• Write a terrain creator/editor plugin
• Write a wysiwyg shader editor
• Write a wysiwyg particle editor

pros: pretty much localized tasks, interesting on their own. the tool system is pretty much in place so they would focus on the interaction with a small part of the crystalspace engine. on top of that, the tools could be run stand alone (no need for “full” ca environment).

cons: the terrain tool should go with the support for the cs view, including paging terrain system. the shader editor could include some node editor (at the moment we consider using http://delcorp.org/~caedes/pygraph for all node editors in the program.

• Write a physics workbench, where we can play with physic properties and experiment

pros: cool! cons: maybe its a bit early for this.

• Make bindings for some libs we could need (file importers, licence analyzers, etc)

pros: yeah, why not :P cons: boring tasks. probably pretty trivial generally.

• Threaded loading and caching of CS meshes; Reduce/eliminate interface unresponsiveness when loading new assets.

The task would cover both threaded loading of single meshes, and paging in an out data as the focus changes on possibly several windows.

pros: really important task, and more or less isolated. cons: requires cs 1.9. would be even better if a system to page metamodel data was added too.

• Platform independant file/directory watcher; Detect changed assets and reload them.

pros: it's a needed part for the asset management system. cons: looks a little bit undefined task.

• A (C)/Python library for tracking files based on heuristics (such as trying to find a filer after it has been renamed, moved etc)

• Transcoding library in (C)/Python for complex assets such as Blender, 3DSMax, Maya, Collada, etc

• A library that tries to find under which licence is a file ((C)/Python)

• Webbased user-interface for asset-management; Assign rating, categorize, assign state of completion, … to minimize/filter 'clutter' in the in-editor asset selection and control which assets can and cannot be used by artists.

pros: This is an interesting task about designing a nice workflow with an editorial part which is very important and very often overlooked. It could be very much focused in the theoretical part.

cons: It's best if the asset manager is advanced by the time the task starts. For best results seamless integration of the caf data library with some web framework would be seeked, like say django seamlessly working on metamodel data (or the other way around).

• Study and collect the needs of the CS users and write a big specification document.
• Make mockups of UIs based on this document

pros: Very important task, not so much about programming, still a big investigation and proposals would be welcome. Something that rarely gets done otherwise :). (we have been chasing someone to do this kind of task for some time)

cons: Not so technical.