I would like to address something that came to my mind (again) while creating the planetary climatic tgd.
functions are still somehow mystic for a lot of people, especially for new users. Even if the functions network is basically "easy", you will have a huge amount of notes pretty fast while most of them don't need to be edited in order to use the network.
for my tgd mentioned above for example, there are a few nodes to get the y values and blend some other shaders with them.
in order to change the climatic zones, you would just need to change some constant scalar nodes.
so... what about a node that delivers an interface, similar to the multi-shader but with the ability to edit the name of the entry?
you could have a name entry, a value to edit within the node and a field that points to the node you want to edit this way.
for example like this:
(http://img262.imageshack.us/img262/1830/interfacenode.jpg)
Question A: What do you think, would this be useful?
Question B: And possible?
best regards,
Goms
Cake shader V7 gets my vote :)
I think there may be a issue when getting one node to actually change the value of another node. Modifying the value of a node, yes, but actualy changing it, I would have thought, no.
There was talk quite a while back of creating a "Macro" node, I dont know how far that got developed.
Richard
Quote from: cyphyr on December 05, 2010, 08:36:35 AM
Cake shader V7 gets my vote :)
I think there may be a issue when getting one node to actually change the value of another node. Modifying the value of a node, yes, but actualy changing it, I would have thought, no.
There was talk quite a while back of creating a "Macro" node, I dont know how far that got developed.
Richard
That's exactly the type of thing I'm after, a node that can directly affect another's position, scale rotations etc.
The interface capsule node like Goms suggests is a good idea, too.
I like the idea as well.....its a good alternative to a group shader/macro.....
I'm not clear on what you're suggesting, but if others are right in saying it's similar to "macro" nodes, then it's something we do have planned for the future. We also have some other related features that will be in upcoming releases, e.g. a "note" node where you can make text notes in the network.
- Oshyan
I think what Goms is describing, is a node where you set up a load of constants that drive a particular function set, ie one node with multiple constant fields. A bit like a tab in a PF where you set the scales etc.
This would be quite useful, it would mean you have one dialogue open where you can change multiple variables that drive your function set, rather than having to double click each constant seperately to change it's value.
That sounds kind of like the sort of things "macro" functionality would enable. See World Machine for a good example.
- Oshyan
Yip, that's what i meant.
I don't even think i would be that hard to implement. one approach i could imagine is to use the input of a constant scalar/color, as its already there but not doing anything, to overwrite the nodes setting. This way you can build your network like always and later create an "interface node" on the top level to change the values from there.
This node only needs to create multiple constant scalar/color nodes of the amount i want with a name i can choose.
The only problem may be that this node needs multiple outputs of different values, which i haven't seen yet in tg2.
the long-term goal should be a macro functionality like worldmachine has, as its just amazing. :)
Houdini's HDAs are a good example if this, and simple channel referencing
can be an incredibly powerful way to work.