Re: [nfsv4] Proposed charter changes (with updates)

[Magnus Westerlund <[e-mail]>]

On Thu, 2020-06-04 at 09:01 -0400, David Noveck wrote:
> > Indeed. But I'm not sure the QUIC WG will take up an upper layer-
> > specific matter, 
> 
> Possibly not.
> 
> > so I am expecting the same rule might apply. 
> 
> I dont'see how we can take a rule designed for the case in which the
> responsible working group doesn't exist and apply it to a case in which the
> responsible working group does not want to take on the work.   That's clearly
> a different situation.
> 
> In that case, the work charter proposal says we "will work with the
> responsible working group to make sure needed extensions are available".   In
> the case in which we want something that the Quic working goup doesn't want to
> do, we would do it, unless it is something that the QUIC working group doesn't
> think should be done.


So from a structural point of view when it comes to RDMA over QUIC I would like
to point to that such a mapping onto QUIC may actually be done outside of QUIC
WG. 

So the QUIC charter has the following paragraph: 

The third focus area describes the mapping between the HTTP application protocol
and the transport facilities of QUIC. This mapping will have performance
characteristics comparable with HTTP/2, and provide extension mechanisms similar
to HTTP/2. Upon completion of this mapping, work to define additional mappings
may be included by updates to this charter, or may be worked on by other working
groups.

Note the last sentence, that it is both an explicit recharter to bring in
additional mappings and that mappings may very well be done in the WG of the
protocol being mapped onto QUIC. However, if one think the mapping is going to
be particular thorny and possibly also require QUIC protocol extensions, i.e.
have requirements that other QUIC users may benefit from it could make sense to
do the work in QUIC. 

For that reason I would recommend that one work through the technical proposal
to better understand what desired propoerties of RDMA over QUIC has, as well as
what is required to realize those from QUIC. Then one can likely make a good
decision if that work should be done in QUIC WG or in this or another WG. 

> > That's why I suggested an AD needs to be in the loop, to review the
> assignment.
> 
> I thought it was intended to assure the IESG, that we are not going to do
> stuff without supervision.
> 
> In any case, as a practical matter, there is no way would do any of this
> without the responsible AD's approval/support.


So, this may calm the IESG. Still the wider review that recharters get from the
review call help avoiding surprises and unfortunate interactions between working
groups. 


Cheers

Magnus Westerlund 


----------------------------------------------------------------------
Networks, Ericsson Research
----------------------------------------------------------------------
Ericsson AB                 | Phone  [phone]
Torshamnsgatan 23           | Mobile [phone]
SE-164 80 Stockholm, Sweden | mailto: [e-mail]
----------------------------------------------------------------------