Leaderboard

Thanks I will have to inspect that then. I am not sure how it tricks them though. If it uses the same string as 0e1 0e2 0e3, then edge storing is supported? Also does it work for all SOPs? I remember a lot of SOPs limiting their input, to points or prims, so even selecting what you want seem to get converted. For instance the Edit SOP, if I move a single primitive that has 50 points, it will store that in the group field, instead of my primitive, which actually makes things hard, because if I want to reselect that primitive after changing the topology, I want to be able to type the index of that primitive and have the same transformation applied to it. I wonder if H12 will have true edge support. I don't think so though because if it did, it would be in the H12 newsletter But if that isn't done when the geometry engine is rewritten from scratch, I can't think of a more drastic change to make this happen

If you recook the node it will be updated. I tend to use `opname(".")` instead.

Take a look at hou.selectedNodes(): http://www.sidefx.com/docs/houdini11.1/hom/hou/selectedNodes Also, you could use kwargs["pane"] to get the current interface pane and call currentNode() to get something. It's not exactly the same as hou.selectedNodes(), but it's a different way about going at it.

Do you need to select these edges procedurally? If so, I've been using this asset. It will create a string of edge loops "0e1 0e2 0e3". etc. etc. and assign it to a detail attribute. I usually just copy/ref the parameter to where I need it. It's been one of those very handy assets I use all the time. You can also use a mask to exclude edges you don't need. I didn't make it. Downloaded it from somewhere. edge_group.otl

See example. The accuracy it rather limited though. Note also, that Mantra rendertime subdivision needs some consideration in use. For example it renders much slower than polygons, it take much more memory, it ruins any instancing optimization. These days almost always subdividing a mesh in Sops works better, as computer can handle dense meshes pretty well. Rendertime subdivision works nice in character pipelines, when you have a dense mesh still not enough good for rendering. For things like screws may be not as useful crease_example.hipnc

I used the Crease SOP a couple of times and it works really well with edges too. You have to select those edges explicitly in the viewport and add the Crease SOP from within the viewport to make it work. Unfortunately the Crease SOP is not edge aware by itself and reselect for current tool does not work as expected. I'm always creating an Polybevel SOP (or any SOP that uses the weird syntax for edge selection) and do my edge selection there and reference its group parameter inside the Crease SOP. In your case I would however also tend to use real geometry to sharpen those edges. It gets cleaner and more predictable results. I tend to use creasing when I want smooth transitions from hard to soft edges where its hard to create a nice egdeflow for newly introduced geometry. -dennis crease.hip

Make it a permanent default and you'll be even faster then typing $OS

Ok guys, this will be the first challenge that I ever created so I hope that it is a bit of a challenge for you guys:P This challenge is different then the ones that I have seen anywhere so far, because it has rules. So here is the Challenge with all the rules and goals: GOALS: -Create the most effective DoP Network possible -there has to be at least one fracture Object in the scene -at least one liquid simulation -one cloth simulation RULES: -you are only allowed to use 3geo nodes on the OBJ level -only on DoP is allowed -as many fluid containers as you need and one ground OBJ -as many VOP and SOP networks as you want INSIDE THE GEO NODES: -you are only allowed to use 3 copy nodes in each geo node to copy the original geo(for modeling purpose you can use as many as you need) INSIDE THE DoP NODE: -only 3 static RBDs -only 3 other RBDs -max of 2 fracture OBJs -only one force(beside gravity) SIDE NOTES: -when you cache out a simulation and layer a different simulation onto that cached simulation it counts towards the coals but it does not count as two simulations in the DoP. I know this is totally different then normal challenges, but I always think that it is to boring if you can do what you always do. I hope that those rules forces some of you guys to think out side the box and learn something new:P Oh and the winner is the person that creates the most creative simulation(every one can vote for their favorite!) HAVE FUN GUYS