As in the title, an ability to decide wheter the part is being rendered or not. This way we can probably increase performance in games.
Ehh wait… Why is this needed. It already exist, it is called Transparency.
Does Transparency disable polygons completely? Or just makes it invisible?
That’s the idea at least, to be able to make a part vanish without deleting it.
The engine doesn’t bother putting it in the render queue from what I understand.
From what I understand in efficiency from greatest to least its 1,0,0<x<1
But you’re not sure
We would need someone who worked on the engine to tell us.
I wouldn’t mind a property of Model that hides groups of parts efficiently
There goes a question: if I set Part’s Transparency to 1 and put somewhat transparent decals onto them, will that be treated in the same way as if there were no decals and normal Transparency on thr Part?
Decals are a separate entity. The rendering of decals logically should be entirely separate from parts. The physics portion of the part is still there and its data is still there which is why the engine can not render a part but still can display a decal on where a face should be.
Yes, but they nicely cover up the surface, it is not just a flat image, keep that in mind
Not sure what you’re saying.
Displaying a decal doesn’t require a object to be rendered so the part is not rendered and the decal is.
I can’t see a use case for a property like this. Why do you want this exactly?
If you want a property that stops an object from being rendered, set transparency to equal 1.
If you’re concerned about parts that the user isn’t looking at, they aren’t rendered to begin with (that wouldn’t make sense and we can observe this behavior experimentally).
The only use case is if you’d like to be constantly (and by constantly I mean a lot) changing an object’s rendering state based on how (?) the user is looking at it. If you’re doing this relatively often (RenderStepped or even Stepped; which I can’t think of a reason you’re doing it for) there is probably a better way to accomplish this.
There are better ways to increase performance. You’ll find that (especially on ROBLOX) users tend not to care about amazingly high detail builds (which is where I assume your problem is coming from) but more so user experience. I can’t imagine why you would want to have a high concentration of parts in a small area.
Use local transparency modifier clientside if you dont want to overwrite the actual transparency of the parts (which is the main reason one might want a Visible boolean)?
Assuming it works…
Can you spot the difference?
1 pair of white-red spheres has Transparency set to 1 with correct decals applied to them, would they still be quicker?
Are you doing this to optimize or for some other reason? The difference in performance in your example is probably very small, so I wouldn’t apply what you’re doing just for optimization (you probably won’t need it). Rendering a texture takes rendering time as well, so you may even be losing on performance here because the other set of spheres is just a solid color.
If this feature is happening it should be because of ease-of-use reasons IMO, not for performance reasons. (i.e. easily hiding (groups of) parts without having to reparent them)
I don’t need it right now tbh. I once asked about this on discord, but didn’t get any serious information. Now I know that setting the Transparency to 1 is working just fine, so this request isn’t really necessary now.
The 2 pairs of spheres image doesn’t really have a purpose in the topic of this request, so I brought it up to not create another thread. I don’t care about such small performance details. Its only purpose was to ask if visible parts are equal to invisble parts with decals put on them (thus making them visible) when it comes to performance.
###I will repeat, this request has no longer any meaning, no need for further posts I suppose
Not sure if I can tag someone to lock it down
You can just mark any (appropriate) post as the solution and that should be fine, like the one where it said that transparency 1 was the solution
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.