• Chapter1Game
• Xna Pong
• Xna Breakout
• Xna Tetris
• Rocket Commander Xna
• Xna Shooter
• Xna Racing Game
• Dungeon Quest GDC

• Chapter1Game: This application is not really a game, but a test project to check out if XNA 2.0 is properly working on both Windows and the Xbox 360. It is from the first chapter of my book "Professional XNA Game Programming". BTW: The second edition of the book is coming out soon, there are 3 more chapters about Multiplayer game programming and a cool new role playing game.
  
  
  • Game Installer: Chapter1GameSetup.exe (210 KB)
  • Source Code and Game Content: Chapter1GameSourceCode.zip (57 KB)
  • Screenshot:
    
    
    
• Xna Pong: Xna Pong is a simple clone of the favorite pong game from 1978. It is just a few hunderd lines of code and should be very easy to understand.
  
  This game is from the book "Professional XNA Game Programming" by Benjamin Nitschke. For more information read chapter 2. (2008-02-10: Now updated to XNA 2.0)
  
  
  • Game Installer: XnaPongSetup.exe (309 KB)
  • Source Code and Game Content: XnaPongSourceCode.zip (258 KB)
  • Screenshot:
    
    
    
• Xna Breakout: XNA Breakout is a simple Breakout/Arcanoid game based on the XNA Pong game from the previous chapter.
  
  It is fully described and covered in Chapter 3 of my book "Professional XNA Game Programming". The code is quite short and should be easy to understand. (2008-02-10: Now updated to XNA 2.0)
  
  
  • Game Installer: XnaBreakoutSetup.exe (383 KB)
  • Source Code and Game Content: XnaBreakoutSourceCode.zip (316 KB)
  • Screenshot:
    
    
    
• Xna Tetris: This is a simple, but highly addictive Tetris game. You can control the blocks with your cursor keys, aswd or a game pad and the game works both on Windows and the Xbox 360. Reaching levels above 5 is really hard. My highest level was 9, try to reach more :) (2008-02-10: Now updated to XNA 2.0)
  
  This game introduces the helper classes (chapter 4 of my book) and makes more use of unit testing and game components in XNA.
  
  
  • Game Installer: XnaTetrisSetup.exe (310 KB)
  • Source Code and Game Content: XnaTetrisSourceCode.zip (277 KB)
  • Screenshot:
    
    
    
• Rocket Commander Xna: XNA port of the famous Rocket Commander game. The game principle stayed the same, but the controls were a little bit simplified to make it more fun on the Xbox 360.
  
  If you want to learn more about the Rocket Commander game, check out its official website www.RocketCommander.com and check out the Video Tutorials on Coding4Fun by MSDN. (2008-02-10: Now updated to XNA 2.0, also supports very big resolutions now and runs faster on the Xbox 360)
  
  
  • Game Installer: RocketCommanderXnaSetup.exe (16.5 MB)
  • Source Code and Game Content: RocketCommanderXnaSourceCode.zip (52.8 MB)
  • Screenshot:
    
    
    
• Xna Shooter: Shoot'n'up game specifically created for my book "Professional XNA Game Programming". It features full HDTV support, runs on Windows and the Xbox 360, 5 weapon types, 5 enemy types, a powerful ship and some power ups. It is quite fun to play and it gets harder and harder the longer you play. Based partly on the Rocket Commander XNA engine, but also features lots of new effects and shaders. (2008-02-10: Now updated to XNA 2.0, also much easier and balanced)
  
  This game and the racing game are the most improved. The game works now much better in high resolutions and on the Xbox 360. But most importantly the game is now much easier, balanced and more fun. Additionally a level percentage is now visible on the bottom and more EMP bombs can be picked up to make it easier at the end of the level.
  
  • Game Installer: XnaShooterSetup.exe (14.9 MB)
  • Source Code and Game Content: XnaShooterSourceCode.zip (32.2 MB)
  • Screenshot:
    
    
    
• Xna Racing Game: XNA Racing Game Starter Kit I wrote for http://creators.xna.com. More information and more downloads can be found on http://XnaRacingGame.com. It runs best on the Xbox 360 in HDTV (1920x1200), but it also runs fine on the PC. (2008-02-10: Now updated to XNA 2.0, driving also improved a lot, better tested on Xbox 360 and fixed some issues).
  
  Following things were improved: Shadow mapping on very big resolutions works now (crashed before), more options for lower quality settings, fixed physics, car now always stays on the road, fixed loopings, cars are much faster now, winning conditions work better now, and fixed several other bugs.
  
  • Game Installer: XnaRacingGameSetup.exe (40.2 MB)
  • Source Code and Game Content: XnaRacingGameSourceCode.zip (95.4 MB)
  • Screenshot:
    
    
    
• Dungeon Quest GDC: And finally the Dungeon Quest XNA Game, which was developed in just 4 days on the GDC 2007 at the XNA Contest. Dungeon Quest GDC is a relatively complex 3D role playing game (at least for just 4 days of work). An early version even supported coop multiplayer on the Xbox 360 via splitscreen. The game was developed by Benjamin Nitschke (abi.exDream.com) and Christoph Rienaecker (WAII). (2008-02-10: Now updated to XNA 2.0). This is NOT the full Dungeon Quest game (see www.DungeonQuestGame.com for that), this is just the GDC version.
  
  Please note that the level was reduced to allow loading on the Xbox 360 (which otherwise crashes with an OutOfMemoryException), the game is not fully playable, only the first part is implemented. You can also press F2 to toggle the Options menu and some minor bugs were fixed. But this game is no longer supported, I will not improve it anymore! Please check out the new Dungeon Quest game from www.DungeonQuestGame.com, which is coming in a month or so.
  
  • Game Installer: DungeonQuestGDCSetup.exe (25.6 MB)
  • Source Code and Game Content: DungeonQuestGDCSourceCode.zip (67.2 MB)
  • Screenshot:
    
    
    

1. Either use the XNA project conversion utility (can be found on the XNA Creators Club website) or just create a new XNA 2.0 project in VS 2005.
   
   
2. If you created a new project, drag all source code files into the project and seperate the content files out and put them all in the existing Content directory (only there the content pipeline is activated). If you just converted a project and the content files did not move, move them yourself to the content directory. Gladly all my projects with more than 5 content files had a special content directory anyway, so no need to change anything content-wise for them. If you don't want some of the files to be compiled to .xnb files, you have to change the build action from "compile" to "content" (and then use the "copy to output directory" switch) or to "none" if you want them to be ignored like for .wav files, which are automatically processed by the .xct (XACT) file for you.
   
   
3. Find the line content = new ContentManager(Services); and replace it with Content.RootDirectory = "Content";. If you do that, get rid of the content manager in your game class since you can now use the build-in Content property to access the underlying Game content manager. In case you don't want to do that or if you need an extra variable, replace the above line with content = new ContentManager(Services, "Content");. Both ways will make sure all the content is now loaded from the content directory instead from the main directory of the application. In more complex XNA games you can also change the BaseGameDirectory to the content directory, but then you would also have to move all other resource files to this directory (config files, save games, levels, etc.). It is usually a good idea to separate the compiled (.xnb) content from the content the user can change (config, levels, etc.), so I suggest just redirecting the content directory of the content manager.
   
   
4. Replace the LoadGraphicsContent(bool) method with LoadContent, remove all the if (loadAllContent) commands (was never false anyway, just let the content of the if loop stay) and also remove the call to base.LoadGraphicsContent(bool) (does not do anything like all the Load or Unload methods in the XNA Game class, they are just empty virtual methods). You can also ignore this and the next step since it will only generate depreciated warnings, but I suggest cleaning up your source code whenever an opportunity like this presents itself. I also added some missing region blocks to the code and some comments here and there were they were missing.
   
   
5. Finally delete the UnloadGraphicsContent method unless it did anything beside base.Unload and base.UnloadGraphicsContent. In my XNA games the UnloadGraphicsContent usually looked like this and can be safely removed now (at least if nothing else is in there):

/// <summary>
/// Unload graphic content if the device gets lost.
/// </summary>
/// <param name="unloadAllContent">Unload everything?</param>
protectedoverridevoid UnloadGraphicsContent(bool unloadAllContent)
{
  if (unloadAllContent == true)
    content.Unload();

  base.UnloadGraphicsContent(unloadAllContent);
} // UnloadGraphicsContent(loadAllContent)

1. First of all make sure the old XNA 1.1 code gets compileable by going though the changes mentioned above (e.g. replacing ResourceUsage with TextureUsage or BufferUsage) and removing everything that does not exist anymore (like ResourceManagementMode.Automatic). If a method is non-existent in XNA 2.0 like ResolveRenderTarget, comment it out and remember where it happened.
   
   
2. You might go through other issues, but you have to come back to the RenderTarget issue. This took the most time in the converting process for me (probably half of all my issues come by something related to changes with RenderTargets in XNA 2.0). For that reason always make sure that rendering to textures still works while you make changing. I always used the TestCreateRenderToTexture unit test inside the RenderToTexture class to figure things out.
   
   
3. Additionally to making some changes in the BaseGame class (loading content via LoadContent, using the base.Content instead of creating a new content manager, etc.) I also removed all the RenderTarget helper methods and fields from the BaseGame class (SetRenderTarget, ResetRenderTarget, etc.) and moved them into the RenderToTexture class. While this makes the code more clean and restructured by making a few more fields private, if you do not call the new InitializeDepthBufferFormatAndMultisampling of the RenderToTexture class the calls to SetRenderTarget and ResetRenderTarget will not work correctly and will not restore the default depth buffer (which has to be remembered first). If you get the following exception it means the DepthBuffer Device.DepthStencilBuffer was set to null, but is obviously still used. In order to fix that make sure the remDepthBuffer variable is set to a correct value in the InitializeDepthBufferFormatAndMultisampling method!
   

