Anticlue

Mach II: The Tipping Point

After reading Peter Farrell's Mach II is dead, I think it is time that Mach II was released to the community instead of being the individual work of Hal and Ben. They have done a great job, but better project management needs to occur or MachII will fade away. First a project team structure needs to be set up, there should be an oversite group which included individuals such as Hal, Ben and Sean. This group ultimately approves what needs to go in what release and what ways to communicate to the general public. My next suggestion is to add a group of developers to do the building. Developers who want to contribute for the good of the framework. Also a team of documentation experts would do well to be apart of the team. A shortfall of the project is the lack of documentation; if the framework was for the community then there should be more documentation. A way for the community to contribute documentation, like MachII.info needs to be recognized.

The next big items needed are a schedule and a change control process. Please correct me if I�m wrong, but there are no schedules for improvement or change control processes. At first the change of the Mach II framework was rampant, almost 4 version changes in six months. Then there was 1 change within 6 months, and now nothing since October 2004. An anticipated release and functionality schedule needs to be shared with the community at least one that will take advantages of what has been offered with CFMX 7 and fix the plugin manager bug.

Thinking of the plugin manager bug, the other item that needs to be shared with the community is the bug list or issues. The issue list needs to capture what will be fixed, what won�t be fixed, and what doesn�t need to be fixed. This will give everyone a good sense of where the Mach-II project is at the current time. JIRA is available to most open-source projects for free. How many issues or bugs are there with the current Mach-II?

The final item is that throughout the project, there seems to be little or no planning to anything outside of the first code development, and this item is causing Mach-II to fail. There is no substitute to planning, and a project without planning will always fail. Good luck to Model-Glue to avoid this risk, just plan a little more communication, a schedule of releases, and a project team with community involvement.