- From: Karen Coyle <kcoyle@kcoyle.net>
- Date: Sat, 14 Mar 2015 12:07:46 -0700
- To: public-data-shapes-wg@w3.org
Peter, I definitely didn't say "natural language alone." In fact, I said the opposite: If >> there are formalisms that must be used, they should exist in addition to >> natural language descriptions. I also am pretty darn sure that not many people who approach standards documents actually read and understand the formalisms. I want to make sure that bright, tech-savvy folks of many backgrounds can understand the standards document. Any formalism used will exclude more readers than the English language will. It can't be the heart of what we have to say if we hope to communicate widely. kc On 3/14/15 10:10 AM, Peter F. Patel-Schneider wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > I think that natural language alone is an extremely poor way to define > anything. It is just too easy to misinterpret natural language statements > even if the language is simplified and stiltified. > > The SPARQL 1.1 query language, for example, has a complex mathematical basis > provided in http://www.w3.org/TR/sparql11-query/#sparqlDefinition. RDF has > a model-theoretic definition. > > peter > > > > > On 03/14/2015 09:55 AM, Karen Coyle wrote: >> >> >> On 3/14/15 8:46 AM, Irene Polikoff wrote: >>>> >>> >>> would and should actually read and understand Z semantics or some >>> mathematical axiomatic descriptions, but would find SPARQL more >>> difficult to understand? >> >> I hope nothing that I have said could be read as support of Z semantics >> or mathematical axiomatic descriptions because I am not familiar with >> either. I prefer that standards be written in clear natural language. If >> there are formalisms that must be used, they should exist in addition to >> natural language descriptions. Many standards take this approach. In >> fact, the only place that I seem to have run into (incomprehensible) >> formalisms is in the OWL documentation. Everything else uses natural >> language (stiltified to be precise, perhaps) and examples. Even the >> SPARQL documentation takes this approach, so I don't understand why >> "formalisms" are assumed to be needed for SHACL. >> >> kc >> > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1 > > iQEcBAEBAgAGBQJVBGtuAAoJECjN6+QThfjzDqYIAM6lMD3F+zMkHkPajzyBxHnQ > KMdbrd2wdhU/sYubsuXmSRJFE9mutXophNOFgtcbsgRA46ILV5kZV4hp43pc+5SP > 7avLLPD/HN+1dNOvs5i3f8OBqqhYxYuRSeD8KCNz9am/TlDdM3EC0/67C9qcZS9O > +utIi1kgeQMVjrhOhdAvWnz0qC7s/02hoMMV3CF3Jk/h93NUamyjFLHHUhqsTvNR > +1wBK+xXWOAjWwpeQWeDyA08OoA9p1oCMDL+h3pfa1c3c91BdWUVxL8Yfq5MDtKY > rrYkb5DUt/dhfWhxCJyM1YAAGcCzka1dnx6tJVzUGUkL4kodNY5LXUM5NC0kgkc= > =Lsxm > -----END PGP SIGNATURE----- > > -- Karen Coyle kcoyle@kcoyle.net http://kcoyle.net m: 1-510-435-8234 skype: kcoylenet/+1-510-984-3600
Received on Saturday, 14 March 2015 19:08:15 UTC