An error has occurred during the Clear operation while trying to clear the depth or stencil buffer, no DepthStencilBuffer surface exists.
System.InvalidOperationException: An error has occurred during the Clear operation while trying to clear the depth or stencil buffer, no DepthStencilBuffer surface exists.
at Microsoft.Xna.Framework.Graphics.GraphicsDevice.Clear(ClearOptions options, Color color, Single depth, Int32 stencil, Rectangle[] regions)
at Microsoft.Xna.Framework.Graphics.GraphicsDevice.Clear(Color color)

• http://www.gameports.net/cms/meldung,medien-kritik-an-arbeitsbedingungen-in-der-spielebranche,2292.html

• http://www.gameports.net/cms/artikel,medien-eine-branche-im-aufwrtstrend,801.html

Could not load file or assembly 'xunit, Version=7.10.25.1028, Culture=neutral, PublicKeyToken=null' or one of its dependencies. The system cannot find the file specified.
    System.IO.FileNotFoundException: Could not load file or assembly 'xunit, Version=7.10.25.1028, Culture=neutral, PublicKeyToken=null' or one of its dependencies. The system cannot find the file specified.
    File name: 'xunit, Version=7.10.25.1028, Culture=neutral, PublicKeyToken=null'
    at Xunit.TdNet.TdNetRunner.RunTests(Type type, MethodInfo method, ITestListener listener, TestRunState& runState)
    at Xunit.TdNet.TdNetRunner.RunMethod(MethodInfo method, ITestListener testListener)
    at TestDriven.TestRunner.AdaptorTestRunner.Run(ITestListener testListener, ITraceListener traceListener, String assemblyPath, String testPath)
    at TestDriven.TestRunner.ThreadTestRunner.Runner.Run()
    
    WRN: Assembly binding logging is turned OFF.
    To enable assembly bind failure logging, set the registry value [HKLM\Software\Microsoft\Fusion!EnableLog] (DWORD) to 1.
    Note: There is some performance penalty associated with assembly bind failure logging.
    To turn this feature off, remove the registry value [HKLM\Software\Microsoft\Fusion!EnableLog].

• Silverlight 1.0 RC (for da browser)
  
• Silverlight 1.1 Alpha Refresh (for da browser)
  
• Silverlight 1.1 Alpha Refresh SDK (some new samples and updated code)
  
• Blend 2 Preview (August CTP, important for designers)
• Microsoft Silverlight Tools Alpha for VS2008 Beta 2 (for Alpha Refresh, July 2007)
• Microsoft ASP.NET Futures (July 2007, also kinda important for ASP.NET Ajax and Silverlight)
  

• http://www.codeflakes.net/blog/
  
• http://silverlight.net/blogs/jesseliberty/archive/2007/07/27/
  silverlight-1-0-rc-and-1-1-alpha-refresh.aspx
  

• Scott Gu's Blog about VS2008 Intellisense
  

• 1001: usually AG_E_UNKNOWN_ERROR (yeah, thats really helpful)

• some element is null while loading a control
  Fix: Check all controls constructors and fields, set breakpoints there!
  

• Namespace duplicate found (usually error code 2254)
  Fix: Check all namespaces, make sure each name and namespace is only set once
  
  
• Control has no default Constructor and can't be instantiated
  Fix: Provide a default constructor with no parameters, it can be empty!
  

• NullReferenceException (or other exceptions) happend somewhere
  (usually in the Control constructor somewhere)
  Fix: Catch the exception or set breakpoints to find it, then fix it.
  Often it is also useful to comment code out and test if it works again.
  
  
• If this all not helps and the 1001 errors keeps coming up when you use the
  control use the following approach:
  
  Fix: Comment out the control, the page should now work.
  Comment it in again and check every single property you set, try removing
  as much as possible and test again. If it still does not work, check out
  the control itself, comment everything out here inside the control canvas.
  It should work now except for some errors of missing inner control names.
  Now slowly comment everything in again and try to see which inner control
  did not work. Often a depreciacted, commented out or deleted property is
  used and causes this error. Again try to work with as little as possible
  control properties and increase until the error pops up again.
  
  This more complex case will hopefully not happen to you in the beginning,
  but the more controls you build, the more likely it is that this one
  will annoy the hell out of you. Keep calm and go forward as methodically
  as possible, it will also help you for future error encounters.
  

• Can happen if a TargetName for a Storyboard (or something similar) cannot
  be found. Make sure the target control name is still valid.
  

• 2005:    ParserError: Unknown namespace xy

• The line should tell you which namespace is unknown and this should
  be quite easy to fix. It usually means you have either forgotten to
  add a namespace or add an element into an unsupported location.
  E.g. adding something to a nested Canvas can sometimes cause this error.
  

• 2024: ParserError:

• Invalid attribute value text/python, usually happens in the main xaml file.
  Similar to 2265, use the same fixes.
  
  
• It can also happen when the mime type
  is not available (e.g. in IIS) and .py files can't be used.
  Fix: Define .py as mime type in IIS as 'text/ironpython'
  

• 2101: can be:

• LostFocus event (or similar) not supported here
  Fix: Move global events to the main canvas, they are only supported there!
  

• 2207: AG_E_RUNTIME_METHOD

• playing media failed, unable to play it, loading probably failed with 403
  or 404
  

• 2210: AG_E_INVALID_ARGUMENT

• Short: In 90% of your cases the embedded resource cannot be found, make
  sure that the name of your embedded resource is correct in the
  GetManifestResourceStream function in your control constructor. Also make
  sure the file is marked as an embedded resource in VS Orcas!
  

• This means the embedded resource could not be found, which means either the
  call to GetManifestResourceStream uses the wrong string OR the resource
  is just not existant because of a wrong content type! make sure the
  control you are trying to access is actually an Embedded Resource and not
  just a Content or Silverlight Page.
  
  Fix: Set control xaml as Embedded Resource and make sure all strings are
  correct! GetManifestResourceStream should succeed! Using the Silverlight
  1.1 API UI Controls is also helpful because they figure out the path
  for you (just leaving you with misstyping the control name or forgetting
  to set the embedded resource type or one of the million errors below).
  
  Note: This is very annoying because SourceControl sometimes messes this up
  and just merges 2 files and setting it as Content or Silverlight Page,
  which make this error hard to track, so always check for the correct
  file build action first, it should always be Embedded Resource for controls.
  

• Another way this error can happen is when inside the control a namespace
  error occurs (similar to 1001 or 2254), you just get 2210 and trying to
  find why it does not work while all names are set ok and the resource is
  found, etc. One example of this failing is that a namespace or assembly
  used inside a control to reference other controls cannot be found or is
  misspelled (e.g. path does not longer fit, can be stupid things like
  ClientBin/Some.dll should be ../ClientBin/Some.dll)
  
  Fix: Check all namespaces in the control too, make sure all names are
  correct and files (especially assemblies) can be found the way they
  are specified!
  
  Advanced Fix: I use reflection to find out about the class name and use
  that same name for the .xaml file, this way I never have to specify
  any name and merging files does not lead to problems as long as the xaml
  file has the same name as the class (which is easy to spot in VS Orcas).
  

• Another VERY crazy way for this to happen is when one of your controls is
  using the partial keyword for its control class and overwrites some of the
  autogenerated background fields (through x:Class in the main Canvas of the
  control), it is probably better to always remove the partial and make sure
  the control is all managed by you (like in the SDK samples).
  
  Fix: Remove the partial in c# and the x:Class in xaml if you get compile
  errors, after that it usually just works :)
  

• Older notes (before I figured this out):
  Seems to happen with perfectly fine xaml code, no c# breakpoint is ever
  hit, so this must happen while parsing the xaml. Often also changes to error
  2254 if some name or namespace is wrong, but then gets back to 2210 if that
  is fixed.
  

• Someone here traced the 2210 back to “InitializeFromXaml” being called
  with an invalid resource stream (GetManifestResourceStream), which failed
  one line above it:
  http://silverlight.net/forums/t/1755.aspx
  

• 2251: ParserError, AG_E_RUNTIME_MANAGED_ACTIVATION

• Probably a security issue, seems to happen on a Max when the xaml file
  could not be loaded or accesses something that is not available (assemblies)
  Fix: Seems to be a Mac bug, will hopefully be fixed in the future.
  

• 2252: can be:

• Canvas Load Error, canvas could not be loaded because something is wrong
  it may help to check out the line and error description for more details.
  Fix: Look through the xaml line by line, annoying, but something is wrong
  

• AG_E_RUNTIME_MANAGED_ASSEMBLY_DOWNLOAD:
  When this happens the control at the line specified failed to load.
  This can be because it crashed somewhere in the constructor, setting a
  breakpoint usually tells you wether it crashed in C# or already in the
  xaml. If it is in the xaml, you have to figure out the error there (try
  to uncomment, check namings, etc.). If it happens in C# setting a
  breakpoint makes it much easier, but the error is still annoying. For
  that reason you should make sure to test your controls as much as
  possible before using them on a complex page.
  
  Fix: Set a breakpoint in the constructor of the control and fix the
  exception! Or if that does not help, take a close look at the xaml
  and comment it out to see what works and what does not.
  

• More help about 2252 from http://silverlight.net/forums/t/370.aspx
  Assembly loading failed, this can happen because of the IIS being
  misconfigured (unlikely if you doing development) or because the
  path to the assembly could not be found (much more likely).
  
  Fix: Make sure both the page xaml and the control xaml point to the
  correct directory for the dll and make sure this works from the
  perspective of the page, not the directories the xaml files are in!
  In my case changing ClientBin/.dll to ../ClientBin/.dll often did
  the trick!
  

• Another issue can happen when you rename the assembly, but not all the
  .dll links are correct. It can also be problematic if there are special
  letters like - in the assembly, which will get renamed to _, but the
  filename is still with -, see here:
  http://silverlight.net/forums/t/370.aspx
  
  Fix: Just check all .dll links and make sure both the name and the path
  are correct.
  

• 2254: AG_E_RUNTIME_MANAGER_ASSEMBLY_DOWNLOAD, can be:

• Control type not found because a namespace is not set correctly, this can
  happen if you just copy a file from another project and do not change the
  namespace accordingly.
  Fix: Make sure all namespaces are correct (same in xaml as in the controls)
  
  
• Namespace duplicate found, make sure that all namespaces and names only
  occur once.
  Fix: See above 1001
  

• 2255: AG_E_PARSER_BAD_TYPE

• An unsupported type was found, which usually means you are specifying a
  control or class for the xaml, which does not exist (anymore). Make sure
  that the x:Class parameter points to an existing class and uses a correct
  namespace.
  

