Attifunel process req test description added#679
Attifunel process req test description added#679attifunel wants to merge 4 commits intoeclipse-score:mainfrom
Conversation
|
The created documentation from the pull request is available at: docu-html |
| Requirements analysis is a testing technique where tests are designed based on the | ||
| requirements assigned to the software element under test (component, module, unit etc.). |
There was a problem hiding this comment.
| Requirements analysis is a testing technique where tests are designed based on the | |
| requirements assigned to the software element under test (component, module, unit etc.). | |
| Requirements analysis is a test derivation technique where tests are designed based on the | |
| requirements assigned to the software under test (unit, component, feature, etc.). |
There was a problem hiding this comment.
I would still keep "SW element" as it's the most generic definition in ISO 26262 of any hierarchical piece of SW
| Requirements should always be tested against their "nominal" behaviour: i.e. "if input | ||
| x = A, the output y of SW component X shall be equal to b". This leads to a test case | ||
| injection value x = A and verifying that output y = B. "Negative" testing technique should | ||
| be also considered: i.e. testing what happens when input x = B (if not already specified | ||
| in other requirements). |
There was a problem hiding this comment.
| Requirements should always be tested against their "nominal" behaviour: i.e. "if input | |
| x = A, the output y of SW component X shall be equal to b". This leads to a test case | |
| injection value x = A and verifying that output y = B. "Negative" testing technique should | |
| be also considered: i.e. testing what happens when input x = B (if not already specified | |
| in other requirements). | |
| Requirements should always be tested against their "nominal" behaviour: i.e. "if input | |
| x = A, the output y of SW component Z shall be equal to B". This leads to a test case | |
| injection value x = A and verifying that output y = B. "Negative" testing technique should | |
| be also considered: i.e. testing what happens when input x = B (if not already specified | |
| in other requirements). |
There was a problem hiding this comment.
I would also change the negative case input to something different than B...
| Note that requirement-based testing derivation is typically covered by other techniques | ||
| mentioned here, such as Boundary Values, Equivalence Classes, Fuzzy Testing, Interface | ||
| Testing and Fault Injection. |
There was a problem hiding this comment.
| Note that requirement-based testing derivation is typically covered by other techniques | |
| mentioned here, such as Boundary Values, Equivalence Classes, Fuzzy Testing, Interface | |
| Testing and Fault Injection. | |
| Note that requirement-based testing derivation is typically covered by other techniques | |
| and methods mentioned here, such as Boundary Values, Equivalence Classes, and Fuzzy Testing. |
Interface testing is verification method, but according to the iso not a derivation technique. Same holds true for Fault injection.
There was a problem hiding this comment.
Mhmhm... I have this opinion that ISO 26262 makes a bit of mixed bag there by mixing "derivation methods" and "techniques". I wanted to stay generic speaking of "techniques" without derivation, but maybe this causes confusion?
Added requirement method with description