You are currently browsing the category archive for the ‘Hacking’ category.

Thanks to the efforts of a mysterious individual, the XOR key has been found and the script has been dumped successfully!

Now onto the slow part.

I will do at least a two pass translation. (not translation then QC but translating each line twice) This will take time and I will probably not update the blog frequently except at major milestones. I might tweet progress so if you’re curious take a look my timeline. If the project dies I will update this blog. Once I have finished the “Wafter” arc I will begin gathering a team for image-editing, patch creation, et cetera.

With the help of all of you I have finally made it to the starting gate.fist-pump Kud

It’s been a week since this projects declaration so I thought it might be beneficial to gather together the progress so far.

No script. No translation. It frustrates me but this is the reality of visual novel translation. This next week is my finals so I would love to start pouring over scripts after that but I still need help. Despite technical ineptitude, I’ve compiled my own binaries of RLDev (with help from AppleHQ) . The last step before translation is solving the XOR encryption.

After trying literally every file, ( I have no other method of attack) I have had limited success disassembling the (possibly?) system-level SEENs  9014, 9015, 9016, 9017 and 9020.

Most notably, I’ve had partial success disassembling the content SEEN 5001, the character list (which RLDev could already see) and the kepago calls were disassembled but the script stops after the first line (a title declaration). The bytecode and that data is available here.

A lurker found some Chinese “SEEN keys” and postulates following:

I used the same pattern as LBEX!/FIVE – looked at key 4 and it is simply key 1 xor key 2 and key 3 is just…well, I think you can figure out by observation alone. I was able to extract the SEEN but got syntax errors on disassembly – I’m not sure if it’s due to syntax changes in RealLive itself or code that wasn’t decrypted properly. But maybe one of the Chinese keys is actually key 4 and key 1 or 2 is key 4 xor the other, I dunno. If I get more time I’ll play around with it.

It looks like you run kprl -G KUDO -o kud_bytecode -x SEEN.txt or something to that effect and it outputs a ton of .RL bytecode files into “kud_bytecode”, which you [could] then disassemble file-by-file into actual scripts using kprl -d (script file).

However, although it doesn’t help that much, you *can* see the text by looking at the resulting RL files in a text editor – although I don’t know if it’s complete or if parts of it aren’t garbled.

The game.cfg entry I used is as follows:

If anyone can experiment based on his information you would have my gratitude. Luckily this is RealLive not some completely undocumented engine and I’m sure someone can crack this nut!

If  I’m to translate this game I need the script. Kud Wafter runs on RealLive which can be hacked with Haeleth’s RLdev kit.However I believe the script to KW is encrypted with a “XOR key” (cypher?)

Fluffy (LB! team) has provided the source to their build of RLDev.

I found hints on how to extract RealLive’s XOR key on the Gemot.

My understanding is if this key is put into game.cfg one can extract the script.

If there is one thing I learned in high school it’s that I can’t code to save my life. Unless anyone wants to help I might spend next weekend learning how to “compile binaries” ;_;

Additional Information

Kud Wafter uses RealLive Version1.6.3.4 and was released on 6/25/2010

These might be relevant but I can’t read Chinese (kudos to a lurker)

Update 4/26

I finally have my own copy of kprl and an amazing lurker may have cracked the code but work isn’t quite done (see comments and his game.cfg below).

Jyuichi on Twitter