I got really excited as soon as I saw Mike Veinder's post in the Selenium LinkedIn group about an upcoming event: Future of Testing Frameworks.
There will be presenters from Selenium, Applitools, Testlio, Shopwareag, Microsoft Playwright, Q2, https://www.linkedin.com/company/trendig/, Slido, KPN, Serenity Dojo, Smartcode Ltd and last but not least Testguild.
As the online event starts in 15 minutes, I decided to formulate my takeaways into an article, so I can share it with you, even if you missed the conference. 😊
Well there were some troubles. First of all I misread the time and the event started one and a half later than I expected. That's might fault. But the event did not start at the provided time, I tried to reach Applitools via chat, via email, and finally I got the info they have technical issues. Luckily after a few dozens of minutes they were able to solve the problem, so the presentations could start.
I won't mention all the presentations, only the ones I found useful at my current stage. It does not tell anything about the presentations, it is only showing my personal preference.
I brought 3 presentations and the final closing remarks to you this time.
So let's start with the first presentation.
The event is hosted by Applitools which is a company creating an Automated Visual Testing tool that uses Visual AI.
Kristel Kruustük (Testio) started the show.
There is a new methodology Fused Testing which is a combination of automated and manual testing. Kristel started as a manual tester where they had to deal with always increasing customer demand.
Kristel started with her story. In 2008 after graduating from high school she has no idea what to do, thanks to her sister she started to learn software development. She did not feel confident enough to start as a developer, so her friends recommended she become a software tester. Her plan was to start as a tester and then transition to become a software developer.
She became a member of a very small agile team. Her first task was to find bugs, but she did not receive any support from it. She examined everything using the exploratory technique.
2012 she graduated from the university and moved from Estonia to London. She just quit her second job as a manual tester, and she decided to become the best tester in the world until she realized she can make work better for all testers. She decided to do her own research. She developed a very deep passion for testing.
After many interviews, she started full-time as a manual tester at a fin-tech company. She got quite a huge freedom at this job.
She had the opportunity to move to automation. At that time she was the only tester in the team, she had no one to learn from, so she started to reach out to experts on the field.
In order to be good in anything, there is only one way, practice.
She got quite frustrated after a few months as the testers were only compensated for the number of bugs they found. With her mindset of quality is everybody's responsibility she did not like that always the testers were blamed for any
AngelHack to build a better platform for other testers around her. She did market research and found out testers need a platform, where they can communicate.
She started her journey with Testino at that point.
Testio is the originator of network testing. It is a philosophy or methodology o
From a team of two, they grow into a more than 100 people company. They had 80 full times employees.
They started Testio as there was a market gap. Her main goal was from day one to improve testing. 2 billion customers use the products they test.
CBS, Stava, Microsoft, Netflix Hotels.com, CBS and athenahealth are their biggest customers.
Thanks to Testio 84% of developers releasing code faster than before.
According to her view, humans will always have a place, some tasks cannot be automated.
They came up with the fuse testing methodology which is a combination of automated and manual testing. Combines the skills of manual testers with the efficiency of automation. It is a combination of shift left and shift right.
Sometimes we do not know whether we found a bug, or we have a failure in our tests. This is where manual testing comes into the picture.
Decreasing Your Design System’s Visual Time to Test - Nina Westenbrink
- She lives with her boyfriend and 2 cats in Amsterdam. One of her hobbies is reading books. One of the book she recommends is Girl, Woman, Others.
- She likes adventure games and RPGs. The first game she played is Indiana Jones which she still has on 10-12 floppy disks.
- She has a vegetable garden.
- She has snowdrops in her garden, she is also frying bees.
In 2012 she started her software engineer career. There were many devices on the table, when she asked them what are they used for, they told her they are using them for testing.
They had one full-time tester. Developers were also responsible for testing. Testing needed a lot of time. They were not able to test the impacts of all the new features.
If you go back 6 years' time can you tell have much time it took to test a feature on all devices?
Well, it took 2,5 hours considering all device and test environments. In case not everything worked well, it took more time.
424 days ago
They have replaced the devices with virtual ones and emulators. They used other tests tools, they usually checked the rendering of the component, but not the style. With this infrastructure, the testing time was reduced to 2 hours.
