When I started mpFx, I was looking for a way to share with my community some of the work I have been doing. I had no idea I would see such interest. While many of them are languishing while waiting for code to be posted, the volunteers have been very helpful in simply calibrating the approach.
Sean Dai, in particular, helped focus the effort considerably. My initial concept, which involved developing a fairly substantial layer of code to integrate PSI data and functionality, has morphed into an approach that provides more of an API than an object model (not that the two are exclusive, but think of the difference between the Win32 API and an Office application's object model). I think that this will better service the community. After all, what I am attempting is a framework that allows PSI developers the opportunity to contribute their functions and data to a framework that reduces recoding. I think a heavy object model eventually results in to much unexecuted code and too much overhead.
Everything is experimental at this stage. I am really interested in setting the groundwork and then letting the developers fill in the blanks.
I am scheduled to do a webinar about mpFx for the Project Server Developer's Community later this month. This opportunity will allow me to articulate the vision and design goals in a public forum, which can only serve to improve the end product.
But, with two kids and a dog, plus a job of my own--work has been slow. As it turns out, my daily work has turned once again to the task of developing solutions with Microsoft Project. Over the next couple of weeks I hope to have made considerable progress with the experimental build.
In addition to the framework, I am developing design goals and some "spec-lettes", those brief write-ups about what should be included. If the team is reading, I would love to see some of your ideas about what should comprise the first release.
More tomorrow. Off to tending to the kids!