IETF standards

["Roy T. Fielding" <fi...@kiwi.ics.uci.edu>]

The IETF standards process is a bit weird in that the revisions
to a proposed standard are listed as Internet Drafts even when
they are known to be better than the original RFC.  During the
revision process, it is necessary for vendors to implement according
to the new draft in order for the new draft to become a Draft Standard
(the step above Proposed Standard, which is where RFC 1738 and RFC 1808
currently rest in peace).  We cannot wait until the revisions become
an RFC.  In other words, we are better off leading the actual standard
rather than following it.

Both of the comments in PR 2553 are correct: the URI supplied by the
user is invalid, but Apache should still be able to handle it.  This
falls under the "be robust in what you accept" principle and is
explicitly allowed by the URI syntax revision, even though generators
are required to make URI all-ASCII for greater interoperability.

.....Roy

Re: IETF standards

[Rodent of Unusual Size <Ke...@Golux.Com>]

Roy T. Fielding wrote:
> 
>                             We cannot wait until the revisions become
> an RFC.  In other words, we are better off leading the actual standard
> rather than following it.

Right ho.  What's that draft name again?

> Both of the comments in PR 2553 are correct: the URI supplied by the
> user is invalid, but Apache should still be able to handle it.  This
> falls under the "be robust in what you accept" principle and is
> explicitly allowed by the URI syntax revision, even though generators
> are required to make URI all-ASCII for greater interoperability.

Thanks for the clarification, Roy.

#ken	P-)}

Ken Coar                    <http://Web.Golux.Com/coar/>
Apache Group member         <http://www.apache.org/>
"Apache Server for Dummies" <http://Web.Golux.Com/coar/ASFD/>