Ralara

Members
  • Content count

    10
  • Joined

  • Last visited

Community Reputation

310 Neutral

About Ralara

  • Rank
    Member
  1. Does anyone know where I can get a download of the 3DSFTK toolkit? All the posts I have found on this forum or around the web that have a link to it no longer exist.   I do realise that the 3DSFTK toolkit is an old piece of software, but I have been tasked with fixing a program that requires it to run.   Thanks in advance, Ralara   Edit: Managed to find http://www.findthatzipfile.com/search-44460564-hZIP/winrar-winzip-download-3dsftk.zip.htm but I can't test it until later. Can anyone confirm that this is indeed the correct files that I require?   Thanks again, Ralara
  2. I'm not 100% on Java so I may be wrong here, hopefully someone can correct me if I am, but although 'currentPiece' is public it is still inside the 'Tetris' class so the 'Piece' class cannot access it directly as you have, but it can access it through a 'Tetris' object. So it would be something like "Tetris.currentPiece", similar to the 'fall()' method where "Tetris.sleep(2000)" is used.   On a side note, there is a way of determining the rotation of Tetris pieces without having to store that location as you have in matrices (some people would argue that using tables is better as there is a limited number of shapes, but I prefer the algorithm personally). You define a block in the shape that will be the centre block that will be used to rotate around. Each other block is then defined in "shape space" relative to the centre, i.e. a block one column to the left will be [-1,0] relative to the centre block. When rotating (CW) the new row is the negative of the current column and the new column is the current row. So from the example above would become [0,1]. (This is still in "shape space", so is an offset from the centre block). For CCW rotation the negative is for the new column instead of the new row. Below is a posting of my code performing the above (note this is in C++ not java); Vector2 Shape::BlockLocAfterRotate(UINT inBlockNum) { // Get the current location of the block Vector2 outLoc = mBlocks[inBlockNum]->mLoc; // Location of the centre block for this shape Vector2 centreLoc = mBlocks[mID_CentreBlock]->mLoc; // The offset of the current block from the centre of the shape int offsetRow = centreLoc.y - outLoc.y; int offsetCol = centreLoc.x - outLoc.x; // The new location (in shape space)... int newOffsetRow = -offsetCol; // New row is the negative of the previous col (still in shape space) int newOffsetCol = offsetRow; // New col is the previous row (still in shape space) // Convert back to grid space outLoc.x = centreLoc.x + newOffsetCol; outLoc.y = centreLoc.y + newOffsetRow; return outLoc; } Edit: Doing the algorithm way means you can just update the x,y components of your piece instead of having to create a new Piece every time and could very solve your problem too!
  3. Second entity moves really slow!

