I applied for a job based on a friend's recommendation. The recruiting team called me the next day. We had a conversation lasting around 30 minutes about my previous workplaces and responsibilities. At this point, everything seemed to be going as expected.

As a next step, I received two online tests to screen my core Java and web development knowledge. Initially, I was a bit skeptical. How would they check my knowledge with a test? I could look up anything on the internet. My next thought was, why use such a test in an online form in the first place? This question kept coming up during those days, and I still don't know the answer. To keep the story short, I gave it a try and started with the core Java test.

My skepticism turned to shock when it turned out that I had to write "program parts" on a webpage with a "compile now" button. The webpage screen was divided into two parts: on the left was a description of the task, and on the right, a large text area with a scrollbar for longer code snippets. The test contained four tasks, and I had one hour to complete them, allowing 15 minutes for each.

The tasks were easy. I had faced harder ones at university during my Programming 1 classes. But to be honest, I couldn't remember the last time I typed a "for" loop, an "if" statement, or a getter/setter pair myself, without using IDE hotkeys. Then, the solution appeared in the form of a hint: "Copy the task to your favorite IDE and then copy back the solution." Thank you, wise one, I will do that. This is where the story shifted from drama to comedy, and I started to have real fun.

Before I begin discussing the tasks, I must say that I cannot reveal the exact tasks because I checked the box to keep the test secrets, and I am a person of my word.

Task 1 - Make this code work.

The solution required writing four to five classes, mostly boilerplate code in one file that also contained a main method. I violated my own principle, the vanguard of quality code, and completed the task in my favorite IDE. Unexpectedly, the code compiled in my favorite IDE did not work in the less-favored web environment. Luckily, the web tool provided some error messages. You can imagine that the problem was not something seen every day, so the fight against time and the environment began. Change something, click the shiny compile button, and hope for the best. Repeat this until the red turns green. I hope you've figured out that this was not the end of the ordeal. As an extra surprise, one of the classes in the task was not something I should have written, but rather a part of Java itself. This wasn't a big deal, but in the web environment, the imports were handled somewhat invisibly, leading to a never-before-seen error message complaining about it. A good engineer never gives up when obstacles emerge, so I finished the first task in 25-30 minutes with a quite high stress level.

Task 2 - Playing around with char to int type conversion.

I admit, I had never worked deeply with characters before, so I started with a quick investigation on the topic and wrote a prototype solution, which passed only half of the test cases. I had already spent over 10 minutes on research and implementation, so I decided to move to the next task and not submit my partially working solution.

Task 3 - 15 minutes left.

I really liked this one. It had a small twist that makes a task a good interview question in a personal interview. My first solution was ready in five minutes in my favorite IDE. Of course, it wasn't working in my non-favorite web IDE, so I had to make some small adjustments. When you work in your favorite IDE, the sheer happiness produces so much endorphin that time flies on light wings. Only the sight of the red flashing clock could bring me back to reality when copying my solution back from my favorite IDE to my non-favorite IDE. At least the code compiled. Then I discovered the slight trick in the task. You cannot sell something before you buy it, right? Five minutes of refactoring in my favorite IDE plus some tests because I did not want to fail again. Then, copy the code back to the correct place. Every time, I was getting better and better at this. But there was no time to compile; I moved my mouse pointer over the submit button and... time was over. Unfortunately, there was a typo-like issue in my solution, so it would not have worked anyway. I fixed it in my favorite IDE simply because I dislike incomplete things and then leaned back in my chair, relaxed. At least I don't have to copy anything into my non-favorite IDE again.

Asked to implement basic tasks in an unusable online programming quiz.