[feedback-report] Feedback types

[Carl S. Gutekunst]

Steve Atkins wrote:

> Doug Otis wrote:
>
>>As this is relatively new, and there could be future requirements that
>>invite modification, then finding some type of convention for extending
>>the form seems useful.
>>    
>>
>The whole concept behind a standard reporting format is that it's, well,
>standard. It's well-defined and can be reliably parsed and the semantics
>of the report can be mechanically extracted and understood.
>
>If there are any elements of the report that my code does not
>understand then I have no way of knowing whether my partial parse of
>the incoming report is an accurate understanding of what the report is.
>
Is it reasonable to give the feedback type the same sort of rigid (and 
restricted) definition as the MIME primary content types? That seems to 
be what Steve is asking for; certainly the arguments are similar.

(Note this argues against fine-grained feedback types.)

<csg>