RE: PATCH vs multipart/byteranges vs Content-Range

Subbu Allamaraju wrote:
>Brian Smith wrote:
>> because (a) text/plain isn't a MIME type for
>> a patch format, and (b) it
>
> You mean, text/plain is invalid for bytes? Why
> would text/plan be invalid for a patch format?

text/plain isn't necessarily invalid, but in order to use that MIME
type, the server would have to parse the request entity just to
determine the format of the patch: is it a Darcs format patch? diff -u
format? Git? something else?
	
>> By the way, I really like the idea of a patch format based on
>> multipart/byte-ranges; I even think that such a format should be a
>> SHOULD requirement for servers that implement PATCH.
		
> Really? As you pointed out above, Content-Range is not specified
> for requests. Secondly, IMHO, patch format does not need to be
> constrained to byte-ranges.

By "patch format based on multipart/byteranges", I didn't mean to use
multipart/byteranges as-is. As Roy pointed out, any kind of standardized
patch format--one that is recommended for servers to support--should
have a syntax for updating headers along with the actual entity. And,
any such general-purpose patch format would have to have a way of
representing insertions and deletions, which multipart/byteranges cannot
do as it is defined today.

- Brian

Received on Saturday, 16 February 2008 21:18:20 UTC