Investment in Hub Developers is a Good Thing

Even if paid OSAF staff can develop Chandler 1.0 in a reasonable timeframe, it will be a better product with additional hub developers. Distributed development requires a commitment to communication, organizational structure and approach to make it possible for people to participate at a deep level. There are many reasons why OSAF has made this commitment, a few are listed below.

Smart People Everywhere. We think we�re smart :-), but we know that there are plenty of other very smart people in the world who can make Chandler a far better product than OSAF staff will alone.

Quality. We want Chandler 1.0 to be top-quality, well-tested product. OSAF staff cannot do this alone. Dedicated early adopters who care enough to notice, track down, and report bugs, as well as assist in managing the bug�s life cycle will be critical to the quality of product we ship.

Specialized Versions. We hope that Chandler will be the basis for specialized versions and Chandler-related businesses. For example, someone might want to create a version aimed at enterprise markets, providing support and other features important to enterprises. Anyone attempting to do so will want to understand Chandler inside and out and be involved in central project dynamics.

Distributing expertise. The product will be far healthier if there are several people with deep expertise in a given area. A group of peers can help each other by brainstorming, vetting ideas, challenging assumptions, reviewing code and generally getting things done. If peers reach a point where it makes sense for them to become the official code maintainer for an area, so much the better. We�ll have to figure out what �official� means, how authority is delegated and so on. This is a victory condition, and will be great issues on which to be focused.

Additional “Chandler-Based-Applications.” Chandler is intended to be both a useful PIM application and a platform for the development of additional information-management applications. These Chandler-Based-Applications will need to develop a data schema that meets the needs of that application and that works well with the Chandler repository, superwidget and other infrastructural pieces. This will be difficult without a deep understanding of Chandler. To encourage Chandler-Based-Applications we must encourage understanding in Chandler technology. In addition, it is notoriously hard to build a robust platform using only a single application to test its capabilities. Additional Chandler-Based-Applications will test and improve the strength of the Chandler platform.

-- MitchellBaker - 05 Jun 2003