- From: <noah_mendelsohn@us.ibm.com>
- Date: Tue, 19 Dec 2006 16:56:47 -0500
- To: Elliotte Harold <elharo@metalab.unc.edu>
- Cc: "Rice, Ed (ProCurve)" <ed.rice@hp.com>, www-tag@w3.org
Elliotte Rusty Harold wrote: > Not everything can be HTTP, but not everything needs to be on the Web either. Right, and when I said earlier I'd try to put a bit more balance into the paper that's in part what I had in mind. There's a sense in which the Web Services initiative is targeting some things that are not and quite possibly should not be "on the Web". It's also true IMO that there are some important opportunities being missed to integrate the two approaches: I.e. to name resources with URIs, to use HTTP GET (possibly returning SOAP envelopes) for safe retrievals that don't require the higher qualities of service offered by the upper levels of the WS* stack (reliable delivery in the face of intermittent network failure, transactions running for several days, etc.) (and in a later note Elliotte wrote): > "balance" is all too often a false ideal. Sometimes one point of > view is right and one is wrong. Facts should not be balanced by fiction. Well, I'm not intending to play all sides just for the sake of being agreeable, but I am glad to try and paint a balanced view. In any case, I think the whole value of a workshop like this is to get a respectful hearing for differing perspectives. Even insofar as the TAG may have more detailed expertise and experience to share regarding Web and REST technologies, I know that we're interested in participating in the workshop as much to learn from others as to promote our own ideas. From that perspective, I'm more interested in what our positions will be coming out of the workshop than going in. Noah -------------------------------------- Noah Mendelsohn IBM Corporation One Rogers Street Cambridge, MA 02142 1-617-693-4036 --------------------------------------
Received on Tuesday, 19 December 2006 21:57:00 UTC