Re: [hybi] I-D Action: draft-ietf-hybi-permessage-compression-28.txt

[Matt Bonneau <[e-mail]>]

Hi Barry and Tobias,

I would be willing to help out with the RFC process on WAMP.

I have made pretty significant investments of time in creating
implementations and working with the spec and also use WAMP regularly. I
would welcome further development through the RFC process.

Regards,
Matt Bonneau

On Mon, Sep 14, 2015 at 5:23 PM, Tobias Oberstein <
[e-mail]> wrote:

> How "much" interest is needed? How should we "proof"? Should I ask all to
>>> speak up on this mailing list?
>>>
>>
>> It's hard to put a number on it, but there needs to be enough interest
>> and commitment to participate that we can rely on progressing
>> documents and getting meaningful review.  We have been having a lot of
>> trouble with that over the past couple of *years*, so it's not just a
>> moderate slowdown.
>>
>
> Thanks for explaining and background. I now understand that real world
> usage and takeup is only one factor, but enough interest to do the concrete
> work on a spec document is what counts. Makes sense. To get a high quality
> spec, there needs to be peer review and discussion. It's part of the IETF
> mantra I guess, and I totally buy that.
>
> That said, if we really *do* have enough commitment, I'm willing to
>> recharter the working group.  I can't speak for the chairs about
>>
>
> Awesome! Would you give us a bit time to organize enough commitment and
> write a draft? Some weeks before closing the WG (if we can't deliver)?
>
> whether they're willing to continue, but perhaps we could get new
>> chairs to take it forward -- Salvatore and Gabriel have chaired this
>> working group for a long time now.
>>
>> So do we need a draft RFC written _before_ we can request to recharter the
>>> WG?
>>>
>>
>> We don't *need* it, though a draft and some reviews of it would be
>> some evidence that there's enough interest.
>>
>
> Ok, we can prepare a draft according to RFC style from this spec document
> ("WAMP basic profile") that we have:
>
> https://github.com/tavendo/WAMP/blob/master/spec/basic.md
>
> That spec has been extensively discussed and has gone through many
> iterations from Mar 2012 to now.
>
> Discussion is on GitHub (94 closed, 67 open issues right now) and the WAMP
> mailing list (171 threads, 151 subscribers).
>
> The point is that implementors are less interested in implementing
>> websocket things, now that http/2 is here.  Layering things on
>> websockets might not be very useful if implementors are moving to
>> http/2 instead, for such functions in the future.
>>
>
> FWIW, WAMP originally could only run on top of WebSocket. And it's still
> the most used "transport". But WAMP nowerdays can _also_ run over RawSocket
> (essentially, minimally framed TCP with some handshake sugar) and
> HTTP-Longpoll, and don't see why we couldn't specify a HTTP/2 "transport
> binding" (that's WAMP lingo for defining a mapping of WAMP to an underlying
> transport).
>
> WAMP-over-WebSocket is definitely the most important one, as it is the
> only thing that (today) can run right into all browsers. It is also the
> only transport that provides full, transparent payload compresssion -
> thanks to the WebSocket compression extension.
>
> But, again, if there really is enough interest in implementing new
>> function on websockets, and we can see that interest in commitment to
>> work on the specs, then that might be fine.
>>
>
> Thanks! Please give us some time before closing the WG. Once the WG would
> have closed, I guess the bar would be much higher .. and where should we
> take it then? I have skimmed through the different IETF art areas, and the
> WGs and charters for the "Applications and Real-Time Area" in particular,
> and it appears to me, that WAMP would fit the Hybi-WG's mission ..
>
> Cheers,
> /Tobias
>
>
> _______________________________________________
> hybi mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/hybi
>