It really is. I've been really excited about this project. Although the client rebuild isn't really progressing yet for above reasons, the alpha release is almost ready. Server is functional and English client is patched and working too. We are still translating the DB and fixing bugs, but nearly there :)Quote:
This is really interesting. MAM was godlike.
So these files are still being tricky little bastards. No variation of the HSV seems to give me correct or expected colors. Thus, I seem to have hit another wall. I haven't yet found the flag which decides vertical orientation / origin pixel position (tells you to draw from top/down left/right), so some of the sprites are still flipped (eww).
Looks like my best bet might really be to reverse engineer the game client to understand how it reads and operates on the RLE files. Although I have had some success in reverse engineering different packet calculations (including a nasty little bug with a battle action packet), these RLEs still insist on being the death of me. I tracked down the code which finds the '.RLE' extension, but it looks like it just results in storing the file in memory, then continues on its merry way doing other non image related operations. So something must pick it up from memory to parse the file and draw to screen. Problem is hunting down WHAT does that.
For any of you who might've worked on reversing conquer, are there any tips you may be able to provide? I know a lot of their assets are in c3(2.5d~) format, but perhaps there are certain sprites or 2d images that get used and drawn using similar methods. I see a lot of references to borland TForms, and various import drawing libraries, but these seem to handle drawing aspects of the forms, and not actual sprites. But perhaps I am making a bad assumption there. Just because there isn't a visible form doesn't mean the monsters aren't drawn on some form element.