• Also seems to happen when there are multiple comment blocks mixed in the
  xaml file or some name contains an invalid character because of that.
  Clear the xaml file and test until it works again, then insert the removed
  code again to figure out where the error is happening. The line/position
  information of this error is usually not very helpful!
  

• 2265: Error: Parser code

• Parsing xaml failed, this has been reported after saving some xaml in Blend
  but it could not be parsed in Visual Studio.
  Fix: Try using only parts of the xaml, use smaller files, dunno.
  
  
• Parsing xmal failed, some syntax error.
  Fix: You should be able to spot the error and fix it if it is a smaller
  file, for bigger files comment most of it out and comment it back in until
  it breaks again. Then fix the issue.
  

• 3001: AG_E_INVALID_FILE_FORMAT

• Often a ImageError, which usually means you are trying to load an
  unsupported image format like the .GIF format, which is not supported
  by Silverlight yet.
  Fix: Use .PNG instead of .GIF, etc.
  

• 3002: AG_E_NOT_FOUND:

• Error type should describe which kind of resource was not found (usually
  ImageError). Check your constructor and Page_Loaded methods and check
  all the files that are loaded there and make sure they exist. This is
  probably the equivalent of FileNotFoundException, just all the helpful
  information (which file, why, etc.) is missing. Sometimes the error type is
  confusing because it may report an ImageError, but in reality just the xaml
  file could not be loaded, always check the xaml file (name, loading, etc.)!
  

• 3010: can be:

• Silverlight installation is not complete yet, restart browser (or computer
  if this error persists).
  Fix: Tell user to restart browser.
  

• XnaProjectChangerSetup.zip (79.76 KB)

• XnaProjectChanger.zip (34.95 KB)

• Ctrl+1, Ctrl+2: Comment code like CR_Commenter (not implemented yet, it is commented out, too much features from CR_Commenter missing).
• Ctrl+3: Build region like CodeRush or CR_Commenter, also commented out, CodeRush's version currently does work much better, but adds comments to #endregion
• Ctrl+4: Collapse or expand the current region you are in
• Ctrl+5: Collapse or expand all regions in the current file. In IronPython it has 3 modes: All uncollapsed, All methods collapsed, All classes and methods collapsed
• Ctrl+6: Regenerate all collapsable blocks in IronPython, this is important because the IronPython language service does not support any collapseable blocks, maybe I will implement that later in there ...
  

•    RegionAddinSourceCode.zip (25.67 KB)
  
  
•    RegionAddinSetup.exe (80.1 KB)
  

• CoolCommands 3.0 adds some nice features to VS, mainly to improve interaction between VS, projects and the explorer or command line
  
  
• MZ-Tools is a great addin for Visual Studio I never have heard of before (and I know a lot of addins).  It has a long feature list, check it out yourself ^^ Sadly it does not support VS Orcas yet, but it works nicely in VS2005 and does not override any behaviour of other addins like CodeRush as far as I can tell.
  
  
• CodeIt.Right and CodeIt.Once I have heard of them before, but I never really installed it and check it out. It provides many useful reformating, refactoring and produtivity features, which allow you to quickly add existing code, reformat it automatically and apply rules for all developers on a project. They have also 3 short videos on their website so you can quickly see what this is about. Again no VS Orcas support yet as it seems.
  
  
• SmartOutline is also an addin from the CodeIt.Right guys, they won the 3rd place in the VS Studio Extensibility Contest. All this addin does is add regions, but it is free to download and use. Not sure why this is great, my Commenter plugin could do that 3 years ago and does it in a much faster way (shortcut: Ctrl+3). CodeRush does support it too (Ctrl+R or Ctrl+3 too).
  
  
• TestMatrix for Visual Studio .NET is a unit testing, code coverage and profiling addin. I'm not sure where the profiling comes in (I got Ants Profiler for that, havn't found a better tool yet. But I would like a better tool because Ants Profiler hasn't changed much in 2 years, which suxx), but the code coverage and unit testing features are nice. Probably not worth buying if you use TestDriven.Net like me. Code coverage can also be covered via NCover (a lot of covering in this sentence ^^).
  
  
• Spell Checker is a really useful addin. Like using Word, when the spell checker is there, everything feels right, but when it is missing, it gets very annoying when you need it. I used the SpellChecker for CodeRush from RindHand, but Spell Checker is also very good.
  
  
• Koders Addin for Visual Studio 2005 is also cool, but like most of the plugins, it only works in VS2005, no Orcas support yet. I have installed it for a long time, but I never use it. Dunno why, the idea is great, it allows access to many billion lines of source code and prevents you from re-inventing the wheel.
  
  
• CodeKeep is very similar to Koders, but it focuses on code snippets. I have not used this addin before, but I will try it out, maybe I will use it more often than Koders.
  
  
• I might also have installed 50 million other addins, but I have no time to talk about all of them, most of them are not worth mentioning ^^ In the past I have also talked quite a bit about addins and tools for Visual Studio, including TestDriven.Net, LineCounter, CodeRush, VisualAssist (for C++), CodeSmith and more. They are still great addins, no reason to mention them again, you know them probably already ^^
  

BTW: If you had an account for the old Arena Wars game you probably got this newsletter containing some cool information about our new "Veteran" rank for the original Arena Wars players. If you had an account for Arena Wars One everyone will see that you are a Veteran in the Arena Wars Boards and in the game itself.

• Wir haben ja schon an die 1000 Studiums-Dokumente und nun kann man von jedem auch eine Flash-Vorschau direkt auf der Seite angucken und ein bisschen drin stoebern. Momentan sind wir sehr stolz auf dieses Feature. Das ganze funktioniert natuerlich auch fuer alle neue Unterlagen, die ihr hochlaed.
• Tag basierte Interessen mit einer huebschen Uebersichtsseite, die immer gleich andere User zeigt, welche auf aehnliche Dinge abfahren.
• Verwalte deine Freunde in "Meine Freunde". Hier kommen auch noch ganz abgefahrene Features in Kuerze.
• Im Studiumsbereich kann man sich Faecher anlegen oder existierenden Faechern beitreten und somit Kommilitonen leicht wiederfinden und mit ihnen diskutieren und Unterlagen tauschen.
  
• Unter Bilder kann sich jeder eigene Alben anlegen und Bilderbuecher erstellen. Hier haben wir ausserdem noch paar coole innovate Features fuer die nahe Zukunft geplant (Note: Momentan wird das Albumsmodul noch ueberarbeitet, also entschuldigt kleine Fehler und Unannehmlichkeiten).
  
• Man kann natuerlich auch das uebliche machen, wie Freunde adden, im Forum rumposten, Gruppen erstellen, Profileinstellungen machen und Gaestebuecher vollspammen, etc.

Yesterday I wanted to put all the samples from the book on my blog, but it is already way to overloaded here with screenshots and games, adding another 10 games will not make anything better. Instead I had a crazy idea to create a XNA Community site in one day. It is called XnaProjects.Net. The idea is for everyone to submit their games and links. News are grabbed with Google Blog Search and more features will come in July 2007 when I got more than 5 minutes time in a row.

Anyway, check out this great new website, submit your games and links and check out whats already submitted by me (10 Games so far, yes, thats a lot of XNA games I did in the past few months):

• XNA Pong
• XNA Breakout
• XNA Tetris
• Rocket Commander XNA
• XNA Shooter
• XNA Racing Game
• SimpleRacingGame
• SpeedyRacer
• Dungeon Quest
• Skinning with Collada Models in XNA

    -or- path is a directory.

    -or- path specified a read-only file.

#region RemoveAllAttributes
/// <summary>
/// Helper to remove all file and sub directory readonly attributes
/// to make SafeDeleteDirectory work!
/// </summary>
/// <param name="directory">Directory to change (will be deleted later
/// anyway)</param>
privatestaticvoid RemoveAllAttributes(string directory)
{
    // Get all files and sub directories.
    string[] files = Directory.GetFiles(directory);
    string[] directories = Directory.GetDirectories(directory);

    // Kill all attributes, make files normal and directories just directories
    foreach (string file in files)
        File.SetAttributes(file, FileAttributes.Normal);
    foreach (string dir in directories)
        RemoveAllAttributes(dir);

    // Finally set the main directory to a normal directory!
    File.SetAttributes(directory, FileAttributes.Directory);
} // RemoveAllAttributes(directory)
#endregion

• AnnoyingFilesRemover.zip (6.1 KB)

• AnnoyingFilesRemoverSourceCode.zip (12.27 KB)

• Open up regedit
• Navigate to HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\VisualStudio\8.0\AddIns
• Search for the CodeRush key and export it
• If you can't find a CodeRush key here try (probably CodeRush 2.x specific): [HKEY_CURRENT_USER\Software\Microsoft\VisualStudio\8.0\PreloadAddinStateManaged]
• If you can't find it there either try the reg installers below!
• Now edit the exported .reg file and just change the \8.0\ to \9.0\ and save.
• Finally doubleclick the file to merge it with your registry (at your own risk btw ^^)
• Start VS Orcas and enjoy the DevExpress menu and the CodeRush functionality :)
  

• Open up regedit again (you have to repeat this probably a few more times if you miss any important reg key)
• Search for your Addin name or navigate to HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\VisualStudio\8.0\InstalledProducts
• Find your product, it might not be just the name of the product, but something similar (like DXCore for CodeRush).
• You need to open up the key and copy the internal package GUID (the long ugly id) like {BD359A86-BF88-4c4e-A7D2-6CDDA36C3B79} for DXCore
• Now search the whole HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\VisualStudio\8.0\ namespace for this key and export everything to external files.
• Merge those exported .reg files and replace \8.0\ with \9.0\ like above.
• Save, import again and start Orcas to test. Repeat until it works ^^
  

• Install CodeRush 2.1.2 to VS Orcas, uses the default paths (english), change if you have installed it somewhere else
• install_coderush2_vs9.reg (3.08 KB)
• Same with german paths (the documents and program files are different in german), change this to other languages if your windows is not english or german:
• install_coderush2_vs9_german_paths.reg (3.08 KB)
• Same for VisualSVN 1.2.1:
• install_visualsvn_vs9.reg (1.23 KB)
• install_visualsvn_vs9_german_paths.reg (1.22 KB)

• A Sneaky Peek At Orcas, short article showing you how to create a new project with lots of screenshots

• LINQ Introduction, very long article, good information, especially if you need a good introduction and overview about all features.

