- From: Mike Meyer <mwm@contessa.phone.net>
- Date: Fri, 2 Feb 1996 11:09:05 PST
- To: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
> I'm not clear here on when you say "...Content-Type...be used to specify > length". What if a CGI script *can't* know the length, such as in a "server > push" situation (which uses multipart types and just keeps sending part > after part until the user closes the connection). Unless I'm really confused, the idea is to provide a more reliable mechanism for a client to find the end of an object than provided by the server closing the connection. Content-Length: does this directly. However, there are situations where the finding the content length in advance is expensive, if not impossible. The other two headers fields deal with that case, by providing an "end-of-object" indicator for the different situations. For example, if you have a multipart objects, then the size of each part is determined seperately and presumably reliably. You reach the end of the object when you get the encapsulation boundary following the last part. That there is no "last part" in a server push situation doesn't change this. In the case of dynamic objects which have an unknown initial length, you use the transfer-encoding of "chunk", which allows you to send an arbitrary number of non-zero length chunks, and you reach the end when you get a zero-length chunk. > As to Larry's suggestion of "strongly recommended", I have a > not-well-thought-out idea: allow "Content-Length: -1" to mean "I definitely > don't (or can't) know the content length and I might use disconnect > instead". This lets 1.1-aware clients know more than just "the server > thinks it knows the length" and to be more prepared for disconnects. While you're at it, make Content-Length default to "-1" so we can just leave it out. <mike
Received on Friday, 2 February 1996 11:23:17 UTC