In Reply to: RE: Nope!, Never had that problem ....................... posted by zacster on September 30, 2020 at 05:09:37:
" a program testing a program needs to have a way to test the program testing the program"
While true in an absolute sense, in practice, automated testing is often very successful in identifying defects in software.
Your statement ignores the reality that the complexity of the code has a huge impact on the likelihood of defects e.g. that the code doesn't do what is intended.
I have written SQL modules of 5K lines (not neccessarily something to brag about I know) and a test would often be one line of code- calling the procedure, passing in a parameter value, and returning a pass/fail report based on the return value from the procedure.
There is no way that there will be anywhere near the quantity of defects in a one or two line test script vs. a procedure of 5000+ lines. *Could* there be? Yes. *Will* there be? Almost never.
In the real world, your inference that using code to test code is not useful does not at all conform to the realities of programming.
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- RE: Nope!, Never had that problem ....................... - LtMandella 09:06:58 10/01/20 (9)
- RE: Nope!, Never had that problem ....................... - Cut-Throat 13:20:28 10/01/20 (8)
- RE: Nope!, Never had that problem ....................... - LtMandella 13:36:26 10/01/20 (7)
- J.River Mac Version Sonic Comparison - MRMB 13:43:33 10/09/20 (6)
- RE: J.River Mac Version Sonic Comparison - LtMandella 18:47:33 10/09/20 (4)
- RE: J.River Mac Version Sonic Comparison - MRMB 09:41:10 10/10/20 (3)
- RE: J.River Mac Version Sonic Comparison - LtMandella 10:26:52 10/10/20 (2)
- RE: J.River Mac Version Sonic Comparison - rivervalley817 10:41:23 10/10/20 (1)
- RE: J.River Mac Version Sonic Comparison - LtMandella 19:48:50 10/10/20 (0)
- RE: J.River Mac Version Sonic Comparison - LtMandella 14:48:56 10/09/20 (0)