Development on STARS was put on hold while the library PySAL was given priority focus. Many of the analytical methods first implemented in STARS were moved into the PySAL library during this time. Until PySAL was stable and had gone through several release versions, the plan had always been to keep STARS on the back burner. Once PySAL was stable, the idea was to rebuild a new version of STARS that rested on PySAL.
Another reason behind the long period since the last STARS update had to do with the thorny nature of heterogeneous data underlying space-time analysis. There were no standards for this data, and much exploratory programming was done by various people around the STARS project in this regard, but nothing ever seemed completely satisfactory.
Fortunately, since this period the library pandas has appeared on the scene and offers a very flexible, efficient and well documented library for dealing with this type of data. As such, the current STARS project is planning on making heavy use of pandas for data management issues which would free up development time to focus on some of the space-time analytical and visualization related issues.
Moving forward, there are short run and long run plans for STARS development. In the short run, there have been many requests for a version of STARS that runs on Windows 7. Because the last stable release relied on numeric (pre numpy days), some effort must be directed at refactoring the last version to meet these immediate needs. As of 2013-08-01 this is largely completed in the source living in the numpy branch. Binary installers are planned for the near future.
In the longer term, however, the focus will be on a completely new rewrite of STARS that rests on PySAL and draws on the pandas library. STARS will also serve as somewhat of a testing ground for new space-time methods that eventually will be moved into the spatial_dynamics library of PySAL.