It makes me sick to the stomach to know we work so many hours of each day, everyday. We hardly spend any time together, we once in a blue moon grab time out for a conversation through the day and why? because we are hard workers, slaves to the taxes and bills, but hang on those who work countless hours, 17+ hrs surely they should be bringing in enough money to survive this costly life right? wrong. It should be that way, but people take advantage of our generosity sometimes, not to mention our time, we'd like to say NO more often but we like to help others, it is our nature to do so. In return we have to spend even more hours from our day to make up for the loss of money earned.

Fact of the matter is, everyone is feeling the same pain. Relationships break down because couples don't share time together anymore, people are dying as Nathan said because they are overworked and stressed out trying to keep up on payments, and my god those payments never end either do they? one thing after another, after another and it goes on, and on, and on and then we look over our shoulders and the kids are all grown up and then we work ourselves hard again to save for our funeral.

That my friends is life as we know it, what do we do...we are forced to carry it through till the end.

What you're describing is something I worry about for you as well, as you know. I'm glad for nothing in this post, but meta-wise I'm glad you're taking the effects of overwork and overstress seriously.

History is full of the dismantlement, fall, and decline of imperial states, but you know, I don't recall the civilian story being told in those histories. It's entirely possible that one of those former-3rd-world nations is the only salvation.

Cheery thought? Of course not. But you know I believe that cultures like yours and mine are tar pits of the soul. I'd be very enthused if you and the fam sought and found greener pastures.

In a project I'm working on (http://lucero-project.info/lb/) we've been creating RDF representations of data from other systems. One example is something called ORO (http://oro.open.ac.uk/) which describes research papers published by faculty members at The Open University in the UK, and in many cases also has a copy of the research paper as a PDF.

I'm now working on another project where we want to assert 'similarity' between research papers. I initially thought that we would be able to use the URIs we'd established in the Lucero project above. However, a quick discussion revealed that this was problematic, as the technology we are using to find 'similarity' (semantic analysis to establish cosine similarity) is comparing pdf to pdf (or document to document) - and some records in ORO have multiple PDFs attached.

I believe that the URIs established do not identify the 'description of the resource' - they definitely identify the resource (as Ed suggested). It's just that it seems that the 'resource' identified isn't the PDF. For those interested in Bibliographic description type stuff, this takes me into the realm of FRBR and SWAP which are intended to model this stuff better in the first place....

I don't want to complicate matters unnecessarily, but It feels like one thing that is missing in the various bindings you presented is the role of the Request. The Request seems to be an essential part of the binding of a <u>

I have trouble with the idea that URIs identify Representations at all. They identify Resources right? Isn't that something that the Document and REST camps have in common?

The most important word for me is "used" here. Because the URI as a string can be used as an globally unique identifier for anything. But if it is used to retrieve a representation, it is always interpreted as a identifier of an information resource.

Although it is a technically solvable problem to dereference a URI that identifies a non-information resource (303, hash), it alters the semantics of a URI. The string "http://dbpedia.org/resource/Moby-Dick" can be used as a globally unique identifier for the concept "Moby Dick". But when I make use of the string by typing it into my browser I always get back information *about* "Moby Dick". That is exactly what I expected. So why 303 here? Let's take a URI in a write scenario where I can update the information. I don't expect to alter "Moby Dick" itself, but the information the owner of dbpedia.org has about "Moby Dick". The point is here: only the party issuing the URI identifying an abstract resource has the authority to alter the information or 303 me to another resource. But there isn't such a concept of "information authority" in the thing itself.

Maybe we should use non dereferencable identifiers by intention to identify non-information resources. And use URIs only if they are a reference to an information resource. Does that break anything? I don't think so, because we only treat the two strings "http://dbpedia.org/resource/Moby-Dick" and "http://dbpedia.org/page/Moby-Dick" as a reference to the same information resource. The reference (e.g. the uuid) to the non-information resource should be enclosed in the "body" of the information resource.

great post. a company that has a new approach to what you are proposing is SBSGRID. they do offer precise graph traversal in a search experience. so its the end of search... but also the start :)

best
bernard

First, the definition of a URI -

"A Uniform Resource Identifier (URI) is a compact sequence of characters that identifies an abstract or physical resource."

and second, a clarification of what a "resource" is.

"the term "resource" is used in a general sense for whatever might be identified by a URI. "

It is just an identifier.

I don't see how a URI could represent a 'source of information'
(note that I consider 'information' as something that is content)

in those following cases

* root of server

* an empty directory

* the dot and dotdot reference

* the path of an executable

even if you don't consider 'information' as being content,
and take a broader interpretation of the term

using expressions like "U can be resolved to an address for X",
"U refers to that information", etc.

to me, it imply a "location"
as what kind of question does U answer: "where it is located ?"

so I agree with
"we always use a URI to refer to a source of information (static or computed), whether it's modifying it, or getting some version (representation?) of it."

but I don't think it works in all situations (eg. File System vs Server/Network)

(cf. http://infoserviceonto.smiy.org/2010/11/25/on-resources-information-resources-and-documents/)