Recent content by glennfishwick

Use NULL instead.

A query such as: select * from sysobjects o, syscolumns c where o.name = 'patients' and o.id = c.id Will return all the columns in a particular table. The colid column gives you a sequential ordering through the columns.

Does anyone have or know of any good (and fast) routines to anonymise data in a table. I need to jumble up customer details in our database for testing purposes. I have a script that does this but it uses cursor's and runs like a dog! All that I need to do is swap names around or update names...

Unfortunately the file that I have is in backup format so that will not work.

Does anyone know if it is possible to restore a database without its Transaction Log file. Like when you can do an attach_db without specifying a transaction log and a empty one is created. Our problem is that an unmanaged site has sent us a database with a massive transaction log that is too...

Is this a good idea: In an order processing type system we will currently create and order & order_lines as follows: INSERT INTO orders... INSERT INTO order_lines... INSERT INTO order_lines... INSERT INTO order_lines... <repeated for number of order lines> < then various other repeated...

I have a query that has been extracted from an application and I have tuned it a little. However I now get drastically differing performance times depending on where I run it (I have used DBCC DropCleanBuffers & DBCC FreeProcCache to clear memory and plans). i.e. Run from my PC it takes approx...

I am in the process of investigating the approach of using stored procedures to hold all our data access and business logic. I can see how you would use stored procedures to control database inserts. I can't see what the best way to control database Updates via stored procedures is. i.e. If...