Proposed new issue 185: clarify relationship between fault name and type of fault data

["Satish Thatte" <satisht@microsoft.com>]

Per the discussion in the TC call today, it seems that there are different interpretations of allowable relationship between a fault name and the type of fault data associated with that name.  My interpretation has always been that a fault is a “package” consisting of a QName and a specific type of data “payload”.  Using the same QName with different data types would amount to “overloading” of a kind.  But clearly, my interpretation is not the only possible one and in any case this is not clearly stated one way or the other in the specification, which it should be, because this is an issue of normative constraints on the usage of fault names/data.

 

I therefore propose that we open a new issue, Issue 185, in order to discuss and clarify this ambiguity in the text of the specification.

 

Satish