Tosca Tips & Tricks: notes on TestCase Design module of TCP

Notes on Powerpoint Presentation by Tricentis as part of the 'TestCase Design' module of TCP:

• Testcase Design section elements (p.11)
• TestCase Design Folder => TestSheet => Attribute => Instance
• Attribute properties (p.12+)
• AttrType:
• Logical (grey top)
• Physical (red top / no coloured top)
• BusinessRelevant:
• No (white top)
• Yes (red top / no coloured top)
• Result (green top)
• Instance properties (p.15):
• Character (F7)
• Valid
• Invalid
• Straight Through
• Position (F8)
• Inner
• Boundary
• Equivalence class (p.16)
• Each representative produces the same error =>1 representative value per class
• Work because the same result is expected for each and every representative of the same class
• Equivalence Partitioning (p.17+)
• Boundary value errors are one-time errors and don't increase coverage (therefore: not part of regression)
• Floating attributes (numerical values)
• Non-floating attributes (enumeration of values)
• Avoid testing for all invalid values, only needed/requested ones
• Straight Through (p.27)
• It is the test case with the highest risk
• It is the test case which can be most easily implemented
• It is the test case which offers the best flexibility to be combined with the other attributes
• Standard TestSheet: subdivide attributes in:
• Preconditions
• The processes under test
• Verifications
• Linear Expansion
• One defined Straight Through
• One test focus per test case
• 
  

Notes on videos by Tricentis as part of the 'TestCase Design' module of TCP:

• First placeholder note
•  Arranging Instances (rightclick attribute => Arrange instances)
• Order: Straight through, Valid instances, Invalid instances, Boundary values
• Generating Instances
• Rightclick on attribute => Create instance (both in left- and right side of Tosca viewing screen - the latter even by just manually typing a new instance value in an attribute field)
• Combining attributes at a higher level
• Create Instance and delete the single empty instance at the desired combination level leaving only an Instance Folder
• Select the attributes to combine
• Rightclick the select instances -> Complete Instances -> Linear Expansion
• Instance Filter
• User the Filter column in the instance selection screen to select for particular attribute values
• Particularly handy when checking for reduntant testcases
• 'Reset Instance Filter' when done comparing
• Redundant testcases
• Just delete them
• Best Practices for Instances
• Naming: separate different aspects of an attribute value and/or instance with pipes
• Verification Attributes (green top)
•  (Often, always?) no need to create an instance at the top attribute (verifications are RESULTS of the instances, not instances themselves)
• Creating Instances on a Testsheet level
• Make sure the instances of the testsheets attributes (if any) are generated
• Rightclick -> Completed Instances -> Linear Expansion
•  Verifications will be ignored when it comes to the combinatorics
• Combinatorics
• Linear Expansion
• Contributions of an attribute with N instances to the total number of testcases of the instance above it: N-1 (with the '-1' being the straight through)
• Verifications
• Choose the value 'N/A' for verification attributes which are not relevant (such as error message for a valid instance or success messages for invalid ones).
• Best Practices
• Create one Testsheet per requirement
• TestCaseDesign vs. Requirements
• You can only drag-and-drop a Testsheet on a Leaf requirement and not on a branch requirement
• If a Testsheet is drag-and-dropped on a requirement a 'Testcase Substitute Link' is created for every Instance of the Testsheet (yellow circular arrow with red  ribbon)