About a month later than I had hoped, OpenCLAWS is publicly available for download from our Subversion repository. I spent about two and a half months scrubbing the source tree in preparation for public use. It was a long process. I had to learn about the design and flow of each component, and then go about refactoring wherever needed. The end result is the removal of a lot dead AMS code, consolidation of duplicate functionality, bug fixes, and standardized coding conventions.

Right now OpenCLAWS is not very useful. Only the central server (CAT), history module, and identity module have been imported. What this means is that it’s possible to perform identity management operations with history tracking. All other functionality will come in the next few months. In the mean time, the existing components need to be thoroughly tested. The handbook should help you build and install OpenCLAWS. I haven’t written the documentation for making CAT and the modules work together, nor have I written any SOAP API documentation. So (for now) you’re on your own for that part.

After considerable thought, I’ve decided to scrap the CLAWS web clients and PHP API. It’s hard to reuse those components since their design is heavily influenced by ITS business processes. I’d rather separate the communications API from the business logic so that it’s easier for other organizations to adopt. Further, I’d like AdminTools to follow an MVC pattern like the new Start client. I expect this portion to take about two months (target end of October-ish).

At the same time, I intend to import modLDAP and write a new Active Directory module. This task will largely depend on how long it takes to write a PHP API and a minimal web client. Once I have a usable client, the two tasks will continue in parallel. I’m shooting for full Active Directory and LDAP support, and a new AdminTools client by the end of November.

For those who might be interested, I’m maintaining a list of compatibility issues between CLAWS and OpenCLAWS. I’d like to maintain compatibility of the SOAP interface for interoperability. At some point, I will flush out all of the API and design documentation, which should hopefully be a good reference for both projects.

I only worked in that position for a few months, but that experience gave me valuable insight into the software’s design and implementation. After returning to the customer support side of ITS, I continued my involvement with the project by testing what I could and reporting bugs and feature requests. Later I successfully built and installed CLAWS in my own environment.

The plan was to develop features and bug fixes on my own, and then contribute patches back to the project. I made significant progress documenting the build and install process. I even had a fairly large patch set that I was getting ready to send upstream. And then… the decision was made to remove public access to the source code. My project has died, and I have since left RIT for greener pastures.

Moving to San Diego represents a new beginning for me, with new challenges, new opportunities, and new interests. Over the last few weeks, I’ve thought a lot about my experiences at RIT. And I remembered why I took interest in CLAWS to begin with: it was a brilliant idea with enormous potential.

I started thinking that maybe RIT has taken CLAWS as far as they can. And maybe the project needs a new beginning to reach its full potential. So I have decided to resurrect my project and steer it in a new direction, completely independent of RIT, with defined goals, and a plan to reach those goals. Also, with a new direction comes a new name that reflects my commitment to the open source community: OpenCLAWS.

I am currently in the process of scrubbing the source tree and applying my patches. As soon as I finish, I will make it publicly available on the OpenCLAWS website. After that… we’ll see where the community takes it.

Read the project mission for more information on the scope and direction of OpenCLAWS.