read the previous posts (from my first).
probably because the student doing implicit skinning is also doing an internship, and will probably be unable to do GSOC 
I think Elastic implicit skinning is an improved version of implicit skinning, so they may implement it.
Honestly, I wish I had time to implement it myself but I can’t, the Blender’s armature deformer performance is not good (cpu side), hope they improve it too.
IIRC, the reason that Elastic implicit skinning would not be implemented (at first at least), is that it doesn’t use any of the existing code in blender for posing the mesh.
The “old” Implicit skinning method uses the posing systems already in place and then preform it’s own mesh surface solving method.
While the new Elastic implicit skinning is more of a simulation of the skin. Think of it as if you applied the cloth modifier on the mesh instead to figure out how it would move. This means that you have to simulate how the mesh moves from the base pose. If your simulation is too simple, then the mesh will do weird stuff.
So it’s a trade off. Do you want the more simple (but also faster) method to pose your charater, or do you want the slower and but higer quality method?
The author talks about these drawbacks for the different methods in his latest paper on this page:
http://rodolphe-vaillant.fr/?p=papers
Thanks ZedDB. I’ll try to find time to read this. But to me, from the Siggraph video demos, elastic implicit skinning looks so freaking amazing and robust compared to implicit skinning that I wouldn’t care to wait for simulations to run at all. In any case when we would animate we would use a proxy using the fast old way and we would use the simulation only at render time, just like they do in big movies with realistic CG characters when they simulate all the anatomic layers.
In any case I have to try karab44’s muscle system, it looks so dope. Pretty rad. Duuude, siiick ! Sorry.
So is the implicit skinning cancelled or not ?
IIRC, yes it has been cancelled.
I just saw the developer meeting notes and had a bit of hope but I probably misunderstood the “All students reported in with a first week report”
I really hope explode modifier gets vertex group support without the viewport updating in the way
Speaking of skinning methods, have you guys seen this? It actually looks pretty straight forward to implement:
https://www.disneyresearch.com/publication/skinning-with-optimized-cors/
@illumin@i , I don’t see anything along those lines in the summerofcode page. What makes you have hope for that ?
@Indy_logic, didn’t know about this method. That would already be a big improvement, although nowhere near elastic implicit skinning.
I also just found this one which seems as good as elastic implicit skinning but apparently performs faster :