WaTor - An OpenGL based screensaver

Table of contents
Introduction
Set of rules

General rules
Rules for fish
Rules for sharks

Screen saver basics
Installation
Configuration
Limitations
License
References
Release notes

Introduction

This article describes the implementation of an OpenGL screen saver based on a discrete simulation of predator-prey interaction. The simulation is called Wa-Tor. It was originally described by Alexander K. Dewdney in the Scientific American magazine.

It simulates the hypothetical toroidal Planet Wa-Tor (Water Torus) whose surface is completely covered with water, occupied by two species: fish and sharks. The sharks are the predators. They eat the fish. The fish exist on a never ending supply of plankton. Both sharks and fish live according to a strict set of rules. This simulation of a simple ecology is highly dynamic as both species are walking a thin line between continuing life and extinction.

The screensaver is based on a dialog based MFC application combined with OpenGL for the 2D drawing. As usual, for OpenGL, using the binary without a hardware or drivers correctly supporting it will result in very low frame rates. The source code requires Microsoft Visual Studio V7.1 in order to compile, MSVC6 is not supported.

Set of rules

In this paragraph, I will briefly explain the basic rules for a Wator simulation and describe the changes necessary in order to implement the screensaver. In general, the simulation is based on discrete time steps. The simulation runs on a rectangular grid. To represent the toroidal world, opposing sides of the grid are connected. If an individual moves out on one side of the simulation domain, it reenters immediately on the opposing side. Fish and shark move every time step (if possible) and interact according to the following set of rules:

Rules for fish

In every time step, a fish moves randomly to one of the four neighboring fields, provided it is empty. Every fish has a predefined "breed time". On exceeding this time, it gives birth to a new fish in one of the neighboring cells, provided this randomly selected cell is free. (If not nothing happens.) Breed time counter of both the original and the descendant fish will be reset to zero. Technically fish never die. They live until they reach the breed time, then they clone and both parent as well as offspring restart their life cycle. (To clarify: according to the simulation the parent dies and creates two offspring individuals, but I don't delete the parent, I just reset it.)

The following picture shows options for prey movement. Arrows indicate possible movements. Fish are not allowed to move to cells occupied by sharks. If there are no free neighboring cells no movement occurs.

Rules for sharks

Sharks move randomly to fields that are either free or occupied by fish. Every round they lose one point of energy. If they enter a field occupied by a fish they eat the fish and gain a defined amount of energy. If the energy level drops below zero the shark dies. If the energy exceeds a predefined value sharks create an offspring in a free neighboring field. The energy is split evenly between the parent and the child.

Modifications made to the original algorithm

In order to turn this simulation into a screensaver, I made some changes to the original set of rules. According to the original algorithm, sharks do not move randomly. Instead they move specifically to cells in their neighborhood containing fish. Testing this is slow, thus I allow only random movement. Surprisingly there is no visible difference. Alexander K. Dewdney suggested an implementation very similar to a cellular automata. If you implement Wa-Tor that way and by looping over all cells in the simulation domain, you are forced to include cells that do not contain any individual in the calculation. Since this is too slow I chose to implement it as a linked list, thus instead of looping over all cells, my algorithm loops over existing individuals only.

Theoretical background

The information provided here is taken from the Wikipedia article on Lotka-Volterra differential equations. For a more detailed explanation I refer you to this article.

Drawing the number of individuals per species versus time yields curves characteristic for the so called Lotka-Volterra equations (also known as predator-prey equations).

Diagram: Population size vs. time for a predator prey model..

This behaviour can be described by two differential equations. There usual form is:

where::

y is the number of some predator (here: sharks)
x is the number of its prey (here: fish)
t represents the development of the two populations against time
alpha, beta, gamma and delta are parameters representing the interaction of the two species.

Although the solution to these differential equations show a periodic behaviour, they can not be expressed in terms of normal trigonometric functions. However, an approximate linearised solution yields a simple harmonic motion with the population of predators leading that of prey by 90%.

Screen saver basics

Technically a Windows screensaver is a normal executable with the extension scr. Additionally it has to have a defined command line interface in order to display itself in configuration mode, in preview mode or in screen saver mode.

