Splitting chains into separate edges



next up previous contents
Next: Storing the (cell Up: Design Tradeoffs and Previous: Using integers instead

Splitting chains into separate edges

This choice makes the data structure bigger, but each data element is now fixed length instead of variable. This is a similar concept to a relational database satisfying Codd's third normal form [ ]. Also the algorithm is simpler with fixed length items, and parallelizes better. Whether this was the correct choice depends on one's tolerance for complexity.



Wm Randolph Franklin
Wed Dec 14 14:03:28 EST 1994