A “merge sculpts” function would be great to optimize the thermo
If you could for example merge two sculpts in a single element that cannot be separated the gameplay thermo would not fill so fast
I guess to deal with the subtract issue they could add a new edit type operation in the list. Let’s call it “reset”. This basically acts as a full stop. Any edits after a reset will only act on current and future operations. So if you added a subtract sculpt into an existing sculpt, it wouldn’t do anything.
But yeah your group idea would be good also! It would make it easier to do stuff like joining of seams of separate sculpts.
My method would save gameplay thermo. And perhaps some performance. It would also make it easier for people to build something out of many component parts without incurring the gameplay cost normally associated with it.
The idea is: you have special-sculpt-group(sculpt A, sculpt B). At load time, the sculpts would be built as normal. Then perhaps the surfaces can all be lumped into the same object for performance.
At run time, each clone will all use the same object in memory to render and for physics shape, and all that stuff. So they're instances of that object. So they only cost 1 block of gameplay per copy, instead of 3 (group, sculpt, sculpt).
This would allow people to work the way a lot of people feel is more natural: make 1 brick sculpt, copy it a ton and group it (in this special way). Because it's using this special kind of group it will be optimised away so it behaves in memory and performance as if it were one object instead of a group of hundreds of objects.
I really wish I could find my original post all about this.
It would not save any thermo, the new object will result in more surface area to compute for the new sculpt. Gameplay thermo would still go up a bit.
They couldn't merge it, which is why that function doesn't exist I'm guessing. How would negatives work? They'd normally apply to *all* previous edits. So if you just slap them at the end of the edits list they'd mess all sorts of things up.
So making such a thing work would likely be a whole load of work trying to get things to work reliably and as expected.
My suggestion of making a special type of sculpt group would mean none of the building process or edit list would be affected at all. It would simply build as normal. But then instance that object as if it were one single object, and allow clones to be one single object, saving on gameplay and making it easier to work with.
Yeah they technically could “merge it”. It would probably have to be called “add” or “append” instead of “merge” though. As Nins points at, it would literally be just pushing the edits from one sculpt on top of an existing sculpt!
As far as I know each sculpture has a list with edits wich where applied to them and in what order.
So to make this work you would just need to stack these lists on top of each other inside the same sculpture.
I would really love it if this gets in.
Yeah... I made this suggestion in early access. The problem is that you cannot *truly* merge two sculpts to make a new sculpt, because of how it needs to rebuild the sculpt from scratch on-load.
But my idea was to make a special "merged sculpt" group that tells the engine that the entire thing can be instanced and nothing inside can be changed at runtime. No response on that idea though. Although I can't find that post right now...