Why the process of software testing is evolving?

Hello folks ! Yes, software testing is moving towards the future! 
And recently these positive developments are going very fast too.

​The number of blogs and articles with excellent​ content on Twitter or LinkedIn recently was really eye-catching! Thanks colleagues and other test professionals.
The best article I read this week had a subject that was -in my opinion- a really progressive way of thinking about software testing. ('checking and exploratory work are not antithetical').

​So why is it that I give this blog the title:  'software testing is moving towards the future!'
​

​​Reason 1:​ Test automation tools are getting better by the day. I've seen more cool features, new functionalities, wider support for more devices. Etcetera. But in general, the quality of test automation tools is improving, and it is improving FAST!

​Reason 2: ​There has been a long couple of years when the market was asking for a certain type of 'generalist'/'all round software tester'.
​But, recently the recruiters are asking better and more specific questions.
New companies supplying CV's of test candidates are ​visibly showing on their websites that they know that a security tester is not the same as a performance tester! Which in my eyes is a very good thing. We need test specialists in many areas.

​​Reason 3: ​(And this is the best reason). There seems to be a growing positive insights (at software building companies) that a computer(tool) will never test in the same way or find the same bugs... as humans.

Now take a closer look at these 3 scenarios:

A) ​Only a human will test the software. This scenario would have a certain effect on the type of bugs that will be found, the amount of bugs found, etcetera. 

​B) ​Only a test automation tool will (re-)test the software. 
​If you'd apply this scenario then this would also have a certain effect on the software bugs that will be found.

​C) ​A human ​and the test automation tool will (re-)test the software.
​This will also have a certain impact on the amount of bugs, the type of bugs and other aspects.

So are these the only 3 ways! Of course not! the possibilities are endless. What to think of these other scenario's:

- let a human test (18 years old),
then let another human test. (85 years old)
​
​- Let 2 humans test at the same time and compare the results.

​- Let the computer test and then let the computer test again.
​
​Be creative and you can go on-and-on...

​​​​

What is it I'm trying to say with the example above?

​What I'm trying to explain is: (metaphorical way of explaining)
There have been some years when people wanted their news, and liked to read it in the morning on the tabled, Phone or computer. 
​('Everyone wanted the latest test automation tools.')

​But lately there is a development going on that reading the news in the newspaper wasn't such a bad thing after all.

​There will always be books.
('We will always need the opinion/insights/judgement of a human when testing software').
​(regardless how advanced the test automation tools will ever get).

​
​Now, here is why you wanted to read this article! (Here is where it gets fascinating!)

Remember I wrote in this blog that I read a blog/article with the content 'checking and exploratory work are not antithetical?'

The reason I think this article is most interesting, is that you can apply that thought/vision/insights on the A, B, C, picture that I made.

​
​So, *​hypothetically you could say (*if they had been antithetical):

• ​Let's put the computer to mainly check, and let the human focus on explore.

or

• Let's put the computer explore, and let the human check+explore.

or

• ​Let's put the Artificial Intelligence software 'check and explore' and the human also 'check and explore'.

or even:

• ​Let's only put the Artificial Intelligence software 80% focus on check and 20% on discovering new things. (or the other way around).

(Imagine this picture next to the upper one:)

I hope this blog inspires you to think about software testing this way in a wider/broader vision.
​Think in terms like: 'What would be the effect if this was the case!'.
 

​​​​Every of the above given scenario's of testing the software would have a different effect on what and how it will be found!

​The next step would be:
​How to decide what the best approach (*scenario) would be?

(*Is it wise to say to a human : your goal/focus is strictly: 'check on the 4 requirements' and dump all other information!/findings ?
Probably not very wise, right?)

And
​How could we best implement/apply this choice in the test process (Agile process)?

​Because if every choice has a different impact on what, how, how many bugs, etc will be found... Then probably every chosen scenario will have a different kind of effect on the ​Product Backlog And /or ​Sprint Backlog. ​(and thus the test throughput time).

​(Stay tuned because my visions on (how best to cope) that part will be soon online too!) twitter.com/testensoftware

​