>>Hi,
>>
>>I am testing how I could implement CA as a public object (or referenced by application object) such that it can be used over and over for different SQL commands. For example, when creating reports I need to send to DB a SQL Select and get back the cursor for the report. So my idea is to use the same CA object for all reports.
>>
>>Here is what I tried:
>>
>>1. Created CA (DataSourceType'ADO'), set all the necessary ADODB properties. Then I call CursorFill() and get the cursor for the report.
>>2. Then I close the cursor (use in ("mycursorname"))
>>
>>To get cursor for another report I do the following
>>1. Set new value of alias for the existing CursorAdapter
>>2. Set new value of SelectCmd for the existing CursorAdapter
>>3. Call CursorFill().
>>I get another cursor with no problem.
>>
>>My question is, do you see any potential problems with this approach?
>>
>>TIA.
>
>In OOP terms the design is supposed to be singular, and reused multiple times as instances at runtime. Your way, you will trade space for slowness, by having the overhead of constantly updating the single instance at runtime.

Thank you for your input. Obviously I don't want to slow down the report generation. And this is why I thought that by keeping the CA object instantiated at all time will speed up getting the data from SQL Server. What I don't understand is what you mean by "constantly updating the single instance"? If not updating the single instance with the change of values of a couple of properties, wouldn't instantiating it again, every time, will be slower?