Scott Long wrote: > Again, I'm not exactly sure how a generic mechanism can handle the > distinction of data vs. metadata vs. journal data. Also, what you I don't care about the distinction at this level - all data is treated equal :)