[
Lists Home |
Date Index |
Thread Index
]
Hello Julian,
Didier said:
> > a) The resource is an abstract entity representing something on
> > the web. Its
> > a metaphysical monad, an abstarct thing associated to a concrete
> > thing: the
> > representation.
> > b) A resource may be identified by a URI. More particularly its location
> > represented by a URL and its name by a URN.
> > c) The outcome of an HTTP GET operation on a particular resource
> > identified
> > by a particular URI is its representation.
> >
> > However, a resource has not:
> > a) several representations
Julian said:
> How do come to that conclusion? The representation returned upon GET
depends
> on a variety of factors, including HTTP request headers, time of day and
so
> on... So clearly a resource can have many representations.
Didier replies:
Yes you are right so let's modify the last item:
a) a resouce may have several representation. For a given URI, the requester
may receive different representation of a particular resource even if foir
different transactions the requester uses the same HTTP request (same URI
same headers). The requester may receive different representations based on
criteria the server uses to package the reply.
Julian said:
>
> > b) several URIs.
>
> Depends on how you define identity of URIs. For particular URI schemes,
two
> URIs may identify the same resource even if they aren't lexically
identical.
Didier replies:
Let's bring more precision to this one too. Within a particular URI scheme,
A particular resource may not have more than one URI.
Thanks for your comments Julian, I forgot about the accepted types that
could be declared in an HTTP header (accept, accept-language,
accept-encoding etc...)
cheers
Didier PH Martin
|
