Kristian believes that there is a lack of leadership and direction that has caused virtually no major work to get done since the Compiz Fusion creation (Beryl merger) in 2007. Compiz hasn't seen any new developers contributing significantly to the core development efforts since that time, but instead are losing their developers.
Blamed for this unfortunate state is the lack of project goals, inconsistent organization, and undocumented code. When it comes to the inconsistent organization, Compiz still has multiple bug trackers, developer mailing list, source-code repositories, etc. Kristian believes that these problems must be addressed before work on Compiz continues. He is also calling for clear goals to be defined for each release.
The Compiz community also needs to decide what to do with the branches of Compiz that are out there already. Last week we talked about Compiz++, which is a rewrite of Compiz in C++, separates the OpenGL and Composite layers, a new plug-in interface, and brings forth several other significant changes.
To read all of what Kristian Lyngstol had to write about the future of Compiz, read his mailing list post.