Joe Breeden [mailto:[EMAIL PROTECTED]] wrote: ... > wondering about alternatives to embedding SQL in to the code > of a program. ... > It would be interesting to know how other people have solved > that problem. One approach is to use something like Ima::DBI, which I'm currently toying with. With Ima::DBI, you still embed your SQL in your perl code, but at least you put all of your SQL into a single module somewhere and you do so in a very structured way. To access the database from the rest of your program, you call methods of your database query object. This is a lot cleaner than whipping up a query string every time you want to hit the database. It's also a lot more flexible. You could, for instance, create different database classes for different database backends, and still keep the programming interface the same. Of course you could do all this without Ima::DBI; roll up your own custom database wrapper classes. But Ima::DBI also handles some mod_perl DBI issues such as guaranteeing one DBI statement handle per process. Michael