AmanoJyaku
-
Content Count
243 -
Joined
-
Last visited
-
Days Won
11
4 Followers
.thumb.gif.6c60790cc662fd9c860e8a379dd1b944.gif)
Recent Profile Visitors
Single Status Update
-
It's been a while, but I figured I'd drop a quick note about the NCS reverse compiler. I have, in fact, been working on it and documenting it. I needed a distraction from all that's happened over the last year and a half, which I won't get into as I don't want to spoil a day of celebration. I have some personal things that need to be done by July, so I won't have any time in May and June.
The good news is I expect to finish soon after. I've solved the problems other people have been stuck on, like recursive functions and finding structures. I need to perform extensive testing, and each test requires documentation, a review of the documentation, and identifying potential tests to perform. Additionally, as I've discovered errors in NCS files (not sure if vanilla or from mods), I also need to write a reverse compiler that can handle those errors to determine if it can produce usable output.
Long story short, I'm targeting a release date in September.
May the 4th be with you, always.
-
Glad to hear you are still chipping away. In your absence, @JCarter426 and I have taken a more brute force approach to things. JC hacked DeNCS to allow it to batch process both games and bypass certain common problems. We then manually reconstructed those that couldn't be decompiled by looking at the bytecode. We have decompiled almost all of K1's scripts in this manner, with only about a dozen or so remaining now (I've attached them if you are interested). The results are up on Github:
https://github.com/KOTORCommunityPatches/Vanilla_KOTOR_Script_Source
Most are just the straight output from DeNCS, but it also includes those that were manually cleaned up in the course of working on the Community Patches and other mods.
Things have stalled a little lately, so TSL is still somewhat incomplete. Although it didn't have quite as many issues as K1 did.
There were some interesting stats to come out of the process. Perhaps the most interesting tidbit was that - for K1 at least - seemingly all the reports of a bytecode mismatch were due to Bioware apparently using an older version of ActionStartConversation with one less variable.
-
