CTE XL - FAQs
Overview
Licenses
Software
- CTE XL does not start >>
Testing with CTE XL
- The system to be tested is very large. How do I prevent the tree from exploding? >>
- Do I have to create a tree even for very simple test objects? >>
- What's so new about this method? I have already proceeded in this way. >>
- Ever since I started applying this method for test case detection, I've had to define much more test cases than before until I'm finally satisfied with the test. Isn't this inefficient? >>
- Does the CTE XL also assist me in test execution? >>
- In which case shall I use test sequences? >>
- Wouldn't it make sense to enter precise data as classes in the tree right from the start? In this way the test cases would already be my test inputs. >>
- Which one is easier: creating the tree or filling in the table? >>
- What is the best way to complete the table? >>
Licenses
I have not received any license keyPlease check your SPAM folder. In case the license key cannot be found there either, your company's firewall might have blocked the email. In this case please get in touch with your support.
top
CTE XL identifies the license key as "not valid"
Please restart CTE XL, which should solve the problem.
top
Software
CTE XL does not startPlease check if the download has been completed and - if necessary - download CTE XL again.
top
Testing with CTE XL
The system to be tested is very large. How do I prevent the tree from exploding?First of all a large system should mentally be split into several test objects. An individual tree is created for each test object. Integration tests can help test the interaction of the overall function. Example: In order to test a vehicle, individual detailled trees are being created for the functions of brake, gas and steering gear; another tree only distinguishes the entire ride, e. g. city ride, highway and freeway.
top
Do I have to create a tree even for very simple test objects?
The classification-tree method is best suited for test objects whose functionality depends on various factors and their combinations. Very simple tests (e.g. if a weblink is still valid) should rather be described in a different way, e.g. by means of a list.
top
What's so new about this method? I have already proceeded in this way.
The method deliberately goes by the same approach a good tester would intuitively use. This method, however, documents the approach in a traceable and standardized way.
top
Ever since I started applying this method for test case detection, I've had to define much more test cases than before until I'm finally satisfied with the test. Isn't this inefficient?
Unlike an ad hoc approach, this method makes one aware of the influences and combinations that are crucial for the test. This might cause you to conclude that more test cases are necessary. The reason for this, however, is not the method itself, but the - previously possibly underestimated - complexity of the test object, which now becomes apparent. In case you have to reduce the number of test cases, you would be well-advised to do so on the basis of the tree.
top
Does the CTE XL also assist me in test execution?
CTE XL is a specific tool for the detection of test cases. Although the textual test case specifications provide a good basis for further test activities, execution tools like Tessy are required for the automatic execution and interpretation. CTE XL, however, provides certain possibilities for documenting the results of manually executed tests.
top
In which case shall I use test sequences?
Generally the usual test cases are used instead of sequences. Sequences can be considered if you have to send an input course to the test object for a test run. But even then it is often more useful to vary the entire process in individual test cases. Sequences can be of use if you require a very fine definition of the process.
top
Wouldn't it make sense to enter precise data as classes in the tree right from the start? In this way the test cases would already be my test inputs.
In general this approach cannot be recommended. Although sometimes exactly one value (e.g. null) is part of a reasonable distinction, for the most part, however, you do not want to test a precise value, but define certain features for the entry (prime number instead of „13“). This leads to a better documentation of the test approach and supports the method. Furthermore, precise values often cannot be combined: „below ten“ and „prime number“ is OK; „8“ and „13“ is not.
top
Which one is easier: creating the tree or filling in the table?
More often than not creating the tree is considered easier, as a new aspect can be included irrespectively of the others. When determining a test case, however, all aspects together need to be considered in order to make a reasonable definition of the combinatorics.
top
What is the best way to complete the table?
To begin with, it is often useful to determine certain groups of test cases including a certain test idea. Subsequently a number of suitable test cases is defined for each group. Generic examples for groups are: regular cases, extreme cases, invalid cases. Another idea is to establish combination rules that will be adhered to in the table (e.g. „classification A shall be fully combined with classification B“).
top
Site recommend
Print
