At one place that I used to work, The Powers-That-Be decided that 1) we were gonna be a Java shop, and 2) we (the programmers) were all gonna be certified in 6 months.

Apparently, I was the only one who read Sun's suggestions for the test.
1) it shouldn't be attempted without at least 1 year realworld experience. (we had 6 weeks of classes, good classes, but 6 weeks)
2) it was going to show you code presented in the worst and most confusing way possible. (great)
3) the first thing you should do is figure out if the code would compile. (isn't that what I've got a compiler for?)
4) the next thing you should figure out is if it will throw a run-timer error. (isn't that what testing is for?)
5) THEN you should figure out what the code will do.

>
>I've listened in on a similar "interview" on Java, which was full of questions totally unrelated to actually getting things done, but rather oriented to knowing the formal differences between this and that, or "will this compile?" questions. Who cares? Formal differences matter when you write a paper, not as much when you write code - in the latter case, it's the different behaviors that do. Will it compile? Easy to check, that's what we have compilers for.
>
>It actually gets me deurinated each time when someone expects a programmer to do this or that so not to irk the compiler. It's a stupid piece of code on a stupid machine, let it do its job, don't force your programmer to do machine's job. He's got better things to think about.