OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

sdo message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Questionable sentence in section 4.1.3.


Hi Guys,

Sorry to revisit an old issue, but I'm not sure that the following sentence in section 4.1.3 (part of the resolution of SDO-56) is quite right:

Generated wrapper DataObjects contain disconnected copies of their corresponding converted property values.

What exactly is a "Generated wrapper"? Does it mean cases where the "stored value" is not a DataObject, that is, cases other than the value of an xs:anyType property or the root object returned from an XMLDocument? If so, I think this is really an implementation specific statement/requirement. What if an implementation chose to store all values in an under-the-covers DataObject - note that that is effectively what a DOM tree does (i.e., even simple values are stored in an XML element wrapper).

This isn't a huge issue, but I think the "Generated" part in this statement is a little ambiguous. I'm also not sure that requiring it to always be disconnected is gaining us anything, but instead is forcing implementations to always return a copy of the value, even if it could have simply returned a live value. Does anyone remember why we decided that we needed to say that it is disconnected?

Maybe the fact that the "MUST" keyword isn't in this statement implies that it's not required by every implementation, but personally I would prefer to see the sentence changed to the following:

Returned wrapper DataObjects may contain disconnected copies of their corresponding converted property values.

Thanks,
Frank.



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]