You know what? It is to the point where I don't want to be neutral anymore. I asked several times in my previous question to ask the answerers to clarify what they meant, and they didn't reply. Fuck you you dare to say they replied to me. They talked among themselves about micro optimization. I asked several answerers and they never responded.
Okay, since you're clearly talking about me, let's look at this.
The thread is here, for anyone interested in checking for themselves:
http://www.gamedev.net/topic/681032-which-way-would-have-better-performance/Your original post has a direct answer from Hodgman which directly addresses the question you asked. That alone makes this entire rant empty and meaningless. You got exactly the answer you wanted. You're angry because you got more than that.
Further down, I replied pointing out that it was a meaningless comparison since the 2 methods don't do the same thing.
You then asked one person what "readable/ understandable code" meant, which is literally the only question you asked that didn't get an answer. Every single other post you made, I responded to directly.
Yes, we discussed more things on the thread than your direct original question. This is for several reasons:
1) It's a discussion forum; we discuss things. You start the thread but you don't own the thread. It lives on for the benefit of other people in future who might have the same question. This means some posts on the thread won't be a direct response to you. Deal with it. If you want a strict Q&A session, StackExchange exists for that.
2) I'm here to help people finish games, not waste time on things that don't matter. The calculation asked about doesn't matter except in some extremely exceptional cases. I wanted to make that clear.
3) Hodgman's answer had several aspects that were worth exploring, such as the use of length-squared vs. length, what would be fastest if you were writing shader code vs. application code, and the issues that should guide how to write your code given that "there wouldn't be too much difference between them". People discussed those aspects for everyone's benefit. This is especially relevant since you provided no context for your question.
And this one guy with an attitude came to my question saying something that is completely irrelevant to my question, and later claimed I stressed micro optimization. Fuck that guy for saying that. I only asked a question to know which method is faster.
What I said was entirely relevant to your question because they are not the same operation. Sphere tests measure against a distance X. Cube tests measure against distances between Y and Y * 1.866 depending on how the cube is oriented relative to whatever approaches it. And X doesn't necessarily equal Y. The cube method also involves up to 6 conditionals which means performance is dependent on how many of those you regularly trigger, and what order they're in. To answer this correctly requires more information, which you did not seem interested in providing.
Look at the all the other questions. You will never see anyone with an attitude go their questions.
You're right, I never answer anyone else's questions. (Apart from the
hundreds, maybe thousands, I've replied to since I started here 16 years ago.)
I sincerely apologise if you feel that my responses upset you, but you need to accept that we're here to discuss things, that sometimes our assumptions will be questioned, and that people will come into a thread with different opinions. As GrumpyOldDude said, we're not paid to mentor you. Everything you get here is free. Take what you can from it, and graciously ignore the rest if you have to.
