| || |
New functionality may include 100's of items that require testing. QA is under severe time restrictions due to Management's desire to deliver. Often, the first pass reveals several bugs. These bugs are immediately reported. As Dev is working on these, QA finds new bugs within this same section. Problem: Dev is complaining that all the bugs should have been found the first pass. Is it realistic to expect ~all bugs to be found before the first report is sent to Dev?
Re: Testing Scope
No. They should be reported as soon as they are found. This way the programmers can start fixing them right away.
Since you are on tight deadlines you do not have time to find all the bugs and then report them. If you do this, QA is sitting around waiting for the programmers to fix everything that was found.
If the programmers keep complaining about this, ask them to do better Unit Testing so QA does not find as many bugs as you are now.
Re: Testing Scope
qatrip, I'd say that you need to consider a few points here:
1. Your Management team's desire to deliver does not have to be translated into an unreasonable Project schedule, where testing is seen as the task that comes at the end of the lifecycle if time permits. If this is the case you may need more strength in Project Management or QA scheduling.
2. As kstawarz said, your Developers really need to stop complaining and start unit testing. They should also be educated by your project manager into the reasons why you don't have a Dev team waiting for the final bug to be raised before they start fixing the first bug.
3. It's imperative that your Developers give a sign off on the functional and technical specs that they have coded from. Many bugs can be caused purely through a Developer not understanding the product that they are expected to build. If you push that Developer to sign off the spec that they will build from, it's amazing how focussed that person becomes.
4. You need to ensure that your test runs are well defined and identical with each test. It's no good having a test run 2 that does not cover the same path as test run 1 (unless of course run 1 was incomplete). I would understand their frustration if you keep moving the goal posts and performing a different range of tests with each run