Re: END_SEGMENT?

True, but sometimes we've run into hardware that just can't cope with
certain chunking, and nonetheless need to support it.
We ran into more of this quite recently as a matter of fact :(

More interesting to me, in any case, is that END_SEGMENT allows for
effective layering.
-=R


On Thu, Apr 17, 2014 at 10:35 PM, Roy T. Fielding <fielding@gbiv.com> wrote:

> On Apr 17, 2014, at 9:25 AM, Jeff Pinner wrote:
> > Consider this use case in HTTP/1.1. An API provides a stream of data
> using HTTP/1.1 chunked encoding. Each chunk contains a single record, and
> the length of this record is indicated by the chunk length. In HTTP/1.1 the
> length of these chunks was unconstrained, but when translating into HTTP/2,
> these chunks must be segmented into multiple data frames to fit within the
> frame size limit.
> >
> > If the chunk delineation was meaningful, then END_SEGMENT allows this
> meaning to be preserved.
>
> FWIW, that won't work in HTTP/1.1 (intermediaries and network libraries
> will consume and coalesce chunks before any application gets to see them).
>
> ....Roy
>
>
>