2009-10-10, 03:43 AM
Hah! Finally some success 
After 3 days of work (including working instead of playing games) I have SOM running in hardware accelerated mode with the following interfaces being intercepted:
IDirectDraw7
IDirectDrawSurface7
IDirect3D7
IDirect3DDevice7
Intercepting them is not the hard part... doing it without breaking anything is wherein the trouble lies. In this case my focus was working around the most fundamental DX graphical object IDirectDrawSurface7.? These things are passed around the DX API like crazy. It represents basically like the screen and textures etc. To make things crazier, DX is really a mess to work around. And I was also working with DX really for the first time in my life. As a software engineer I really can't grasp why Microsoft code is so expansive. I think to myself in the real world, maybe things are really so complicated (I'm looking at the COM interface for example) but I have some serious doubts. I mean I build systems which appear to be much more sophisticated than Windows and the amount of knowledge required to work within such systems is just dwarfed by Windows. And I'm just talking about the Win32 APIs... much less whatever the hell happens underneath the hood.
Anyway, I'll attach the logs for the first successful run. Some of the PANIC! lines are possibly cause for alarm (which I will need to look into) but since it's working I don't think the cause for panic is being raised by SOM. Like during the Flip() lines, I can probably easily draw anything over SOM before sending the Flip on to the real DX libraries. Actually Flip() needs to be removed and replaced with different calls if SOM is to run in windowed mode. Drawing stuff to screen is a little complicated by realizing whether SOM is in normal play or in a menu (or the on screen displays for that matter) but throwing up debug info shouldn't present any challenges.
I don't know if I will spend more time experimenting with windowing SOM or try my hand at drawing some info to the screen (like what's in SOM's memory/where) first. But the one thing I know for sure, is I'm glad to be over that mountain / probably I will relax for a while. When an objective presents unexpected challenges I tend to obsess over it until I'm in the clear... then pat myself on the back and try to makeup lost my recreational time
After 3 days of work (including working instead of playing games) I have SOM running in hardware accelerated mode with the following interfaces being intercepted:
IDirectDraw7
IDirectDrawSurface7
IDirect3D7
IDirect3DDevice7
Intercepting them is not the hard part... doing it without breaking anything is wherein the trouble lies. In this case my focus was working around the most fundamental DX graphical object IDirectDrawSurface7.? These things are passed around the DX API like crazy. It represents basically like the screen and textures etc. To make things crazier, DX is really a mess to work around. And I was also working with DX really for the first time in my life. As a software engineer I really can't grasp why Microsoft code is so expansive. I think to myself in the real world, maybe things are really so complicated (I'm looking at the COM interface for example) but I have some serious doubts. I mean I build systems which appear to be much more sophisticated than Windows and the amount of knowledge required to work within such systems is just dwarfed by Windows. And I'm just talking about the Win32 APIs... much less whatever the hell happens underneath the hood.
Anyway, I'll attach the logs for the first successful run. Some of the PANIC! lines are possibly cause for alarm (which I will need to look into) but since it's working I don't think the cause for panic is being raised by SOM. Like during the Flip() lines, I can probably easily draw anything over SOM before sending the Flip on to the real DX libraries. Actually Flip() needs to be removed and replaced with different calls if SOM is to run in windowed mode. Drawing stuff to screen is a little complicated by realizing whether SOM is in normal play or in a menu (or the on screen displays for that matter) but throwing up debug info shouldn't present any challenges.
I don't know if I will spend more time experimenting with windowing SOM or try my hand at drawing some info to the screen (like what's in SOM's memory/where) first. But the one thing I know for sure, is I'm glad to be over that mountain / probably I will relax for a while. When an objective presents unexpected challenges I tend to obsess over it until I'm in the clear... then pat myself on the back and try to makeup lost my recreational time