Re: NEW ISSUE: status of multipart/byteranges

On mån, 2007-11-19 at 05:47 +0100, Bjoern Hoehrmann wrote:

> I can write some client code that can parse such messages, but I would
> pass the request through some HTTP library which may then well consult
> a proxy and so on, and it's quite likely that somewhere on the line we
> have an implementation that does not determine the message length from
> the self-delimiters or does so incorrectly and perhaps even forwards a
> possibly problematic response.

Forwarding of 206 responses isn't such big problem if the client asked
for it. If there is an intermediary/client component not understanding
the response format it will result in the response hanging on completion
until the server times out the connection. In worst case the client
won't understand how to deal with what it requested, which will be
notices quite quickly. There is some exceptions if pipelining is
involved, but those too is not very critical.

The main problem is if this media type is used in other contexts than
206 responses, most notably for requests where it quite likely gets
wrongly parsed by most implementations.

I would suggest the specifications is narrowed down to only allow this
method of message delimiting for responses, not requests.

Regards
Henrik