TE-Code Roadmap

The current plans for the two editions of TE-Code are to continue to stabilize them in preparation for a 3.0 release. Originally, this release targeted for Q4 2004, but several things have happened which are likely to cause this date to not be met. The rest of this page discusses the plans for the near future of the two editions.

The Java Edition

At the present time, the main outstanding tasks to be completed before the 3.0 release are (in no particular order):

• Extend the JUnit test suites
• Complete the port of the configuration classes from the .NET edition
• Complete work on the validator package
• Update the existing JavaDoc documentation
• Ensure compatibility with J2SE 5.0
• Merge te-code and te-trace into one distribution
• Create additional examples

Other than the inclusion of the new configuration mechanism and completing the work on the validator package, no new code is expected to be added to the library. It is not yet clear, however, how long it will take for this to happen as I am only able to work on new development outside of normal work hours.

The merge of te-code and te-trace should be straightforward, but it will also result in some of the classes using alternative mechanisms for debugging to be migrated to use the trace package. I do not foresee this being a large amount of work.

The current target for a final 3.0 release of the Java edition is Q1 2005. Hopefully, it can be closer to the middle than the end.

The .NET Edition

The current code in the .NET edition is stable, but it does not yet have a feature parity to the Java Edition. That being said, some aspects of the .NET edition are more advanced (the configuration system, for example) and have caused changes to be made to the codebase of the Java edition.

The key outstanding points to be addressed before a 3.0 release are (in no particular order):

• Adding support for "native" XML configuration providers
• Adding support for Windows Registry configuration providers
• Implement the Filter package from the Java edition

If the above items could be completed, while not being exactly feature-for-feature compatible with the Java edition, the core functionality of the library will exist across both editions. Once this has been done, the following items are high on my priority list for additions to the .NET edition:

• Examine the persistence package and determine if it makes sense to implement for .NET
• Add a generic layout manager system similar to that described in Layout Managers in C#
• Provide a similar, configuration-driven method of defining user interfaces as in the Java edition

I don't believe that implementing the remaining functionality for the 3.0 release will take that much time. However, I don't think it will happen before Q1 2005 unless I get a lot more free time than I currently have. The code that is there is stable and should be isolated from changes during the completion of the 3.0 feature set.

The Python Edition

The Python code which exists hasn't been touched in quite a while. I thought that I was going to be using Python again for a small project which would cause some of the rest of the library to be implemented, but this project has not materialized. Until I actually have something to do with Python, I don't think the scope of the library will change. Currently, it only contains an implementation of the Traceable interface an a version of ErrorTrace from 2002.