- From: Eduardo Casais <casays@yahoo.com>
- Date: Tue, 22 Sep 2009 06:13:57 -0700 (PDT)
- To: public-bpwg@w3.org
> FWIW I am not sure I can remember where the X-
> notation comes from
RFC822
RFC2076
> nor can I now find any reference saying that extension
> header fields must or should be registered.
They actually should not. So far, registering X-prefixed
fields has been exceptional. The reason: X- is originally
supposed to mean eXperimental, not eXtension.
For the purpose of the CTG, I assume that the intent
behind restricting extra X-fields is this: for the purpose of
dealing with transformation proxies conformant with the
guidelines, neither clients nor servers (nor possibly other
proxies) should have to be aware of and deal with other
non-standard fields introduced by a proxy, nor compensate
for or take into account their possible impact on other
parts of the HTTP header.
E.Casais
Received on Tuesday, 22 September 2009 13:14:42 UTC