i have a temprature control application, that is written in java , it has mujava also which is used to test the program using mutation testing one last thing that i need which i can get my head around is the following ,
have three copies c1, c2, c3 of your code (I will call the original c).
Mutate each (but in different ways) to form c1', c2', c3'
Have a 3-version application c4 that uses c1', c2', and c3'
(i.e. it runs c1', c2' and c3' and uses a 'voting' policy).
Test c4 to determine what proportion of tests kill it
(by distinguishing between c4 and c).
Ideally repeat this process with different choices of c1', c2' c3'.
If you do this then you should have estimates of the reliability of
3-version programs in which the three versions are different and all
are faulty. You can compare this with the estimates you have of the
reliability of individual mutants (i.e. Faulty programs where you have
not used n-version programming).
Making all three versions faulty seems 'fairer' when it comes to the
also very important which is already included are some n version programming in the c1 c2 and c3 but i have the classes that have n version programming
i will also need a report with some test results and some justification
8 freelancere byder i gennemsnit £271 på dette job
I am a professional with roughly 6 years exerpience and a college level degree in the field of web development. For an understanding of my skills please check my portfolio at [login to view URL]