Howdy all, I'm starting to look into KineFX for character rigging and animation. So far I like what I see. Once you get over the initial learning curve I find that making skeletons, editing joints and keyframing stuff is pretty fun and easy. But! People are telling me that KineFX isn't ready for primetime and that the more I use it the more limitations I'll run into.

What limitations are there? How far can I get with character animation in KineFX before I run into problems? What do the old Houdini rigging shelf tools and/or Maya do better? Thanks a lot!

I've been using Kinefx in production on Indie game, it's definitely ready. To be honest, it needs few workarounds, but you can check some tools which I did for it: https://github.com/kamilhepner/kinefx_tools [github.com]
It's totally ready for the primetime. I've talked with people using it in production, and they more than happy with it.

If you running into limitations share them with us here, there is probably more than one solution to every problem.

I didn't like the old Houdini approach for rigging with those quirky python scripts it smelled like Maya's approach. I've rigged in Maya for many years and my personal answer to question what Houdini does better than Maya is: Everything.

depends on what you call primetime, it's definitely useable and as with any houdini nodes you can use them in production for whatever you want, but kinefx on it's own (even wrapped rigs) is far from user friendly for animators for example
blendshapes seem to be only linear
also rigging wise it's a bit limited, uses more traditional concepts and building individual simple constraints is easy and powerful, but it works in more or less linear fashion where you have to keep stacking sops to get updated transforms, to go beyond and define bi-directional constraints, efficient history dependent constraints or also constraints that go beyond just affecting transforms, you'd be out of luck for out of the box solutions, motionclips also as far as I know store and blend only transforms so its still far from what you need for production ready character pipeline, and will not avoid building your own workflows around it in such cases until it matures
lot's of that comes from limitations of VOPs and SOPs and how they are evaluated which would be difficult to overcome without SESI's support

Thanks for the comments guys, very interesting!