• 101 LINQ Samples: Very useful resource to get started and see some useful examples of the new language features.

• LINQ Samples Update for VS Orcas March CTP

• Download VS Orcas March CTP (self-extracting install) 4.5 GB

• Download VS Orcas March CTP (Virtual Image for Virtual PC or Virtual Server in case you are a pussy) 6.1 GB

• Download VS Orcas March CTP (Team Foundation Server) 860 MB

Let's get started. This is the train ride to the west entrance. It will not be that empty on the way back.

The CeBIT fair ground (photo was taken a little bit later). There were about 25 filled halls with a lot of exhibitors and over 200 000 expected visitors.

The first booth we saw. Not really spectacular, but we started with some boring halls in the beginning and left the entertainment halls for the end.

RFID chips are used to track packages here. The sensors in the front are used to check if a package has left or entered a storage facility.

Heise.de had a booth too and we saw a couple of Linux penguins in that hall too. The usual get 2 CT editions for free abo could be ordered here too. They also had a Sony Play station 3 running Linux on the other side. The funniest thing was some other booth in that hall where some woman announced that you can win a "Sony Play 3 Station ". We laughed our asses off.

Some Chebaba dude or chick was handing out candy.

The big companies SAP and Microsoft almost filled the whole hall 5 together.

Microsoft did not have anything new to show because most of its products have been released in the past months. It will take another 2-3 years for the next cycle. There were only some smaller handheld devices with Windows XP (like the Origami from last year) that looked interesting and a lot about boring business and security. I would have liked to see some info about Longhorn Server.

An Asian woman wanted to take this Audio sport car with her.

This is the data storage server for the whole Microsoft booth. There are 2 rack servers in here and about 24 disk drives. Some guy from Microsoft came over and explained to us that the server got 4 TB data storage and runs several services, all Windows based. But we smiled a little thinking of our little server at home, which is just a normal PC running Windows 2003 x64 (you will need that for any drive bigger than 2 TB), but got 4 TB on a single RAID 5 already, with some other backup and boot drives too. And it is not as noisy as the servers we saw here at the CeBIT.

Some cool relaxing chairs near the AMD business booth. There were a lot of water fountains in hall 1 and 2.

Another server with about 64 disk drives running a Windows Cluster service and there were showing virtualization here and explained how cool it is to have multiple virtual servers, which can balance each other. I did not wanted to spoil them by telling them that using virtual servers will just throw away almost 50% of your performance away. Using it for performance critical applications did not make much sense to me. Again, just 6.1 TB on this data server, not much if you ask me. In a few years everyone will have more memory on their home computers ^^

A blue haired girl at the ASUS booth. I was not sure if she is really from ASUS as she was handing out something to win some Intel stuff. Later I found out all the blue haired girls were from Intel and they were at 8 different booths, which you had to visit to win something boring. I would guess the total number of people how did that was zero. BTW: Did I mentioned that my mainboard that stopped working last week when I came back from the GDC was from ASUS? :-(

A Porsche! There was a PC inside, but for some reason the car was more exciting (and expensive). You could even get a price if you ask, but you know, when you have to ask, you can't afford it anyways.

Racing games all over the place. Why does everyone going to a fair want to play racing games? And why does everyone just show racing games? Man, this is boring! Ups, I did a racing game too, maybe I should be more exciting about this ...

The Samsung booth in hall 1. Samsung is a cool company anyways with all their support for e-sports and their range of cool affordable products, but everyone at this booth was extremely friendly. Most other booth in hall 1 were just focused on business and did not even want to be asked about anything.

A new laptop, which got the new Nvidia GeForce 8000 Go card in it. Coming out in 1-2 months. Pretty exicting. Makes my new laptop with a GeForce 5900 GS look much less impressive.

Here are the specs. If even such a small laptop can have a GeForce 8000 and do DirectX 10 stuff, bigger laptops will have some cool graphic boards too in the future.

The new 30 inch screen from Samsung. It has a better image than the 30 inch Dell, but it is still very expensive and the Dell monitors are much more affordable (especially the 24 inch). The resolution on a 30 inch is great: 2560x1600 ^^

Some guy was zooming in on girls with this new camera, but I was watching the watcher.

Two girls from Intel (again and again, they were all over the place) checking out each other. Not really hot, but getting to it ..

A sphere of monitors at the very unfriendly IBM booth. They had some business people explain the marketing ideas of IBM in Second Live, but they had no idea about the application and their "sims" were empty like no ones else. But who cares, as long as people keep pumping money into "3D worlds and presentations" they will keep going. IMO the Second Life technology is not really perfect for virtual presentations. There is no easy way to login and most of the required technology is not even implemented yet (Second Live does not even have voice support).

This is how the sphere monitors are controlled by the way. Each one has a laptop on its back. I'm not sure this is the best way to do it, but we are still on the IBM booth, no reason to be clever here.

A little later we saw Commodore. Every year someone new buys the name and brings out a couple of shitty products, when will it ever stop? For example they got some "new" 4GB mp3 players with some simple display, which costs 250 bucks. By the way, other companies like Samsung and Creative had similar shitty mp3 players and marketed them as "new" too. The walkman was something new, but everything after that is pretty much the same. I just do not get the point when people start speaking about "taking your music anywhere with you". Are you freaking kidding me?

For that you get a Zune or iPod with 30GB and much better technology. I got a Zune from the GDC by the way and it does not work here in Germany yet. First of all you are not allowed to download or pay for anything when you are not in the US or Canada (the only countries where the Zune is released yet), and then you can't even install the Zune software on a German operating system. There are some hacks to get around some issues and having direct access to the Zune drive (very useful), but they only work on a English windows version or XP. Here is the link if you are in the same trouble.

Did you notice that no booth chicks will look at me when I'm trying to take a picture? I will keep trying till the end of the day ^^ Please also note that we entered the entertainment halls and everything will look more like the Games Convention now.

An Asian RPG game, which has a really cool kill counter in the lower right. The game looked very manga like, but still had realistic graphics. I have to check back what this game was about later.

R2D2 was there too. Actually it was a pretty cool robot that is already available in Japan and it has cameras, can move around and it can blink and beep like the "real" one from the movies.

Hmm, this might look like a completely empty booth, but actually the very exciting dude there was showing invisible devices. Amazing, isn't it?

More cases and a lot of different displays on them. They all cost around 40 bucks, which is not much for computer cases anymore and the quality gets better and better.

The World Cyber Games 2007 Euro Championship was in hall 23 of the CeBIT. But it only was played from the 15. to the 18. March and it was all over when we were going into the hall.

Holy macaroni. This girl from Razer was throwing something at me while I was taking this picture. We could maybe count that as a girl looking at me while I'm taking a picture, but she was out to get me. This does not count, let's go on. And no, I did not catch the bandage.

AMD and ATI (also AMD) were in hall 23 too. Try to find the laptop facing in the wrong direction on this picture!

A cool Formula One racing car, which reminds me of the Speedy Racer Mod of the Racing Game. Ups, I already said too much ...

The ATI chick "Ruby" looks good as always.

And some cool water cooled gear with AMD stuff in it.

And finally some mainboards with AMD chipsets on a wall. Not sure if I would agree with AMD's new slogan "Smarter Choice" as they are currently very much behind the Intel Core 2 Duo CPUs and the Nvidia 8800 GPUs are also much faster than their ATI counterparts!

At the XFX booth the famous Quake world champion "Fatality" was playing against random people. Those two police girls where protecting the gamer area. Some people even wanted be be arrested .. strange people these gamers.

And another picture of those two nice girls. They had to be alert all the time, so I'm ok with them not noticing me taking pictures.

And this is Fatality playing against some random guy. They were playing Quake 4. The announcer had a pretty boring job:
00:00: The match begins ..
00:06: 1 to 0 for Fatality already
00:10: 2 to 0 for Fatality
..
00:36: 8 to -1 for Fatality ..
etc.

Oh no, we are at the Games Convention, aren't we? Just for a t-shirt everyone is going crazy again!

Now that is a pretty cool mouse. Mine can not do anything like that, I'm happy when it even works at all.

More cases, much more cases!

Power supplies and cases, almost like in a computer shop, just the prices are missing.

In hall 24 we saw a lot of hardware vendors, mainly for mainboards, graphic cards, power supplies and coolers. Zalman has really grown big. They got so many new products.

They do not only have coolers, power supplies and water cooling systems, but also cases like this one: The Z-Machine!

G.Skill memory. DDR3 and faster memory with 900Mhz, 1000+ MHz and more coming soon.

A pretty cool looking server mainboard for 2 CPUs and 16 memory slots, uffa!

Something is missing here? A nicely build PC with corsair ram and LEDs on them is there. A keyboard is there too and even speakers, but where the hell does the gamer look at? At a white wall?

DDR3 memory is coming. No, it is nothing really new, graphic cards have that already, but now the big memory sticks for the mainboards are going to be faster too. Also notice the cool looking cooler and the 8800 GTX build in here. I guess it is a Core 2 Duo system too, probably the X6800, which is still very expensive. The prices will drop in April however. Exciting times to come.

More stuff from Gigabyte. The hall was filled with hardware vendors, but we were tired and had not much power left walking through every single booth. For a more complete view about the halls 23-25 check out Guru3D's review of the CeBIT 2007. They have a lot of cool pictures too and more technical details.

HAH! Another Intel chick and this time I got her just as she was quickly moving forwad. Not only did I get her sharply while the world around her blurs away, she is also looking at me and smiling. Mission accomplished for today.

Professor Intel explains why everyone needs a X6800 quad core CPU for games today, when over 95% of all games do not even care if you have two cores. But still, even a single core from a Intel Core 2 system is much faster than anyone else can give you.

