so xna...

[nordwindranger]

I got started with c# game development with tom miller's books, and have been lurking this site for quite some time.

As I already have a small fortune in "c#+dx9" books, and have no real interest in the xbox 360, can i keep doing what ive always been doing, or has xna killed regular c# game development. Honestly Im really not at all interested in switching over to something new, as Im just beginning to understand conventional c# and dx9 development.

has xna killed the future of regular c# game development, or is it a seperate entity that I can happily ignore?



thanks

andrew

[Chr0n1x]

XNA has not killed it, firstly XNa is still in beta, so it is still going to go through some changes. Secondly the semi-required content pipeline hasnt been released yet, and thirdly the documentation is still kinda sparse. So for now stick with MDX1.1 and learn throught that, then when XNA comes out properly, you can easily migrate to that, and you will have skills in both!

[SeeMe]

Hello,?

At first glance, even if the XNA framework is in fact "MDX2", the concept behind him is really close to MDX1.
But
Here are my personal view on the XNA framework in its current state :
XNA +Side
- All the Device creation/handling is completly done begind the scene by XNA, it really well done !
- Same thing with the Game loop, completly hided.
- DirectInput managed

XNA -Side (compared to MDX1)
- You have less freedom, mostly concerning the "batch control" of your application.
- No possible anymore to use pointer to quickly move data around like in MDX1 or direct write acces to Graphic card memory (GraphicsStream class).

In fact, If I could resume the global though on the microsoft forums around XNA : It's already a really nice piece of framework, but for XBox security reason the low level acces to the device have been removed at the risk of a lower global performances.

I'm working on an engine myself, writen in MDX1, all the managed behind the scene things like device manager, game loop, directinput are inside my engine, so I have for the moment little, or no benefit to go to XNA. Btw I tried to transform a part of my current engine in XNA, the performances are good until the number of batch raised where you can't avoid the "behind the scene" Lock/unlock after each xna draw call. The only solution around this would be to create in XNA a "Cache" buffer, where when you need to render thing you send to data to the cache, and when the cache is full you flush all the data to the device with only on draw call.

Let's see how the framework is advancing !

[BrophyJ]

More downsides to XNA - all of which are attributed to XBOX compatability:

* Will be stuck at DX9 technology.  Newer versions of DX will appear for the PC (although apparently only for Vista or newer versions of Windows) but will not be available for XNA games.

* Will not support any input devices beyond Keyboard, Mouse, XBOX controller.  No joy sticks, flight simulator sticks, etc.