When the test can be stopped

When the new version is to be shipped, I am often asked if the test is complete? Are you confident in quality? What are your standards?
I think many people will find it difficult to answer this question. Basically, you can decide whether to end the test based on the following five conditions.
1. The boss has the say
Basically, the boss is invincible. He says when and when. I don't think everyone will, and I dare not disagree. XD
2. The team has a consensus to stop
If the team decides when to stop the test after the discussion, it can also end.
3. The current price is too high
If you want to find the price of the next bug, it will exceed the loss caused by this bug. It is indeed unnecessary to test it again, so it can end immediately.
4. If the proportion of bugs found drops to the expected target
Sometimes you will observe how many bugs are found in each period of time. If you find that the bug has been falling down and is lower than your goal, you can ship it. for example, less than five buckets/per day, and none of these buckets are serious buckets.
5. If the expected coverage rate is reached
How can you measure your test cases and know how many items are covered? Your test scope is insufficient. such as 90% line coverage and 75% branch coverage. when you reach the set goal, you can also say that the test is over.
At present, it seems that there are only two of them. The comparison has data for reference. The first three comparisons are free evidence. in fact, these are all safe statements. because you only need to give time and resources, you can still find the bugs.
Because <1 bugs/per day or 100% line/branch coverage is not guaranteed. at most, the book gives you confidence. after the shipment is caught, QA is still waiting to be scolded. unfair, but fact. XD
From: 51testing software testing network http://www.51testing.com/html/46/n-863746.html