Hello
I am studying this method...
indeed i get good results on block processing (ogg, png)
i've also realized that softs like oggre or lolz compress a large block better than several small files
(for example, for oggre, deduplication is better on a large file than on individual small files)
in short, I like it, it's time-consuming to build and takes a bit longer to decompress, but it works well.
I'll take this opportunity to share the little program I compiled from source code this weekend.
It's sfk's partcopy function alone.
It just saves a few MB on the final archive by not copying sfk in its entirety.
to use it :
Code:
partcopy re_chunk_000.ok -fromto 991941787 1309235054 1.dat
(remove sfk at begin and -yes at the end...
@For L33THAK0R:
this instruction is for copying a part of a big file (re_chunk), we take a portion begin at 991941787 and finish at 1309235054 and export this bloc to a new file called 1.dat
in simple words, we concatenate small files in a big one , compress with your favorite program, because you concatenate same kind of file, optimized compressor will give better results
then during install decompress then separate files
best regards