- From: Graham Klyne <Graham.Klyne@MIMEsweeper.com>
- Date: Mon, 29 Oct 2001 09:56:32 +0000
- To: Pat Hayes <phayes@ai.uwf.edu>
- Cc: w3c-rdfcore-wg@w3.org
At 04:29 PM 10/26/01 -0500, Pat Hayes wrote: >>> http://www.w3.org/2000/03/rdf-tracking/#rdfms-replace-value >> >>[I missed this when it was raised] I like this proposal, as it makes >>much more intuitive sense to me, and leverages existing developer >>understanding of toString() in existing programming languages. Assuming >>it's consistent with any decisions about datatyping, I think there would >>be value to our current charter in adopting this. >> >>To deal with backward compatibility, I'd suggest that rdf:value be >>retained as a "deprecated" form, equivalent to rdf:toString, which >>becomes the preferred form. >> >>Thus, my vote is AGAINST deciding to defer this issue to a new WG at this >>time. I think the proposal should be accepted or rejected by this WG. > >If it is being considered, then I vote to REJECT it, for reasons that I >hope will become clear when I get the MT datatyping document finished. >(Basically, the appropriate interpretation of rdf:value in the new MT is >that it is simply identity, and the 'toString' name is then actively >misleading. This view of the meaning of rdf:value has built into it a view >about datatyping that I want to argue against, basically one which assumes >that all literals occurrences are strings. I think that it only makes >intuitive sense if one agrees to read all literals as referring to >themselves, which, I will argue, embodies a use/mention confusion which is >neither semantically productive nor necessary.) This would be fine by me (modulo actually seeing and understanding the proposal). I think Aaron said it better. I don't think this is an issue that should be postponed. #g ------------------------------------------------------------ Graham Klyne MIMEsweeper Group Strategic Research <http://www.mimesweeper.com> <Graham.Klyne@MIMEsweeper.com> ------------------------------------------------------------
Received on Monday, 29 October 2001 07:54:15 UTC