Re: parsing URI (references) according to RFC 3986

On 2011-06-23 18:03, Boris Zbarsky wrote:
> On 6/23/11 3:51 AM, Julian Reschke wrote:
>>> The handling of URIs like "#foo" (and in particular what's used as the
>>> base URI for them) has been an interop problem in the past, as I
>>> recall... Some browsers use the document URI, some browsers use the base
>>> URI, some do a mix depending on various other considerations... or
>>> something. For some browsers I haven't been able to figure out what
>>> they're doing.
>>
>> I agree. But this is an HTML-issue (deciding what the base URI is), not
>> a URI/IRI issue. Yes, it should be solved, but it should be done in the
>> HTML spec.
>
> You seem to misunderstand. Last I checked, the URI specs called for
> "#foo" and "foo" to be treated differently in terms of resolution with
> respect to a base URI.... Something about local vs non-local references
> or the like.
>
> HTML does decide what the base URI is, of course.

Yes, they are treated differently, one is a reference only consisting of 
a path, the other one only consisting of a fragment identifier.

Their resolution is IMHO clearly described in RFC 3986; could you be a 
bit more specific about what your concern is?

Best regards, Julian

Received on Thursday, 23 June 2011 16:16:05 UTC