Show Posts
This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Clear your CMake cache and retry.
tim@debian:~/Downloads/LaurentGomila-SFML-8868350$ make
[ 14%] Built target sfml-system
make[2]: *** No rule to make target `/usr/lib/x86_64-linux-gnu/libGL.so', needed by `lib/libsfml-window.so.2.0'. Stop.
make[1]: *** [src/SFML/Window/CMakeFiles/sfml-window.dir/all] Error 2
make: *** [all] Error 2
trying to build it as runnning it doesnt work.
I get this on the final build:
Any chance we could get a makefile for linux, or even better codeblocks project
CXX := g++
Divided: TinyXML AnimateSprite.o helper.o Assets
$(CXX) -o Divided main.cpp tinystr.o tinyxml.o tinyxmlerror.o tinyxmlparser.o AnimateSprite.o helper.o sfx_jump_ogg.o Door.cpp Player.cpp PlayerW.cpp SplashScreenWorld.cpp SettingsPanel.cpp MainMenuWorld.cpp LevelWorld.cpp EndingWorld.cpp CreditsWorld.cpp TextOverlay.cpp Spike.cpp Lever.cpp Block.cpp MovingBlock.cpp FallingBlock.cpp -lsfml-window -lsfml-graphics -lsfml-audio -lsfml-system
TinyXML: obj/xml/tinystr.cpp obj/xml/tinyxml.cpp obj/xml/tinyxmlerror.cpp obj/xml/tinyxmlparser.cpp
$(CXX) -o tinystr.o -c obj/xml/tinystr.cpp
$(CXX) -o tinyxml.o -c obj/xml/tinyxml.cpp
$(CXX) -o tinyxmlerror.o -c obj/xml/tinyxmlerror.cpp
$(CXX) -o tinyxmlparser.o -c obj/xml/tinyxmlparser.cpp
AnimateSprite.o: AnimateSprite.cpp
$(CXX) -o AnimateSprite.o -c AnimateSprite.cpp
helper.o: helper.cpp
$(CXX) -o helper.o -c helper.cpp
Assets: obj/sfx_jump_ogg.cpp
$(CXX) -o sfx_jump_ogg.o -c obj/sfx_jump_ogg.cpp
Did you do fps-lock?
the game is about 50-70 precent CPU!!
Cool game.
You are using pixel positions as matrix indexes...
Try it like this:
for (int bx1 = left; bx1 <= right; bx1++)
{
for (int by1 = top; by1 <= bottom; by1++)
{
player.TouchGround(blocks[bx1][by1]);
for (int ei = 0; ei < enemies.size(); ei++)
{
if (InCameraView(enemies[ei]))
{
enemies[ei].update();
enemies[ei].TouchGround(blocks[bx1][by1]);
}
}
}
}
In case of tile based sprites printed from a matrix, if you have fixed size tiles, you can know exactly where to begin and where to end printing.
Take the left border X position of your camera and divide by the tile size to get the X index of the first tile to print on the matrix. Divide the X position of your right border by the tile size and add 1 to make sure it goes past the screen, and you have the last X index you are going to print.
Same goes for Y.
Then you have just to iterate on these positions and print them.
std::map<int, std::map<int, sf::Sprite> > testBlocks;
...
for (int bx = CameraView.Left; bx < CameraView.Width; bx += 32)
{
for (int by = CameraView.Top; by < CameraView.Height; by += 36)
{
App.Draw(testBlocks[bx][by]);
}
}
You could use a higher-level structure to sort your blocks, like a quad-tree (Google it!).
It will allow you to quickly find blocks of interest (for drawing or collision tests) without iterating on all of them.
I wrote a little just about spatial partitioning here: http://www.sfml-dev.org/forum/viewtopic.php?p=35285#35285
(Where Quad-tree is an algorithm/data structure in)
Also found where I've discussed grid's which is a simpler and maybe more beginner friendly than Quadtrees. http://www.sfml-dev.org/forum/viewtopic.php?p=32767#32767
I use a display list - simply, a list of things which I want displayed. Drawing my main game screen, which only has maybe 15 or 20 items at any one time, takes next to no time.
bool right, left;
float x, y;
switch (event.Type)
{
case sf::Event::KeyPressed:
if (event.Key.Code == sf::Keyboard::Right)
{
right = true;
}
if (event.Key.Code == sf::Keyboard::Left)
{
left = true;
}
break;
case sf::Event::KeyReleased:
if (event.Key.Code == sf::Keyboard::Right)
{
right = false;
}
if (event.Key.Code == sf::Keyboard::Left)
{
left = false;
}
break;
}
sprite.SetPosition(x, y);
if (right == true && left == false)
{
x += offset;
}
if (right == false && left == true)
{
x -= offset;
}
void WorldLevel::draw()
{
for (int bi = 0; bi < blocks.size(); bi++)
{
if (InCameraView(blocks[bi]))
{
App.Draw(blocks[bi]);
}
}
for (int ei = 0; ei < enemies.size(); ei++)
{
if (InCameraView(enemies[ei]))
{
enemies[ei].draw();
}
}
}
void WorldLevel::update()
{
for (int bi = 0; bi < blocks.size(); bi++)
{
if (InCameraView(blocks[bi]))
{
player.TouchGround(blocks[bi]);
for (int ei = 0; ei < enemies.size(); ei++)
{
enemies[ei].update();
enemies[ei].TouchGround(blocks[bi]);
}
}
}
}
You should use a fixed timestep.
I do it something like this:Code: [Select]
const float UPDATE_INTERVAL = 10f; //10ms -> 100 ticks per second
float restTime = 0f;
while(app.IsOpened())
{
restTime += app.GetFrameTime();
while(restTime >= UPDATE_INTERVAL)
{
restTime -= UPDATE_INTERVAL;
Update(UPDATE_INTERVAL); //here you calculate the offset and gravity and so on with a fixed timestep.
}
Draw();
}
The problem might be if you have functions which are proportional to time^2 and not to time then the interval matters and depending on the pc the interval can change. So you just need to set the update interval to a fixed time step (I usually do 100 ticks per second, you can adjust it to your needs).
offset = App.GetFrameTime() * 0.25f;
grav -= ((App.GetFrameTime() * 1.5f) + 0.5f) / factor;
grav -= ((jMax * 3) + 0.5f) / factor;
//jMax = 1.3f;
I don't know what's the purpose of your "offset", neither why the framerate limit would matter when drawing the HUD.
Draw your HUD on another view, which you will never move.
sf::View Camera;
sf::View HUD;
...
App.SetView(Camera);
App.Draw(...objects impacted by the camera);
...
App.SetView(HUD);
App.Draw(...objects not impacted by the camera);
App.SetFramerateLimit(200);
float offset;
...
offset = App.GetFrameTime() * 0.05f;
hud.SetPosition(Camera.GetCenter().x - 445, Camera.GetCenter().y - 295);
You could create a new custom class containing a SFML sprite and one rectangle. And update both of them at the same time, even better, you can update the rectangle using sprite class methods for getting coords.
For checking to exclude some blocks from the vector, you could calculate the distance to your player, or check if they are inside the view. But, anyway you have to check ALL blocks.
But if you do a collision check against every block in your world, you'll get a slow program. So you need to priorize and partition the world space. Use an orthogonal 2D grid or BSP trees. (http://en.wikipedia.org/wiki/Binary_space_partitioning)
Now I haven't time, but maybe I'll do a little example later.
Let your imagination fly..