Command line optionn	Saver action
/P:###	Start the screensaver in preview mode. Treat ### as the decimal representation of a `HWND`. Popup a child window of this `HWND` and run in preview mode inside that window.
/C:###	Open the screen saver configuration dialog. If ### is not present `GetForegroundWindow()` should be the parent of this dialog. Otherwise ### should be treated as the decimal representation of a `HWND`, which should be used as a parent.
/S	Run as screen saver in full screen mode. When there is some kind of keyboard or mouse activity in the screen saver mode, the screen saver is ended.

Table 1: Command line options.

Installation

Download the archive with the binary and copy the file WatorSaver.scr to your Windows/System322 subdirectory. I do not provide a setup program due to the simplicity of the installation process.

Configuration

After moving the screen saver to your windows/system32 directory, you can edit its settings like any other screen saver settings. Go to the Windows desktop, right click, and in the context menu, select the display options entry.

The configuration dialog allows setting up the simulation parameters described above as well as additional settings for the display options.

Screensaver parameters

All changes made to the parameters will be reflected instantly in the preview window. The following table lists the parameters and their meaning for the simulation process:

Prey breed time	The number of time steps necessary before a prey can create an offspring.
Energy per hunted prey	The energy a predator gets for each prey individual it eats. In each time step, the predator loses one energy point. It dies when the energy falls below zero.
Energy to create offspring	The minimum amount of energy a predator must have in order to create an offspring. When creating an offspring, the energy is evenly split between the newly created offspring and the parent.
Pixelsize	The size of an individual in pixels. Reducing this number will require very high performance!

Limitations

Like most screensavers, this is a program featuring a controlled amusing misuse of system resources. It is currently a beta version and you use it without warranty of any kind!

Currently the screen saver does not support multiple monitors. Moreover it will only save your screen if you disable all static text displays in its configuration dialog. This won't change in future versions but I don't consider this a problem since today's TFT Displays don't need a screensaver anyway.

You use it at your own risk, no warranties of any kind!

Licence

Wator Screensaver
Copyright 2007 Ingo Berg

This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA

References

A more detailed introduction on screen saver programming and issues related to this article can be found at the following places:

Writing an OpenGL screensaver for Windows by Rachel Grey.
Probably the simplest possible OpenGL screensaver. All you need to know on OpenGL initialization and basic Windows screensaver design (highly recommended).
How to write a 32bit screensaver by Lucian Wischik.
A good starting point to learn more about screensaver programming. Very detailed information but some of the information is outdated (Windows 95/98 specific).
Fight the dialog units.
This screensaver uses code written by George Yohng and published at CodeProject that simplifies Windows dialog scaling related issues. It saved me a great deal of work!
Java applet of a Wator simulation.
The page is in German, but it is the best looking applet available and you don't need to understand the text anyway since it will tell you nothing that you won't find on this page.
Another Java applet of a Wator simulation.
Since I took some of the explanations from this site, I'd like to give it credit here.
Wikipedia article on Lottka Volterra equations.
Find out more about the theoretical background of predator prey simulations at this Wikipedia article.
OpenGL tutorials by NeHe Productions.
Excellent page with a lot of OpenGL tutorials and their source code for a variety of platforms and languages.
Another Wator screensaver (with source code).
This screensaver does not depend on 3D hardware. Thus you might want to give it a try if you don't have accelerated 3D graphics. (And believe me you don't want to use my version without hardware support for OpenGL.)

Release Notes

Rev 0.90 (Beta): 31/07/2005

First version of the screensaver released. This is a beta version released for testing purposes only. There is much room for improvement like adding multiple monitor support.

Rev 1.0: 18/08/2005

This release contains bug fixes and slight modifications that make the screensaver look prettier (nothing spectacular).

Fixed bug that prevented the screensaver from running on ATI graphics cards.
Added a background pattern.
Changed title bitmap.
Statistic is now scaled between its minimum and maximum.

Rev 1.01: 21/01/2006

Support for multiple monitors added.
The screensaver can now be deactivated by mouse moves.

