By English definition, “Offset” means to move something out of predicted line. Which means moving something, basically.
Position here also changes / moves / repositions the position of an object.
From what I see in Geometry Nodes, it seems that Offset means adding whatever value that it’s given. That’s why when I inputted Position Node into Offset, the Cube got twice larger because all vertices’ locations got added / moved by their own locations.
But Position could also do the same by getting twice larger when I added the vertices’ locations to themselves.
So, why would I need Offset, exactly?
My questions are:
– What’s the distinction between using Offset and Position in Set Position in Geometry Nodes?
– What’s the reason for adding the concept of “Offset”? Wouldn’t “Position” be able to do exactly whatever “Offset” could do?
The final position of the object is position + offset. If position is not given an explicit input, it defaults to “position”, like from a position node.
It is only a convenience. You could conceivably use a vector math/add node and a position node instead of ever using the offset input.
As an example of someplace where offset is convenient, consider recreating a displacement modifier: normal->vector math/scale->offset->output. Easy to put together, easy to understand. If you added it to existing position explicitly, it would get half again as big.
This makes me wonder, how about only Position? Or only Offset?
“Position+Offset” means repositioning something and then repositioning it some more, in one operation.
Are you asking what happens if you use only one? If you use only offset, you’ll get position as if from a position node, plus input offset. If you use only position, you’ll get the input position, plus a 0,0,0 default offset.
Are you asking, could GN nodes be made without using both inputs? Yes, they absolutely could. If that’s something you’d like, feel free to use only one of those inputs. Whichever one you’d prefer.
Sure, it’s repositioning, and then repositioning some more. One of those repositions is absolute in object space and one of those is relative to existing position. Sometimes, absolute is convenient and sometimes relative is convenient. It’s unlikely that somebody is going to use both inputs in the same node, but they could if they wanted, to combine a couple of repositions that they were thinking about differently (an absolute shrinkwrap followed by a relative displacement.)
Sorry, I actually was not asking any question that you assumed. But thank you for trying to help and for the useful information.
What I’m asking is, if both Offset and Position do the same thing, which is repositioning the object, then why include both functions which do the same job, in the same node? Actually, why including the concept of “Offset” at all? Both “Offset” and “Position” reposition something.
Position repositions an object. Then Offset repositions it again. So the Set Position Node does the same job twice? If this is the case, then why didn’t the Developers just include only one function?
I assume that your whole last paragraph would be your answer to my question above?
This text is 50px from the top of the blue container. This is the offset, or relative position. This text is 85px from the top of the page, and 98px to the right. This is the absolute position, or position.
The Offset is still 80px from the top of the blue container, but the actual position of the text in space has changed to… 115px from the top and 298px from the left.
