Please start any new threads on our new
site at https://forums.sqlteam.com. We've got lots of great SQL Server
experts to answer whatever question you can come up with.
| Author |
Topic |
|
AskSQLTeam
Ask SQLTeam Question
0 Posts |
 Posted - 2001-12-16 : 11:58:02
|
| Tim writes "SQL Server 2000 If anyone is familiar with the ACT Contact Management System database, they use a 12 char field as the primary key for most records.This key is generated using the current date/time and some other random component.Example: R/M%! (WE!YI decided I prefer this to an integer identity because it isn't easily predictable, and removes the sequential nature of identity fields. They also use this format for date fields: W=L Z (same levelof detail as DATETIME datatype). I don't know why they do this though.Anyway, not having much experience with CS algorithms, has anyone got some pseudo-type code with enough detail to build some TSQL or C code to generate this 'random' key?I also welcome your thoughts on the use of this as a primary key in itself (the records have no natural key/s).Tim" |
|
|
robvolk
Most Valuable Yak
15732 Posts |
Posted - 2001-12-16 : 13:50:30
|
| graz has a section on "Randomness", and if you search SQL Team for "random":http://www.sqlteam.com/SearchResults.asp?SearchTerms=randomYou'll get a few ideas that you could adapt without too much trouble.You could also create a GUIDCOL, which generates a unique identifier. This is a 16 byte binary value that is guaranteed unique. Books Online will have details under "GUID" and "uniqueidentifier". Since this is built into SQL Server it will work with much less hassle than other methods, however the GUID values are completely cryptic (don't think you'll mind, just an FYI).The datetime format they use is probably a binary representation of a datetime value. I don't know if you'll be able to convert it to datetime, if it doesn't convert automatically. |
 |
|
|
|
|
|
Generating 'Random' String based Primary Keys