However, I am considering starting
https://dis.4chan.org/read/prog/1361556470/36-41,45
If I get a usable starting point for the engine, I will post what I have somehow.
Design:
1. User provides thread id.
2. Page is downloaded from shiichan
3. Page is parsed, posts extracted
4. >>N references are analyzed as replies and used to infer pseudo identities
5. Pseudo identities are assigned to loli characters in a school room.
6. OP loli walks up to group of other lolies, introduces herself as NAME if not Anonymous, and says Subject and Comment.
7. Lolies then speak to each other or to the group, using the >>N tags to direct speech at certain lolies.
Partial magic: Camera dynamically pans to capture the speaking loli and loli(es) being spoken to.
Very Magic: Content is analyzed for emotional content, be it anger, humor, etc. Tone is introduced into the speech synthesizer.
Very Very Magic: Add action and plot. Use more complex settings.