News

New Oracle SQL Data Modeler Integrates Subversion

• By John K. Waters
• January 31, 2011

Oracle on Monday released a new version of its SQL Developer Data Modeler. Version 3.0 of the free, light-weight tool adds collaboration support by integrating the popular Subversion open source version control software and bumps up its flexibility by letting users write rules based on corporate standards or design-specific criteria.

While used by everyone from data modelers to developers to architects, according to Sue Harper, senior principle product manager at Oracle, the tool is often used by data architects who build the high-level abstract analysis models known as entity relationship diagrams (ERDs), which are transformed into relational models.

The big upgrade in this release is the aforementioned Subversion integration. Subversion is a widely used, open-source, centralized version control system licensed by the Apache Software Foundation. It lets developers working in distributed teams keep track of current and historical versions of files and source code via a repository system.

"Most data modelers today want to work in teams where a number of people are updating the same data model," said Sue Harper, senior principle product manager at Oracle. "Integrating with Subversion allows users of the tool to put the entire design into a Subversion repository, to which others have access. That makes it possible for everyone on the team to update it. Subversion is very powerful, and many of our users were requesting this support."

This version of Oracle's DB modeler also adds another often requested upgrade: new support for user-defined rules. Version 3.0 also includes a large set of pre-defined design rules that developers can useto help identify any errors or inconsistencies in the design. "In this release we're allowing people to create their own design rules using a scripting engine, and possibly some JavaScript, to run through the whole model and see if it has conformed to their rules," Harper said.

Along with this new ability to create user-defined rules, this release adds support for user-defined transformation scripts. This feature makes it possible for developers to make global changes or updates to a design -- things like adding a set of standard columns to all tables, or adding or removing prefixes to tables or columns in a model, Harper said.

"We think a lot of users actually point the modeling tool at the database, which allows them to connect to an existing schema and then reverse engineer or pull in the definitions to create the diagram," Harper said. "You might have an application developer, for example, who's not necessarily updating tables, who wants to see what the relationships are between the tables. So they would reverse engineer to create the diagram, update the diagram, and then forward engineer it to produce the script to update the database. That's something this tool does very well."

SQL Developer Data Modeler was first introduced by the company in July 2009 as a direct response to demand from users of SQL Developer, Harper said. "We tried to dissuade them," Harper said. "We said, we're just working with SQL Developer. We do queries against the database and that's it. But they were insistent, and they specifically wanted support for entity relationship modeling."

The core technology for the tool came with the 2008 acquisition of a Belgian company called Ikan.

Oracle SQL Developer Data Modeler 3.0 is designed to be used with Oracle Database 11g, IBM DB2 for Linux, UNIX and Windows (V7 and V8), IBM DB2 for OS/390 and z/OS, and Microsoft SQL Server 2000 and 2005. The tool also imports data models from Oracle and non-Oracle products, such as Oracle Designer and CA Erwin (Data Modeler 4.x and 7.x).

The code is available now for free download from the Oracle Technology Network Web site here for customers with an Oracle database license.