- From: <paul.downey@bt.com>
- Date: Fri, 21 Apr 2006 19:48:47 +0100
- To: <jon.calladine@bt.com>, <public-xsd-databinding@w3.org>
> 1) Section 1. Add the following text to Para 3 > "While it is impossible to guarantee that schemata > produced using these patterns will work with the > universal set of databinding tools, these patterns > have been based on real experience with many tools, > covering a wide range of languages and platforms." Ok, I went with: """ Whilst it is not possible to guarantee that schemata produced using these patterns will give a good user experience with the universal set of databinding tools, the patterns contained in this specification have been tested with a number of differnet tools covering a variety of different programming languages and environments. """ > 2) typo: section 1. 4th para lest=least done > 3) Section 2.4 typo in incomplete sentence. > "Several different patterns are provided for expessing > the absence of a value, however minOccurs and nillable." > should be: > "Several different patterns are provided for *expressing* > the absence of a value, however minOccurs and nillable > are not recommended in combination." done > (N.B. going forward a lot more detail is required here > I believe....) > 4) Section 2.4 "Type serializers summarily convert type > members with NULL values to xsi:nil declarations, > which can cause unintended data changes on the server" > should be qualified: > "Some type serializers ......" Section needs a rewrite this following the resolution of ISSUE-7 (there's an ednote to that effect) > 5) section 2.8 The patterns are listed out of numerical order. OK, fixed. that's the magic of spec-refs > 6) Section 2.9 is empty and should be removed for the purposes of this snapshot. added an ednote as the section comes from the input document. > 7) We have some examples in section 3 for which there > are 'health warnings' for in section 2. A nicety would > be to have a link back to the relevant design consideration . I think we need to number design considerations and attach an XPath/Schematron assertion to each pattern, especially if we're going to generate them as 'warnings'. I'll track this as a separate issue. Paul
Received on Friday, 21 April 2006 18:49:24 UTC