Twist across multiple box prims?

Visual: Take a normal 10x10x10 box prim, and twist the end by 27 (I am attempting to get a photo into this message)

My question is:

Any good way to have a twist flow across adjoining prims? I am thinking of a building footprint (say, 20x20, or 40x40, to use round numbers). I want each story to smoothly twist up from the one below it.

(in other words, if I were merely doing a 10x10 building, I'd just make it hollow, stack them, and be done.. but I need a much bigger footprint)

Well, two obvious answers come to mind:

1. Copy the twisted prim straight up, rotate it 27 degrees
   
2. Copy the twisted prim straight up, add 27 to the B and E Twist parameters
   

The first is easier and more likely to work, since the Twist parameters are rounded off strangely and might not increment as smoothly as you'd like.

However, there may be a noticeable "crease" between prims, even though they align perfectly, since the surface normals of the twisted faces won't unify across separate prims. This crease will be more obvious the more severe the twist is.

Thanks Deanna.. when I say 'adjoining', I mean the base...
The stacking part is cinchy

Let me put it this way:

* plunk down 4 prims.. base of 20x20
* now.. get the 4 of them twisted to match the same effect as one by itself...

basically.. if you had a megaprim that was 20x20x20, and did the twist...

Just had a quick play. I think you can get pretty close using shears but it's not going to be perfectly lined up at the top with the precision currently allowed on the shear parameters, and the vertical twisted edges aren't going to line up nicely.

thanks SS - I see what you mean about shears. They make an interesting set of shapes!

Unless Cadroe has something up his sleeve, I can't think of way to do this (without tons of prims)

OHhhhhhh... now I get what you mean, 4 in paralell, not 4 in series. (I thought the question was a little too easy )
Yeah, what Seifert suggested is about the only option, essentially perform the same twist on all 4, then try to shear the tops so that the 4 corners align as the corners of the bases do.

[Edit: Quick n' dirty test. Left is a 1 x 1 x 1 cube with 27 twist. Right is four 0.5 x 0.5 x 1 cubes with 27 twist, shear -0.28x, 0.17y]

thanks - the shears are good at small scales, and start breaking up at 10m

the exercise did lead off to a good design direction though!

oh god...please don't use mega prims!

Did you know that not only are they frowned upon by Linden Labs and by most people, they also contribute more lag to the sim?

sure, if used tastefully, LL might look the other way, but they still add extra lag.

Not only that, but you're twisting them which creates even MORE work for a graphics card.

Nope, I would not use them! I was just trying to summon the example of a 20x20x20 object for visualization.

One can imagine the Wild West atmosphere that would ensue if they were suddenly made legit.

Do you have any references for these statements? I've yet to see any evidence that megaprims add to lag in any way, but I'd be very interested to see some.

while this may not be the concrete evidence that you're looking for, it does make sense from a logical perspective.

/327/e6/175017/1.html

post #12


Well, I have not cared whether they got used or not in the past, this opened my eyes and i wonder who the original source was that provided this information.
Sure, we could just say "don't believe everything we read on the internet", we all know that these mega prims are no longer allowed in SL except for grandfathered prims and ones that do not cause complaints from neighbors...in other words....no, they SHOULDN'T be used, but we can't enforce them is LL statement.

As I said, the logic behind the original statement makes sense.

Well... "deep think" to my knowledge is a term used only in the context of the physics engine trying to work out how to deal with interpenetrating prims. The physics engine runs on the server and has nothing to do with the client rendering. This makes me suspect that at best some statement had got garbled, but most likely this is just unfounded rumour.