More cases .. Wasn't Apogee the software publisher behind Commander Keen and older jump and run games? That logo does look quite different.

Some big piece of cooling equipment. Not sure if that even fits in a normal desktop. But it is all passive and while mainboards are cooled passively too and the power supplies are very silent the only noisy part in your PC will probably be the hard discs. Haven't seen silent ones of those in a long time.

More cases .. and what the heck. Another girl is looking? The day can end now. It was not all that bad, but most pictures I took where blurry and when I gone through the photos I have selected for this post, I saw that all the girls from the first half were looking away ^^

Oh no, this poor guy just wanted to take a closer look at these half naked chicks and now he is in real trouble. I better keep going.

And there is the reason to buy a 1000W power supply (e.g. the Galaxy one from Enermax). Ok, you need about 24 hard discs, 4 CPUs and 2 8800 GTX graphic cards to even get to 933W, but does not everyone have such a PC at home? The raid controller here is looking cool, just 6 ports which 4 sata cables coming out of it. I did not know the vendor (BMC or something), but other RAID controller suppliers like Highpoint had similar cards too.

That's it for me. Time to get some rest. Next big fair will be the Games Convention in Leipzig in August, but there are a few smaller conferences till then and I will keep shooting photos :-)

Download the game installer here (40 MB): DungeonQuestSetup.zip
Alternative download (slow! from this site): DungeonQuestSetup.zip

Note: This game requires a Shader Model 3.0 GPU as it was build for the Xbox 360 and optimized that way. For lower end PCs it pushes way too many polygons each frame anyway so that it will be too slow. If you have a low end SM 3.0 GPU (like the Nvidia GeForce 6600), make sure you run in a lower resolution (see below for details).

Here is a little video from the current version of the game. Enjoy!

Screenshots:

Ok, let's get started. What is this guy doing there. He doesn't look very friendly, maybe it is better to hit him in the head!

First quest completed. This key is important to open the door to the second level.

Nonstop fighting action. Well, at least if you don't lose your way.

This is how it looks like after I got angry. I told those guys, don't start any fights!

Ohh no. I did not use all my skills (see right side) and this Ogre killed me too quickly. I guess I have to start over.

The credit screen when the game ends. Notice my book :)

No reason to stop, let's start over and try again. The fight is continuing. The game is not that hard, you just have to use all skills and avoid being hit by fireballs or big enemies when you have low hit points.

Improvements:

• Improved performance a lot by optimizing some internal code and using all ps3.0 shaders now (sad thing is it only runs with ps3.0 hardware now).
  TIPP: If you have a low spec ps3.0 graphic card (like the GeForce 6600 or lower) use a smaller resolution, the game will run much faster. Use the DungeonQuestSettings.xml file and enter a smaller resolution there (e.g. 800x600). The game is heavily GPU bound (like all my games) and it pushes several million polygons each frame (scene alone has almost a million polygons, but it has to be rendered 3 times for all shadow mapping and post processing effects). Runs on XBox 360 or GeForce 7x hardware very good with 100fps and more.
• Fixed parallax mapping for the cave and improved the specular effects, looks much nicer now. Also increased view distance and light affect range.
• Improved the music and sound effects, also added a lot more events and little sounds.
• Added many text messages to help the player understand the game better
• Check if computer can do ps 3.0, else shows big message on screen.
• Support for 64 bit systems (did not start before)
• Reduced glow in post screen (see comments of beta version)
• Fixed game logic and quests (did not work at all in the beta)
• All game actions get a nice message now in the center (killed ogre, got key, quest complete, new weapon, etc.)
• Implemented several new animations for hitting and dying
• Fixed text size in 3:4 resolutions (did only look correct in 16:9)
• Improved gravity (much stronger now) and fixed several collision bugs
• Fixed collecting key, added it to the UI and also allowing dropping weapons now
• Allow switching weapons (right/left shoulder buttons or Q/E or Mouse wheel)
• Selected enemy ring to help you see which enemy is attacted
• Rumble Xbox 360 controller when hit/getting hit
• Added end screen and credits with book link, credits
• Improved weapons, damages and made the game more balanced.
• Show high score list when game over
• Mouse support for dungeon quest (shooter like, just asdw for moving)
• Block door when we don't got key! Also added sphere collision for monsters (they don't intersect anymore)
• Fixed shadow mapping and added a lot of cool shadow effects and pseudo point light sources.
• Added nice looking vista compatible icon, also tested the whole game on vista, runs fine :)
• Added some grunt sounds and an ambient theme in the background!
• And many more ..

The Lobby of the Moscone North Building, which was also the XNA Lobby and the place we stayed all day to finish our game. The Gears of War theme music was getting annoying after hearing it a few times, but after hearing it for 4 straight days I guess I need a few weeks of silence before I can ever play Gears of War again ^^

Two of the other teams working hard on their XNA games.

Gabe Ahn from Sony introducing us at the COLLADA round table I was speaking at on Wednesday (left to right): Shaun Leach from Zipper Interactive, Ken Normann from EA (working on the Sims team) and me (some XNA guy from exDream in Germany? ^^)

The panel discussion was good IMO, hopefully more and more people get attracted to COLLADA and start using it for their tools.

The outside world at day 4, it might not look very bright to you, but after 2 days with just 2-3 hours of sleep this looked incredible bright and it was hard to open the eyes out here. But the weather was nice all week and it was warm enough to wear t-shirts all the time except at night.

A picture from Microsoft's XNA Party at Thursday evening. It was still early and we prepared our presentation of the XNA game we did in 4 days: Dungeon Quest. We had many problems in the last two days and just skipped some of our features, but we got it done and it looks really good, especially for the short time frame we had to develop the game.

Christoph and me standing proudly in front of our game at the XNA Party. Many people came and played and congratulated us on the great result.

One final picture from Saturday morning before we were leaving to the airport to head back to Germany. The trip was 25 hours for me and after that I felt as tired as Friday, I probably need to sleep for a week now .. but I got work to do ..

If you want an real early preview of the version so far, you can download it here (with installer and everything, about 38 MB). We will fix some issues today and make an update this weekend. There are many more features we had implemented (splitscreen, shadowing, quest solving), but we removed all that to just make the version run because there were still some issues. Anyway, enjoy the game so far. Thanks for all the nice comments.

Update: Fast Download from new Server: Dungeon Quest Beta (38 MB) (Day 4 GDC version)
Original Server (slow): Dungeon Quest Beta (38 MB) (Day 4 GDC version) (WARNING: slow download from this website)

Tomorrow there will be an update and all the beta issues will be fixed + more features will come, stay tuned!

Tomorrow I will post another screen for today and talk about what happened today (I spoke at the Collada panel, that was a lot of fun and we took some new pictures).

Our Dungeon Quest Game still looks pretty good and we advancing a lot, at least with the graphics. Keep in mind that we started with nothing, no real engine, I have never done point lights before and all the texture and 3d models were created in the last 2 days.

Some parts of the game are also implemented, but we will still have a lot of work to get all the quests, monsters and AI working properly and there are still some heavy issues like implementing shadow maps indoor with point lights (uff) and I still have no idea how to get the split screen stuff working efficiently.

Here is the screen shot for day 2 (yesterday), it shows the player model and the cave.

Stay tuned, more to come!

Our game is called Dungeon Quest and we are absolutely insane because we want to create a Multiplayer Role Playing Game in just 4 days and we haven't even thought of all the distractions here. I will try to give you a screen per day and tell you want we have done so far.

Yesterday we worked purely on the environment, the level and the general game idea and some smaller details. As I said before we started with nothing, I just created a new project and Christoph did just open up 3DS Max and started working, there are no textures or models we have already done. And we both have never worked on a role playing game before.

In the morning we still had a lot of problems getting our computers up and running (3DS Max wasn't starting, I needed lots of plugins and tools, stuff like that). It usually takes a day or more to set up a development machine in our opinion. Probably a week until you got all the important tools installed you forgot in the first day. Anyway, we finally started to build the cave, create new custom collada exporter for that and placed lights. While Christoph was working on the 3D cave I was getting the engine up and running and noticed that there are many things that have to be done:

• Like just rendering some text on the screen,
• Finding a good font for that,
• Creating a bitmap font,
• Also making other UI elements (selected monster, health, level up, etc.)
• Rendering shaders and playing around with point lights
• Working on many other smaller issues like effects, billboards, vertex formats, etc.
• Most unit tests require a good camera and testing the cave level itself was very hard, so I already created the ThridPersonCamera class we going to use for this game.

Today we will focus on building the main character, getting some collision detection working and some basic game logic (keys, doors, monster state machines, player behaviour and animations). Hopefully I can provide you with a nicer screenshot tomorrow, just an empty level is pretty boring.

And here are some pictures from yesterday for your enjoyment.

This is the lobby of the Moscone North building at the GDC. The GDC Expo hall is below us and on Monday the Serious Games Summit was here, but we hadn't had any time to even check out who is speaking at the actual GDC. At wednesday I will be at some Collada round table speaking about Collada and issues we had with it so far ^^

There are many art galleries here in San Francisco and the Asian population is also huge. It was not hard to find an Asian Art Gallery.

Here you can see an ad for Microsoft's Vista Operating System directly in front of the big Mac building here (seems a lot of people got one here). There is nothing interesting about a Mac for a game developer IMO.

Alcatraz at day time. I already showed you a picture at night earlier.

A view from my desk at the Hotel over San Francisco. Its still impressive every day you wake up ^^

And some picture of a place I don't know the name of. Usually I would look it up, but I have to go back to work. Cya tomorrow!

Let's go, this is Germany at night, very exciting, isn't it? We started our journey at 4:20 am

If you have read my blog before you already know how the Hanover Airport and Amsterdam airports look like. We did not take much pictures there anyway. The plane ride was long as usual (11 hours alone from Amsterdam to San Francisco) and it was a really long day (~33 hours). This was our plane, a pretty big Boeing 767.

We arrived at our hotel at 3:00 pm and were pretty amazed about the 36 floors.

But then we saw this room. There must be something wrong, we are not the President. The registration gave us the wrong key to a conference room, but we stayed there for a while, took pictures and enjoyed the room and bar ^^

Our rooms were not bad either, pretty nice view over San Francisco.

