Nokia has announced that they will discontinue Qt Jambi development and turn it into a community managed project. How this will affect the Moonlight Application Framework and Moonlight|3D remains to be seen. The good news here is that the license will change to LGPL. This will enable us to offer the sources and the binaries under the LGPL. Up until now, only the source was under the LGPL, but the compiled binares were licensed under the GPL. This was a bit dodgy, but we can finally do away with it.

The last release of Moonlight|3D contained a few quite severe bugs that appeared when one used certain combinations of the Duplicate Objects, Merge Meshes and Delete Objects tools. These are mostly fixed now in the current nightly builds. The remaining fixes will come in the next couple of days.

Also, Duplicate Objects can now rotate and scale objects while duplicating and it can duplicate all children of the original scene element if so desired. This creates true copies of the scene elements rather than instances.

There is now a simple snap to grid feature available for the Transform Selection tool that snaps translation, rotation and scale to fixed units. Internally, the program is capable of much more sophisticated constraints for snapping and these will be implemented gradually.

These changes will probably make up most of Moonlight|3D 0.2.3, which will be released in the coming weeks. And there are already features in the pipeline for later releases.

Florian Niekiel alias rageon has been working hard on a set of icons that will become part of a completely overhauled toolchest in Moonlight|3D. In this new revision of the toolchest, the only text that will remain is the category headers. Everything else will be converted to nice, big, easy to hit buttons. Here are some examples of the icons that will be shown in the toolchest:

These icons are from left to right: create cylinder, create light, create simple material, delete selected, duplicate objects, make edge, select object and triangulate. Florian has also drawn an early mockup showing how the new toolchest will be structured:

In unrelated news, the Moonlight Application Framework is due to get a complete overhaul for its Property and Value classes and everything related, which will effectively be a major rewrite. The goal is to move away from a hardcoded list of available types and allow property types which are provided by plugins, removing a major limitation of the current implementation. This makes it possible to store much more sophisticated data types in those properties and property sheets will display the correct editors by default more often.