- From: Al Gilman <asgilman@iamdigex.net>
- Date: Sun, 26 Aug 2001 10:03:59 -0400
- To: gis-wg@gridforum.org, uri@w3.org
Let me experiment with that ANNOUNCE slug to discourage cross-posting in reply. ** URI folks: The Grid Information Services Working Group has under consideration a proto-RFC known as the Grid Notification Framework (GNF), see Grid Forum Areas and Groups http://www.gridforum.org/groups/WG.html The general class of applications or services served by this framework has widespread use for the combination of the two "information fields" producedBy=service asOf=dateTime And that combination is perhaps (we need to nail this semantic bit down, yet) a sufficient key for obtaining any available further information related to the transaction which yielded the information passed in the notification. Sounds like probable cause for having a URI form for this tuple as a unit, or at least considering one. Since the vast marjority of URI uses transiting the Web today can be viewed just as naturally as service-request parmeters as they are viewed as resource-identification dereferences, it is likely/desirable that the 'service' identification in the above remain conformant to or interoperable with reference by URI. I get overly academic. The point is that The Grid is a user group which _will_ be using 'duri' semantics as discussed in draft-masinter-dated-uri-00.txt available by e.g. ftp://ftp.ietf.org/internet-drafts/draft-masinter-dated-uri-00.txt ** GIS-WG folks: Please see http://lists.w3.org/Archives/Public/uri/2001Aug/thread.html#26 for ongoing discussion of this Internet Draft. Please use discretion in cross-posting replies. Consult with Gregor or Larry Masinter if you feel you need to sharpen your discretion. Al
Received on Sunday, 26 August 2001 09:43:52 UTC