I have a bunch of tables that I'm joining in a view. I need each resulting row in the view to have a unique id...not necessarily a guid, but some value that uniquely can identify each row...and further, can be decomposed into the corresponding primary keys that were used to construct the unique value.
At first I was using the formula suggested here: Formula for Unique Hash from Integer Pair
...and that's worked well, but now I've run into the problem that Ids may run outside of the integer address space...and further may even run (slightly) outside of the bigint address space.
Can someone offer a suggestion as to how to work around this (perhaps by creating a composite guid instead of a numeric value)?