On Sat, 28 Jun 2008, Phil Blundell wrote:

> Why so? RFC 2616 is fairly explicit:
>
> "A client MUST be prepared to accept one or more 1xx status responses prior
> to a regular response, even if the client does not expect a 100 (Continue)
> status message. Unexpected 1xx status responses MAY be ignored by a user
> agent."

I agree with Phil here that this is rather explicitly mentioned in the above
context. Note also that this isn't about "unknown" codes but rather
"unexpected" and for libcurl everything 101 - 199 is unexpected.

It's interesting to see that the HTTP WG still uses pretty much the same
phrasing in their latest draft for "httpbis":

(see section "9.1 Informational 1xx")

http://www.ietf.org/internet-drafts/draft-ietf-httpbis-p2-semantics-03.txt

-- 
  / daniel.haxx.se