Elysian Shadows

Light-Dark wrote:Ive noticed a inconsistance with your post, you said C++ is your language when the engine source file is .c? which is a C source file opposed to the .cpp, but hey might be your IDE or something, looks pretty good so far, keep up the good work man. Oh and im assuming you used a lazyfoo tutorial?, your Player class and collision detection look like his

dandymcgee wrote:Looks like a great start man. The only constructive criticism I can give you based on that screenshot is to CLICK HERE NOW.
You will never regret it, I promise.

EddieRingle wrote:Looks like a neat start, but I heavily suggest you don't re-load the textures in the Dot class on every key press (dunno what your load_image function does, but I can only assume the worst).

EddieRingle wrote:Looks like a neat start, but I heavily suggest you don't re-load the textures in the Dot class on every key press (dunno what your load_image function does, but I can only assume the worst).

EddieRingle wrote:Looks like a neat start, but I heavily suggest you don't re-load the textures in the Dot class on every key press (dunno what your load_image function does, but I can only assume the worst).

superLED wrote:

EddieRingle wrote:Looks like a neat start, but I heavily suggest you don't re-load the textures in the Dot class on every key press (dunno what your load_image function does, but I can only assume the worst).

One of my early mistakes... "What the fuck, 1 000 000 kb memory usage for this shitty small game?"

lalacomun wrote:

EddieRingle wrote:Looks like a neat start, but I heavily suggest you don't re-load the textures in the Dot class on every key press (dunno what your load_image function does, but I can only assume the worst).

My load image, load's a bmp image then optimize it to a png image, after that it free the previous image (the bmp one)
so as you saw on the screenshot:
load_image(c:\...player.bmp);
i load that image , then optimize it, and then free the bmp image

void initDot() {
  SDL_Surface * dot_img_up = load_image(dot_up.png);
  SDL_Surface * dot_img_down = load_image(dot_down.png);
  SDL_Surface * dot_img_right = load_image(dot_right.png);
  SDL_Surface * dot_img_left = load_image(dot_left.png);

  int dot_facing = 0; //0 = up, 1 = down, 2 = right, 3 = left
}

void drawDot() {
  if(dot facing == 0) {
    draw_surface(dot_img_up, dot_x, dot_y);
  }
  else if(dot facing == 1) {
    draw_surface(dot_img_down, dot_x, dot_y);
  }
  else if(dot facing == 2) {
    draw_surface(dot_img_right, dot_x, dot_y);
  }
  else if(dot facing == 3) {
    draw_surface(dot_img_left, dot_x, dot_y);
  }
}

void moveDot() {
  /* your movement stuff here */
  case SDLK_UP:
    facing = 0;
    break;
  case SDLK_DOWN:
    facing = 1;
    break;
  case SDLK_RIGHT:
    facing = 2;
    break;
  case SDLK_LEFT:
    facing = 3;
    break;
}

void cleanUpDot() {
  SDL_FreeSurface(dot_img_up);
  SDL_FreeSurface(dot_img_down);
  SDL_FreeSurface(dot_img_right);
  SDL_FreeSurface(dot_img_left);
}

superLED wrote:

lalacomun wrote:

EddieRingle wrote:Looks like a neat start, but I heavily suggest you don't re-load the textures in the Dot class on every key press (dunno what your load_image function does, but I can only assume the worst).

My load image, load's a bmp image then optimize it to a png image, after that it free the previous image (the bmp one)
so as you saw on the screenshot:
load_image(c:\...player.bmp);
i load that image , then optimize it, and then free the bmp image

But when you already have loaded the image once to memory, why remove it and load it back in again?
You should just load the image once, when you init your engine, and reuse that image throughout the whole game, then you free the image when the game ends.
Think about it. Now, each and every frame, the program opens the image, read every single pixels from the image, and stores in to memory, and frees the surface.

Code: Select all

void initDot() {
  SDL_Surface * dot_img_up = load_image(dot_up.png);
  SDL_Surface * dot_img_down = load_image(dot_down.png);
  SDL_Surface * dot_img_right = load_image(dot_right.png);
  SDL_Surface * dot_img_left = load_image(dot_left.png);

  int dot_facing = 0; //0 = up, 1 = down, 2 = right, 3 = left
}

void drawDot() {
  if(dot facing == 0) {
    draw_surface(dot_img_up, dot_x, dot_y);
  }
  else if(dot facing == 1) {
    draw_surface(dot_img_down, dot_x, dot_y);
  }
  else if(dot facing == 2) {
    draw_surface(dot_img_right, dot_x, dot_y);
  }
  else if(dot facing == 3) {
    draw_surface(dot_img_left, dot_x, dot_y);
  }
}

void moveDot() {
  /* your movement stuff here */
  case SDLK_UP:
    facing = 0;
    break;
  case SDLK_DOWN:
    facing = 1;
    break;
  case SDLK_RIGHT:
    facing = 2;
    break;
  case SDLK_LEFT:
    facing = 3;
    break;
}

void cleanUpDot() {
  SDL_FreeSurface(dot_img_up);
  SDL_FreeSurface(dot_img_down);
  SDL_FreeSurface(dot_img_right);
  SDL_FreeSurface(dot_img_left);
}

I wrote this code fast and dirty, just so it could be easily understandable. But I hope you got the drill.

Light-Dark wrote:Noooooo why are you creating surfaces for each direction D:, you should nicely clip a entire sheet, im positive theres a lazyfoo tutorial on it :P, hell im 100% sure in the TileSyste.cpp source file in my SMB engine theres a 32x32 tile clipper(you can make adjustments to dimensions).

GroundUpEngine wrote:Cool project, bit like the early version of elysian shadows!