Eliot Lear wrote:
Thanks, Julian, for your response. Please see below:
The media type of the payload of the patch request is completely
orthogonal to the media type of the thing being patched. Yes, that
means that we need to define one media type per patch format; but
that's by design in HTTP, methinks.
Ok, I understand. And so you just re-register a new media-type for
the different PATCH method (assuming the type doesn't already
exist). Thus the new header. Okay. I withdraw my major concern.
Um, which new header? Accept-Patch?