Re: rel links in HTTP and HTML

On Wed, Nov 23, 2011 at 16:34, Timothy Lebo <lebot@rpi.edu> wrote:


> 1) If the rel links are returned in HTTP, should they also appear in the
> HTML?
> 2) Should the rel links provided be equivalent in the HTML and HTTP?

Headers specified in the HTML <head> usually overrides what is
specified in the HTTP header. So if I say
<META http-equiv="Expires" content="Tue, 20 Aug 1996 14:25:27 GMT"> it
is interpreted as if an additional header line was inserted in the
HTTP response:

Expires: Tue, 20 Aug 1996 14:25:27 GMT

Now Link: is different because you can have multiple instances of
Link. However I am not sure if you would be allowed multiple links
with the same rel - http://tools.ietf.org/html/rfc5988 says:

   This specification does not place restrictions on the cardinality of
   links; there can be multiple links to and from a particular IRI, and
   multiple links of different types between two given IRIs.  Likewise,
   the relative ordering of links in any particular serialisation, or
   between serialisations (e.g., the Link header and in-content links)
   is not specified or significant in this specification; applications
   that wish to consider ordering significant can do so.

So from this you could have as many Link rel="provenance" as you like.
That means that the HTTP links and HTML links can be equal or
different, or just appear in one of them.


I guess it depends on the application and server environment what
makes most sense. For instance if I have not got much control of the
HTTP server I would use the HTML-method. If I have loads of resources
that are non-HTML or that I don't want to edit, then I should use the
HTTP-method. If I want to support users to save the HTML and still see
the provenance (the page is more like a document than a web page, or
might be mirrored on multiple locations) then I should use both.

It might also be that they provide complementary views of the
provenance, for instance the server might be a content management
system (CMS) that in the HTTP provide link to metadata provenance
about when the document was uploaded, etc, while the author has added
HTML links to provenance he has gathered with his local authoring
tool.


Now a good question is - if two such header appear, and they specify a
different anchor - ie. they talk about two different entities - is
there an implied prov:wasComplementOf() relation between those two
URIs? Are there two implied prov:viewOf relations (if this relation
existed) from the entity URIs to the "actual URI" where the document
was downloaded from?

-- 
Stian Soiland-Reyes, myGrid team
School of Computer Science
The University of Manchester

Received on Thursday, 24 November 2011 09:27:40 UTC