Re: Rechartering HTTPbis

Hi Henrik,

On Tue, Jan 24, 2012 at 09:05:02PM +0100, Henrik Nordström wrote:
> tis 2012-01-24 klockan 09:48 +0100 skrev Willy Tarreau:
> 
> > Maybe HTTP/2.0 could be designed to take advantage of SCTP.
> 
> Heck, even HTTP/1.1 as-is can take advantage of SCTP, and with very
> little transport change/mapping it can utilize it very well for framing
> as well getting rid of many transport & framing issues of HTTP/1.1.

Possibly, I don't know in fact.

> > HTTP/1.1 has a number of issues that make the current spec very
> > heavy and implementations complex (eg: remember you can't fold
> > set-cookie, the issues with multiple content-length, etc...).
> > Taking the opportunity of a new version to clear a few of these
> > old issues would be nice.
> 
> But SPDY do not.. well sure it addresses the folding issue (there is no
> folding in SPDY), but multiple content-length is still an issue. Less so
> than in HTTP/1.1 as it's not transport related any more, but still.

Which is why I'm talking about HTTP/2.0 instead of SPDY. SPDY is something
which exists, HTTP/2.0 is something we must define. Maybe in the end both
will be the same but I'm making a strong difference between "define" and
"use" :-)

[BTW I don't like the name "HTTP/2.0", it reminds me of the marketing term
 "Web 2.0", but that's just a detail]

Regards,
Willy

Received on Tuesday, 24 January 2012 20:31:56 UTC