Rev 1.02: 03/03/2007

License changed to GPL-II.

License

This article, along with any associated source code and files, is licensed under The GNU General Public License (GPLv3)

About the Author

iberg

Software Developer

Germany

Member

Article Top

Poor

Excellent

Add a reason or comment to your vote: x
Votes of 3 or less require a comment

Comments and Discussions

You must Sign In to use this message board. (secure sign-in)

Search this forum
	Profile popups Noise Layout Per page

Third Leg, Plankton

RDABC

10:39 11 Apr '07

Would it be worth the effort to add plankton to the frenzy as a live or die entity. This could go full circle adding a new set of rules for the Plankton.
* Plankton feed on the excrement of sharks and fish (more points from sharks).
* Fish still feed on the plankton but like the sharks they too are having to search for food.

Plankton will also be following the fish and sharks.

Other changes could show energy levels of of the fish/shark by the hue or how bright it's color is. This would provide more study as to how sharks are moving and attaining food. We did this years back with the Life simulation where a green square was a noobe, blue was living, red was dieing off. This created a more interesting colorful pattern.

Folding for Team 32

Re: Third Leg, Plankton

iberg

7:52 13 Apr '07

RDABC wrote:
Would it be worth the effort to add plankton to the frenzy as a live or die entity. This could go full circle adding a new set of rules for the Plankton.
* Plankton feed on the excrement of sharks and fish (more points from sharks).
* Fish still feed on the plankton but like the sharks they too are having to search for food.

I thought about this a while ago and I think it could be interesting. Unfortunately adding a third species to this code would be difficult.

RDABC wrote:
Other changes could show energy levels of of the fish/shark by the hue or how bright it's color is. This would provide more study as to how sharks are moving and attaining food. We did this years back with the Life simulation where a green square was a noobe, blue was living, red was dieing off. This created a more interesting colorful pattern.

I once had this implemented and it looked pretty unspectacular. Clearly not what one would expect. The problem is that the motion that you see on the screen does not primarily originate from movement of fish or sharks but from permanent creation of new offspring. The most individuals (approx. 99%) don't move at all. They cant move because they have no free space in their neighboorhood to move to so they sit there waiting to die.

Enhancement: Filing populaton in and out

tbrammer

3:18 4 Dec '06

I love the WaTor saver - except for the fact, that it always starts with a random population. Many times all sharks die, leaving one single shark behind. It takes a while until a "stable dynamic population" is established.

So I implemented filing of the population. Saving and loading is quite fast. Much less than a second even for large populations.

Here is how:

CWatorGL::InitPopulation() and CWatorGL::OnDestroy() must be modified. We load the population in InitPopulation() and save it in OnDestroy(), if we are not in the preview mode.
And we need 5 new methods in CWatorGL:

-------------------------------------------------------
In "WatorWnd.h":



class CWatorGL : public CWnd

{

private:

	...

  bool GetPopulationFile(CString &csFileName);

  void writeFish(FILE *fp, Fish *pFish); 

  void writeState(FILE *fp);

  void readFish(FILE *fp, Fish *pFish);

  void readState(FILE *fp);

	...

};

-------------------------------------------------------
In WatorWnd.cpp:



...

#include "WatorApp.h"

 

void CWatorGL::InitPopulation(int a_iNumPred, int a_iNumPrey)

{

  FILE *fp = NULL;

  CString csFileName;

  if (GetPopulationFile(csFileName))

		fp = fopen(csFileName, "rb");

		

  if (fp)

  {

		readState(fp); 

		fclose(fp);

		return;

  }

  ...

}

 

void CWatorGL::OnDestroy()

{

  if (!theApp.isPreview())

  {

    FILE *fp = NULL;

    CString csFileName;

	if (GetPopulationFile(csFileName))

 		fp = fopen(csFileName, "wb");

	  

	if (fp)

	{

		writeState(fp);

		fclose(fp);

	}

  }

  ...

