Crystal Space
Welcome, Guest. Please login or register.
October 24, 2014, 03:01:08 pm

Login with username, password and session length
Search:     Advanced search
9011 Posts in 2044 Topics by 8815 Members
Latest Member: Nickymo
* Home Help Search Login Register
  Show Posts
Pages: 1 [2] 3
16  Crystal Space Development / Support / Re: Loading 3D Models on: June 06, 2010, 01:54:53 am
I'm neither familiar with the OSG viewer nor 3ds (I use Blender), so I don't know how much help I'll be.

Is the OSG viewer compatible with CS map/world file formatting? Did you export it from 3ds Max with the exporter? (I'm really not sure how well the exporter is maintained. I think most people use the blender exporter and even that is nearly poorly maintained.)

In any case, you should be able to use viewmesh to see if the mesh will load. In addition, walktest makes use of vfs.cfg, as do all CS apps, so you'll want to make sure you have files where they need to be.

Generally, using a world file to load meshes, textures, materials, etc. is easier and more flexible. There shouldn't be much difference between manually loading from code and using xml and letting the engine load it for you. As long as you indicate where the texture files are located, and make use of vfs.cfg, you should be fine.

As for rotating, use the 'matrix' tag. Something like this:
<meshobj name='Cube'>
      <plugin>genmesh</plugin>
      <params>
        <factory>genCube</factory>
        <material>no</material>
      </params>
      <zuse/>
      <move>
        <v y='0' x='0' z='0'/>
        <matrix>
          <roty>-1.23502</roty>
          <rotz>-1.04741</rotz>
          <rotx>1.41551</rotx>
        </matrix>
      </move>
    </meshobj>
17  Associate Projects / CrystalBlend Discussion / Re: b2c XML parser not found (Updated) on: May 02, 2010, 05:44:34 am
Oh, well now the link works.

Anyway, I downloaded and installed it...This time the console did not complain about not being able to find an XML parser, but instead complained of not being able to import various files from C:\python25. Gave messages like 'import xml failed' 'xml no registry found' etc. I spent a lot of time monkeying around throwing files in and out, redirecting paths, uninstalling/reinstalling. Everything I did seemed to cause more problems than it fixed.

Thanks anyway kickvb, I guess I'll have to learn python over the summer and implement my own xml parser. It can't be too hard (famous last words), I've done it in the past w/ C++.
18  Crystal Space Development / Support / Re: iCollection XML/map file on: April 29, 2010, 01:02:50 am
You seem to be correct. The info has been removed from the Development version documentation. I wonder why. Apparently, XML and I are not getting along as of late. Thanks anyway.
19  Associate Projects / CrystalBlend Discussion / Re: b2c XML parser not found (Updated) on: April 29, 2010, 12:47:55 am
The new link takes me to http://pyxml.sourceforge.net/

I don't believe any pyXML2.5 can be dled from there.

It seems the only available pyXML on SourceForge is the 2.4 version; that will not do. http://sourceforge.net/projects/pyxml/

What are my options here, install linux and use 4Suite? B2C has lost support for Windows.

I am relatively unfamiliar w/ Python. Is there some way to maybe modify the b2c script to recognize pyXML 2.6.x? (see UPDATE from original post). I'm honestly not sure 4Suite is still being maintained either. Problematic...
20  Crystal Space Development / Support / iCollection XML/map file on: April 28, 2010, 03:30:36 am
CS version 1.4; else irrelevant for this question.

So collections are in and regions are out. I'm having a little trouble getting the collection to work in the XML map file.

see: http://www.crystalspace3d.org/docs/online/1.4/manual/XML-World-Definition-Section.html#0

I write it exactly as it tells me to in the XML:
<world>
  ...stuff...
  <scene...>
     ....
  </scene>
  <collection name='kitties'>
    <meshobj>meowmoew1</meshobj>
    <meshobj>meowmoew2</meshobj>
    ....
  </collection>
</world>

I run the .exe and get the error: crystalspace.syntax.badtoken: Unexpected token 'collection'!
I put the collection in a few other spots, such as inside the scene nodes, but nay.

