I'm doing some research on how to improve the coverage and quality of unit level testing, a bug that was found during development with unit testing is easy to fix, but a bug that was found after development in system level testing especially stress testing, can be extremely painful to fix. I'm looking at a lot of bug reports with bug fixing code from everywhere about the bugs that was found in system level testing, and trying to come up with a standard, a systematic way to determine whether the bug can be found locally. The reason I'm trying to find a systematic way is that deciding whether a system level bug can be found at unit level or not by just reading the bug report, it is really subjective. Different people have different opinions, it's like whether a restaurant has good food or not, but with a meaningful reasonable standard like a check list, it would be more convincing.
Any suggestions?
Aucun commentaire:
Enregistrer un commentaire