Lists Home |
Date Index |
> Ramin Firoozye wrote:
> > Actually, even if you do a GET to the database, if it goes through an
> > intermediary function (like a servlet) it will end up changing the
> How does going through a servlet change the resource state?
If what is identified by the resource's URI is the application then no, this
latter is not modified (at least on most of the application I saw :-). But
this application may change the state of something stored on the server. If
the URI represents an abstract name space then let's say that the resource
server (I am just using an outrageous example here ;-) is the object
associated to this abstract namespace identifier (i.e. domain.com or "crème
brule"). Since the namespace may represent about anything and since this
hierarchy is an abstraction representing the resource the game is open
governed by the URI owner. The URI semantic is left to the resource owner.
So, yes going through a servlet may change the resource state if the
resource is an object based on a hierarchy of objects. If URI can represent
abstract resources then the namespace interpretation is left to the resource
owner therefore based on this premise, the servlet may change the state of
the resource because you don't know what exactly the resource is, you may
have received just a fragment of it as a response.
Gee... this mental game reminds me of the fun we got about the bugs created
by URIs during the Apple Project X and its language that preceded RDF. If
a name space is lacking any semantics then the URI inrepretation monster may
awake you in the middle of the night. Please don't talk about the URI
interpretation monster to my daughter, since Project X she don't like it :-)
Conclusion: yes going through a servlet may change the state since you don't
know what is associated to the resource. You just get fragments of it.
Moreover, the returned fragments may be as numerous as !param. Who ever saw
an atom? who can say for sure what is a resource? a program? a document? an
elephant? a bug? a bank? an abstraction? god? all these answers? But if the
URL just mean its a location then its OK, its just a location. The guy
living at this location may be a good guy or a monster who knows?
Didier PH Martin
 At that time MCF wasn't based on an XML syntax. it was based on a syntax
like attribute:value. Guha did a first version, we then produced Guha and me
and second one, then Guha went to Netscape after Apple Research closed its
doors, MCF evolved into an XML syntax version, then Guha and al. created RDF
under the W3c umbrella. <End of historical recap>