How does the SQLite DB work ?
Posted: Fri Jun 22, 2018 09:42
A question to you, core developers :
After working with SQLite(for MTPC), I have learned from my mistakes, and am feeling quite safe and have learned much. Now, I'm asking myself whether it would be good to modify the SQLite DB from outside(Java) to make map ops not in LUA, but from outside program; more specific, I want to have - Java program which creates Schematics from 3d models -> writes them directly into map.
So, I know how it basically works : You got indices & blobs. Each indice(int) stores blob(chunk) position in map and is primary key. Blob is mapchunk(which size ? 80 ?). How do indices work(which order ? x+y*mapsize+z*mapsize² ?). And now, the final question : How do the binary large objects work ? Of course, every node has content ID. So, are you just storing the content IDs in giant array, indiced by x+y*chunksize+z*chunksize² ? Where's the metadata ?
After working with SQLite(for MTPC), I have learned from my mistakes, and am feeling quite safe and have learned much. Now, I'm asking myself whether it would be good to modify the SQLite DB from outside(Java) to make map ops not in LUA, but from outside program; more specific, I want to have - Java program which creates Schematics from 3d models -> writes them directly into map.
So, I know how it basically works : You got indices & blobs. Each indice(int) stores blob(chunk) position in map and is primary key. Blob is mapchunk(which size ? 80 ?). How do indices work(which order ? x+y*mapsize+z*mapsize² ?). And now, the final question : How do the binary large objects work ? Of course, every node has content ID. So, are you just storing the content IDs in giant array, indiced by x+y*chunksize+z*chunksize² ? Where's the metadata ?