Bonknuts you're right in that the buffer shouldn't matter but for X2 at least it depends if you get lucky
When decompressing, X2 will load it's prefilled buffer and somehow it'll get messed up when it decompresses my compressed version that uses 0x00. Not sure how it's doing it as I can decompress mine and theirs fine regardless but it's some quirk that I must not have included in mine. So I get around it by just not compressing any of it and just making it look like it is. Figure that'll work for 99% of the cases hopefully
Anything is better then doing a buffer for each block (500+)
Edit: Just thought about that you would need to know the original buffer IF they end up using the original/latest LZSS algorithm (I've seen an older one where it doesn't use binary trees). The original LZSS has a quirk/unintended side effect (I suspect) of RLEing data that matches what's in the original buffer setup. So in that case, yeah you would need to know how the original buffer is setup
LentFilms, what's bad is I have so many I tend to forget about which ones need what work
