I was trying to find a way to describe my difficulty with ICE modeling, you pretty much summed it up. I have no problem with the math side of using ICE, but even the simplest tasks sometimes become very frustrating when something goes wrong. When nodes go red they often lack any meaningful clues about how to fix whatever is wrong. I am sure there is always a good reason for whatever fails, but it sometimes seems - and in practice actually is - impossible to figure out what actually needs to be done to fix it. While I have no problem attributing most of the frustration to my lack of experience with the system, I can't help but think that it should be possible to improve the error feedback.
As a programmer I can appreciate how hard it would be to improve the situation all that much given it is essentially a visual programming environment and I suspect errors from low level systems just need to percolate up and there just isn't much to do other than drill down and figure it out.. I think perhaps the best thing we can hope for is more thorough ICE documentation, including examples of troubleshooting of common problems and debugging techniques. So much of ICE feels so barely documented that much of what I have tried to do so far has been difficult at best and borderline good luck when I do manage to figure it out. ICE rigging for instance has been a real challenge. The current docs & samples are IMHO not nearly as comprehensive as they deserve to be and I can only hope they get more attention. -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Andy Nicholas Sent: Tuesday, July 17, 2012 4:13 AM To: Rob Chapman; [email protected] Subject: Re: ICE modeling I find the most frustrating thing with ICE modelling is the lack of feedback when something goes wrong. For example, if I feed an array of point positions and indices to make some polygons using the Create Topo node, and something's wrong in that data, it's extraordinarily difficult to figure out what's gone wrong as there's no feedback or error message from the ICE node. Proper support for 2D arrays as a basic type would also be very useful. Particularly when trying to retrieve array type data from arrays of locations. That's a really common stumbling block for doing complex effects. On 17 July 2012 at 10:48 Rob Chapman <[email protected]> wrote: > Agreed, also how to 'grow' meshes from curves or animation data, on > these newly generated meshes to then select polys by area or closest > to a pointcloud ,merge surrounding polys to current selection, extrude > this etc. maybe how to blend between some instances and a ICE > modelling mesh. A controllable (by proximity, CAV map,weight,) > localized subdivision would be great. I have no clue how to get all > this to work in one ICE modelling tree together - have managed to > bodge together two different examples - a repeated extrude from a > random selection, but I cannot control the extrusion very well - but > this is the only examples we have so far that are not crowds or fences > :) > > https://vimeo.com/40452537 > > http://xsisupport.wordpress.com/2012/05/03/ice-modeling-extruding-a-ra > ndom-polygon/ > > Cheers > > Rob > > > > > > > > On 17 July 2012 10:21, Chris Marshall <[email protected]> wrote: > > What I'm finding is simply that almost everything I try either > > doesn't do what I expect or doesn't do anything at all. > > Some fundamental basic workflow examples would be useful, just to > > get the general logic clear. > > > > Understanding why repeat loops don't work and how to work around that. > > How to get animation of particles from a point cloud to build an ice object. > > I spent a day trying to get that to work with no luck. > > I'd like to be able to control the positioning of elements by > > translating or rotating them in local space. > > > > Thanks > > By lunch today would be useful!? > > ;-) > > > > > > > > On 16 July 2012 18:05, Grahame Fuller <[email protected]> wrote: > >> > >> What sort of tutorials do you want to see? Creating shapes from the > >> ground up, or something else? > >> > >> What I have found is that the topo nodes themselves are not that > >> complicated. The hard part is the math and logic you need to control them. > >> So maybe math tutorials are needed? > >> > >> gray > >> > >> From: [email protected] > >> [mailto:[email protected]] On Behalf Of Chris > >> Marshall > >> Sent: Monday, July 16, 2012 09:44 AM > >> To: [email protected] > >> Subject: Re: ICE modeling > >> > >> Yes!! > >> On 16 July 2012 11:50, adrian wyer > >> <[email protected]<mailto:[email protected] > >> om>> > >> wrote: > >> +1 on the request for ICE modelling 101 > >> > >> had a glance at the digital tutors one, but it's more about cloning > >> than actually 'modelling' > >> > >> this feature REALLY needs some in depth tutorials!! > >> > >> a > >> > >
