Our company is moving from waterfall to scrum.

So now testers and developers are really working together. And there is an idea to use checklists and test-cases created by testers as stuff that will help programmers to develop software.

I really feel that something is wrong about it. E.g.: developers might misinterpret requirements documented by testers in test cases. Is there any disadvantages?

UPDATE: after 9 month of scrum I can say that developers actually have no time to double check info (from documentation and from test-case). So I can not even say whether it has bad or good influence on product quality. Actually definitly not bad, as it would be noticed.

  • use check-lists and test-cases created by testers as stuff that helps programmers to develop software, ---- so no more design documents? – Yu Zhang Jul 14 '17 at 11:00
  • Do you already know how these checklists and test cases are going the be formulated? Simply written text? – beatngu13 Jul 14 '17 at 11:00
  • @YuZhang, from wiki "the Scrum Product Backlog is simply a list of all things that needs to be done within the project. It replaces the traditional requirements specification artifacts. These items can have a technical nature or can be user-centric e.g. in the form of user stories." – Ivan Gerasimenko Jul 14 '17 at 11:04
  • @beatngu13, I don't think the way checklists and test cases formulated can change anything to the question. We keep test cases in HP ALM, checklists could be of any kind – Ivan Gerasimenko Jul 14 '17 at 11:10
  • Sure it does; for instance, if testers specify the test cases in a format you can generate code from (e.g. UML, Cucumber, Spock …), it's a quite different workflow compared with plain text documents (I'm not saying it's necessarily better). Moreover, checklists can be part of a DoD or represent a rule in static code analysis tool. – beatngu13 Jul 14 '17 at 11:22

[…] there is an idea to use check-lists and test-cases created by testers as stuff that helps programmers to develop software.

Personally I have found that checklists and test-cases are both helpful to developers as long as they are seen as helpful tools and not fixed requirements. Checklists are a great way to not have to remember a dozen different things to check. Tests cases can be used in a TDD/BDD style to be written before the code and to fail until the app code is written. Existing tests should certainly be part of the developers toolkit that they run as part of their development so that they can get immediate feedback.

Some of the tasks that they can help with:

  • Developing test plans
  • Checklists of what devices to test on
  • Checklists of different users to test with
  • Checklists of what functionality to check
  • Developing test cases (i.e. failing cases)
  • Code review on unit tests as ability allows
  • Guiding UI test writing to use Page Objects
  • Checklists of complicated scenarios to run through
  • Analysis of what functionality UI tests should cover
  • Ensuring tests cover happy, sad and optional paths

Be aware also that having automated tests would help out with many of the otherwise manual tests

| improve this answer | |
  • Most of the listed items is common stuff that testers DO, the question is "Is it OK for devs to use checklists and test cases as guide for the development?" – Ivan Gerasimenko Jul 14 '17 at 11:43
  • Good poiunt @IvanGerasimenko I see that distinction clearly in waterfall. In agile however I feel these become more 'team' activities that everyone should be doing. QE specialize in them and guides good practices but tasks are done by all. Given TDD, BDD and ATDD more of these activities should happen up-front with input from both. – Michael Durrant Jul 14 '17 at 12:33
  • Updated answer (first paragraph) to better answer question – Michael Durrant Apr 25 '18 at 11:25

It's fine if developers execute the test cases and/or use checklists before handing over the application to testers for testing. However, the benefits will be evident if the developers actually use these

I would suggest, provide as much support as you can; to the developers in any way possible.

There is no risk or loss. It is good in a way that QAs will get the better quality code. And in case if testers don't find any bug since they are re-running the same test cases then they can plan ad-hoc testing and/or monkey testing or anything. You will have more time to experiment and find new ways to test the application.

There is nothing that can go wrong with this approach. The only thing is that testing team might be able to find lesser defect as they will get clean code.

Generally, because of the delivery pressure on developers, they don't get enough time to harness the benefits of these helping documents. But, developers should actually test the code thoroughly even for functional scenarios before handing over the code to testers for testing.

From my perspective, it doesn't matter how many bugs testing team raise; the only thing that matters is how many bugs are leaked.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.