Re: Terminology: fragment identifier part of an URI?

If you continue reading RFC 3986 to section 4.1, you will find

URI-reference = URI / relative-ref

We did not want to exclude "relative-ref" from the kinds of fragment
URI references that we wanted to refer to. Thus, a "URI" is actually a
subset of "URI reference".

Regards,
Silvia.


On Thu, Apr 15, 2010 at 5:03 AM, Sami Vaarala <sami.vaarala@codebay.fi> wrote:
> Hi,
>
> A minor terminology nit:
>
> The current draft (http://www.w3.org/TR/2010/WD-media-frags-20100413/)
> states in Section 2.1 that:
>
>    According to RFC 3986, URIs that contain a fragment are actually not
>    URIs, but URI references relative to the namespace of another URI.
>
> Based on my reading RFC 3986, this seems incorrect.
>
> RFC 3986, Section 3:
>
>    URI           = scheme ":" hier-part [ "?" query ] [ "#" fragment ]
>
> Section 3 also gives an "example URI":
>
>    foo://example.com:8042/over/there?name=ferret#nose
>
> The "URI" production refers to the generix syntax for a URI.  The
> "absolute-URI" production does not include a fragment identifier, but
> is not intended to be the only URI format.  RFC 3986, Section 4.3:
>
>   Some protocol elements allow only the absolute form of a URI without
>   a fragment identifier.  For example, defining a base URI for later
>   use by relative references calls for an absolute-URI syntax rule that
>   does not allow a fragment.
>
>      absolute-URI  = scheme ":" hier-part [ "?" query ]
>
> This also implies that absolute-URIs are only a subset of all URIs.
>
> Best regards,
>
> -Sami
> --
> Sami Vaarala
>
>
>
>

Received on Thursday, 15 April 2010 01:20:15 UTC