This talk will discuss the various development decisions made by xTuple (formerly OpenMFG) over the past six years with its PostgreSQL-centric applications, and look to the next-generation architecture currently in the planning stages. We will discuss:
- Putting business logic into the database via PL/PGSQL and constraints
- Where/how best to handle errors
- Designing updatable views for the API and data import
- To normalize or denormalize?
- Why build Yet Another Report Writer?
- Tools to check for data problems before upgrading
- GUI or Web: does PostgreSQL care?
- Scaling with PostgreSQL: where's the ceiling?
- Future-safe scripting that lives in the database
- A more extendable, modular future
John is the Director of Product Development, and Gil is a Senior Software Engineer for xTuple. They are both intimately involved in both overall architectural planning and day to day coding. John's background is in ERP and operations; he came to xTuple from a customer site, where he implemented OpenMFG and contributed several major code enhancements to the product. Gil has a long history in database software, including leading roles at Informix Software and middleware vendor Brokat Technologies.