Re: Does it make sense to have 2 unique keys in one table?

In article <45bgr9$9dr@toon.ctp.com>, ctwom@ctp.com says... > >In article <4569de$6j9@News1.mcs.net> "Burton M. Strauss III" <BStrauss@MCS.NET> writes: >>dchan@dragon.cpe.ulowell.edu (Empty) wrote: >>>if the columns are partially duplicated, like >>> >>>unique key1 unique key2 >>>column1 column3 >>>column2 column4 >>>column3 >>>column4 >>> >>>does this make sense? >> >> > >Creating 2 unique indexes on a table can make perfect sense depending on >your situation. The last company I worked for allowed certain processes >(from a client company) perform transactions against one of our tables, >and insert their own unique identifier upon (successful :-) completion. > >The client would then be allowed access to that information using their >own index structure/account id. > >Each installation has unique requirements - and if yours requires 2 >unique indexes to properly identify a record, so be it. > >Chris > >---------------------------------------------------------------------------- > Christopher Twombly Voice : 617-374-2056 > Cambridge Technology Partners Email : ctwom@ctp.com > Cambridge, Massachusetts 02139, USA #include <std_disclaimer.h> -- I agree with Chris - as long as you're these two identifiers identify the same thing, which sounds like what Chris's company did. If, however, the table is keeping two types of entities within it, and therefore two different id's, you have to ask yourself why they're together in one table. -Chris Krahe, Columbia, MD "A car does not the driver make." -cpk '88