
>
> In both discussion points I see that the discussion is going in circles.
> My personal conclusion is that on a technical level it will be real hard to come to an agreement.
>
> On a more non-technical matter, I have argued (rightly or wrongly) about the connection to work in 3GPP, admittedly so far only proposed work. Others have pointed out work related to WiFi and DOCSIS, how much should this weight in here ?
>
> Are the chairs any wiser ? :-)


Hi Ingemar, just speaking for myself and not any other co-chairs ... 
here are a couple of thoughts:

Yes, I'm getting a few positive things out of the threads.  For 
instance, there are some points it seems like everyone might agree on, 
including:

  * work on testing and publishing results has been very helpful
  * importance of RTT fairness
  * impact of the TCP Prague (or other end-host/transport) behaviors on
    the dual-queue performance
  * maybe some other things

I think where there are big issues, confusions, and disagreements 
causing back-and-forth, seems to be around:

  * entangled evaluation of L4S with TCP Prague (I see this badly in the
    threads ... From what I can tell, the public test Prague code
    doesn't yet meet all the Prague requirements in the L4S draft, and
    that's caused some explosion of emails.)
  * maybe lack of understanding about how and where people are hoping to
    use/deploy L4S?  (like as an operator, what else would be there that
    mitigates some of the concerns people have, and that supports
    gathering experimental results, etc)
  * concern that there can only be one of L4S or SCE happening
  * maybe some other things

It probably would be good if we all focus messages on making important 
new points and observations, not picking apart each sentence in replies 
or re-iterating points several times.  That would be in the spirit of 
working together towards the best outcomes, in contrast to just holding 
a public debate.  There are 700+ subscribers on TSVWG, and posters 
should continue to keep this in mind.