Honestly, it reminds me of the same problems I had in WC3 with unit facings at different angles.
Well, I put them on Explicit Rotation with IsLocal enabled and having the tentacles simply upside-down. Seems to keep the tentacles' positions nicely, but their facing is off. It seems when rotating the overseer 360 degrees, the tentacles actually rotate twice.
Could they be getting facing information from two different sources somehow?
When IsLocal is turned off, the tentacles face in the same direction as the overseer, even though they are using Explicit Rotation.
I have trouble thinking in spatial coordinates, but it seems to me that changing any value in the Up vector other than Z will cause adverse effects when rotating the unit.
I gave up on Variance Rotation, so I also applied a few local offsets in addition to the Explicit Rotation positions on the underside. Interestingly, when I changed the facing vector, the explicitly rotated tentacles were held in that facing. Using IsLocal corrected that issue.
First of all, let me say, great tutorial ProzaicMuze! I've just started playing around with the editor and I must admit the data editor was quite a daunting thing at first. Now I think I'm starting to get a feel for how these different data types work together to produce the desired effects.
Using your tutorial, I was able to make a great Uberlisk. :)
However, I decided to try something similar using site operations to attach two spine tentacles to an overseer, underneath the belly where the other tentacles hang down.
I've gotten the two tentacles to appear, but unfortunately I'm having a lot of difficulty getting them to stay in the right spot. I figured that I should use the Variance Rotation to achieve the effect. So I first made two site operation actors, SOpAdjustSpineL4 and SOpAdjustSpineR4 (I already have L1-3 and R1-3 set for the Uberlisk), with Up Angle set to 315 and 45, respectively. I also made a site operation actor, SOpAttachSpineC, which has its Attachment Query + set to Target 4 (which I believe is the central underside of the overseer).
Unfortunately this put the spine tentacles on the top of the overseer, and in seemingly unrelated positions.
I thought that perhaps Target 4 was upside-down, so I switched the angles and it almost seemed to work, but as I experimented further, it seemed like the tentacles appeared in random positions, even when the same Up Angles were used. I noticed that changing the scale of one of the tentacles also produced strange effects.
In addition, when rotating the overseer, one tentacle seems to oscillate widely back and forth, while the other either remains in place, or oscillates just a little.
I am at a complete loss as to what is causing this to happen. Certainly, I don't fully understand the different Site Operations, so I could be making a mistake that is causing the problem.
I have both spines using SOpAttachSpineC as an anchor. Could using the same object or target be causing these problems? Should I try something other than Variance Rotation? Any insight anyone can provide would be a great help.