On 2011-09-28 15:40, Barry Leiba wrote:
I'd like to see more participation in this thread, besides just from
Mike and James. What do others think?
Barry, as chair
...
I agree that if a protocol element needs to support non-ASCII
characters, the spec should be *crystal clear* how it's going to be
transported over the wire.
Note that before James' mail I wasn't even beware that the same the
problem I noticed for error_description also affects much more relevant
protocol elements.
With my HTTPbis-editor hat on:
- use a consistent way of encoding in all parts of the header fields,
- and try to avoid raw UTF-8. There are alternatives, such as applying
another round of percent-encoding, or using RFC 5987 encoding in the
first place.
Finally, what's best in other wire formats may not be the best in header
fields; insisting to use the same encoding will cause pain.
Best regards, Julian
_______________________________________________
OAuth mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/oauth
