**********
01/18/2008
**********

Super Robot Wars Complete Box came yesterday. Tonight I thought I’d do a little preliminary work and get acquainted with the Playstation a bit better. This marks my first foray into the world of Playstation hacking. This is something I’ll have to keep in mind in the coming days/weeks as I learn the hardware.

New hardware. New game. Where do I begin? I want to get the script dumped as soon as possible. That way Suzaku can be chipping away at the script while I work on implementing all of the changes that an English translation calls for. The most important of which is a VWF, but I’ll worry about that later. For now I just want to find the script and get it dumped.

Since this is a new system, I’m spending quite a bit of time reading up on the Playstation. As far as literature goes I’m currently reading: “Everything You Have Always Wanted to Know about the Playstation But Were Afraid to Ask”; “MIPS32 Architecture For Programmers: Volumes I-III”; and various documents by doomed that cover a variety of topics. I’ll probably spend a few hours this weekend surfing the web for additional information. Remember: there is no such thing as too much information.

At the moment my toolkit includes YY-CHR for graphics hacking and exploration and pSX as my emulator of choice. I’ve followed pSX off and on since it came on the scene and I’m really impressed with its debugger. The only feature that I miss is the trace logger, but I can probably make do with the debugger. If worse comes worst I can use PCSX’s tracer.

I want to find the text routine and I don’t have any idea where the text is on the CD or anything else about the text. I’m not going into this completely blind, however. I’m going to make a few (fairly safe) assumptions about how this game works.

First, the game has to read the text and then draw the corresponding tiles that represent each character of text. That means at some point the tile data for the font characters drawn will have to be loaded. Second, it’ll be a lot faster if instead of reading each tile off of the CD every time, the font were stored some where in RAM where it can be easily and quickly read.

With that in mind, I fire up the game in pSX and proceed to the first string of dialogue in SRW2. This happens to be one of Bright’s lines. Using pSX’s nifty debugger I dump the entire 2MB worth of RAM to a file, which I then open in YY-CHR. From there I scroll through the RAM dump until I find the font.

Now that I know where the font is at, I just need to find the first kana in Bright’s dialogue. From there I can set a breakpoint and work backward until I find the text routine. The reason this will work is because first a character is read from the string. The hex value of the character is then used to determine the address of its corresponding font tiles. In between point A and B, the hex value of the character will most likely undergo a number of changes. So, you find the address of the font tiles, determine where the address is coming from and work backward until you find the starting point, i.e. the point that the character is read from the string. From there you can find all sorts of things: the pointers, control codes, and so on and so forth.

That’s the idea anyway. Right now I’ve only managed to find the font and determine the point at which the kana is loaded. This is usually where I bust out a trace logger. I let the tracer log the instructions up until the tile is read and the work through the instructions leading up to the load until I find the text routine. Fortunately I learned how to crack games without tracers, so this isn’t a big deal. I’ll just have to be creative, but that will have to wait until tomorrow.