Re: [tag] Test Assertion Modeling - comments, etc

[stephen.green@systml.co.uk]

So far I count three main spec/requirement item types

structural (e.g. instance or fragment A must be valid by schema B,  
e.g. markups)
semantic (e.g. content of element A must be data B, e.g. markups)
behaviour (e.g. event A must result in behaviour B, e.g. services)

corresponding test assertion types

structural - lends itself to predicative TAs
semantic - lends itself to prose TAs
behaviour - lends itself to event-behaviour TAs


Steve


Quoting Dave Pawson <dave.pawson@gmail.com>:

> On 18/08/07, stephen.green@systml.co.uk <stephen.green@systml.co.uk> wrote:
>
>> Tool creators do need their own validity test but I gather you are
>> proposing that the essense of such tests should be schema validity of
>> XML output where applicable. This concept is kind of growing on me :-)
>
> Where applicable. Often it is not enough?
>
> Schemas were never meant for 'business rules.'
>
> Such as Schematron can extend the validation.
>
>
>>
>> The concepts of event-behaviour need to be overlaid on this though for
>> standards such as messaging ones where the system behaviour is of the
>> essense.
>
>
> Added to scope too.
>
>
> regards
>
> --
> Dave Pawson
> XSLT XSL-FO FAQ.
> http://www.dpawson.co.uk
>