}

 

//Here are the new methodes:

//

// This method just gives us a path to a population-file. It will be the path 

//to the saver-executable with the extension set to ".pop"

bool CWatorGL::GetPopulationFile(CString &csFileName)

{  

  CWinApp *pApp=AfxGetApp(); 

  TCHAR filename[MAX_PATH+5];

  DWORD dwResult = GetModuleFileName(pApp->m_hInstance,	filename, MAX_PATH);

  bool bResult=false;

  if (dwResult>0)

  {

	TCHAR *p = filename + _tcslen(filename) - 1;

	for (; p!=filename; p--)

	{

		if (*p==_T('.'))

		{

			_tcscpy(p, _T(".pop"));

			bResult = true;

			csFileName = filename;

			break;

		}

	}

  }

  return bResult;  

}

 

// Here we read the population from a file

void CWatorGL::readState(FILE *fp)

{	

	Fish rFish(tpPREY, 0, 0, 0, m_iBreedTime), *pNewFish, *pLastFish=NULL;

	SeaStates kind=tpPREY;

	

	for ( ; ; )

	{

		readFish(fp, &rFish);

		if (rFish.x < 0)

		{

			if (kind!=tpPRED)

			{

				kind = tpPRED;

				pLastFish = NULL;

			}

			else

				break;

		}

		else

		{

			if (rFish.x < m_iDimX && rFish.y < m_iDimY)

			{

				pNewFish = NewFish(kind, rFish.x, rFish.y, 

						rFish.energy, rFish.maxage); 

				pNewFish->age = rFish.age;

				pNewFish->next = NULL;

				pNewFish->pred = pLastFish;

				if (pLastFish)

					pLastFish->next = pNewFish;

				pLastFish = pNewFish;

				if (!pRoot[kind])

				{

					pRoot[kind] = pNewFish;

					pRoot[kind]->pred = NULL;

				}

			}

		}		

	}	

}

 

// Here we write the population to a file

void CWatorGL::writeState(FILE *fp)

{

	fish_type *pFish,

			  EndFish(0,0,0,0,0);

	EndFish.x = EndFish.y = -1;

	

	for (pFish = pRoot[tpPREY]; pFish; pFish=pFish->next)

	{

		writeFish(fp, pFish);

	}

 

	writeFish(fp, &EndFish);

 

	for (pFish = pRoot[tpPRED]; pFish; pFish=pFish->next)

	{

		writeFish(fp, pFish);

	}

 

	writeFish(fp, &EndFish);

}

 

// Writing a single fish

void CWatorGL::writeFish(FILE *fp, Fish *pFish)

{

	fwrite(&(pFish->x), 1, sizeof(int), fp);

	fwrite(&(pFish->y), 1, sizeof(int), fp);

	fwrite(&(pFish->age), 1, sizeof(int), fp);

	fwrite(&(pFish->energy), 1, sizeof(int), fp);

	fwrite(&(pFish->maxage), 1, sizeof(int), fp);

}

 

// Reading a single fish

void CWatorGL::readFish(FILE *fp, Fish *pFish)

{

	fread(&(pFish->x), 1, sizeof(int), fp);

	fread(&(pFish->y), 1, sizeof(int), fp);

	fread(&(pFish->age), 1, sizeof(int), fp);

	fread(&(pFish->energy), 1, sizeof(int), fp);

	fread(&(pFish->maxage), 1, sizeof(int), fp);

}

Enjoy
-- tbrammer

Re: Enhancement: Filing populaton in and out

iberg

11:21 4 Dec '06

Did it really take that long to find a equillibrium? Normally it takes only a few seconds.

Nevertheless I like the idea that people enhance the code. And there is quite some room for enhancements... I'm see the remote possibility to crash the saver when loading data belonging to a simulation witch a different pixel size. This should be dealt with accordingly. It shouldn't happen when the saver terminates normally but since i just removed a bug potentially crashing the saver i would be cautious. I'll see if I add this or something similar, currently I'm quite busy.

Re: Enhancement: Filing populaton in and out

tbrammer

21:55 4 Dec '06

>>Did it really take that long to find a equillibrium?

