• Advertisement


  • Content count

  • Joined

  • Last visited

Community Reputation

216 Neutral

About b3rs3rk

  • Rank
    Advanced Member

Personal Information

  1. I suggest to learn a bit of shader programming. Changing the y value of each vertex in the shader code will increase the performance enormously, allowing you to increment the number of vertices in your buffer ( more vertices -> better quality)
  2. maybe you could find this useful : // This is allocated once, at the class-level scope readonly List<Potato> expiredPotatoes = new List<Potato>(); public void Update() { // Standard update foreach (var potato in potatoBag) { if (potato.Expired) expiredPotatoes.Add(potato); } // Removing pass foreach (var expired in expiredPotatoes) potatoes.Remove(expired); expiredPotatoes.Clear(); } (taken directly from this, item 5)
  3. Master's Thesis & Research - How?

    from my experience, most of the "research work" is scrubbling the net searching papers and more papers, till you find something you like (step 1). Once you find it, step two is searching MORE papers bout what's the State-Of-The-Art, history, bibliography and so on.... it's a hard and time consuming work..unless you start with something already in your mind. If so, you can jump step 1 :D When I got my master degree, i spent 4 months searching for an "area" and other 3 months searching for papers about that. Once I was pretty sure of which was the best algo, I started trying to find a way to improve it (other 2 months), and a way to implement it (1 month) :D In any case, if you really decide to take this path, you will spend AT LEAST one whole year studing,crying, studing,crying, studing,crying....till you find a solution, so be really SURE that this is what you want :D
  4. You don't. Keep in mind that you should concentrate on making your code work and later perform well, but don't waste your time fighting stupidity. You can't do anything if someone makes wrong use of your libraries.
  5. Xna spriteBatch trouble =(

    I suppose you're using a regular grid, not isometric projection... Anyway, in my tile renderer i have a "base" layer that contains the background, a top layer (yah!) and a middle layer. This one contains a list of objects (sprites). Every frame (or every time the camera changes or some object moves), I sort this list by the Y value (centroid of the bounding rect is a good starting point), and render accordingly. Hope this helps :)
  6. Isometric Tiles Culling

    it's a bit more complex than that. Of course, once I'm able to perform World to Screen space transformation easily, I could cycle on the tiles list and check if they're visible (each frame or every camera update), but I suppose there's a better way. With regular rectangular grids, one could easily compute the current visible rectangle on the grid and get the visible tiles id... could I do the same with iso-grids? maybe transforming viewport bounds on the isoGrid forth and back... Suggestions ?
  7. Isometric Tiles Culling

    Quote:Original post by rxa How are you doing World to Screen space transformation? Still have to work on that :D It's the first time I look at this, I really don't know where to start :/
  8. I've already posted this in the Iso Forum, but got no responses... Anyway, I've started a small isometric renderer and now I would like to reduce a bit the draw calls... does anyone know a good tutorial on how to check which iso-tiles are on screen? Thank you :)
  9. Hi! I've started a small isometric renderer and now I would like to reduce a bit the draw calls... does anyone know a good tutorial on how to check which iso-tiles are on screen? Thank you :)
  10. my fault, I always tend to make things simpler than what they should be :) Anyway (to @stianstr), why don't you take a look at some pre-built library? RakNet is very good for such things :)
  11. dunno if this works, but should be something like this (WARNING, pseudo-code ahead): int len = sizeof(yourStructure); void* pData = (void*)yourStructure; if (SDLNet_TCP_Send(sd_tcp, pData , len) < len) { ErrorLog("SDLNet_TCP_Send: %s\n", SDLNet_GetError()); } (anyway thank you, I stopped writing c/c++ code a year ago when I started working for the company I'm actually in, here we use only .NET.... it's always a pleasure :D )
  12. frustum culling a tiled terrain

    As oliii suggested, if your objects are on a regular grid, you should use some type of spatial structure. Usually Quad Trees / Octrees are good choices for terrains.
  13. that's exactly what I was looking for! thank you my friend!
  14. Quote:Original post by harveypekar well, positions are easier reconstructed in view space. So it follows that incident vectors are easier in view space too, hence the need for view space normals. In the end, it doesn't matter at all. Pick a space, and stick to it. All the rest is optimization, and is best done at the end, when you see all the implications. is there a way to reconstruct position in screen space?
  15. Deferred shading problem

    Another question (even if nobody still answered to the first...): for point lights should I render a fullscreen quad or a sphere mesh for each light source?
  • Advertisement