Re: [apps-discuss] Feedback about "Update to MIME regarding Charset Parameter Handling in Textual Media Types"

[Julian Reschke <[e-mail]>]

On 2012-02-22 14:25, Henri Sivonen wrote:
> In reference to
> https://svn.tools.ietf.org/svn/wg/appsawg/draft-ietf-appsawg-mime-default-charset/latest/draft-ietf-appsawg-mime-default-charset.html

Henri,

thanks a lot for the feedback.

I believe that one thing that makes the current spec a bit confusing is 
that part of it sound like we're changing the defaults for existing 
types that specify their own defaults. We are not; *but* we are changing 
the framework so that new types can be defined in a saner way, and also 
so that (maybe) the definitions for text/xml and text/html can be tuned 
in the future. (Not sure about text/plain)

So, maybe add this as a clause to the abstract; and also clarify the 
prose where we mention text/xml and text/html?

Also, I see that you would like us to sanction whatever algorithm the 
HTML5 spec has come up with for handling text/html; potentially 
including content inspection and out-of-band information.

I think out-of-band information is a non-starter; it just doesn't work 
for many environments, and breaks basic assumptions about message 
semantics; if you need to do that for text/html so be it, and label it a 
"willful violation", but by all means don't let this hack leak into new 
media type definitions.

Another change I want to do is to add to:

"New subtypes of the "text" media type, thus, SHOULD NOT define a 
default "charset" value. If there is a strong reason to do so despite 
this advice, they SHOULD use the "UTF-8" [RFC3629] charset as the default."

a sentence that if it's not UTF-8, it MUST be an ASCII-compatible encoding.

Feedback appreciated,

Julian