Well - sometimes, if you use high resolution and special settings. It often happens, that all sharks die except for the last one, because the food density is to low. This poor last shark has to wait until the fish population has grown right to his teeth. Than the sharks population explodes, eating up all fish - except for the last one. This kind of extreme oscillation works on a couple of times until a more "natural" equilibrium is reached.

Besides this, it is quite nice to watch what happens to a well established population, when the parameters are changed. It's hard to explain. Just try Smile | :)

.

>>..possibility to crash the saver when loading data belonging
>>to a simulation witch a different pixel size..

You mean the size m_aiField[m_iDimX][m_iDimY] of the grid I suppose? My enhancement of



	void CWatorGL::readState(FILE *fp)

takes the pixel sizes into account by



  	if (rFish.x < m_iDimX && rFish.y < m_iDimY)

  	{ ... //insert new Fish to population }

If you load a population into a larger field, the new extended area remains empty.
If you load into a smaller field, the fishes on the cut-away area are just ignored.
I did only few testing but it seems stable to me.

I'm also thinking of other enhancements:
+ the population may react to "islands" on the background picture.
Islands could be recognized by their colour.
+ fish population depending upon krill (or other food) density

Re: Enhancement: Filing populaton in and out

iberg

11:56 5 Dec '06

tbrammer wrote:
This poor last shark has to wait until the fish population has grown right to his teeth.

Ok you'r right this may happens. Sometimes the fish almost fill the complete screen resulting in a slowdown until the sharks spread out. Usually i made settings allowing gigh framerates even when the screen is completely filled (Pixelsize 3 to 4).

tbrammer wrote:
You mean the size m_aiField[m_iDimX][m_iDimY] of the grid I suppose? My enhancement of

That's the one i meant.

tbrammer wrote:
I'm also thinking of other enhancements:
+ the population may react to "islands" on the background picture.
Islands could be recognized by their colour.
+ fish population depending upon krill (or other food) density

Reacting to islands? Could be interesting to watch, The "islands" could define different sets of simulation parameters. Interesting perspective for a fish to enter a zone and suddenly it lives longer or can create more offspring. The fish population depending on a third species is interesting I once started a version with three species but never finished. If you succedd i'd be interested to see your version.

BugBug

Highcooley

12:45 13 Nov '06

Hi there

Played around with the Screensaver and was deeply impressed how much fun it is. Really good work with the Open GL realisation...and nice graphics too.

But now to my problem: I once tried to set the pixelsice to a high value (can't remember how high...but was for sure >50)...just for testing out, how the fishes and sharks would react on such a small space. Now i can't go back to the settings screen again...it just does not want to open anymore. I can do the settings by start the screensaver manually with the .scr file...changes also take place when I start the screensaver over the .scr file. But the real installed screensaver over windows does not change the pixelsice back...i tried to uninstall and reinstall the screensaver but no chance...it won't work again.

Does anybody have any Idea, where i can set this value back to normal? Any dll oder other hidden setting files?

Thx for your Help

Re: BugBug

iberg

21:01 13 Nov '06

Hi, settings are stored in the registry. Deleting the node: HKEY_CURRENT_USER/Software/Wator Screen Saver should delete the settings. The maximum pixel size does not seem to be imposed correctly. I will fix it with the next update.
Sign In·View Thread·Permalink	5.00/5 (2 votes)

Re: BugBug

Highcooley

22:56 13 Nov '06

Hi, thx a lot...setting the node back to a lover value solved the problem go on coding complex simulations!
Sign In·View Thread·Permalink

What Dewdney meant, maybe.

Tadeusz Westawic

16:28 3 Feb '06

I could easily be wrong, but...

I think what Dewdney meant by cellular automata approach is different from your interpretation. I think he meant it in the sense of using a blitter to process the shark and fish data. Blitting lends itself well to parallel processing cellular automata.

I am not sure we as programmers have any direct access to bltters anymore even though I think they must exist on our graphics cards. Does anyone know how to access one?

Does anyone know what I'm talking about? I'll explain if invited.

Anyway, I didn't mean to crash your article. I, too am a longtime fan of Scientific American and have written WaTor using both the list approach and the blitter approach on an Amiga platform.

