Presentation on theme: "Comments on 32N1238 ISO/IEC WD24707 Japan ISO/IEC JTC1 SC32/WG2 2005.4.15."— Presentation transcript:
Comments on 32N1238 ISO/IEC WD24707 Japan ISO/IEC JTC1 SC32/WG2 2005.4.15
2005-04-13 東京電力株式会社 システム企画部 岡部雅夫 2 Folded map is necessary We think one of the significant features of CL is type-free. For example, some name (or non-logical symbol) can be interpreted to Domain of Discourse and the relational extension of the Domain of Discourse at the same time. Then, to ensure the consistency among them, folded map is necessary. This is what we learned from the previous version of CL WD. But, we cannot find any description about folded map in this version of CL WD.
2005-04-13 東京電力株式会社 システム企画部 岡部雅夫 3 Folded map XVOVRXVOVR int I (X) U I rel I (X) Rel I Note: Rel I : relational extention of U I int I rel I relation I : folded map such that relation I (int I (X) ) : rel I (X)
2005-04-13 東京電力株式会社 システム企画部 岡部雅夫 4 Semantic Conformance should be defined dependent on Domain of Discourse In new version, the term ‘domain of discourse’ is used as a (subset of ) of range of interpretation, a subset of ‘universe’. This is a good idea. As far as we can understand, this can be related to context, subject area etc. in XMDR etc. To have more semantics, it is better to describe ‘7.1.2 Semantics’ conformance, dependent on D, a domain discourse.
2005-04-13 東京電力株式会社 システム企画部 岡部雅夫 5 7.1.2 Semantics’ conformance’ Annex A, B, C does not say nothing about the dialect’s formal semantics. But, for example, how do we know, (and (Person Harry) and (Organization ISO)) is a ‘boolean sentence of type conjunction and component C 1 … C n ’ Intuitively, it is obvious. Some more rigorousness may be required. Syntactical definition of formal semantics of each dialect may be necessary. So, we can say nothing about whether J(T)=I(T) because we do not know T correspond to what type of sentence at table1. And then, ‘weakly semantically conformant ‘ is not necessary because any text T of the dialect is weakly semantically conformant as far as it is satisfiable in its formal semantics because a formal semantics of the dialect is a extension of CL.
2005-04-13 東京電力株式会社 システム企画部 岡部雅夫 6 Editorial (or structural) comment (1 of 2) ‘5.2 Design Overview’ should be moved to‘Annex(informative)’ or ‘Introduction’ because this part is almost tutorial and not normative at all.
2005-04-13 東京電力株式会社 システム企画部 岡部雅夫 7 Editorial (or structural) comment (2of 2) The following annex should go to the main text ‘Annex A(informative)’ (KIF) ‘Annex B(informative)’ (CGIF) ‘Annex C(informative)’ (XCL) because if we look at ‘7.Conformance’, these are the core of this standard. ‘6.5 Summary of the core syntax’ should go to Annex(informative) since this is not normative or may disappear because ‘Annex A(informative)’ (KIF) come to the main text.