Lecture 5

To discover and fix the bugs To determine whether or not the system is usable in an operational situation.

Software Inspection(static): analyze, review and examine the code to find errors. Software testing(dynamic): the system is executed with test data and its behavior is observed.

You can pick up errors that mask other errors, incomplete version of a system can be inspected, pick up problems in terms of the quality of the code, pick up problems you cannot get from testing.

<p>To force a program to work incorrectly. As a software tester, your skill will be measured by the number of bugs you can find so you have to force the app to crash. To show that the software works correctly. For example, fault based testing to show that certain types of faults do not exist and to test that the program does not divide by zero so the test data would be zero.</p> Signup and view all the answers

<p>Developers, an independent test group (in a large project), the customer</p> Signup and view all the answers

<p>Development testing is done by the developers, unit testing is testing individual parts of a program such as functions or methods and classes, system testing involves checking if the whole system works by testing all the parts to see they interact correctly.</p> Signup and view all the answers

<p>testing all methods and functions, checking and modifying the variables and testing the object under all conditions.</p> Signup and view all the answers

<p>The first is the normal input to check if the program works as expected. Next is the abnormal input for if they are properly processed and does not crash the program.</p> Signup and view all the answers

<p>White box testing: examine the source code, black box testing: testing the product to see if it works as expected. Design and come up with the test cases without looking at the code, rely on use cases and function specifications to come up with that test case.</p> Signup and view all the answers

<p>Random testing is generating inputs randomly. All you need is to know what the program is supposed to do and what the function is supposed to do and it can come up with random inputs. Systematic is trying to select inputs that are actually valuable by choosing representatives of the classes that are &quot;to fail&quot;, &quot;often&quot; or &quot;not at all&quot;.</p> Signup and view all the answers

<p>It generates too many test cases, it will take ages and a large computational power to complete the test cases, it is also expensive.</p> Signup and view all the answers

<p>Equivalence partitioning: dividing the input data into sets of data from which the test cases can be derived. For example, a test case of a set of data between 1 to 10 is created. If any one of the value from the set passes, then the whole set is considered valid.</p> Signup and view all the answers

<p>Boundary value analysis is testing the boundary values, values just inside and values just outside of the input ranges. For example, in a range between 1 to 10, boundary data would be 1 and 10, values just inside would be 2 and 9 and just outside would be 0 and 11 and only those numbers would be tested.</p> Signup and view all the answers

<p>Statement coverage: each line of code gets executed at least once. decision branch: each decision(like an if-else statement) returns true and false at least once. Condition coverage: each condition within a decision branch should return all possible outcomes at least once. path coverage: all possible paths inside the code should be tested at least once.</p> Signup and view all the answers

<p>Integration testing is combining all the unit test and checking if they interact correctly when they are integrated together.</p> Signup and view all the answers

<p>functional testing: testing all the functions, performance testing: testing all the non function requirements, for example, reliability, ability, etc. acceptance testing: testing for validation with the system customer.</p> Signup and view all the answers

<p>test driven development is creating a test case before coding and developers will implement with the goal to pass the test case. How it works is that it starts by identifying a new functionality that can be done with a few lines of code. Then write the test case for that code and implement it as an automated test. Then you will run the test. You cannot implement new functionalities until the previous ones pass the test cases.</p> Signup and view all the answers

<p>Every function has a test case, if change in code breaks the code, the test case fails, otherwise, success, because of the test case, you know which function is causing the bug, tells you what the function is supposed to do.</p> Signup and view all the answers