Primitive naming

[Ronald Vallenduuk]

If I want to do a prism primitive similar to tri3a1, but with one right-angle side and the diagonal side, what should that be called?
In general, what is the naming scheme for primitives? Looking at all the various cube-based primitives I haven't found the logic yet...

[Michael Heidemann]

There is no rule for all possibilities. If the name you choose extensible for other similar primitives, I am pretty sure we will use it.

[Chris Dee]

There is a logic to thebox primitives:

Code:

box
   n    = number of faces [i]present[/i]
    x   = used to differentiate the different configurations of faces (- = the "default" adjacent, o = 4-face ring, u = 3-face "U"-section)
     e  = number of faces [i]absent[/i]
      p = used to differentiate different configurations of the missing edges (a=adjacent, p=parallel)

You don't say which edges you plan to include, but from what you describe, your new primitive should be tri2.dat or tri2-n, if one or more of the 9 edges are omitted.

[Ronald Vallenduuk]

It should be a tri3 as it'll have the base triangle as well as two sides. Here's what I mean:

tri3 by Duq, on Flickr

[Max Martin Richter]

In my eyes this prim is not needed, because you can get the result by simply deform tri3a1 in it's rotation matrix.
If you need help with deforming feel free to ask.

/Max

[Ronald Vallenduuk]

If you could point me to a tutorial about deformation matrices that would be great, thanks.

[Philippe Hurbain]

I agree with Max, this new primitive is not needed.
There's a detailed page here about this kind of transformation. It's in French, but hopefully Google translate is your friend

[Ronald Vallenduuk]

Thanks for the link. I'll first have a go with dusting off my school-French. If that fails I'll ask Google to help out ;-)