Re: Fragmentation for headers: why jumbo != continuation.

In message <CAP+FsNeuFNErbPzgNtyss3S53FV_Ha2oYRO4D5dkcz7-+aA+Fw@mail.gmail.com>, Roberto Peon wri
tes:

>There are two separate reasons to fragment headers
>[...]
>2) Reducing buffer consumption and latency.

Roberto,

I keep begging you to give me an example of what real-world
applications benefit from being able to fragment headers, and you
keep ignoring my request.

Given that the high-traffic chokepoints for HTTP traffic will benefit
a LOT from knowing the length of the compressed header-set up front,
you need to provide at least an equally compelling reason why being
able to fragment headers brings benefits elsewhere in the HTTP grid.

Where is the beef Roberto ?

Why won't you tell us ?

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.

Received on Thursday, 10 July 2014 23:20:33 UTC