Comments on: AngelXNA v1.0

By: Intelligent Artifice / Back from Game Forum Germany 2011

Intelligent Artifice / Back from Game Forum Germany 2011 — Sun, 30 Jan 2011 22:50:22 +0000

[...] and best practices in data-driven development. He had shown me some of what he’s done in AngelXNA and it’s definitely a cool [...]

By: Playing with Prototyping Tools « The Toolsmiths

Playing with Prototyping Tools « The Toolsmiths — Mon, 28 Sep 2009 00:25:51 +0000

[...] evaluating the BoyLib framework, Jeff Ward suggested that I compare it to Angel and AngelXNA [Jeff on AngelXNA], both of which were new to me. This is also read through the lens of my current efforts to [...]

By: Jeff

Jeff — Mon, 04 May 2009 00:48:46 +0000

So, memory leaks aren’t really an issue unless they’re so big they cause your program to crash, but memory trashing 90% of the time causes crashes, and the other 10% of the time causes really weird behaviors. 100% of the time, those bugs are REALLY hard to diagnose and remove, and if you’re doing something for a game jam, where the end result is a demo, this can rob you of a lot of time.

By: Marc

Marc — Sun, 03 May 2009 22:30:53 +0000

“C# allows me to program faster and worry less about things like memory leaks, memory trashing, and weird side effects. ”

But why would you care about memory leaks and memory trashing in a small prototype?

By: SJML

SJML — Fri, 01 May 2009 18:54:17 +0000

This is something we wrestled with a lot when designing the original Angel. It’s a tough line to tow, but when it came time for any decision of “should we do it X way or Y way” we used the razor of “What would make things easier on someone who has no idea how the engine works?” That doesn’t necessarily mean only doing things simply, but it means hiding the complexity as much as possible and providing nice handles to the user.

In the initial architecture, I was inspired (in spirit, if not in style) by Rails, with the attitude of “do what 90% of people want to do by default.” If people want to delve deeper, we did our best to document things and make the code as clear as possible.

Of course, that means sacrificing some efficiency and doo-dads. Angel doesn’t use a scenegraph of any kind, and thus would be hard to scale to massive amounts of Actors. Including shader support would also be hard. We only support one type of controller, but hopefully it’s the kind most people have access to. At the end of the day, it’s kind of freeing building something to *just* be a prototyping engine, because you get to optimize for understandability instead of framerate or ultimate flexibility.