Re: [tsvwg] New Version Notification for draft-morton-taht-tsvwg-sce-00.txt

["Holland, Jake" <[e-mail]>]

Hi Greg,

Thanks for clarifying, that's very interesting.

I applaud the release of a reference implementation of L4S suitable
for independent evaluation and real-world testing, so thanks and
nice work!

However, it's surprising to me to hear that support for a system
that was presented to the IETF as a proposed experiment is
mandatory--I would have thought support would only be applicable
to nodes explicitly participating in experiments?

If support for this is going to be very widely deployed, is the
scope of the experiment documented anywhere?  (Or has any thought
been given to anything like a pre-emptive sunset date, after which
the use of L4S and the L4S semantics for ECT(1) should be stopped
unless a later document extends the experiment or standardizes it?)

I couldn't find anything in those docs, so I thought I'd ask for
a pointer.

If there's no such thing, can I ask for more information about
other steps you've taken to address the responsibilities outlined
in Section 8 of RFC 8311 with regard to this experiment?
https://tools.ietf.org/html/rfc8311#section-8

(I've seen a lot of the lab work presented, with many thanks to
Bob, Koen, and others, and it looks promising. But I mean to ask
about mitigations that address potential issues that are so far
undiscovered that may crop up in deployment or in externally run
experiments?)

Thanks and regards,
Jake

On 2019-03-11, 19:38, "Greg White" <[e-mail]> wrote:

    I stand corrected.  They are experimental. Nonetheless, they are not proprietary specs for the cable industry. 
    
    The DOCSIS implementations of L4S are entirely in firmware and can be enabled/disabled by the operator via configuration settings.
    
    -Greg
    
    On 3/11/19, 6:08 PM, "Holland, Jake" <[e-mail]> wrote:
    
        On 2019-03-11, 16:50, "Greg White" <[e-mail]> wrote:
        > It is a standards-track draft here at IETF
        
        I thought the L4S drafts were experimental, not standards-track:
        
        https://tools.ietf.org/html/draft-ietf-tsvwg-aqm-dualq-coupled-08
        https://tools.ietf.org/html/draft-ietf-tsvwg-ecn-l4s-id-04
        
        Was there one that's standards track?
        
        Thanks and regards,
        Jake
        
            
            -Greg
            
            
            On 3/11/19, 5:33 PM, "Dave Taht" <[e-mail]> wrote:
            
                
                
                Greg White <[e-mail]> writes:
                
                > On the item below, I am intending to finish up and submit a TSVWG
                > informational draft today that describes the (new) mandatory support
                > for L4S AQM and dual-queue requirement in DOCSIS 3.1 (CM & CMTS)
                > equipment. The draft will essentially be a reformat of this document:
                >
                > https://cablela.bs/low-latency-docsis-technology-overview-february-2019
                >
                > These requirements were added in December, and will roll out as
                > firmware updates to existing DOCSIS 3.1 equipment (nearly 100% of
                > CMTSs, and a growing fraction of CMs) over time.
                
                I was unaware of this notice, and as I note, it's only march...
                
                Using up the last ECN bit in this way, for the sole benefit of the cable
                industry, strikes me as a poor standard for the internet.
                
                >
                > -Greg
                >
                > From: tsvwg <[e-mail]> on behalf of Bob Briscoe
                > <[e-mail]>
                > Date: Monday, March 11, 2019 at 10:29 AM
                > To: Dave Taht <[e-mail]>, "[e-mail]" <[e-mail]>
                > Cc: Jonathan Morton <[e-mail]>
                > Subject: Re: [tsvwg] New Version Notification for
                > draft-morton-taht-tsvwg-sce-00.txt
                >
                > L4S is far from dead. It's merely been working differently from how
                > you're used to. Those working on an L4S AQM (at least those in the
                > cable industry) had to have a private WG for the last ~18 months, but
                > now we're allowed to publish and talk openly again.