Our Hotel (The Argent) from below.

And a view over San Francisco at the evening. We spotted the Golden Gate Bridge (right from this picture) and the Alcatraz Island.

And finally San Francisco at night. We were pretty tired by now and I still had to install stuff on my laptop, after merging the 2 HDDs all my data was lost and I had to download and install everything again, but at least the HDDs are faster now (yeah, Raid 0 kamikaze). The Chinese had some parade below us and made some noise, but we were very tired and went to sleep early.

More pictures tomorrow.

Here are some screenshots from the version. More to come till the end of march, when we will be done with the game and all new effects, etc.

New snow texture pack and heavy glow, may need a little weaker post screen effect ^^

Fight in Mission 1, water looks cooler now and some new effects, but I was to lazy to play more and go into a later level.

Content

• Video, Screenshot and Downloads
• Introduction and why Collada?
• Class Overview
• Loading collada files
• Problems with the animation data
• Optimizing vertices
• Optimizing the Optimization
• Adjusting the shaders for skinned meshes
• Post screen shaders for the final result
• References and Links

Video, Screenshot and Downloads

Screenshot of the same scene (better quality since youtube videos just suck quality wise):

Before you start reading this article I suggest you download the executable and source code first in case you want to try it out directly while reading. The whole project is based on the RocketCommanderXNA engine and just adds the ColladaModel.cs class, which does all the amazing stuff you see in the video ;-) The video above shows shadowing from the XNA Shooter engine, which is not yet released and their fore not included in the downloads. The rest of the code is the same, just the shadow map rendering was removed. More details about shadow mapping in XNA can also be found in my book (plug plug plug ^^).

• The source code is about 2.6 MB and contains projects for XNA Game Studio Windows and the Xbox platform, as well as a project for Visual Studio 2005. It was tested on Windows XP and Vista 32 and 64 bit as well as on the Xbox 360.
  SkinningWithColladaModelsInXnaSourceCode.zip (2.6 MB)
• The executable contains the output (32 bit Windows, but runs on 64 bit and Vista too) from the project in case you just want to try it out or do not have XNA Game Studio Express installed yet. You still need the XNA Framework for this program, download it here, if you do not have it installed yet! You also need a Shader Model 2.0 capable graphic card.
  SkinningWithColladaModelsInXnaExecuteable.zip (2.5 MB)
  
  Update 2007-02-26: Fixed time steps, in release mode with more than 200-300 fps the animation was getting too slow. Now works flawless :)
  

Introduction and why Collada?

There is a nice project on CodePlex called XNA Animation Component Library, but it only support FBX and ASCII X files (and recently BVH, ASF and AMC formats) and I could not get any of my test models to work with this library. The main thing missing here is good shader support and it also has a lot of problems with complex animated models. Other than that exporting 3D Models as X files is really a pain in the ass, no matter which exporter you use (Panda Exporter for 3dsMax was good a year ago, I still use it, now kiloWatt X file Exporter for 3dsMax is better, and the Microsoft X File Exporter for Max always sucked). Simple models might work, but the more complex 3D models get and the more meshes and animation data is stored in a scene, the more problems you will have. Sometimes it is not possible to reconstruct everything correctly on your importer side.

Anyway, at the very beginning of the Racing Game development there was no content pipeline in XNA (it was Beta 1) and I implemented loading 3D Models with help of the collada file format, which is basically just XML and very easy to read. For that reason it was relatively simple to get some 3D data loaded and displayed in the early XNA versions with help of vertex buffers. There were some problems with shader settings and I had to try many different exporters and ended up with the one from Feeling Software. Back then it had still some problems loading shaders and using the correct techniques, but the recent version (3.0.2) is much better and works like a charm.

After Microsoft had implemented the content pipeline and made it possible to load X and FBX files indirectly by going over the content pipeline I had to remove most of my collada code and re implement the model loading with the new framework. Loading and displaying 3D models was much simpler this way and especially some early unit tests were really simple, but as soon I tried shaders and loading tangents there were a bunch of problems. I reported many bugs back then and it has gotten a lot better, but I still had to fix several issues in the Racing Game and Rocket Commander XNA myself like finding out the correct shader technique and fixing tangent data with help of a custom model processor.

Some things like the level loading in the Racing Game just do not work with X or FBX files because they use splines, which are not exported at all in these formats. Collada came as the rescue again because it is really no big problem for this format. Later versions of the Racing Game removed the collada level loading and introduced a binary format for the levels, but the importer still accepts collada files.

Recently I wanted to test a couple of animated models and use skinning since it can often be more useful than static animations and it usually looks much better, especially for organic 3D models. As I said above I could not get anything working with the XNA Animation Component Library and I especially do not like the way they still use the content pipeline and the project is too complicated for me anyway (I just do not need 6 different file formats, I just want one and it should work perfectly with all the features I need).

After some searching I saw some guy called remi from the collada forum was working on importing collada models too and I posted some thoughts there too (maybe this was a mistake, I got many emails asking me about tips ^^). Here is the thread about that in the collada forum. He has provided a test project with some models and it works nice for static meshes without shader information, but that was not really what I was searching for.

This was a month ago and I had not much time working on any of these issues, but after restarting my blog earlier this week I thought this would be a nice topic to talk about. I'm still pretty busy with Arena Wars Reloaded, but I worked a couple of hours every day on this little test program for the past few days. The rest of this article explains the project and into which problems I ran.

Class Overview

I wrote the ColladaModel file from scratch, but I could reuse some of the static mesh loading code I had done last year. All of the bone and animation loading code was just try and error and I only used the collada specifications as a source of help, but most stuff had to be tested with the unit tests at the end of the class many many times.

The main program just calls the unit tests TestGoblinColladaModelScene or TestShowBones, there is no real game here, its just a test program. The unit test then calls several Standard Engine classes for doing all the post screen processing, rendering the ground plane, etc. More importantly the ColladaModel class itself basically just provides a constructor and a Render method, everything else is private and will be handled automatically for you. Most people might ignore this, but this is always the most important thing about my classes, the use should be as simple as possible and when I look at the projects mentioned above I really ask myself why people sometimes think so complicated.

The internal Bone class inside ColladaModel is used to store all the bones in a flat list, but each entry has a parent and a list old children bones. This way the list can be used both in a simple for loop, but you can also go through it recursively (which is obviously slower and often more complicated). We will talk about the loading process in a minute.

All the mesh data is stored in vertices, which is just a list of SkinnedTangentVertex structs. The SkinnedTangentVertex struct is very much like the standard TangentVertex struct used in Rocket Commander XNA, but it has 2 new members: blendIndices and blendWeights. Both are in the form of Vector3 and their fore can hold 3 values allowing us to interpolate up to 3 bone influences for each vertex in the shader. More is often not required and we have to re normalize all bone weights anyway, so skipping the least important bone weights is not a big deal. My test models use mostly max. 2-3 influences. Please also note that the vertex shader has now a lot more work to do with all that skinning and you should really optimize it as much as possible. Both the number of vertices we have to process is important (we will talk about optimizing that in the optimizing part of this article) and also the number of instructions the vertex shader has, both numbers should be as low as possible. The GPU is really fast processing this data, but if you do not have animated geometry with bones, there is no reason to let it process all that data (which can make the vertex shader 2-3 times longer and slower). The ground of our test scene does not use a skinning shader as an example.

And finally there are some additions to the ShaderEffect class. First of all we got a new shader called "SkinnedNormalMapping", which does the same thing as the normal mapping (or parallax mapping) shader, but it has an array of 80 bone matrices we can use for skinning. These matrices are set with help of the SetBoneMatrices method in the Set Parameter region of ShaderEffect.cs.

Loading collada files

Before we go into the details of the loading process, lets make sure we read the summary of the class first because it clearly states what we can do and can't do with this class. This is just a test project and I wanted to make things as simple as possible for both you as the reader and for my requirements.

   /// <summary>
   /// Collada model. Supports bones and animation for collada (.dae) exported
   /// 3D Models from 3D Studio Max (8 or 9).
   /// This class is just for testing and it will only display one single mesh
   /// with bone and skinning support, the mesh also can only have one single
   /// material. Bones can be either in matrix mode or stored with transform
   /// and rotation values. SkinnedTangentVertex is used to store the vertices.
   /// </summary>
   class ColladaModel : IDisposable

OK, with that said let's go directly into the loading code, which is located in the constructor of this class. All variables used in these class are just for internal use, all you need to know are the vertices and bone lists, which I have already mentioned, and the vertex and index buffers, which are used for rendering. All the rest of the variables are just there to help us loading the collada file (don't worry, there are not many variables anyway and most methods are short too).

    #region Constructor
    /// <summary>
    /// Create a model from a collada file
    /// </summary>
    /// <param name="setName">Set name</param>
    public ColladaModel(string setName)
    {
        // Set name to identify this model and build the filename
        name = setName;
        string filename = Path.Combine(ColladaDirectory,
            StringHelper.ExtractFilename(name, true) +"."+
            ColladaExtension);

        // Load file
        Stream file = File.OpenRead(filename);
        string colladaXml =new StreamReader(file).ReadToEnd();
        XmlNode colladaFile = XmlHelper.LoadXmlFromText(colladaXml);

        // Load material (we only support one)
        LoadMaterial(colladaFile);
        
        // Load bones
        LoadBones(colladaFile);

        // Load mesh (vertices data, combine with bone weights)
        LoadMesh(colladaFile);

        // And finally load bone animation data
        LoadAnimation(colladaFile);

        // Close file, we are done.
        file.Close();
    } // ColladaModel(setFilename)
    #endregion

As you can see first of all the filename is constructed and we just load the file as a text file and throw it to the XmlHelper.LoadXmlFromText helper method (which just uses the existing XmlDocument functionality to load xml from a string). We now get the main collada node, which contains all the children nodes we need for loading the materials, bones, meshes, etc.

Next all the materials are loaded, but we are only going to use the first one we find because we only support one single mesh anyway. The LoadMaterial method goes through all used textures and shader effects from the collada file and constructs the material at the end of the method with help of a new constructor in the Material class itself. While this is cool and a lot easier than loading material data from x files, it is not very exciting code, so let's move along.

