- From: <Patrick.Stickler@nokia.com>
- Date: Thu, 21 Oct 2004 13:15:49 +0300
- To: <dehora@eircom.net>, <www-tag@w3.org>
> Can someone clarify what is lost when the concept of information > resource is taken away? Or what is gained when it is present? I can at least state my view, which may coincide with the views of others (or not). The importance of the concept of "information resource" is central to the issue httpRange-14, such that, if the range of http: URIs is defined to be a class of information resources, well, then you should obviously define what that class is. Fair enough. But if the range of http: URIs is taken to be unconstrained, then any particular class of resources is irrelevant to the fundamental machinery of the web -- which is agnostically concerned simply with the accessibility of representations of any kind of resource via its identifying URI. Particular classes of resources, such as information resources, will be relevant, even central, to particular applications of the web machinery, but that does not make such classes essential components of the web machinery itself. I would argue that even the significance of a class of information resources to the hypertext web is debatable. Present, typical user perceptions and expectations -- limited pretty much to just wanting consistent behavior when traversing links -- does not depend on URIs identifying information resources. Typical users don't care one way or another what a given URI actually identifies, only that its resolution behaves in a consistent manner. Those URIs could identify any kind of resource, including non-information resources, and if resolution of the URIs exhibit consistent, reliable behavior, then users will be happy. And in fact, that is the case for many appliations which already use URIs to denote non-information resources. IMO, the definition of information resource could be entirely omitted from AWWW without any notable losss of utility, clarity, or coherence. Go ahead, try it. Delete it from the latest draft and change all instances of "information resource" to "resource" and see what impact that has. It's an enlightening experiment ;-) Now, those who wish to constrain the range of http: URIs to such a class of resources will certainly object to any such deletion, but that too, is enlightening, as it illustrates how closely tied that definition is to issue httpRange-14, as those who will most strongly object are (I predict) strong proponents of the more restrictive side of that debate. Cheers, Patrick
Received on Thursday, 21 October 2004 10:23:26 UTC