Re: Giving the Framing Layer a real name

[Mark Nottingham <[e-mail]>]

On 28/02/2013, at 9:06 AM, Martin Thomson <[e-mail]> wrote:

> On 27 February 2013 13:59, Mark Nottingham <[e-mail]> wrote:
>> To me, a distinct name of a major, separable part of a protocol is only good spec hygiene; in HTTP we already have "representations" (nee "entities"), "resources" and so forth. They allow people to talk about different things that are happening with clarity; right now, people are using the term "HTTP/2.0" very, very sloppily, and that's a concern. "It means what I mean" is not a great basis for communication.
> 
> I don't agree that it's separable in any real sense, which has been
> the source of the contention.  However, I appreciate the value of a
> handle by which we can identify important "things".  I always imagined
> that in the few places it was necessary to refer to the concept,
> "HTTP/2.0 framing" or "HTTP/2.0 framing layer" would suffice.
> 
> That depends on the scope of what you are referring to, which is - I
> believe - an important part of what we need to clarify.  Does this
> include the creation and use of streams as well as the use of frames
> to convey data?


Yes, good point. We're already seeing the effect of this blurriness in terms of what people consider the "framing layer."

To me, EVERYTHING about mapping HTTP-specific semantics is in the "HTTP" section; everything that's generic -- which includes streams and stream management -- is "below" on the framing "layer."

I'm OK if we choose to use "HTTP/2.0 Framing" and stick with it -- it's just that it's used inconsistently now. However, this may not be the best name, because it has "HTTP/2.0" in it, and then we go and talk about using HTTP/2.0 *on* it. OTOH I'm not thrilled about introducing Yet Another Acronym...

Cheers,


--
Mark Nottingham   http://www.mnot.net/