[Bloat] the "ietf" accord bof

[Bloat] the "ietf" accord bof

[Dave Taht]

I rather enjoyed the "accord" bof, which was trying to find diffserv
models to fit 3gpp (5g) behavior.

The analogy jari made about fq_codel being the "zero bit" alternative
to diffserv was quite good, but I missed his actual presentation. The
session had a really good, basic, introduction to 3ggp's design,
also... Then the ecn-ish debate started... The session was closed by
ted hardie's refresher of how rfc791's original use case, but the
slides don't do it justice.

You can find the slides at:

https://datatracker.ietf.org/meeting/95/materials.html/


--
Dave Täht
Let's go make home routers and wifi faster! With better software!
https://www.gofundme.com/savewifi

Re: [Bloat] the "ietf" accord bof

[Jonathan Morton]

> On 8 Apr, 2016, at 05:51, Dave Taht <dave.taht@gmail.com> wrote:
> 
> The session was closed by
> ted hardie's refresher of how rfc791's original use case, but the
> slides don't do it justice.

It’s perhaps worth remembering that RFC791 defined the TOS/Precedence bits as end-to-end stable, a property explicitly removed by Diffserv.  Of all of Diffserv’s failings, I think this has become the most troublesome.

 - Jonathan Morton