W3C home > Mailing lists > Public > ietf-http-wg@w3.org > January to March 2014

Re: draft-montenegro-httpbis-uri-encoding

From: Zhong Yu <zhong.j.yu@gmail.com>
Date: Fri, 21 Mar 2014 10:36:36 -0500
Message-ID: <CACuKZqFjHXxzmO8onrggPDn7V18DRsMap2USsxPFA8KHDGYjig@mail.gmail.com>
To: Gabriel Montenegro <Gabriel.Montenegro@microsoft.com>
Cc: HTTP Working Group <ietf-http-wg@w3.org>
I don't think you need two headers; "URI-Path-Encoding" seems useless:

1. It's improbable that the origin server uses separate encoding
schemes for path and query. If the encoding scheme for the query part
is known, it can be assumed for the path part too.

2. If no html form is involved, and the browser simply sends a URI it
got from the server in a document, the browser has no idea of its
encoding scheme. It cannot be inferred from the document encoding - it
is not unlikely that a server app uses different encoding for document
generation and URI generation/parsing.

URI-Query-Encoding might be useful:

1. If a browser submits a GET form, it may declare the character
encoding used for the query part, in case that's useful to

2. If a browser submits a POST form with
enctype=application/x-www-form-urlencoded, it may declare the
character encoding used for the request entity body. But we'll need a
better header name than "URI-Query-Encoding". Note that we cannot add
a "charset=" parameter to the Content-Type header for this purpose -
servers are not expecting it; they do not need it; they may be broken
because of it.

3. In any case, the server will NOT read nor honor this header. The
server is the sole authority of how to parse the request.

Zhong Yu

On Thu, Mar 20, 2014 at 11:36 PM, Mark Nottingham <mnot@mnot.net> wrote:
> One of the things we didn't get time to talk about in London was Gabriel's spec:
>   http://tools.ietf.org/html/draft-montenegro-httpbis-uri-encoding-00
> In a nutshell, this offers a way for a client to declare what character encoding was used prior to percent-encoding.
> I've heard hallway feedback about it that wonders if we just want to allow one value ("UTF-8"). Beyond that, folks seem generally neutral-to-positive about it, AFAICT.
> What do people think about adopting this as a WG item, keeping in mind that we can change it in process if there's some particular aspect you don't like?
> Cheers,
> P.S. Just to be clear, this would be completely separate from the HTTP/2 work item.
> --
> Mark Nottingham   http://www.mnot.net/
Received on Friday, 21 March 2014 15:37:04 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:14:25 UTC