Tadzu

Tadeusz Westawic

An ounce of Clever is worth a pound of Talent.

Re: What Dewdney meant, maybe.

iberg

3:22 4 Feb '06

Tadeusz Westawic wrote:
I think what Dewdney meant by cellular automata approach is different from your interpretation. I think he meant it in the sense of using a blitter to process the shark and fish data. Blitting lends itself well to parallel processing cellular automata.

I dont know what he meant by his original implementation but according to his article it was quite slow. There is no mention of parallel implementation or using a blitter in his original article. This implementation is different from his original suggestion but there is not much of a visual difference. The list approach was chosen for performance reasons only.

I guess one could calculate the whole thing entirely on a modern GPU which would result in an excellent performance. I have seen wave equations and Navier Stokes equations solved with several hundrets fps on state of the art NVIDIA/ATI GPU.

Regards,
Ingo

Re: What Dewdney meant, maybe.

Tadeusz Westawic

6:32 5 Feb '06

Yes, yes. Using a GPU is what I think he meant. But not a modern one as at the time of his writing it was still pretty much a MacIntosh world and the Mac, to my knowlege, did have an on-board (motherboard) blitter in its chipset. At the time I owned the other platform with a blitter on the m'board, an Amiga. Tadzu Tadeusz Westawic An ounce of Clever is worth a pound of Talent.
Sign In·View Thread·Permalink

Re: What Dewdney meant, maybe.

iberg

13:24 11 Feb '06

You may be interested in the following articles describing making use of modern GPU's for solving PDE's:

http://www.sci.utah.edu/~cscheid/pubs/smac.pdf
http://wwwcg.in.tum.de/Research/data/Publications/sig03.pdf

I don't have the time to to convert Wator to use the GPU, I'm not even sure I have the right one to do it. But it would surely be fascinating. I saw a Navier-Stokes equation beeing solved with 400 fps in realtime using a graphics card. This is just amazing. Especially because until last year I worked at the University and wrote a flow simulator there and I was happy to calculate one frame in 10 sekonds. Although you cant quite compare it seeing real time flow simulated on a computer is amazing.

Regards,
Ingo

Re: What Dewdney meant, maybe.

Jasmine2501

18:52 27 Nov '06

Yes, QuickDraw had a BufferSwap() function. Performance was pretty bad back then. OpenGL was just coming online for Macs in 1998, but it was slower than QuickDraw, and useless for games. I don't know when the original article was written.

"Quality Software since 1983!"
http://www.smoothjazzy.com/ - see the "Programming" section for (freeware) JazzySiteMaps, a simple application to generate .Net and Google-style sitemaps!

Like it!

Arinir

22:47 24 Aug '05

This is definitely not one of those "lengthy source code printing articles" which are increasing lately on codeproject. Really nice. Got my 5! Arinir
Sign In·View Thread·Permalink	3.00/5 (1 vote)

Re: Like it!

iberg

23:07 24 Aug '05

Thank you very much. Originally I had a bad feeling for not posting a single line of code in the article but I wanted to get it out finally and just had no time. Besides this everything worth explaining already was explained by someone else in seperate articles (for instance using OpenGL and MFC).
Sign In·View Thread·Permalink

Re: Like it!

Jasmine2501

8:51 29 Nov '06

I really like this too. It runs pretty fast on my machine, about 35 FPS with a pixel size of 3. My monitor is 1920x1200, so that's pretty good. At home I'm running it with a pixel size of 1, at the same resolution, but it gets only about 14fps there... still pretty good. My machine here is a 2.8GHz, with 1GB of RAM, and the Radeon 128MB video.

"Quality Software since 1983!"
http://www.smoothjazzy.com/ - see the "Programming" section for (freeware) JazzySiteMaps, a simple application to generate .Net and Google-style sitemaps!

