I need to write a test framework from scratch. My boss (who isn't an expert on testing or Java) required to use Modules, reducing reliance on the classpath.
To be honest, I also have the urge to try and use the newest, most elegant and theoretically correct solution (overly rigid nature of my character). I mean, it's really weird to
see some guy show up with a project using an SDK from 2009 and no interest about what happened in Java in the last 10 years!
That's also the problem: I can't find guidance or examples online on how to build Selenium+Cucumber projects using features newer than Java 8. Actually, all such projects
use =< 1.8 compiler and the related deps. When you code in it, it's all "Deprecated" and god forbid if you need to introduce a new package, you'll be fishing for ancient deps
Anyway, could you please share some ideas from projects you've worked on, or just plain theory from your beautiful mind, about how to leverage on modules in a
Selenium-Cucumber (with page object models and JUnit/TestNG) test framework?
Diagnosed with autism. Please excuse me if I miscommunicated anything.
So is time and money not a consideration at your place of business then? Because what you and your boss are proposing sounds like a colossal waste of time and money. Unless, of course, you're in the business of writing testing frameworks for maybe 5% of all Java users.
Most of the testing frameworks in use today are open source so it's not hard to find where they are hosted. Where is JUnit source code
Before committing to the undertaking, I suggest you look at what you might be getting yourself into. Yes, JUnit was originally written by just two people (Kent Beck and Erich Gamma) on a trans-Atlantic flight, meaning in less than a day. That was merely the spark that lit the forest fire though. Countless hours have been contributed to that product to bring it to the state it's in today, roughly twenty years after that first back-of-a-napkin version.
The comment about a waste of time and money, while admittedly a bit blunt (sorry if that came across wrong), was meant to question the practicality of such an endeavor given that there are many other people who are already way ahead of you on this. If you are going to go ahead and make the call, at least consider forking an existing open source project that already has most of the features you may be thinking of providing in your framework anyway.
PS: we are a "profit with a purpose company" and we definitely look at things differently.
The article looks interesting, thanks for sharing the link to it. It's great to work for a socially responsible organization (I'm assuming that's what is meant by "profit with a purpose") and I wish you luck and success in whichever direction you eventually choose to go.