    I'm not 100% sure on how your timer works, but I would imagine the get_ticks() function returns how many ticks there have been since the last time it was called?   If so, then you should only call get_ticks() once per cycle (at the beginning of the main while loop in your case) and store this in a variable. Then use this value for both the player and the enemy.   As you are calling get_ticks() for each entity, there are less ticks happening between the call for the first entity and the second entity than there are between the second and the first, because of all the other stuff happening in between,   Let me know if this works or not!
  4. Ah, turns out my Z was getting set to 0 and I hadn't noticed! I can't test my code right now but if I change it so the vertex Z is set to the near Z and use my original values, would I be correct in assuming it would work?   Thanks for your response though, makes a lot of sense. Always the simplest things that can be the hardest to spot when they are right in front of you  !!
  5. Quick question regarding the use of the XMMatrixOrthographicLH function.   After spending several days trying to figure out why I couldn't render any 2D objects in my game I finally realised I had the incorrect inputs to this function (all fixed and working now!), I'm just unsure why my original inputs were wrong.   I currently have a camera with a near z of 0.1 and a far z of 1500 just as default but when using these values to create the orthographic matrix nothing would be rendered on screen. I now use the same function but with values of 0 and 1 respectively and everything works.   If someone could explain that would be appreciated as documentation / google searches just say that the near z and far z values are required.   Thanks!
  6. The transformation matrices are just to work out how one sub-mesh is relative to its parent mesh so its not essential and you should at least be seeing something on screen. I'm at work at the moment, but when I get home tonight I'll edit this post with the code I am using / a link of where to download it from. [edit] - The current version of the source code I am using can be found at [url="http://www.adamsmith-demos.co.uk/wp-content/uploads/downloads/2012/10/Source.rar"]http://www.adamsmith-demos.co.uk/wp-content/uploads/downloads/2012/10/Source.rar[/url]. This is only the code, if you want you can build it to debug it but its a directX11 project so you'll need the sdk and an 11 card. If you want you can link to your code and we can both see if we can spot the problem?
  7. I was just about to explain why I thought you were only loading one because I could have sworn I'd seen when debugging that the aiScene only had a single mesh, but I just ran it again and it had 6... going mad I think!
  8. Good news and bad news.... Good news is I can load the models using my loader (I tried the Bob.md5mesh and the BoarMan.md5mesh) so it is possible using Assimp. Odd thing about the BoarMan mesh is that there is not texture listed in the file (that I could see) so that model appears just black until you go far enough away to see it against the background: [url="http://www.adamsmith-demos.co.uk/wp-content/uploads/2012/10/BoarMan_Full.png"]http://www.adamsmith-demos.co.uk/wp-content/uploads/2012/10/BoarMan_Full.png[/url] When the Bob mesh first loads because of its size it appears like this: [url="http://www.adamsmith-demos.co.uk/wp-content/uploads/2012/10/Bob_Loaded.png"]http://www.adamsmith-demos.co.uk/wp-content/uploads/2012/10/Bob_Loaded.png[/url] and once I've moved around it looks like this: [url="http://www.adamsmith-demos.co.uk/wp-content/uploads/2012/10/Bob_Full.png"]http://www.adamsmith-demos.co.uk/wp-content/uploads/2012/10/Bob_Full.png[/url] Bad news being that it doesn't make it any clearer as to what you're problem is... The difference between our code that I can tell is that you load the meshes that aiScene provides (which in the case of Bob.md5mesh is one) whereas I set a hierarchy of sub-meshes up using the aiScene's root node. This way I get 6 meshes for the individual bits of Bob. I don't think just loading the one mesh from aiScene would make it not appear, I think it would just make the first mesh in the file appear which is his body so you should get at least that.... Only thing I can think is that maybe you're not loading the texture properly and the mesh is covering the whole screen and appearing black? <shameless_plug> You can see the way I do my loading at [url="http://www.adamsmith-demos.co.uk/"]http://www.adamsmith-demos.co.uk/[/url] </shameless_plug> Hope this helps!
  9. This does sound odd. I'm not sure what the problem could be. If you could link to the model that you are using I can try and see if my loader works when I get home from work. I seem to be doing something very similar to you at the moment because I've just finished writing the loading part of my engine/framework and I'm moving onto making games!
  10. This for loop looks a little strange to me: [quote name='Corvwyn' timestamp='1349481323' post='4987286'] for(unsigned int i=0;imVertices->x, aiMesh->mVertices->y, aiMesh->mVertices->z); [/quote] The semi-colon at the end of the loop means it won't actually loop through the code, it will only process it once, and what is 'imVertices'. But the reason I don't think you're seeing anything on the screen is you aren't getting the vertex positions when loading (I'm wondering whether there is a mis-copy with the aforementioned for-loop, because you do use aiMesh->mVertices->x / y/ z). If the for-loop has been mis-copied then the following advice won't help and its possible something is going wrong when you are rendering, otherwise the loading code I use for assimp is as follows: [source lang="cpp"]for (UINT i=0; i < numVertices; i++) { const aiVector3D* pos = &(inMesh->mVertices[i]); const aiVector3D* normal = &(inMesh->mNormals[i]); const aiVector3D* texCoord; if (inMesh->HasTextureCoords(0)) texCoord = &(inMesh->mTextureCoords[0][i]); else texCoord = new aiVector3D(0.0f, 0.0f, 0.0f); Vertex v; v.mPos = XMFLOAT3(pos->x, pos->y, pos->z); v.mNormal = XMFLOAT3(normal->x, normal->y, normal->z); v.mTexCoord = XMFLOAT2(texCoord->x, texCoord->y); vertices.push_back(v); }[/source] Ralara