I made a configurable scantron that outputs XML that can go into the web-based statistics system I made last year (it was for robotics). Now, I'm going to remake my Chip8 emulator!
[QUOTE=ZomBuster;20211203]You can static link the libraries if you don't want to hassle with dll's.
just add -s like so sfml-graphics[B]-s[/B](-d).lib[/QUOTE]
I'm doing that already. Does it have any disadvantages except that the exe is bigger?
What I puzzled me is why there are special .libs (what are they, Google only finds articles about the standard library :confused:) for debug builds?
[QUOTE=Robber;20211972]What I puzzled me is why there are special .libs (what are they, Google only finds articles about the standard library :confused:) for debug builds?[/QUOTE]
The debug builds were compiled with all the debug symbols intact. Try comparing the sizes of the debug libs to the release libs.
[QUOTE=i300;20202690]That is pure awesome.[/QUOTE]
People who disagree'd:
The reason I love that so much is because i'm a complete retard when it comes to programming on anything but Windows and iPhone. So yeah. Its awesome.
[QUOTE=windwakr;20174688]Bah, I tried jumping on the bandwagon, but I tripped and smashed my face on the ground:
[img]http://filesmelt.com/dl/mandelfail.PNG[/img]
:saddowns:
[editline]...[/editline]
Damn filesmelt, stop going down. No one can see my fail when you don't work.[/QUOTE]
Programmers never fail, they create new features. I like how you only rendered a small portion of your pattern [B][I]on purpose[/I][/B] to create an artistic side panel.
"it's not a bug, it's a feature" :eng101:
Is SFML 1.5 supposed to work with Visual C++ 2010? A little console test app with only std::cout and sf::Clock worked, but when I try to open a render window I get 19 unresolved externals. I added sfml-window-s-d.lib, sfml-graphics-s-d.lib and sfml-system-s-d.lib and have all the includes I need. :confused:
Did you compile SFML by yourself in VS2010? If not, do so
[QUOTE=zyxxyz;20213482]Did you compile SFML by yourself in VS2010? If not, do so[/QUOTE]
Compiling SFML itself gives me errors too. I'm downloading the svn snapshot now.
[b]Edit:[/b]
Thanks, even though compiling SFML didn't work, my project runs now :biggrin:
Jumping on the bandwaggon, made this 2 years ago in C++ and opengl:
[IMG]http://img5.imageshack.us/img5/9945/38125656.png[/IMG]
[IMG]http://img693.imageshack.us/img693/2765/28459388.png[/IMG]
[IMG]http://img683.imageshack.us/img683/9247/18552181.png[/IMG]
[QUOTE=Technopath;20214050]Jumping on the bandwaggon, made this 2 years ago in C++ and opengl:[/QUOTE]
holy sheit thats cool.
I'm having a new problem :frown:
[cpp]#include "Game.hpp"
#include <SFML\System.hpp>
Game::Game():game(sf::VideoMode(800, 600, 32), "SFML Graphics")
{
sf::Thread t(&run); //"'&' : illegal operation on bound member function expression" in this line
}
void Game::run(void* data)
{
}[/cpp]
[b]Edit:[/b]
Ok, this doesn't work with member methods. I used a workaround that works now.
Is it ok to allocate an object that I'll need as long as the game is running and never free the memory again?
It's not really leaking any memory since I only allocate the memory once, is it?
[b]Edit2:[/b]
I'm doing it different now.
[b]Edit3:[/b]
God fucking damnit. What is this supposed to mean!?? [code]1>MSVCRTD.lib(crtexew.obj) : error LNK2019: unresolved external symbol _WinMain@16 referenced in function ___tmainCRTStartup[/code]
[b]Edit4:[/b]
Fuck yes, it's working again.
But I can't decide if I should use VC++ 2010 or 2008. While 2010 is a [b]lot[/b] better than 2008 I don't want to run into some unexpected problems and waste a lot of time finding out it's not my code's fault.
[QUOTE=Robber;20210196]After being stuck with Java for 3 years I want to expand a bit and learn C++ (properly). We've been using it in school for 1.5 years, but I learnt pretty much nothing from that.
