Dealing with one of those. You know the guys. 4-6 years of experience and gets some feedback from a user about a bug they swear doesn't exist. I've seen the bug personally happen, and yet, out of fear I think of being seen as bad, they refuse to admit it. How have you dealt with this and what can I do? In the past this person has reacted almost violently to suggestions like this, using company resources in an attempt to persecute. I can't get them fired, but I can teach them a lesson.
I've seen a number of this type of developer over the years. They're always right, no matter what, and arrogant about their ability.
In truth, there is only one way to deflate them, and that is to prove that they're wrong.
...a user about a bug they swear doesn't exist. I've seen the bug personally happen...
If the bug is seen, then you need to reproduce it. Prove that the issue exists and make sure the steps to do so are well documented in your bug tracking system. This will firstly formalise the bug, and secondly make it apparent that the developer isn't quite as perfect as he may seem.
his person has reacted almost violently to suggestions like this
Have you feared for your personal safety? Have others? If this person is acting in an unprofessional manner like this, then it needs to be flagged with your manager. Keep records of when and where, under what circumstances. This is important as this evidence may be necessary at a later date if an incident occurs.
Really, the best advice I can give is to just track everything, both of bugs being denied and bad behaviour. If he's arrogant for no good reason, then he'll come unstuck. If he's got behavioural issues to go with it, then you absolutely need to document every time you are fearful and why.
On one hand, there are bugs that are very hard to reproduce.
On the other hand, with 4-6 years he is a newbie, and I know from experience that whenever I was sure that a bug couldn't possibly exist, it did. There's the first rule of debugging code, which apparently he hasn't learned yet: "Whatever you know about this bug, it's wrong". That applies first of all to his statement "this bug cannot possibly exist".
In many cases, reproducing the bug is the hard bit. Once you know how to reproduce it, the fault in your code is often either obvious or very easy to find. So if he claims that a bug isn't possible, tell him from me that he just isn't a good developer.
And there's the second rule of debugging: "The problem is there because there is something wrong with the code". That's an absolute essential. People who don't understand this rule waste everyone's time including their own by claiming a bug is not possible, it is a hardware problem, it is a compiler problem, it's just gremlins that don't like you, instead of concentrating on finding and fixing the problem.