Hi. I’m not sure what you mean by a different entity or grid. For example, a subdivided plane and working with verts or faces instancing?
It is tedious to configure, it is not as flexible as Array modifier to work, also you cannot animate the parameters that you can animate with Array modifier, like offset or count, and you cannot Array on another Array. I am talking to Blender by default, I don’t know animation nodes.
I think I had used some workaround to work with heavy objects or groups like instances parent to a plane or something like that, then array the plane and make the plane not visible (I don’t remember exactly how). But workarounds are not solutions and are still tedious to set up.
As far as I’m concerned these are different use cases. For flexible instancing, grass, motion graphics etc you want the “copy to points” method @zeauro is talking about (hopefully in geometry nodes), but for simpler stuff, regular repetitive patterns such as motifs of stairs you’ll want an array because it’s simpler to setup, and doesn’t necesitate an additional object.
the developers that are most active on the new Website could be Ranked as “Most Active Devs” on the front page of website with an option to tip them some small amount of money (like restaurants style tips)
But yes I agree it would be very hard to manage and set up. I am all out of ideas
You can do that with Instancing methods, by default.
If a grid is parent of a cube, offset between copies of a cube may be animated by scaling the grid.
You can create shapekeys.
You can animate count by using a build, a mask or a boolean modifier.
You can animate anything in parent object. Parent can be any type of mesh.
You can add a parent to parent to instance it and create an Array of an Array.
Or you can use particles.
I am not saying that is not tedious to animate parent. I am just saying that, by principle, other software are using a similar “copy to points” method with a dedicated object exposing dedicated settings to manage this easily. And I was reminding that Animation nodes addon had efficient nodes for that.
So, we could expect same efficiency in future nodes.
First of all I’d like to point out that I really like Jan’s quick Blender tutorials and I learned a lot watching them, and I do appreciate the tip.
That being said, this seems even more convoluted than the hacky object offset thing one has to use with the current array. There are heaps of addons that do it in a simple manner, but this really should be an option for the modifier. For anything complex we can wait for modifier nodes for all I care, but things as simple and common as this… don’t know, maybe it’s just me.
I’ll just keep my fingers crossed for Cody’s revision to be accepted.
I guess people understand that the discussion is not just about whether it is possible to achieve it in Blender.
Sometimes I have the same feeling of those users who come to Blender from other softwares and wonder why some things have to be so difficult or complicated in Blender.
Yup. It’s not rocket science. It’s like working at a factory (which working in production is very much like at times) - you want it to be as simple as possible for common tasks. Making it more streamlined and simple wont hurt anyone Blendering for fun, and one click less can make all the difference for professionals under constant time pressure.
I wholeheartedly agree. I struggled with array, a lot, and it’s not user’s fault but simply something that is lacking in Blender. Which only means there’s room for improvement, and that’s a good thing as can only make Blender even better.
Things like these should be easy to handle as they are in other software without having to resort to scripting or other more complex ways to achieve the same result. There’s a reason and a place for everything: modifiers, particle systems, nodes. Everything is potentially doable with nodes, but that doesn’t mean everything has to be done with nodes.
So true we should definitely NOT be forced to control everything with nodes. While nodes offer complex possibilities to compose things, nodebased editing is by far not the best option for every task. Generally their versatility is often rather an excuse not to come up with better solutions.
The problem is, developers and a few artists have good math and programming skills. For these people with that privileged mind, nodes is easy, the most versatile and the best option.
You often read in discussions at developer.blender.org that when someone proposes complex settings for a feature, developers argue that it is best to keep things simple because Blender is primarily a software for Artists. So when all nodes come to Blender I hope that developers will continue to support the idea that Blender is software intended primarily for Artists, and we have simple alternative options to complex nodes.
But who knows, as example, although we recently had a Principled BSDF shader provided by an external developer, in Blender beginner artists still have to deal with nodes for materials, instead of having a simple asset manager GUI and built in material library.
I think most developers now have a good understanding that even the most complex/atomic parts of a software (nodes here) need a high-level interface to be usable by everyone. That’s where node groups come in, and Jacques (sole developer of “everything nodes” so far) has demonstrated several times now he was very well aware of that necessity. We can’t all reinvent the wheel and program an array by ourselves using nodes, so it’s expected that “geometry nodes” will come with useful node groups that are simple to manipulate but can still be opened and changed as the user sees fit.
I really hope that there will be node presets. Even if you know your way around node noodling, it’s still a hassle to keep reinventing the node wheel, so to speak.
I completely agree and I think you can rest assured : there are already presets for particles nodes that Jacques included in his branch, he even made an addon for loading them while the asset manager is still being worked on (it’s likely that in the future the AM will be used to save and share node groups). Geometry nodes are still some time away but it was hinted at in the discussion (on d.b.o) that there’s nothing preventing using both node trees and regular modifiers on a single object, so eventually we might just be able to use an array and wrangle some geometry nodes on top of that.
