Working with layouts/shapes! Some feature ideas

Hi there,

I am working with layouts (or shapes) a lot because at the moment I think they are a great device to properly shape or mask terrain manually. In fact the layout device is the reason for me why WM is my terrain generation tool of choice. I can easily create a node network and give it to someone else and they just have to work with the layouts and understand how those work instead of understanding how the whole node network functions. It’s great! But as much as I love and use the layout device it also has its limits.

Of course the main reason I would like to see improvements here is tidyness of my projects. At some point having so many layouts can get confusing and having one layout influencing another layout can get very cumbersome and even more confusing. But I think they could be even more powerful than they already are.

Some feature ideas:

  • Easier layout transfer:
    Just a simple function in the layout device that lets you copy/cut a selected layout shape or a group of layout shapes to another layout device. I know there is the import/export function but it’s very tedious.

  • Make value key-points work properly
    There is a feature where I can give individual layout shape vertices an individual height value (right click drag on a vertex) but at the moment it seems like this isn’t working or it doesn’t do anything.

  • Layout combiner/layout grouping:
    Lets say I have a layout for mountain areas, beaches and cliffs. Currently I have to edit them separately. I would love a device that can combine multiple layout shapes so I can work on all of them in one layout view at once.

  • Alternative to layout combiner - Layout output separation and flagging
    Similar to combining layouts maybe a layout device can have multiple output ports and I can just flag which layouts goes out of which output ports so I can have layouts for different purposes in the same device and thus view.

  • Layout manipulation parameters/nodes
    At the moment scaling/rotating layout shapes can only be done manually within the layout device. Of course I could take the output and manipulate that but then I am already manipulating the heightmap which might result in loss. I thought maybe it would be nice to have parameters/nodes that can manipulate the actual layout shapes.

  • Better breakup or distortion controls
    Breakup just feels very basic and raw at times and I think it could be improved a little with more distortion options and a finer scale of how it can be manipulated. Maybe the layout device could have a distortion/displacement input. Similarly to what I wrote about layout flagging and splitting outputs you could have the same for inputs and distort individually flagged layout shapes.

Thanks for reading all this, I might come up with even more ideas :slight_smile:



These sound like good suggestions. May I add one concerning working with layouts

. right click context menue → combine shapes respecting their role ( union or substract )

This would help reducing the clutter produced by drawing multiple shapes to get the right mask

The layout / shapes device is a very useful and popular device and I definitely want to see some improvements come to it.

These are all great ideas, thank you for suggesting them. You’ll be happy to know that a number of them are already on the docket for inclusion when the layout device gets some attention.

I’ve added a featureupvote for this improvement; please vote for it if you’d like to see it sooner than later, as I use the featureupvote results in many cases to prioritize development.


This topic has been around a while, and I agree with the ideas here. The layout/Shapes node could use a bit of love! One feature that I would really appreciate, to add to the list here- the ability to select more than one vertex of a shape, and transform groups of verts at once. I do not think its possible to use box select or shift click to get a selection of multiple verts at once right now. you can select and transform multiple shapes, but transforming a group of vertices is tedious because you have to move every point individually. Unless I am missing something!

Also, the control mapping for transforming shapes with move, rotate, scale, and also the control mapping for changing the elevation of a vertex, seem a bit unintuitive. I think a photoshop style transform box might be easier to use than the current hotkeys. Its also a bit unwieldy to confirm edits with the right click button. This is not how similar things are done in most programs, and its a bit hard to discover information about how that and other non standard hotkeys work in this mode. I think a lot of improvement to manipulating shapes could be had by simply remapping hotkeys to more expected and intuitive keys.


Agree with all of that! I’ve added a few of those suggestions to the feature-upvote item.

The next major release is targeting a different set of improvements, but I expect that the one after that is going to tackle all things visual editing - this includes improvements to Shapes, a possible new Sketch device, and in general finishing fleshing out the dream of the visual editing view to be what it can be.


Another feature that just came to my mind: At the moment a shape is just described by a vertex outline. So if the value key-points would work we could have a shape outline have different heights. But now wouldn’t it be cool if shapes would be described with more points than just the outline? Lets say it is a matrix (or mesh?) of points where I could also manipulate the points within the shape’s area. This would give much more shaping freedom and would reduce the amount of shapes and combiners needed to achieve a certain result.

I think this is more suited for this upcoming sculpting feature, as having to manipulate each matrix cell separately is not something fun to do :stuck_out_tongue:

1 Like

True, I just think that shapes are less destructive than manually painting/sculpting so I hope that this feature doesn’t replace them and they still might get some improvements :slight_smile:

Another idea that I came up with:
If we get a working value key-point I think it would add even more possibilities if the falloff curve and falloff distance could be set per vertex so we can really do a lot of different shapes just by using the vertices.

1 Like