422 days ago
She joined a new company. In that company, the test of a feature took 1 hour on average.
A design system is a collection of reusable components and rules which can be used by other teams in their digital environments.
Designers design the components in Sketch and developers converted them to storybooks and convert them to packages so other teams can reuse them.
Has your company got a design system? How many teams are using it?
August 2021
She is regularly organizing knowledge-sharing events. One topic was Aplitools which was very powerful and easy to use. Using Applitools they were able to reduce the testing to 3-4 minutes.
Lessons learned
- We are all in this together
- Choose browsers and devices on which you want to test before you start testing
- Create sustainable story names considering testing
A Journey to Better Automation with the Screenplay Pattern - Sarah Watkins
Sarah Watkins started her presentation by introducing the Page Object Model (POM). She also stated one problem with POM is not having an official version, it is more like a convention.
Immediately after this, she followed up with the Screenplay pattern which has greater usability and scalability. I really enjoyed her presentation, learning more about the Screenplay pattern is definitely the main takeaway for me from this event.
Sarah introduced the Boa Constrictor which is a .NEt implementation of the Screenplay pattern developed by Andy Knight.
Her main takeaway is: Actors use abilities to perform actions!!!
She provided some C# code examples:
%[Actor (name, logger)
actor.Can(BrowseTheWeb.With(new ChromeDriver());
actor.AttemptsTo(Navigate.ToUrl(SearchPage.Url));
actor.AskingFor(ValueAttribute.Of(SearchPage.SearchInput)).Should().BeEmpty();
actor.AttemptsTo(SearchDuckDuckGo.For("Evee"));
actor.WaitUntil(Appeariance.Of(ResultPage.ResultLink), IsEqualTo.True));]
She suggested for Screenplay pattern using Java and JavaScript you should check Serenity BDD.
Expert Panel: A Fireside Chat with Popular Framework Advocates
I found this part the most insightful. Ramona, Pavel and Simon were talking about the current biggest challenges in test automation
Ramona
The complexity rising complexity of everything in tech. Feature code, testing code. A good challenge for all of the frameworks is to take away the effort and make testing easy. Prioritizing test cases is very important especially as the execution time of tests grow continuously. How can you quantify quality?
Pavel
Writing tests are easier. It brings us to a world where we pile up lots of tests. More tests need more people. Shifting left is happening. Developers like writing tests, they only do not like to maintain them. 😊
Simon
Parallel running is very important. Capacity is a problem. People when get the golden hammer (automation framework) consider everything as a nail. Proper unit tests and smaller integration tests, fewer and fewer E2E tests. The biggest challenge is getting developers to write tests. 😊
Sit and work with the developers and do pair programming.
I missed the last few minutes as I took part in István Marhefka's Carrier Compass meetup's group mentoring session lead by István.
Summary
To sum it up I really enjoyed the closing remarks and the journey of Kristel Kruustük and Sarah Watkins's speech about the Screenplay pattern, but all in all, I was a little bit disappointed by this online event. First of all, Selenide was not on the list.
Why should Selenide be there?
Selenide is an E2E test automation framework built on top of Selenium WebDriver having the same functionality and much more. It has easier syntax, advanced selectors, no more StaleElementReferenceExceptions, built-in profiler and more.
Maybe next time.
The other cause of which I was said as I felt the title "The future of testing frameworks" was a little bit misleading. Well, there was almost nothing promising about the future. The biggest improvement which was mentioned is Selenium 4 which was already released 4 months ago. I do not how you are with it, but for me, that is the past, not the future.
I expected to see AI appearing more in testing frameworks. I know Applitools use AI, but still, for me the greatest advancement in testing frameworks in the last months was Healineum: https://mszeles.com/the-future-is-here-healenium-a-self-healing-selenium-test-automation-library
I hope you enjoyed reading this article. 😊
UPDATE: You can find the presentation videos here.
Have you seen the presentations? What was your biggest takeaway?
What do you think is the future of testing frameworks? I would be happy to hear your thoughts.
Share this summary if you found it useful.😊
#FutureOfTesing
