- From: Larry Masinter <masinter@parc.xerox.com>
- Date: Wed, 18 Aug 1999 16:28:05 PDT
- To: kugler@us.ibm.com, "Yaron Goland (Exchange)" <yarong@exchange.microsoft.com>
- Cc: "Roy T. Fielding" <fielding@kiwi.ics.uci.edu>, http-wg@hplb.hpl.hp.com
> My point is that responding to Transfer-Encoding chunked by sending 411 and > Connection close is HTTP/1.1 COMPLIANT behavior. By which I mean that it does > NOT violate HTTP/1.1. ... > My mileage does vary. I have encountered product groups which refuse to take > the development and testing hit to make a server capable accepting chunked > POSTs, when it can legally reject them with 411. We might make replying '411 and connection close' to a chunked encoding 'conditionally compliant' rather than 'compliant' because otherwise, as has been pointed out, there are inteorperability problems. This isn't an errata, but it belongs on the issues list. On the other hand, a server can legally also reply '200 OK' to every POST, and just ignore the data. A server can legally reply '404 Not Found' to every GET, too! So the fact that there are some stubborn implementors doesn't mean we should change the standard. There is still a 'call' for an HTTP implementation guide; it isn't in the charter of the HTTP working group, but this kind of advice might belong there.
Received on Wednesday, 18 August 1999 16:31:43 UTC