I have noticed a strange behaviour with the match_size node.
If i use it without a secondary input geometry it suposedly use a 1*1*1 bounding box to do the justification, but when I in put a 1*1*1 box in the secondary input it does not give the same result ????
In the case with no geometry attached it seems that the blue guide geometry is moving depending on the settings of the nodes
Hi, I had a little dig and I think it is just a very confusing UI/UX issue (not exactly a straight up bug).
When there is no reference geometry from the second input, internally the translation goal becomes a zero sized bbox. So essentially the internal translation and the scaling cancels out each other for the no-second input case, and the Target Size parameter does references the zero sized point at the origin for the any translations including x and z axis. Which is very confusing with the blue bounding box still.
It is technically documented below. From the Target Position help: "It represents the center of the box unless the relevant axis' justification is Min or Max, in which case it represents the edge of the reference box." And from the Second Input entry in the Justify With parameter: "If there is no second input, a zero-sized bounding box at the origin is used."
Although I agree that it is weird especially from the UI/UX perspective and with the jumping blue reference box. So it could be good to RFE.
I wrote to the developers about this a long time ago, and they replied that it wasn't a bug but was actually intended—a response that completely baffled me. I can't imagine any practical benefit to this, whereas the downsides and confusion it causes are obvious. So, do write to them; perhaps a large number of users will convince them to change their stance.