Even through the bones are located at the end of the collada file, we have to load them first because all our other loading methods, specifically LoadMesh and LoadAnimation need the bone tree structure and the overall bone list to work. All bones are loaded in sequential order because we want to make sure that we can use the animation matrices later in an easy way without having to check the parent order all the time. Only this way we can be sure that going through our flat bones list we still respect the internal tree structure and always initialize the parents first because the children bones matrices are always multiplied with the parent bones.

    foreach (XmlNode boneNode in boneNodes)
        if (boneNode.Name == "node" &&
            (XmlHelper.GetXmlAttribute(boneNode, "id").Contains("Bone") ||
            XmlHelper.GetXmlAttribute(boneNode, "type").Contains("JOINT")))
        {
            // [...] get matrix
            matrix = LoadColladaMatrix(...);

            // Create this node, use the current number of bones as number.
            Bone newBone =new Bone(matrix, parentBone, bones.Count,
                XmlHelper.GetXmlAttribute(boneNode, "sid"));

            // Add to our global bones list
            bones.Add(newBone);
            // And to our parent, this way we have a tree and a flat list in
            // the bones list :)
            if (parentBone !=null)
                parentBone.children.Add(newBone);

            // Create all children (will do nothing if there are no sub bones)
            FillBoneNodes(newBone, boneNode);
        } // foreach if (boneNode.Name)

Problems with the animation data

Getting the animation data was not so easy. First of all I never had done this before because my collada files for the Racing Game were all just static meshes and I really did not need any animation there. Everything that is actually animated in the Racing Game was done directly in XNA, not in 3D Studio.

The first problem is the many formats that animation data can be in. You can have rotations around any axis or translations and even scalings, but most of your bones will only use one or two of these if they are animated at all. Alternatively all the animation data can be computed directly by the exporter plugin in 3D Studio Max and this way you can make sure that all the animation data is in the correct format. It makes testing certainly a lot harder and if you don't even know the format the matrices are in or how to apply them to each other in which order, you are in a world of trouble.

This is exactly what happened to me, I had most of my test models with rotation animation data only, but the Goblin above from my friend Christoph was done with another technique and the exporter could only export the matrices, so I had to support that too. After some try and error I managed to get the basic animations for my test models working. They are all in the project, feel free to load and test them. To test the bone animations I used the following unit test:

    #region Unit Testing
    // Note: Allow calling all this even in release mode (see Program.cs)
    #region TestShowBones
    /// <summary>
    /// TestShowBones
    /// </summary>
    publicstaticvoid TestShowBones()
    {
        ColladaModel model =null;
        PlaneRenderer groundPlane =null;
        // Bone colors for displaying bone lines.
        Color[] BoneColors =new Color[]
            { Color.Blue, Color.Red, Color.Yellow, Color.White, Color.Teal,
            Color.RosyBrown, Color.Orange, Color.Olive, Color.Maroon, Color.Lime,
            Color.LightBlue, Color.LightGreen, Color.Lavender, Color.Green,
            Color.Firebrick, Color.DarkKhaki, Color.BlueViolet, Color.Beige };

        TestGame.Start("TestLoadColladaModel",
            delegate
            {
                // Load our goblin here, you can also load one of my test models!
                model =new ColladaModel(
                    //"Goblin");
                    //"test_bones_simple_baked");
                    //"test_bones_advanced_baked");
                    "test_man_baked");

                // And load ground plane
                groundPlane =new PlaneRenderer(
                    new Vector3(0, 0, -0.001f),
                    new Plane(new Vector3(0, 0, 1), 0),
                    new Material(
                        "GroundStone", "GroundStoneNormal", "GroundStoneHeight"),
                    50);
            },
            delegate
            {
                // Show ground
                groundPlane.Render(ShaderEffect.parallaxMapping, "DiffuseSpecular20");

                // Show bones without rendering the model itself
                if (model.bones.Count == 0)
                    return;

                // Update bone animation.
                model.UpdateAnimation(Matrix.Identity);

                // Show bones (all endpoints)
                foreach (Bone bone in model.bones)
                {
                    foreach (Bone childBone in bone.children)
                        BaseGame.DrawLine(
                            bone.finalMatrix.Translation,
                            childBone.finalMatrix.Translation,
                            BoneColors[bone.num % BoneColors.Length]);
                } // foreach (bone)
            });
    } // TestShowBones()
    #endregion

The most important call here is the call to UpdateAnimations, which goes through the list of bones and updates the so called finalMatrix from the Bone class for each of the bones. In earlier versions this code was horribly complicated and still had a lot of problems, but as soon as I removed all the rotation, translation, scaling, etc. animation support and just allow loading the correctly baked matrices from the collada files, the code has become much simpler (the actual code does have some optimizations in it, but it is basically the same as the posted code here):

    #region Update animation
    /// <summary>
    /// Update animation.
    /// </summary>
    privatevoid UpdateAnimation(Matrix renderMatrix)
    {
        int aniMatrixNum = (int)(BaseGame.TotalTime * frameRate)) % numOfAnimations;

        foreach (Bone bone in bones)
        {
            // Just assign the final matrix from the animation matrices.
            bone.finalMatrix = bone.animationMatrices[aniMatrixNum];

            // Also use parent matrix if we got one
            // This will always work because all the bones are in order.
            if (bone.parent !=null)
                bone.finalMatrix *=
                    bone.parent.finalMatrix;
        } // foreach
    } // UpdateAnimation()
    #endregion

For the loading itself we just have to make sure that the animationMatrices are the correct ones. Collada saves them in a absolute mode. Earlier code from me constructed relative matrices (relative to the initial bone matrix), it was easier to construct relative matrices from the rotation, translation, etc. animation data, but much harder to use these matrices later for the animation. Having these absolute matrices makes the UpdateAnimation code so much easier, so make sure you always use them this way.

However, when rendering the vertices later we can't use the absolute matrices because the vertices have to transformed first to get into a relative space to the bones, rotations should not be around the origin, but around the bone positions. Luckily for us (and you should have seen my face when I finally found out that these matrices already exist in collada and I did not have to create them myself in my own over complicated way ^^) collada stores the so called invBoneSkin matrices for each bone. By applying these matrices we can easily get the bone matrices we need for rendering, these are directly passed to our shader (as compressed 4x3 matrices BTW to save shader constants, the code for that is a little bit more complex, please check out ShaderEffect.cs and the SkinnedNormalMapping.fx shader itself for details).

    #region GetBoneMatrices
    /// <summary>
    /// Get bone matrices for the shader. We have to apply the invBoneSkinMatrix
    /// to each final matrix, which is the recursively created matrix from
    /// all the animation data (see UpdateAnimation).
    /// </summary>
    /// <returns></returns>
    private Matrix[] GetBoneMatrices(Matrix renderMatrix)
    {
        // Update the animation data in case it is not up to date anymore.
        UpdateAnimation(renderMatrix);

        // And get all bone matrices, we support max. 80 (see shader).
        Matrix[] matrices =new Matrix[Math.Min(80, bones.Count)];
        for (int num = 0; num < matrices.Length; num++)
            // The matrices are constructed from the invBoneSkinMatrix and
            // the finalMatrix, which holds the recursively added animation matrices
            // and finally we add the render matrix too here.
            matrices[num] =
                bones[num].invBoneSkinMatrix * bones[num].finalMatrix * renderMatrix;

        return matrices;
    } // GetBoneMatrices()
    #endregion

And this is what you finally get after executing the TestShowBones unit test. I had not implemented mesh loading or the shader itself at this point. I just was loading and testing the bones itself.

Optimizing the Optimization

While this is all nice and dandy and we just optimized the rendering code by 20-30% (I tested with 9 goblins and using 3 passes for them, 2 for shadowing and 1 for the rendering), but the loading now takes painfully long. The main idea here is to not compare every single vertex against every other possible vertex because it does not make much sense, most of the vertices (>99,9%) will always be different. We only need to check the ones that share at least the same position.

I started with comparing neighboring vertices, but since the vertices are stored in index order, they are totally messed up, vertex 1 and 4383 can be equal, but if we just check -10 to +10 we are going to miss it. Instead we have to know which vertices come from the same position data, which we know since collada saves unique positions. All we have to do is to keep a list of all vertices that share the same position and then we can optimize the comparsions later on. Usually only up to 4 to 6 vertices share the same position, this way the whole comparison process just needs 60,000 * 6 comparisons, not 60,000*60,000 anymore.

    // Initialize reuseVertexPositions and reverseReuseVertexPositions
    // to make it easier to use them below
    reuseVertexPositions =newint[trianglecount * 3];
    reverseReuseVertexPositions =new List<int>[positions.Count / 3];
    for (int i = 0; i < reverseReuseVertexPositions.Length; i++)
        reverseReuseVertexPositions[i] =new List<int>();

    // We have to use int indices here because we often have models
    // with more than 64k triangles (even if that gets optimized later).
    for (int i = 0; i < trianglecount * 3; i++)
    {
        // [...] vertex construction

        // Remember pos for optimizing the vertices later more easily.
        reuseVertexPositions[i] = pos / 3;
        reverseReuseVertexPositions[pos / 3].Add(i);
    } // for (ushort)

And then finally the fast vesion of OptimizeVertexBuffer, which uses that data:

    #region OptimizeVertexBuffer
    /// <summary>
    /// Optimize vertex buffer. Note: The vertices list array will be changed
    /// and shorted quite a lot here. We are also going to create the indices
    /// for the index buffer here (we don't have them yet, they are just
    /// sequential from the loading process above).
    ///
    /// Note: This method is highly optimized for speed, it performs
    /// hundred of times faster than OptimizeVertexBufferSlow, see below!
    /// </summary>
    /// <returns>ushort array for the optimized indices</returns>
    privateushort[] OptimizeVertexBuffer()
    {
        List<SkinnedTangentVertex> newVertices =
            new List<SkinnedTangentVertex>();
        List<ushort> newIndices =new List<ushort>();

        // Helper to only search already added newVertices and for checking the
        // old position indices by transforming them into newVertices indices.
        List<int> newVerticesPositions =new List<int>();

