db4o is an open and community-driven product. The main sponsor, db4objects, has subscribed to the notion of agile programming to enable instant reflection of new learnings within the developer and the user communities. At any time, the community can agree to adopt the most effective way to deliver value to the product's users, rather than sticking to a pre-formulated product roadmap, that ultimately may or may not reflect the needs of a market moving as fast as ours, once the product is actually shipped.
As a result, the db4o team has adopted an entirely open, user driven, and agile approach to determine, communicate, and to change its roadmap - accessible to all registered community members at any time.
Under this URL, db4o's core developer team shares all its development tasks, milestones, bug reports and other development-project relevant information, and shows the status of completion of all major enhancements, bugs and other tasks going forward. The tool selected is Jira, which is a non-intrusive project management tool that has become widely adopted in open source and agile development projects, e.g. in Apache, Spring, and Hibernate.
As of 7/20, Jira showcases the current roadmap planning for db4o in 2006. Based on the db4o User Survey, in which more than 1,000 users participated, and the discussions at the db4o User Conference in London, the core team has now come up with a balanced order of engineering priorities.
In response to user demand, the reminder of 2006 will be dominated by performance related work. The next projects will be carried out on fast B-Trees for field indices, OR queries for faster joined queries, and fast defragmentation. A core split and refactoring will enable the seamless implementation of transparent activation and extensive improvements on db4o's client/server mode going forward.
In addition, db4objects has committed to deliver against the most frequently requested wish - an improved ObjectManager/Database Inspector, based on user suggestions and a design session at the dUC - as well as updates on the dRS, including a version for .NET. Last and not least, db4objects will significantly increase its investment in documentation, which should result in further productivity enhancements and shorter learning curves even for most advanced db4o users.
Other Input Collections
Aside from the db4o User Survey, these input collections have helped us to formulate the roadmap plan as of July 2006.
A list compiled out of the view of the core team:
- Performance, performance, performance (ongoing and always with high priority)
- Lower runtime memory consumption
- BTrees on field indeces
- Fast defragmentation
- Improved documentation, samples
- Transparent activation
- ObjectManager 2.0 / Database Inspector
- .NET version of dRS based on NHibernate
- J2ME CLDC
The big wishlist ("no constraints") collected at the dUC: