Re: draft-montenegro-httpbis-uri-encoding

On 2014-03-21 15:26, Poul-Henning Kamp wrote:
>
> I am of the opinion that HTTP/2's success will be determined by
> what it eliminates from the HTTP-mess, rather than what it adds to it.
>
> This topic/discussion is a very good example why I think so.
>
> We can focus om making HTTP/2 simpler to implement correctly, by
> choosing a simple and unambiguous rule, probably "URI's are UTF-8"
> or we can add another tortured heuristic resolution flow-chart, which
> attempts to guess what people though they understood when they read
> the unclear description of all the point-less options in the standard.
>
> What happened to the KISS principle ?

It died when we agreed that we want to be able to gateway between 1.* 
and 2.*, and not ever break existing semantics.

Essentially, HTTP/2 is only a new wire format, and it inherits all 
legacy not related to the wire format from 1.1.

Best regards, Julian

Received on Friday, 21 March 2014 14:37:03 UTC