On Sat, 2001-10-06 at 17:17, Bob Campbell wrote:
Can you be more specific?
The application being built is a fairly complex loan input system. Most of the application will follow a step by step sequence through a series of forms. There is no existing data. All of the data is completely generated by the user. The users will input dozens of fields with information through the first 2-3 forms and then part of that information will be queried and visible in many of the other forms. With the amount of fields and forms, the application is slow. Very slow. All of the coding is done with ZSQL methods and dtml. The database is in mySQL and set up as an ODBC data source.
Hmm. So you are collecting the data from the user, storing it in an RDBMS, and then querying that same data to generate subsequent forms? What do you do with the data at the end of this proccess? What do you do with this data if the application is abandoned at various stages?
2. A number of people have asked me about session tracking and caching. I don't know enough about either one to give an intelligent answer. Are they applicable in this type of situation?
Absolutely, if I understand your situation correctly. Storing the interim data in a session object will keep it in RAM as long as the user session continues. If at the end of the application proccess you need to store the info (or some subset of it) in an RDBMS, you should be able to do that with a single RDBMS transaction. This will have the effect of increasing the performance of your application, as it eliminates the unneccessary storing and retreiving of data in the RDBMS. HTH, Michael Bernstein.