Before I could throw more light on the title of this post, let me come to the role of Automation Testing in the Software Development Process. Though, I am not personally against Test Automation, still I tend to doubt the effectiveness of test automation in the Software Testing Process! Though, I have worked in many Projects involving Test Automation, there were only few cases where Automation could yield the desired result. And in most cases, test automation was a nightmare rather than a moral boost for the test manager!
Whenever a new automation tool is lunched, the advocates of the new tools claim that the applications give in-house testing professionals more control, including the ability to perform tests that are repeatable with ease. However, critics of automated tools consider it as a poor substitute for a thorough and nuanced manual test that a skilled practitioner performs. Most experts agree, however, that an automated test in the hands of an untrained novice could do more harm than good.
"A fool with a tool is still a fool," said Bill Harrod, a
security management consultant at CA, formerly Computer Associates.
The main purpose of adopting test automation is to gain ability in performing test executions faster and more accurately than a manual tester. But in practice, it becomes very hard to achieve. I personally believe that only test automation is not good enough for the over all health of a software. However, it can be used as a supplement along with skilled manual testing. But test automation in itself is not at all sufficient as far as a testing approach is concerned. If at all, we are using any automation testing tool, then also the effectiveness of the tool largely depends upon the skill of the person using the tool. So at last, what matters is the skill of the tester and not the tool. A tool might be with many great features to test an application. But as long as the person using it is not well trained, it is going to do more evil than good. More over, to train a novice takes much time and effort. If the same time could be allotted to skilled manual testing (exploratory testing!), then I think that could yield better results.
Your thoughts please…
Hi Debasis!
ReplyDeleteYour opinion might be true in some cases like
1)when the project is small in size
2)When the changes are not much
3)when there are no skilled automation testers
Actually i too agree with you, as iam also not using any tools. but in my opinion tools are very much useful if we train the skilled manul testers and allow them to use when 1)When proj size is bigger
2)More changes occur frequently
3)Excessive Regression testing needed projects
4)More data to be tested on the same page/ field
Tools are like some advanced weapons for a soldier called Skilled Manulal Tester to find and kill Bugs as many.
@ Purna
ReplyDeleteThanks Purna for your Comment. I do agree with you that automation tools can do wonders if used effectively. But in practice, that is seldom the case. Usualy, the cost associated with the tool, the effort needed to train the testers about the tool and the competency level required to design a good test script using a tool outweigh the positive points of using a tool.
How ever, there are certain areas of testing which can't be covered without using a tool (like Load testing a Website/Application, regular Regression testing which are done almost on a daily basis, testing Complex Projects/Products etc).
Tools can be used as *advanced weapons" to find and kill Bugs. But be carefull not to *shoot at your own feet* while trying to do so. :)
The key of a proper testing approach can be the right combination of Manual and Automation Testing. But the sad thing is that, this formula is yet to be discovered.
Any way, thanks once again for sharing your views with us.
Regards,
- Debasis
I accept 70 to 80 % of Purna's comment because it is so much helpfule when the project is too large and frequent changes occurs.
ReplyDeleteAnd have to follow the performance testing, not able to do manually.
And some clients or companies may not able to use the automation tools..that may be due to budjet problems or some other things....
As per the Debasis, this is also acceptable 20 to 30% only.
Thanks,
Chandra
@ Chandra Mohan
ReplyDeleteThanks Chandra Mohan, for dropping your comment. But, the points you have mentioned are already discussed in my reply to Purna's Comment.
Here, I needed your own views/comments on the post. Not what you think about my or Purna's comment.
Any way, thanks for spending your time on leaving a comment.
Regards,
- Debasis.
Actually i think that only automation is not that much useful,
ReplyDeleteautomation should be implemented with predefined strategy and approach, with a proper frame work keep in mind, depends upon project's requirement and nature.
some characteristic of framework should be maintained while implementing
- modularity
- maintainability
- structure