Pivot as microsoft have developed it is dynamic. It has to be because it needs to create a list of the fields. You don't see this But it is there. These temp proc , tables etc.. exist because they were used as part of developing the SQL language. Its the way that the tables are created in SQL they are hardwired.

I started earlier on by pointing out that Microsoft use these in their own MACROS and Functions especially dynamic code.

I think that much of the middleware and processing should be at the SQL level! You protect access and make this very restricted ( I tend to use one Proc as an entry point) and pass XML data which is then processed and drives the todo's. The XML will have been created from User interface selection and will be precise as to what is required and what is returned. Basically store the key data as from away from the front end as possible and never with express direct user rights! Always building in Audit! Ever time the proc is used the user and request is stored with time and date!