Yeah - using ob_start() is like pinpointing the buffering?

I'd like to know if whitespace was the actual cause or not. Is whitespace the only culprit?

Yeah, but Javascript redirects don't work if the client has Javascript disabled.

Just an FYI in case anyone has issues.

You may run into issues in Visual Studio 2010. Yes, you can put your resources in the same directory as your executable, but that means you have to update your Debug and Release directories every time you add or modify resources.

So, in the project, I place my images into a project directory - and it doesn't matter what build I'm using. Now, I bet I'll have to copy my resources to the executable directory if I want to distribute. But my point is that the references to resources in the project can be problematic in Visual Studio and to be mindful of it.

No, it's not an art question. It's a programming question.

Ok - so the answer is, "Color Keying is faster".

I'm just now getting into SDL. I want to be able to display as many sprites as I would ever want on screen at any given time. I just noticed that displaying transparent .png images used a different function and figured there had to be a performance difference.

I'm not working on any specific game right now, just learning to manipulate images and surfaces and pocketing things away in my own functions and classes. I thought it would be fun to start with something like a Galaxian clone and learn how to make neat looking visual effects that classic arcade techniques used.

So, this is good information. Normally, I would want to use color keying, but I can always use .png alpha transparency for cool effects like glass or other transparent effects. Meh, I dunno. If I'm replicating the same image over and over again, like lasers or whatever, it may not matter I suppose.

No, I think you're missing the point. This is not a matter of looking up controls and implementing some classes. I know what Google is. Every inch of the way, I've tried to implement common web development techniques and .NET seems to not want you to do anything that's not already built into a control. Things I can do in 20 minutes manually are seemingly impossible in .NET.

Session timeout is a good example. I had to notify the user when the session would expire - a very common implementation. I looked to find a .NET mechanism for this. All I got from the .NET community was, "Why would you want to do that??". So I had to write my own Javascript anyway. And it's like this for EVERY little thing. Between the Script Manager and Viewstate exploding - I can't deal with this.

I'm not going to fight with .NET for the rest of my career. I need to know how the Black Box works.