        // Go over all vertices (indices are currently 1:1 with the vertices)
        for (int num = 0; num < vertices.Count; num++)
        {
            // Get current vertex
            SkinnedTangentVertex currentVertex = vertices[num];
            bool reusedExistingVertex =false;

            // Find out which position index was used, then we can compare
            // all other vertices that share this position. They will not
            // all be equal, but some of them can be merged.
            int sharedPos = reuseVertexPositions[num];
            foreach (int otherVertexIndex in reverseReuseVertexPositions[sharedPos])
            {
                // Only check the indices that have already been added!
                if (otherVertexIndex != num &&
                    // Make sure we already are that far in our new index list
                    otherVertexIndex < newIndices.Count &&
                    // And make sure this index has been added to newVertices yet!
                    newIndices[otherVertexIndex] < newVertices.Count &&
                    // Then finally compare vertices (this call is slow, but thanks to
                    // all the other optimizations we don't have to call it that often)
                    SkinnedTangentVertex.NearlyEquals(
                    currentVertex, newVertices[newIndices[otherVertexIndex]]))
                {
                    // Reuse the existing vertex, don't add it again, just
                    // add another index for it!
                    newIndices.Add((ushort)newIndices[otherVertexIndex]);
                    reusedExistingVertex =true;
                    break;
                } // if (TangentVertex.NearlyEquals)
            } // foreach (otherVertexIndex)

            if (reusedExistingVertex == false)
            {
                // Add the currentVertex and set it as the current index
                newIndices.Add((ushort)newVertices.Count);
                newVertices.Add(currentVertex);
            } // if (reusedExistingVertex)
        } // for (num)

        // Finally flip order of all triangles to allow us rendering
        // with CullCounterClockwiseFace (default for XNA) because all the data
        // is in CullClockwiseFace format right now!
        for (int num = 0; num < newIndices.Count / 3; num++)
        {
            ushort swap = newIndices[num * 3 + 1];
            newIndices[num * 3 + 1] = newIndices[num * 3 + 2];
            newIndices[num * 3 + 2] = swap;
        } // for

        // Reassign the vertices, we might have deleted some duplicates!
        vertices = newVertices;

        // And return index list for the caller
        return newIndices.ToArray();
    } // OptimizeVertexBuffer()
    #endregion

With this optimization loading is now pretty fast and rendering performs also nicely thanks to the quick shaders we will discuss below. When running ANTS Profiler over the new project the slowest line of code becomes the actual text parsing, actually the conversion of the long strings for all vertices data into the actual vertices float data, especially the positions array. But we can't do anything about that without loading the data binary and not parsing them ourselfs. It takes maybe half a second for a 50k vertices model to load, not great, but ok for our little test app.

Adjusting the shaders for skinned meshes

Now we got a bunch of vertices we can render, but as you know we need a shader to do anything in XNA, there is no fixed function pipeline. There are also no animation helper classes like in DirectX, but they won't help you anyway if you want to render with shaders. Transforming the vertices on the CPU may sometimes be a choice if you do not have many vertices and not many skinned models overall or if they all have to be updated the same way. But generally it is a much better idea to transform all the vertices on the GPU, which costs a little bit more instructions in the vertex shader, but the rest of the rendering stays the same. Most graphic apps are pixel shader bound anyway and I use fairly complex pixel shaders too. Additionally on Shader Model 4.0 cards like the GeForce 8800 with up to 128 parallel unified shader units you can do very complex vertex shaders and use simpler pixel shaders and it will just use more units for the vertex shaders automatically :)

We already have defined the bones matrix array for skinning above and we limited it to 80 bones per mesh, which is quite a lot. Even if you would spend 3 bones per finger and 20-30 bones for the body you would still have plenty of bones left for complex animations. Sure modelers will now say "thats not enough sometimes" ... well, you can always split up the mesh into several meshes with up to 80 bones each if you really need more. My graphic artists are happy with 80 bones ^^

If we want full Shader Model 2.0 support we can only be sure that we got at least 256 shader constants. NVidia has usually more (1024), but you still want your game to run on ATI cards too, especially older ones. Each constant can hold a float4 and we need 4 constants for each 4x4 matrix. This means we can only have up to 64 matrices with this limit and we still need some constants for the world, view, and projection matrices, the light and material values and anything else we want to do in our shader. You should reserve 10-20 constants for that and now we are down to less than 60 matrices, which might sometimes be too little.

Instead of splitting the mesh or providing a different code path in case the GPU can do more constants (I couldn't get that do work in XNA for some reason, not sure if there is some limit or if I made a mistake, my GPU should be able to do at least 1024 constants, and even 2048 for the 8800), there is a trick to save only a 3x4 matrix. The last colum is always 0, 0, 0, 1 if we have correctly applied the invBoneMatrix and the animation matrix (see GetBoneMatrices in ColladaModel for details and the order of the matrices). But saving 4 float3 values still needs 4 shader constants per matrix so we have to save it as 4x3 matrices instead.

My first idea was to grab the .w values and reconstruct the translation part of the matrix this way. This worked, but the resulting shader had 80 instructions (from abount 20 without skinning), which is not good for the resulting performance when rendering many skinned 3D models.

float4x4 RebuildSkinMatrix(float index)
{
    return float4x4(
        float4(skinnedMatricesVS20[index*3+0].xyz, 0),
        float4(skinnedMatricesVS20[index*3+1].xyz, 0),
        float4(skinnedMatricesVS20[index*3+2].xyz, 0),
        float4(
            skinnedMatricesVS20[index*3+0].w,
            skinnedMatricesVS20[index*3+1].w,
            skinnedMatricesVS20[index*3+2].w, 1));
} // RebuildSkinMatrix(.)

A better idea is to use the 4x3 matrix as a 3x4 matrix by just reversing the order we call mul. This involves some changes to the vertex shader code and looks a little bit confusing sometimes, but if you make sure you transform the world matrix the same way and use it in this reversed mul order too, everything works just fine. Another thing that can be optimized is to pre-multiply the indices of blendIndices for each vertex at the loading time. These indices never change and they don't really care if they are 0, 1, 2 or 0, 3, 6, etc. We save one instruction per matrix we are going to reconstruct (the rest is optimized out by the compiler). This is the much easier version of RebuildSkinMatrix, for more details take a look at the SkinnedNormalMapping.fx file:

// Note: This returns a transposed matrix, use it in reversed order.
// First tests used a 3x3 matrix +3 w values for the transpose values, but
// reconstructing this in the shader costs 20+ extra instructions and after
// some testing I found out this is finally the best way to use 4x3 matrices
// for skinning :)
float4x4 RebuildSkinMatrix(float index)
{
    return float4x4(
        skinnedMatricesVS20[index+0],
        skinnedMatricesVS20[index+1],
        skinnedMatricesVS20[index+2],
        float4(0, 0, 0, 1));
} // RebuildSkinMatrix(.)

This results in a vertex shader that has almost half as many instructions as before and their fore is twice as fast :) Good work. Performance tests showed that I could increase the framerate from 220 fps to 270 fps just by doing that (test scene with 9 goblins). The following code is actually more or less the only part you have to replace in an existing shader if you want to make it skinnable (plus providing the helper method and the skinned matrices too of course).

    // First transform position with bones that affect this vertex
    // Use the 3 indices and blend weights we have precalculated.
    float4x4 skinMatrix =
        RebuildSkinMatrix(In.blendIndices.x) * In.blendWeights.x +
        RebuildSkinMatrix(In.blendIndices.y) * In.blendWeights.y +
        RebuildSkinMatrix(In.blendIndices.z) * In.blendWeights.z;
    
    // Calculate local world matrix with help of the skinning matrix
    float4x4 localWorld = mul(world, skinMatrix);
    
    // Now calculate final screen position with world and viewProj matrices.
    float4 worldPos = mul(localWorld, float4(In.pos, 1));
    Out.pos = mul(worldPos, viewProj);

All the skinning code and bone transformations are nice once you get them done, but without a cool model and some post screen shaders to make the scene look more cool, it is only half the fun. Good think I got the Goblin 3D Model, thanks to Christoph (WAII) again. I also had a couple of other test models and another more complex 3D Model (big evil monster ^^) which was good for some stress testing, but the material just did not look that cool.

As you can see on the following image 6 render targets are used to accomplish the final image. Most shader passes do several things at once and the list of operations (see right side of the image) is longer than the list of used passes. Most of this was just trying to get the best looking values together quickly. If you are an experienced artist you can probably do much better than me, I'm just playing around with the shader parameters until I get bored and then I leave it the way it is. The sceneMap (render target nummero three) shows the unmodified scene without applying the post screen shaders. It does not look half as cool as the final result.

I hope you enjoyed this article and that you are not as tired as I am writing this all at once (uff). It was a fun project, I have already another one in mind for next week ;-) Take care. If you have problems, post a comment. Please note that not all collada models will work, you have to follow the rules of ColladaModel or improve the class a bit for other use

References and Links

• SkinningWithColladaModelsInXnaSourceCode.zip (2.6 MB) from this article
• SkinningWithColladaModelsInXnaExecuteable.zip (2.5 MB) from this article
• XNA Animation Component Library
• Panda Exporter for 3dsMax
• kiloWatt X file Exporter for 3dsMax
• Official Collada site from the Khronos Group
• Feeling Software Collada Plugins for 3dsMax and Maya, best exporters ever!
• Public Forum of Collada.org
• Wikipedia on Collada
• Collada Forum post about remi's Collada for XNA project
• COLLADA 1.4.1 Specification (PDF)

Well, I was posting less and less on this blog and this gonna change now. From now on I will try to post often, maybe even daily. The blog posts will not be as long as before, I will just talk about whats on my mind each day. Initially this blog was started as a little diary for me and I will return to that initial idea.

As you can see I've updated my blog, added a couple of features and changed some pictures and styles. At the top you see a little collage of current game projects I'm working on. Most graphics are from Arena Wars Reloaded and the XNA Racing Game. I hope you like the overall design :-)

I will talk more about my current projects (see right side) in the next couple of posts and catch up with a couple of topics I have on my mind but not blogged about yet (FxCop stuff, CgFX, using Collada in XNA, etc.).