If I remember correctly, I was never able to stick regions in the map file either.

Regards,
21  Associate Projects / CrystalBlend Discussion / b2c XML parser not found (Updated) on: April 22, 2010, 06:06:07 am
Windows 7; stable CS/CEL 1.4; dled b2c April '10

First of all, the link given in the 'manual' (on the b2c site http://leapingcat.org/blender2crystal/index.php/Main_Page) for 'RegularInstall' for pyXML 2.5 is broken. It takes me to the old hosting site 'delcorp' and a page not found error follows. I cannot seem to find anything other than pyXML 2.4 elsewhere on the net. pyXML 2.5 was the install that worked on my other system.

Now I've got py2.6 installed and I don't know if pyXML 2.5 would even work for that.

So, I dled the python 2.4 and the pyXML 2.4, and Blender wouldn't load the b2c export script. The console gave me something in the realm of 'E: XML parser not found. Please install pyXML (for Windows) or 4Suite.'

I then tried with 4Suite, installing the win32 version, but b2c wasn't recognizing that either.

I've found the py script where the error emanates. It's a script called 'default.' It has a check for the XML parser that is set by another script called 'csxml', but my Python is lacking and I couldn't track it farther than that.

From what I gather, pyXML is no longer supported, and I wonder if b2c could support the win32 version of 4Suite?

I'm not in desperation to get this working, I've still got my old system w/ the old install working for b2c, but I would appreciate any help when you have the time. Ug, You're probably going to make me learn Python and reinvent the wheel by creating my own blender export script that doesn't rely on XML parsers! That sounds like too much work... too little time... someday I would like to get involved with the development. Maybe after I become versed in Python.


UPDATE: this odd fellow claims to provide a version of pyXML for 2.6.x, but I still get the same error.
http://www.somethinkodd.com/oddthinking/2009/10/31/windows-installer-for-pyxml-0-8-4-for-python-2-6-x/
22  Associate Projects / CEL Discussion / Re: Running CEL in visual C++ on: April 15, 2010, 05:40:26 am
I'll give it a shot.

Given that you're on Windows (b/c you're using msvc++),
your environmental variables should look like this:

Crystal Space:
Name: Crystal
Value: C:\CS (or C:\...whatever)

CEL:
Name: CEL
Value: C:\CEL (or C:\...whater)

In VC,
Tools-->Options-->Project and Solutions-->VC++ Directories:
Under Include:
C:\CS
C:\CEL
Under Library:
C:\CS\out\release9\libs
C:\CS\out\debug9\libs
C:\CEL\out\release9\libs
C:\...

Note that the directory where you have CS installed is not necessarily that but wherever you have it located. Same is true for CEL.
Make sure to include the appropriate Preprocessor definitions and the appropriate Additional Dependancies
23  Crystal Space Development / Support / Re: Missing .dlls when executing 1.4 app outside of CS dir on: April 11, 2010, 08:47:36 pm
Happy day.

I tried rebuilding CS w/ static libs and it works great now. I wouldn't have thought about that without your input.

Oftly Jonny on the spot with the responce, for which I'm greatful. Now that I've got this working I can devote time to what's important and neglect that 'calculus homework' nonsense, jk.

One more thought: for anyone with the same situation, after building with static libs, make sure you do NOT include the preprocessor 'CS_BUILD_SHARED_LIBS'
24  Crystal Space Development / Support / Missing .dlls when executing 1.4 app outside of CS dir (SOLVED) on: April 10, 2010, 11:31:14 pm
Windows 7
CS 1.4.0
MSVC++ 2008

I downloaded and compiled everything just fine.
It is only when I compile my own project -simple1- and try to execute it that errors occur.

The error shown is 'The program can't start because crystalspace-1.4-vc9.dll is missing from your computer. Try reinstalling the program to fix this problem.

