So I am trying some new stuff here. I see many "user friendly" software use stacks instead of Nodes. This is possible in this case, and you would still be able to go in and change individual operations inside ICE... Here are some pros and cons and I got a little question for users at the end. Oh and I also posted pictures of both workflows...

Stacks
The Ups.

Just double click an operator directly from the explorer to inspect it
Reordering of operations through the stack is easier
You can always go in ICETree mode and customize it yourself

The Downs:

No multiple Roots (for having different point generation and modification inside the same object)
User might have a simplified vision of things and miss oportunity of greater customization

ICETree
The Ups

You are always working inside the place where you can customize the effect the most
Separate multiple points generation and modification with a Root node
It is really the default way to work with point clouds in SI this days

The Downs:

Reordering is not that great

So my question is:
Do you see yourself more as an artsy or techy guy?
Would you rather interect with Motion Graphic tools inside Softimage through regular operators or ICETrees? (I've posted a pic of both options)

Ps: Im leaning towards the stack approach

I'm on the techy side myself. One thing I find problematic with the stack based approach is the shear number of operators you'll end up with and won't be able to remember what part of what ICE Tree you're after. Having 1 ICE Tree with group/comments to dictate each area would be better in my opinion.

+1 for staying inside one ICE tree for as much as possible. Reordering might not be as simple as drag n' drop, but still simple enough to justify the benefits of being inside the tree. I can easily bring in other ICE nodes and play/test things out.

cheers,
Daniel

Artsy. Would still prefer nodes, but not sure which of those two solutions would be better, tbh.

Whether we like it or not, ICE Tree is not the fastest way to setup the stuff around. For example, i can easely see a user not very familiar with ICE (or softimage even) giving it a try and going
"Right! So i just get this node, and this one..."
"And there are bunch of those, i dont know what they do."
"I'll connect this..."
"No, thats not going in there."
"Ill just connect this than..."
"Why the hell is it red now? Whats going on?"
"..."
"What am i even doing?
"..."
"Someone... Please help!"

Compare that with a "press button, acquire results" approach and you're at the root of all that was ever wrong with ICE from an artist point of view.

Well, im of no help at all, im afraid...

They shouldn't be using ICE then.

True, true... Than again, you are possibly writing off a lot of extremely gifted people, Eric. After all, motion graphics are not about the way you get to the effect, but about the end result itself.

Just playing devils advocate, i guess.

I see your point but people go into ICE thinking hey I'm an artist I can use it without knowing anything technical. Which is prettys illy since the application that hosts it is a pretty technical piece of software. I think that the interface for the motion tools needs to be easy and intuitive, but not dumbing things down so much that you have 1 node for every single possible thing someone without a technical understanding could use. Its a balance for sure.

I think if you look at the new scatter tools toolbars and workflow its aiming pretty on target.

Hey guys, thanks so much for the nice and fast feedback

I thought everybody would be cool with the stack approach, so Im glad I askerd... Certanly everybody here is used to ICE by now, so I guess we are all a little biased, but the arguments are solid anyways. It is a tough one...


The idea is to have everything acessible trhough menus (either way), so that should not be (a major) problem.

Anyone coming from C4D, Modo, others?

Thats me, word for word.

But seriously, i think its easy to underestimate how big of a change it can be for people to start thinking in nodes versus operators. Im not hiding the fact that im not the most technical person, but im convinced everyone can learn and use ICE. (Thank you, Paul Smith.) It can just be strange enough to discourage people and drive them away too soon.

Youre probably very right about striking the balance, though. It would be nice to know if people have found new ICE toolbar (left hand side) helpful. I rarely use it, but thats what softimage team came up with when asked to make ICE more user friendly.


My point exactly. Im not a motion graphics artist but i have no doubt i will use your tools. So from a personal standpoint, im very much in favor of keeping it graph form, nice and expandable, but thats only after a year of getting comfortable in ICE. Knowing some after effects and c4d guys, im under impression they wouldn't be too trilled about it.

One more thing to think about: ICE Tree takes up a lot of screen space. By forcing a user to always keep it up you are requiring him to have a huge resolution monitor or even a two monitor setup. I always hated this fact and heard other people complain as well.

first of all, i vote for keeping it in one ICE tree as well.

making some toolbar/menu is nice, but when i was making my scatter tools, i found it really difficult to cover more complex things, because nodes are so flexible and its impossible to cover all possibilities with buttons...

as far as i know from some c4d people, ICE seems really complicated to them. (they have xpresso in C4D but most of them don't use it). they just want to drag&drop things, use 3d manipulators (effectors) in viewport etc...
i think softimage would need some easier to build custom UI stuff, that could make life for these people easier...
of course once you get to some level, you start really enjoying the flexibility ICE offers, but for some basic stuff that can be done in c4d with 2 clicks and then tweaked with tool in viewport, you have to build some nulls+nodes system in ICE and that's just slower and too complicated.
would be nice to give it more thought...