On the 30th of July, Angie Jones will spend an hour on The Club tapping away answering any of your awesome questions related to Automation. Ask your question before 7pm UK Time to get it answered!
Hi, Angie here! Let’s talk about test automation. Happy to answer questions on:
Test Automation Strategy
UI test automation
Visual test automation
API test automation
all the things test automation!
Go ahead and reply with your questions before 30th July at 7pm UK Time, and I’ll answer as many as I can in the Power Hour. Talk with you soon!
How can I best raise my organization’s understanding of the value of test automation? Are there specific steps I can take or helpful reporting methods? Today, we have a fully functioning UI test suite and I’m working toward an API test framework. This was all done with the direction and blessing of the company. However, I’m allowed so little time to work on automation that it’s extremely difficult to move forward.
A few months ago when I was on my 2nd day of a new contract, I had a meeting with my new manger. He wanted me automate EVERYTHING regarding a new mobile app that was coming to market (including edge cases) in react-native (with a detox and jest framework)
I told him ‘you can’t automate everything’, but we can automate on a business critical criteria and I showed him your (angie) automation criteria framework. He loved it, but got really annoyed when I put in a criteria ‘don’t test’.
So my questions are:
How do you deal with a manager who such unrealistic expectations ?
Hi Angie, for a company with no automation other than unit tests, what would be the first few steps you would take in identifying what to look at / focus on / investigate first? Thanks
What’s a recommended way to organize an API testing framework? Are there recommended standard design patterns for API testing the way there are for UI tests with page objects? (just for context- we’re currently using php/phpunit for api testing).
I sometimes use Record and Playback to create automated tests quickly, but always adapt them afterwards to make them more maintainable and robust. I also find that Record and Playback is a great way to introduce new testers to test automation development (but make sure they are aware that we must not rely solely on Record and Playback).
What are your thoughts on the use of Record and Playback tools for test automation development?
Do you have any recommendations for how to get some automation set up for a vendor application without adding additional staff? Our organization spends vast amounts of time testing (and because we have a complex application with custom development and unique business processes we find plenty of critical issues). The body of what we must test keeps growing, and no additional testing resources. We only have one full time tester.
I use selenide as a java based selenide wrapper to develop my automated checks. For me as a non-developer it’s really helpfull. But as I develop more and more checks, code duplication increases. I try to refactor some code pieces i.e put generic stuff into separate classes.
But where should I put such generic classes? For example I have a CustomWebDriver class. Does it make sense to put this under /src/main/java or under /src/test/java for example as part of @Before(Class)?
And is there a commonly used structure where to locate such generic classes? As I have seen examples where they have a utils package under src/main/java.
Thanks,
Mark
Btw keep up the great work you’ve done on Test Automation University
Hi Angie! We are moving to a microservices architecture and the number of APIs to test is growing rapidly. When we just had a few, manually using Postman was manageable. Do you have any advice for an automated API testing framework that can grow into the thousands?
Amazing work with Test Automation University. Huge thanks again your way.
Quick question. We have a Java Selenium automation framework that was built years back back and it has never been code reviewed by any developers. Seems everyone thats worked on it has done something different interms of how they’ve structured their code. So
1 - What would you say is the best approach to refactoring it? I don’t believe are team have great automation testers, our manager knows Java from a Computer science background. Not much involvement from someone with a developer background with our framework. How would you work with legacy framework?
2 - We have one testing repository where all the automation code for the majority of our projects go. Each project is different. One issue we have is people change things eg upgrades versions or shared utilities that works for their project but breaks for other projects. We were initially thinking of having a core repo for shared code and seperate repo for the different projects. How would you tackle this? What are you thoughts around this problem?
I’ve listened to Angie provide some great advice on risk factors.