I had an interviewer at Microsoft look at me like I was insane when I mentioned that I thought some amount of manual testing was necessary. He seemed to believe that everything should be automated. He was a development manager, not a tester.
What job were you applying for? I fully understand why a development manager would want all their testing automated, in our software house we have introduced test driven development, so all unit testing is automated and what we refer to as development integration (so running all the unit tests on a continued build process) is therefore all automated.
If the job was a testers position then I think the answer to the question posed is yes, automation can replace manual testing. However you should then qualify the statement that it is not practical or even desirable for all the manual tests to be replaced by automation due to cost of setting up verses the gain it provides.
As Joe pointed out it is actually very hard to automate all tests, I trust our testers to use their instincts as well as test scripts, and that is very hard to automate (all though arguably not impossible)
I am totally agree with bansaltarun.Manual testing is the part of automation testing.So Manual testing is much important.I would like to share one thing that when we do manual testing then getting more bug then automation testing. <font color="orange"> </font>
[ QUOTE ]
<font color="brown">Manual testing is the part of automation testing. So Manual testing is much important. </font>
[/ QUOTE ]
I agree that manual testing is important.
[ QUOTE ]
<font color="brown">I would like to share one thing that when we do manual testing then getting more bug then automation testing. </font>
[/ QUOTE ]
Does that statement minimize the importance of automated testing or suggest that it is less efficient at detecting defects than manual testing? I do not think so. I think in most cases you would expect that established automated tests would find <loosely stated> fewer defects than automated tests. I think it also important to understand why. In most cases the process of automating tests required manual testing during the creation of the automated tests. During that process many defects are typically found. Hopefully they are corrected. Let us say for example that in the process of automating application X, 50 defects were found and all corrected. In theory the existing automated tests should find no more defects on the version X where all 50 were corrected. If manual testing discovers 20 more, then that is good, but it does not reduce the significance of the automated tests. Those tests proved their worth by showing no regression and that app X worked correctly in the areas covered by the automated tests. Not only did the automated tests do the testing of those areas but hopefully they tirelessly and faithfully repeated tests and saved time and therefore money, thus meeting the purpose of automated testing.