Originally Posted by vectorcatch
The biggest difference is the block size, if I remember correctly jmfs is doing everything in either 512byte or 4k chunks. That means the code reads it and the writes it for each chunk, so there is a ton of overhead. Most modern devices deal internally with much larger block sizes (bandwidth is king over latency here)
Actually if you run the dd part of jmfs manually though the terminal you can control the block size and speed things up significantly.
Typically the block size is small with something like dd if you have a bad drive and you want it to get as much data as possible so you read small chunks, so it won't fails large sections.
Technically I think that's the
part of jmfs, and not the original
although maybe he included both on the cd, the way MFS Live includes both
(which is not to be confused with