Hi there
I have a series of statements that needs to be carried out as a single unit (transaction). (If one fails, all should roll back)
I intend to use ADO and command objects on the mdb backend.
What is good practice?
As I build and execute the various appends and updates I need to do some SELECTs in between and I don't know if this will give me trouble. Should I open another connection from which I do all the reads and only execute the database changes on the other connection object, or can I just go ahead and use the same connection for reads and writes?
I will need some of the autonumbers created along the way (Select @@Identity) , which is why I can't wait executing the database changes until all reads are done.
How about locks? Could that be a problem?
Suggestions to a nice approach greatly appreciated.
I have a series of statements that needs to be carried out as a single unit (transaction). (If one fails, all should roll back)
I intend to use ADO and command objects on the mdb backend.
What is good practice?
As I build and execute the various appends and updates I need to do some SELECTs in between and I don't know if this will give me trouble. Should I open another connection from which I do all the reads and only execute the database changes on the other connection object, or can I just go ahead and use the same connection for reads and writes?
I will need some of the autonumbers created along the way (Select @@Identity) , which is why I can't wait executing the database changes until all reads are done.
How about locks? Could that be a problem?
Suggestions to a nice approach greatly appreciated.