User blog:RottenBlock/System 1 Compilation, Part 6

Short version: updated version of Sys0Decompiler, and have done a first pass translation of Little Vampire Chapter 1 (of 3). That's a 33% of the game by most metrics (chapter count, file count, Japanese walkthrough, it's kind of freaky, actually), or 40% by file size. Game will play with the version of Sys0Eng that I included in the previous post, no updates required for that. No H in chapter 1.

Definitely going to give it all an extensive rewrite later on so keep that in mind. Have made a few minor programming fixes to the game that I'll document in the launch post, not important right now. Downloads are at the bottom.

Created by Automated Tool?
There are several signs in the game's code that it was produced via an automated tool of some sort, rather than the pure code approach used in much later Alicesoft games.
 * Every closing quote mark in the entire game is proceeded by a shift back to white-coloured text, even those that were already white to begin with and could have rightly been written in just one line of code instead of three.
 * In several sections of the game where different menu commands have identical responses, the game uses inefficient coding, redirecting each action through a clutter-filled block of label jumps instead of going straight to their destination, as in cleaner, better-optimized parts of the game. This seems to have happened out of an effort (manual or automated) to make the results of each menu combination appear sequentially in the code, even if they were identical. I try to clean up these blocks, not so much for efficiency's sake (it was the 80s programmers who should have gave a damn about that!) but just to clear out the clutter.
 * Further evidence of automation exists in a few lines of code that clearly intend to say "0" (typically: because the party does not have a certain item) but say "(1-1)" instead, which makes no sense in general and even less sense in a boolean case like this. Honestly not sure how even the automated program would have done this.
 * Conversations with multiple steps have index numbers that tell the game what step you're on. But if one of the middle steps has its own variable-based branch (e.g. asking Tina about different dimensions), the index number will skip ahead based on the number of branches. No human would do that.

Version Differences
A Japanese walkthrough informs me that the final chapter is fairly different on the PC-88 (presuming this DOS port is based on the later PC-98 port). I'm a little curious how they mean, and might go to take a look at it. I've been keeping drafts for the version history and cut content wiki pages for the game, and it seems silly to miss something so big.

Talking about the PC-88 and worrying about MSX-like version differences reminds me that I need to try Gakuen Senki again once I'm at the end of this process, just to see if any of my decompiler errors were to blame for it turning out nonsense. I still think the game is some sort of proto-Sys1 game or a Sys1-"like" game rather than a full Sys1 game, but there's no hurt in giving it another try.

Downloads
Sys0Decompiler 0.4: MediaFire

Little Vampire - Chapter 1 First-Pass compiled data: MediaFire