My first plan was to make a really simple Direct2D app, but that after seeing that the initialization code was twice the size what some of my programs are I decided to use SFML. From what I've seen so far it's much easier. What I really don't like is how there are so many extra steps to make everything work. There are billions of compiler and linker options. Instead of calling a method to open a console window there are over 10 compiler options to open a console window and to disable that the easiest method is to make a whole new project. And to use SFML there isn't a single DLL like I'm used to from programming Java where all the libraries consisted of one single .jar, but instead there are two folders I have to copy into my IDE installation and after #including the hpps I still have to add the .lib files to the linker.
Anyway, first result:
[media]http://www.youtube.com/watch?v=6LppALTpOtE[/media]
:razz:[/QUOTE]
THAT IS PURE AWESOME!
[sp]/sarcasm[/sp]
[QUOTE=Robber;20214387]I'm having a new problem :frown:
[cpp]#include "Game.hpp"
#include <SFML\System.hpp>
Game::Game():game(sf::VideoMode(800, 600, 32), "SFML Graphics")
{
sf::Thread t(&run); //"'&' : illegal operation on bound member function expression" in this line
}
void Game::run(void* data)
{
}[/cpp]
[b]Edit:[/b]
Ok, this doesn't work with member methods. I used a workaround that works now.
Is it ok to allocate an object that I'll need as long as the game is running and never free the memory again?
It's not really leaking any memory since I only allocate the memory once, is it?[/QUOTE]
Common OSs will clean up, but there's seldom a reason why you should rely on that.
To call member-functions in a sf::Thread you could try std::tr1::mem_fn.
If you run-function does not access any member-data of Game, you can declare it static and then you should be able to obtain a pointer to it.
Should you only use malloc to allocate a certain number of bytes as opposed to space for a variable in C++?
[QUOTE=s0ul0r;20214685]THAT IS PURE AWESOME!
[sp]/sarcasm[/sp][/QUOTE]
THANKS!! [sp]/sarcasm too[/sp]
Seriously, we all have to start somewhere. It's my SFML equivalent to the console's Hello World.
[QUOTE=Overv;20215138]Should you only use malloc to allocate a certain number of bytes as opposed to space for a variable in C++?[/QUOTE]
You should use the operator new or new[] in C++ if you need to allocate memory dynamically or something so big that it could cause a stack overflow.
For dynamic-sizing arrays you can also use the standard containers, such as std::vector, std::deque and std::list.
I finally have the basics down for a 2.5D platform shooter I'm working on
[img]http://img638.imageshack.us/img638/9108/iwascreen.png[/img]
[QUOTE=s0ul0r;20214685]THAT IS PURE AWESOME!
[sp]/sarcasm[/sp][/QUOTE]
:fuckyou:
I got all sad when someone told me my mountains looked like something from a teenagers face so I made myself some real mountains! :v:
[img]http://www.facepalm.se/blogg/wp-content/uploads/2010/02/mountains5-1024x640.png[/img]
Here's a bunch of pictures I took when I was exploring: [url]http://facepalm.se/img/mountains/[/url]
[QUOTE=dezek;20217280]I got all sad when someone told me my mountains looked like something from a teenagers face so I made myself some real mountains! :v:
-big image that doesn't fit my 1024x600 res :(-
Here's a bunch of pictures I took when I was exploring: [URL]http://facepalm.se/img/mountains/[/URL][/QUOTE]
I think you're doing the terrain the opposite of how you should. Why not try to follow the generation process of a mountain, not generate the product. I don't think you're going to get very realistic terrain doing it the way you are now.
Keep the noisemap for smaller scale height difference, but generate the more prominent features with at least some intelligence - mountains should occur in ranges, with ranges have differing smoothness based on their age.
Like I said before, you need some sort of erosion algorithm.
[QUOTE=Ortzinator;20218704]Like I said before, you need some sort of erosion algorithm.[/QUOTE]
I think so as well.
I don't know how the erosion algorithms work but I heard they were hard to implement with procedural terrain but I'm gonna do some reading about them and see.
Right now I'm only using ridged multifractal but I'm planning on combining other methods, because now it's hard to get other terrain types than mountains to look good. But I'm going to continue with the algorithm for the generation later because I don't want to get stuck trying to make it perfect when I'm having a deadline.