Of course, that .dll is located in the location that I installed CS in. It doesn't seem to matter that copy that  dll and put it in with the .exe b/c then the error says that another .dll is missing.

All of the sample programs (inside the cs directory) work fine. In addition, when I move my .exe to the CS directory (where all of the other sample apps are located), it works fine.

This indicates to me that maybe the CRYSTAL environmental variable is not being accessed by my .exe? Somehow, the CS directory isn't linking up.

My .exe is located in C:\...Visual Studio 2008\Projects\...Release\
My CS directory is C:\CS
Environmental variable is:
name - CRYSTAL
value - C:\CS


For the past year or so I've been using 1.2.1 on a different computer and it worked fine.
25  Crystal Space Development / General Crystal Space Discussion / Re: 64-bit Building on: March 31, 2010, 11:34:00 pm
Excellent. Just want I needed to know. Thanks guys.
26  Crystal Space Development / General Crystal Space Discussion / 64-bit Building on: March 28, 2010, 09:01:36 pm
Version of CS – CS 1.2.1
Version of CEL – CEL 1.2.1
Version of winlibs package (if on windows) – Corresponding 1.2_002
Operating system – Windows XP Win32
Compiler – MSVC++ 2005
Video card – ATI Radeon X1650 PRO
Driver – Version 8.282.0.0

Hello, again.
Glad to see that Jorrit hasn't gone through with deleting everyone's account who didn't email him Smiley

As time has gone on my sturdy 32-bit P4 / XP has started seeming a bit dated, so I decided to kill part of my life savings and pick up a new machine. It will be a 64-bit running Windows 7.

I have heard some talk on the mailing list about (res?) a 64-bit compatible version of CS, but other than that I haven't heard anything.

I really don't know how compatibility issues work w/ the 32/64-bit dilemma. I like using msvc++ and I need to know if either I can build 32-bit CS applications on a 64-bit OS or maybe I could give the 64-bit CS a try. Btw, I also prefer stable releases of things. I'm not sure why, I suppose its b/c I just love when the application crashes and I'm not sure whether it was some error of mine or one hidden in the depths of CS.

I am intentionally posting this question on the forum, so that others can perhaps benefit later (plus the mailing list is frankly intimidating with these CS deities (I mean devs) roaming around). It might not hurt for some of the CS discussions to happened a little more on the forum and a little less in the mailing list. Seems like it would give the community a bit more of a chance to interact and make the forum seem a little less..inactive. Just throwing that out there Smiley.

Cheers,

27  Crystal Space Development / Support / Re: Odd maximize/restore down window errors Win32 on: December 11, 2009, 06:50:59 am
Oh, I like that solution: In order to avoid the problem, avoid the situation in which the problem occurs.

Anywho, I narrowed the problem down to a single piece of code: camera->SetFOVAngle(angle, width);

After that function is called, I have to either restore down or maximize the window in order for the meshes to display correctly. Because this happens whenever I start an app, it also suggests that what is initially displayed on the screen is not the default camera degree angle, but a false one, until the, I guess, display calculation is reset by maximizing/restoring down the window.

I can call camera->GetFOVAngle() and output the float angle to screen, and it will give me the number that I set it to, but it won't really display that angle, not until the window is changed.

It's a little disheartening to be not able to change the camera degree angle without problems, but I can live with it for now.
28  Crystal Space Development / Support / Odd maximize/restore down window errors Win32 on: December 09, 2009, 07:41:05 pm
Version of CS – CS stable 1.2.1
Version of CEL – CEL stable 1.2.1
Version of winlibs package (if on windows) – Corresponding 1.2_002
Operating system – Windows XP Win32
Compiler – MSVC++ 2005
Video card – ATI Radeon X1650 PRO
Driver – Version 8.282.0.0

