The traditional QA mentioned here refers to the testing work in the waterfall R & D process.
In the industry, the biggest impact to testing work over the years is automated testing and agile processes. Automated Testing challenges the old black box testing and increases the tester's barrier requirements. Agility further increases confusion, many testers lament whether QA is useless in agility, and thus have doubts about QA's career.
These attacks hit many industry-leading companies, such as Google. You can see many Google tests on YouTube.
If we compare Google with traditional QA, we can see some clues for improvement.
First, we divide the test object into unit test, Interface Test and function test.
Unit testing refers to function-level testing. Commonly used tools include JUnit. API testing refers to the testing of Apis exposed by components or modules. Function Testing refers to end-to-end function testing. The following is a comparison:
Google testing vs traditional QA
|
Unit Test |
Interface Test |
Function Testing |
Google |
Strong |
Strong |
20-40% automation, other outsourcing |
Traditional QA |
Weak |
Moderate |
Strong, large |
In terms of efficiency and quality, Google's practices are more in line with the current situation of Fast fish and slow fish in the Internet age. Where is the greatest difficulty in the transformation of traditional QA? The following are common problems: 1. A large number of legacy code without unit testing. First, mock is very difficult, and the workload is huge, so it is difficult for the management to make up its mind. some developers resist unit tests because they prefer to write functions. testers are cold-eyed because their main skills are accumulated and confidence experience is in functional testing, and there is no trust in the success rate of unit testing and interface testing.
To achieve transformation, a large number of training and attempts are essential. The determination and perseverance of the management layer are often the key to success or failure.