> That makes a lot of sense. How practical is it? At minimum, there > would have to be a dedicated Venti to isolate "contrib" from other > sections, plus the arenas would be large: not everyone has broadband > access.
only the differences in the arenas would need to be copied. by definition, the differences in the arenas are not bigger than the differences in the files. bonus: you don't have to calculate the differences in the files. this is something i am doing with ken's fs. i think the outline is the same for venti. though given superblocks n, n+1 on the fs, one knows which data to copy as superblocks are labeled by w address (block number). the starting lba is waddr(n)*(RBUFSIZE/512) and you need to copy (waddr(n+1)-waddr(n))/(RBUFSIZE/512) disk blocks. not quite sure how that translates to venti. - erik