1) The first situation occurs immediately after I start up a CS app, like those demos in the CS folder, and appears to happen regardless of whether the entire level is hard coded, like with the Mazing tutorial, or the level is provided in a map file. The app initially starts in the Restore Down window and something is not right with the meshes and their proportions, or the way the camera is looking at them. It almost seems that maybe the camera degree angle is different than its set to be. In any case, when I maximize the window, by pressing the little button, the meshes seem to correct themselves and appear at the right distance/proportion/etc. This problem still occurs when the app is started in Full Screen mode. (As a note, when I do minimize a Full Screen window and restore it, I get numerous graphics errors on the command line and the whole window is screwed up). Further more, every time a new map/world file is loaded, this problem reoccurs and I have to maximize and restore down the window to get the meshes looking right again. This is a problem for me, as I mix 2D and 3D into a single app and the 2D is not affected by this problem, so they become non-uniform. Any ideas? Maybe it's my graphics card, or maybe this problem has been fixed on 1.4.

2) Now this one is just crazy! So apparently, when using iMeshWrapper::HitBeam(...), whether or not csHitBeamResult::hit is true or false DEPENDS ON WHETHER OR NOT YOUR WINDOW IS MAXIMIZED OR RESTORED DOWN. Of course! How could I not see that? I mean, the screen size and whether or not a hit beam made its mark are completely related... CS cracks me up sometimes, in that sneering, wanting to strangle you, kind of laughter. No hard feelings though =).

Regards,
29  Crystal Space Development / Support / Re: camera and csColliderWrapper on: September 18, 2009, 04:24:44 am
I would suggest studying the Mazing Game tutorial part about the player collider which, of course, could be where you already learned from, but can't hurt to check: http://www.crystalspace3d.org/docs/online/manual/Mazing-Player-Class.html#0

You basically want to represent the player/camera as a box because of the limitations of Opcode cdsys and because it is more efficient for calculations. See the tutorial for the specifics of doing that.

Next, use cdsys_->Collide(object_collide, &object_trans, player_collide, &player_trans) to test for collision. The location of the player_box is irrelevant because you only need its geometry along with the location of the camera via camera_->GetMovable()->GetFullTransform();

A few notes:
Make sure to include CS_REQUEST_PLUGIN ("crystalspace.collisiondetection.opcode", iCollideSystem), in App::OnInizialize(...).
Make sure to initalize the collision system with cdsys_ = csQueryRegistry<...>(...);
Make sure to initialize all of the collider wrappers you intend CD to happen with.
Make sure to reset collsion pairs before testing for collision.

I hope that helped.
30  Crystal Space Development / Support / Re: Texture Display and/or Compression Question on: August 12, 2009, 05:17:42 am
Hey Vincent, thanks for taking the time to reply.

I took your advice and went with the, shall we say, 'primitive' way of getting a background image by just putting the image on a mesh and fitting the mesh to the size of the screen. I've spent so long trying to figure out a way to do it without using a mesh that I am happy getting something that works and I'm taking what I can get. The image quality is still not quite as good as a billboard, but it is good enough.

I don't mean this as a put-down, but I am a surprised CS has no standard means of displaying background images. I would have thought that this would be a desirable feature in platformer or side-scroller type of game.

In any case, I have no need (for now) of keeping the mesh at a constant distance away from the camera because the camera is stationary. I'm not sure if you see what I am trying to accomplish. Think about it like this: hypothetically, you take a picture of the inside of a room and set that picture as the background. You want to have a character move about the room (which is just a 2D picture) so the character has to be in the foreground. The problem in using a mesh with a material containing the image is that it has to be set far enough back that the character you're moving around the room does not collide with it, then, size it to fill the screen. Setting and resizing the background image mesh isn't that much trouble, but it's not the most straightfoward means.

Note: Probably something done wrong on my part, but when I added the <camera /> flag to the <meshobj> my mesh disappeared. Not a big deal since I am not using the flag.
Pages: 1 [2] 3
Powered by MySQL Powered by PHP Powered by SMF 1.1.2 | SMF © 2006-2007, Simple Machines LLC Valid XHTML 1.0! Valid CSS!
Page created in 9.271 seconds with 15 queries.