Re: [apps-discuss] Feedback on multipart/form-data

[Dave Cridland <[e-mail]>]

Julian Reschke wrote:
> On 2013-09-19 15:29, Dave Cridland wrote:
>>     Julian Reschke wrote:
>>>>
>>>>     I know of RFC 2047 and RFC 2231, as well as unencoded UTF-8 and
>>>>     unencoded ISO-8859-1 (I think).
>>>
>>>     No surprise here
>>
>>     Yes, but one wonders if the unencoded UTF-8 actually does any harm at
>>     all in a 8bit/binary-safe protocol like HTTP.
>>
>>     If most processors accept this - and I have no idea - it might even be
>>     useful to document.
>
> The risk here is that some UAs send ISO-8859-1 (or something based on the 
> locale), and servers parse based on the User-Agent. We have the same problem 
> with the Content-Disposition response header field.
>
> It's hard to change these kinds of workarounds in deployed code, and that's 
> why a new mechanism that can't be confused with the old syntax may work 
> better (thus RFC 5987).

Yes, agreed, though UTF-8 is very difficult to accidentally mimic, from what 
I've found.

I wonder if an EAI expert might suggest something here? Or if a top-level 
parameter or header might be useful to indicate UTF-8 header content? I can't 
help but feel it's a simpler construct to adhere to.


Sent with [inky: <http://inky.com?kme=signature>]