From: John Bell on 20 Jul 2010 16:32 On Mon, 19 Jul 2010 13:40:42 -0700 (PDT), Passiday <passiday(a)gmail.com> wrote: >Hi, > >> I am not sure why you aren't using GUIDs/uniqueidentifiers? >> Even though they are horrible, they are probably better than what you >> are currently doing! > >This project has long legacy, it's been ported through different >database platforms. In the environment where it's core structures and >code was developed, there was no such thing as reliable GUIDs. >Besides, I can't agree that the current ID generation would be so >"horrible" - generating new ID at code level is quite simple. > >Pavils I can't believe that you don't have scalability issues with the solution you outlined. John
First
|
Prev
|
Pages: 1 2 Prev: Using Float Type Variables for Primary on SQL 2005 Next: data type equivalent to double |