Makes sense. Then you can add some kind of index to the file so you can jump to the parts you want. Just track a list of chunks, so they don’t have to be contiguous.
Yeah, and we could give that huge file some kind of internal and logical structure to find those chunks, maybe something hierarchical with human readable names!
Makes sense. Then you can add some kind of index to the file so you can jump to the parts you want. Just track a list of chunks, so they don’t have to be contiguous.
Yeah, and we could give that huge file some kind of internal and logical structure to find those chunks, maybe something hierarchical with human readable names!
Yeah! Now we’re cooking! I bet we could really improve resiliency and seek times with different data structures, too. B trees?
Or maybe we could keep the flat structure and have some kind of syntax for querying the data… some kind of structured language.