It occurs to me that this might not actually be all or nothing. QUIC is being optimized for Web browsing. It has got to a point where the idea of Web Services over HTTP/2 over QUIC is really not something I am interested in.
But does it have to be the same stack for Web Services and for Web Browsing?
Web Services need almost no features of HTTP unless they are Web Services that basically reduce to content retrieval. There is the firewall bypass capability (i.e. port 80/443) there is message framing and there is the extension in the port number space we get from .well-known.
Running JSON over a stripped down QUIC over DTLS is much more interesting to me.
On Wed, Mar 7, 2018 at 12:58 PM, Eric Rescorla <ekr@xxxxxxxx> wrote:
Replying to myself....Dear All,Thanks very much for your comments. The IESG will be reviewing themclosely and reaching out to a few of you to follow up. We will comeback to the community once we've done that and had a chance to discussthe way forward.-EkrOn Thu, Mar 1, 2018 at 8:01 PM, Eric Rescorla <ekr@xxxxxxxx> wrote:Hi folks,The IESG has heard some concerns from participants that they would likenot to be photographed. In response to those concerns, we have developedthe attached policy which we intend to put in place going forward.Please send any comments by 3/8/2018.-Ekr-----The intent behind this policy is to balance people's legitimate desirenot to be photographed with the IETF's ability to document activitiesand enable remote participation. In order to enable that, we areproposing the following policy which applies to all IETF events,including WG meetings, plenaries, and the hackathon.LABELLINGThe IETF will make available a mechanism for participants to labelthemselves as desiring not to be photographed. The secretariat willdetermine the details in consultation with the IESG.OFFICIAL PHOTOGRAPHYAny photographer engaged on behalf of the IETF should not photograph individualsdisplaying the "do-not-photograph" label, should make reasonableefforts to avoid photographing small groups with one or more membersdisplaying the label, and should not publish small group photographswith such individuals in them. Photographs of large groups may containincidental images of such individuals and we will not attempt toredact those. Specifically, photographs of panels and the like (e.g.,the IESG/IAB plenary) are expected to contain all individualsregardless of labelling.Working group meetings are generally video recorded and broadcast, andno attempt will be made to avoid recording individuals. However, ifthe IETF publishes still frames of these videos, individualsdisplaying the labels should not be shown.Note: the use of "should" above is intended to reflect that althoughthis is IETF policy, it is a best effort service and some mistakeswill likely be made, perhaps because someone's label is not noticed orvisible. Individuals can contact XXX to arrange for redaction of theirimages, or YYY to report abuse.UNOFFICIAL PHOTOGRAPHYMany IETF participants also engage in photography. We ask that thoseparticipants respect the above policies and avoid photographingindividuals who have asked not to be photographed. Although werecognize that mistakes will be made, repeated intentional violationsof this policy may constitute harassment and could be brought to the attentionof the ombudsteam, per RFC 7776.
