- From: Pat Hayes <phayes@ihmc.us>
- Date: Fri, 5 May 2006 10:45:19 -0500
- To: "Booth, David (HP Software - Boston)" <dbooth@hp.com>
- Cc: <public-swbp-wg@w3.org>
> > From: Pat Hayes >> . . . >> Seems to me that in cases like this we can just agree to identify a >> piece of text with any *constant* function from some domain to that >> piece of text, particularly if the only access anyone can have to the >> function is to call it and get the value of it, i.e. the text, >> delivered as a result. Even mathematicians routinely identify >> constant functions with their values. So whether a web page is >> 'really' a text or is 'really' a constant function from the set of >> times (or whatever) to that text seems to me something we can just >> agree to disagree about, without it mattering. In fact, even the >> publisher of the text and the reader of the text might disagree about >> this, without it mattering to either of them. > >I'll buy that . . . mostly. Except that if some "information resources" >are functions from time to data, while others are merely data, I'm left >with a nagging question in the back of my mind: I wonder what *else* >could be an "information resource"? Given an infinite variety of >resources, might there not be a few more kinds that we have neglected, >which should also be permitted into this exclusive club called >"information resources"? How do we know if we've got them all? Maybe we never will. But those are the breaks when you start applying neat definitions to a scruffy world :-) And after all, its not as if 'information resource' is a natural kind: it is a term of art, so to some extent we can make it mean what we collectively want it to mean. >On the other hand, if I know that *all* "information resources" are >functions from time to data (and nothing else), then I can rest much >easier. But some non-information resources might satisfy this description as well. The Delphic oracle was a function from times to data, and so is my long-case clock. I think the WebArch definition is quite good, even if it could perhaps be be tightened up: an information resource is one that can be completely characterized by a bitstream (to paraphrase slightly). This is actually quite a neat way of putting it, since it allows for the conceptual distinction between the resource and the representation of it which gets transmitted; it doesn't get involved with gritty issues of types versus tokens or temporal sensitivity, etc., but it also has the consequence that information resources have to be in some sense 'made of symbols', since anything physical or natural can never be completely characterized by a symbol sequence. And it leads very directly to the observation that when talking about information resources, you can treat the map as being the territory, for all practical purposes. Pat Hayes > >David Booth -- --------------------------------------------------------------------- IHMC (850)434 8903 or (650)494 3973 home 40 South Alcaniz St. (850)202 4416 office Pensacola (850)202 4440 fax FL 32502 (850)291 0667 cell phayesAT-SIGNihmc.us http://www.ihmc.us/users/phayes
Received on Friday, 5 May 2006 15:45:38 UTC