Re: [tag] revised "Anatomy of a TA"

["Dave Pawson" <dave.pawson@gmail.com>]

On 28/08/2007, Durand, Jacques R. <JDurand@us.fujitsu.com> wrote:

General comments.

The acronyms and abbreviations need removing or clarifying.

antecedent and consequence seem to be borrowed from a long way away?
Pre-test and post-test conditions is more usual?
TA must identify....
The actual test and expected test result are omitted.
Need to differentiate between the test outcome and the
altered state of the test object (post-conditions)




TA-101 is time bound. ('related to a previously sent PO).
That's a pre-condition. Unless the association is part of this
test, there should be no mention of it.
The post-condition mentioned, 'validates' is a test result, not
a post-condition. I.e. it seems to be the outcome of the test?
A post-condition normally relates to the abnormal state in which
a test leaves the test object, e.g. The car window is left open
after test, which is not the neutral state.

No mention of version of the schema?
The pre-condition shouldn't contain questions. They should be
resolved prior to specification of the test.

Re 101b

Very unclear.
Rather than having an object to test, it is the sender of the message
that is being tested?  No test is stated? Again confusion between
the post-condition and the test result.

re 102.
  Makes for a very unclear example. The item under test is the
purchase order receipt. The test is that it has been received
within 24 hours from the transmission of the order.
Again no test stated. The time of transmission of the order is
a pre-condition.





-- 
Dave Pawson
XSLT XSL-FO FAQ.
http://www.dpawson.co.uk