I have read countless different threads/posts about everything nodes, that project is pretty big, it will roll out as small chunks starting with:
1- Particles nodes (was planed for 2.9 but is not fully functional yet, got delayed to 2.9.1+, but you can start playing with it right now if you activate it in experimental section).
2-Object/Geometry nodes (for parametric/procedural generation/manipulation/transformation of vertices/polygons.
3-other parts of blender that would benefit from node setups (NLA ? fluid/rigid body/hair/cloth sims ? other stuff ?)
The end result is something similar to what we already have with material nodes, you create a new material, pick the type from the drop down list “presets” then tweak the settings right there in the material properties tab and THEN “IF” you need more options/control/freedom you dive into the node editor.
So yeah you should expect lots of node “presets” with already chewed functions commonly used in the typical everyday scenarios, plus the option to look behind the scene and participate in the chewing process.
I really hope you’re right and all this becomes true.
Here are my concerns though.
Time. Animation Nodes, and then Everything nodes, it’s been under development for quite a while now. With this I don’t mean the development it’s slow, I cannot even imagine how complex such a task can be. What I mean is that it might still take a lot of time before we can start seeing something. That said, time can be irrelevant, nothing so complex is built overnight. Still, at least for me, it remains a big question mark when this will be production-ready, meaning, when all this will includes those node groups, a mature interface, and so on. Even now, if you look at the tutorials online it’s a jungle. I tried to follow some tutorials but this thing keeps changing and 3 minutes in I couldn’t follow along as nodes were different or non-existent. Too bad there isn’t any plan (at least to my knowledge) to improve or add some modifiers.
Assuming all from point 1 gets done in a timely manner and it will be up and running in a way that people can start using it rather than dealing with bugs, missing features, tweaks, and so forth, I still have doubts about those node groups/pre-built setups. The reason is, how many situations a node group can cover before the intervention by the user is requested? I’ve seen this happening with XSI compounds, which worked on very similar logic. That’s the beauty of modifiers. They work more like Lego. Even a non-technical person can mess around with them, create combinations that will always return something, and if the user gets creative, can push them to new limits. Take the screw modifier: with tips shared by Chipp Walters, or in the videos from MasterXeon I learned about clever uses which they were indeed pushing the use of that modifier.
Even Blender’s particle system could have been pushed, a lot. The principle and the way it works is basically similar to what you can experience with something more advanced like X-Particles. I understand that going with the nodes the door to a myriad of other uses is open, but that also means that they either require skilled technical people or we’re left with the hope that advanced node groups and setups will pop up to make the life of the rest of us easy. Instead, if you take X-Particles, you see the plethora of things you can achieve without the need of understanding one single math node. Of course, there’s a big difference. With X-Particles you do what the name implies, particles. With Everything nodes, well, you do everything. There lies the point for me: having something advanced that anyone can use (XP) or something advanced meant for some (EN) with the hope (yet to be proved) that also other users will benefit from it — somehow.
I’m eager to see how this works out in practice. I wonder if a relatively simple polygon-modeled object with some extrusions, bevels etcetera will not rapidly result in node clutter, in which you can’t easily find back certain operations in order to adjust them.
Everything Nodes shouldn’t be dumbed down for those who don’t understand simple maths.
Everything Nodes doesn’t mean you have to do everything with nodes it means everything is connected by nodes. If you’re a basic user who is more comfortable working with Modifier stacks that will continue to be possible and remain unchanged but there is greater power for the more advanced user who wants to work with nodes and connect modifiers, particles and vertex maps etc and change their values based on procedural associations to other things happening in the scene.
If you try Sverchok you can get more of an idea of procedural modelling. Yes for simple modelling tasks nodes can be more work than conventional modelling. The real value of procedural modelling is the ability to create a range of different assets from a few simple rules e.g. if you’re an architect and you want to experiment with different types of roof structures and replicate them over different buildings it’s ideal.
Node clutter can be managed but it’s a side effect of having a huge amount of flexibility anything the devs can do to help managing complex node graphs will be welcome. Blender’s nodes may have to be worked on to reduce the current amount of dead space to make better use of the space.
Great.
And yes, it 's for advanced workflow speed up - to solve complex problems faster (alike modern GPU algos where you don’t see any speed up on simple scenes but when you start to tackle large quantities (geo, frames) and scene complexity, then it starts to shine.
On other note, simplified Blender 101for the dumb & lazy is still waiting for a small team to fill the niche. Just like e-cycles did Dough is made but someone needs to spice & bake it.
Dough is made but someone needs to spice & bake it.