- From: Roy T. Fielding <fielding@liege.ICS.UCI.EDU>
- Date: Sat, 01 Jun 1996 02:52:01 -0700
- To: Paul Leach <paulle@microsoft.com>
- Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
> Section 14.36 says that support for Range is optional. For clients I
> agree, but I think that for servers there's a problem with Range and
> PUT. If a client sends a PUT with a Range:, and the server doesn't
> support Range:, then the whole entity will be clobbered.
How so?
> It would be better if 1.1 servers had to at least recognize Range and
> produce an error message on methods other than GET if they don't support
> the Range retrieval semantics. (This includes unrecognized range-units
> -- see my previous message).
>
> Proposal:
>
> Change
>
> (However, not all clients and servers need to support byte-range
> operations.)
>
> to
>
> Clients MAY support byte-range operations. Servers MAY ignore Range
> headers on GET requests, and on all other methods MUST either support
> byte-range operations or return a 501 (Not Implemented) error if the
> request contains a Range header or a range-unit that the server does not
> recognize.
I think you are missing something. Range is a request header that
applies to the entity returned as the result of the request. I could
understand sending Content-Range in a PUT request (which would have
the effect of putting just a range of an entity), though I wouldn't
suggest it for 1.1.
...Roy T. Fielding
Department of Information & Computer Science (fielding@ics.uci.edu)
University of California, Irvine, CA 92717-3425 fax:+1(714)824-4056
http://www.ics.uci.edu/~fielding/
Received on Saturday, 1 June 1996 03:07:01 UTC