Re: Like it!

iberg

15:16 1 Dec '06

Thank you. Nice to see people like this kind f stuff.
Sign In·View Thread·Permalink

Great presentation

Martin.Holzherr

21:10 24 Aug '05

Hey, you really invested a lot into this article. It is real fun to read it. Martin
Sign In·View Thread·Permalink

Re: Great presentation

iberg

21:32 24 Aug '05

Great to here you like it. Thank you very much.
Sign In·View Thread·Permalink

ATI card problems

iberg

10:59 16 Aug '05

The version available for download here will not work with ATI cards. I believe the reason lies in the way the ATI display drivers work if MFC is involved. Using MFC and OpenGL probably brings you to the dusty corners of Windows programming where strange things can happen.

The main problem occurs when using the standard OnPaint event handler for the drawing of the screensaver. Thus it can be fixed by removing OnPaint completely and putting the drawing logic in the OnTimer function instead. Thus the CPaintDC call that causes the problem can be omitted safely.

I will publish a new version within the next days/weeks that will solve these problems.

Re: ATI card problems

Ryan Beesley

12:32 23 Aug '05

I know from my managed DirectX I've been working on, I have to declare the window style as SetStyle(ControlStyles.UserPaint | ControlStyles.AllPaintingInWmPaint, true);. By seting these two control style bits, it fixes the infighting that comes from painting somewhere else and invalidating the form, triggering the base OnPaint over and over.

If what you're saying is the bug, something like this would fix the problem, I should think.

I must admit, I haven't looked at the code recently, and I don't have MSVS installed on my laptop right now either, so you may have already done something like this. If not, this is something that you're going to want to do anyway... even if other graphics cards aren't causing you problems.

Re: ATI card problems

iberg

1:42 24 Aug '05

Thanks for your reply. I'm not sure whether it is a bug or an incorrect Windows API Flag or style. Since it ran flawlessly on any other graphics card I blamed it on ATI. In fact the hint you gave seems to be exactly what I was looking for. I posted to several newsgroups and gamedev.net but gut no reply that could solve the issue. As a last resort I removed OnPaint completely and put it in the timer callback thinking by myself that this must then be the only way how OpenGL and MFC can work together flawlessly on ATI cards.

If I can fix it with the SetStyle as you suggested I will use this since I guess this is the better way to do it. In fact you did me a great favour by pointing this out. Since I had no idea what the original problem was. The whole issue left a bad taste, since although the fix I applied is a clean solution and probably has even fewer overhead than a OnPaint call, I just hate it to work around problems without finding out what went wrong in the first place.

So anyway a version that runs on ATI cards should appear here in the next few days since i submitted it last week.

Kind Regards,
Ingo

"Thus you have an infinite world on a finite grid."?!

PedroMC

3:43 11 Aug '05

Thus you have a finite world on a finite grid. Just because the "Wator" has no boarders does not mean it is infinite. "Wator" is a closed finite world. Regards
Sign In·View Thread·Permalink	2.00/5 (1 vote)

FISHes

Jouat

1:05 11 Aug '05

Hi, I remember having a copy of A.K.Dewdney's article about WaTor, buried in distant time and former place, and this morning I found your implementation. Thanks for that nice screen saver, but please: The plural form of "fish" is also just "fish", so would you please correct that in the program and on the web site. Many thanks again. Jouat
Sign In·View Thread·Permalink

General News Suggestion Question Bug Answer Joke Rant Admin

Use Ctrl+Left/Right to switch messages, Ctrl+Up/Down to switch threads, Ctrl+Shift+Left/Right to switch pages.

Layout: fixed | fluid

Hot News: Windows 8 boots too quickly, first world OS problem

The Code Project Insider. Free each morning.

Licence	GPL3
First Posted	4 Aug 2005
Views	156,730
Downloads	5,034
Bookmarked	77 times

WaTor - An OpenGL based screensaver

Table of contents

Introduction

Set of rules

Rules for fish

Rules for sharks

Modifications made to the original algorithm

Theoretical background

Screen saver basics

Installation

Configuration

Screensaver parameters

Limitations

Licence

References

Release Notes

Rev 0.90 (Beta): 31/07/2005

Rev 1.0: 18/08/2005

Rev 1.01: 21/01/2006

Rev 1.02: 03/03/2007

License

About the Author

Comments and Discussions