project primitives in R3 into R2

[bergey]

Some enhancements that would help with this:

#76 z-order
#108 projective transforms

Per #108, points and linear segments are easy.

[bergey]

More discussion from IRC:

<pr> to be able to work with 3d diagrams (using R3 i guess), you have to work
     with the povray backend, don't you? or are there already projections(?)
     that would allow one to use the SVG backend, similarly to how tikz allows
     3d coordinates?  [13:44]
<bergey> pr: Right now the povray backend is the only 3D backend.  [14:12]
<bergey> I'm not familiar with tikz.  I'd like to hear what you're trying to
         do, so I can work on supporting it.  [14:13]
<pr> consider for example the answers to:
     http://tex.stackexchange.com/questions/28284/how-to-draw-a-cube-with-tikz-where-all-faces-have-a-distinct-color
                                                                        [14:19]
<pr> what I want to do is really simple, doesn't even require working with R3
     (would make it more explicit, tho). i just want to draw cubes (like it
     says in the link: draw them like you would draw them on paper) with
     different colored faces, and output it to SVG for example.  [14:21]
-diagramsbot- [diagrams-core] byorgey comment on commit 1e2b9f1: Ah,
              true. http://git.io/caQtYQ  [14:23]
<pr> would be a nice way to confirm that there are in fact 57 different
     (i.e. rotational symmetry) colorings of the sides of a cube, when you're
     allowed to use one of three colors for each face  [14:24]
<byorgey> bergey, pr: yeah, I've done that too (re: infinite diagrams) =(
                                                                        [14:26]
<bergey> Cool.  I'm pretty sure there's nothing currently in Diagrams to take
         R3->R2, but we could start adding it soon.
<bergey> There are some related feature requests open:
<bergey> https://github.com/diagrams/diagrams-lib/issues/76  [14:27]
<bergey> https://github.com/diagrams/diagrams-lib/issues/108
<byorgey> yeah, projections from R3 -> R2 would be super cool
<byorgey> martingale: I am 99% sure that two consecutive Act nodes cannot
          happen in practice  [14:28]
<bergey> I think it will be a while before we can handle the native 3D prims,
         but handling points and Paths seems not-too-hard, especially if
         you're willing to accept (temporarily) the lack of occlusion
         calculations.  [14:29]
<byorgey> bergey: I just tried out the metafont stuff in contrib yesterday, it
          didn't work the way I expected =)  I will see if I can whittle it
          down to a concrete bug report  [14:30]
<bergey> byorgey: Glad you're using it.  I look forward to bug reports!
<bergey> I used the metafont a little bit the other day, and found a bug I
         haven't reported, also.  [14:31]
<byorgey> bergey: well, I'm not really using it in anger yet, just trying it
          out, but certainly hope/expect to use it in the future.  [14:32]
<bergey> I just want people banging on it to find bugs / interface warts.
                                                                        [14:33]
-diagramsbot- [diagrams-contrib] bergey opened issue #17: cycles with
              directions are not symmetric  http://git.io/n_HmHA  [14:34]
-diagramsbot- [active] byorgey pushed 1 new commit to new-semantics:
              http://git.io/c6hSKQ  [14:36]
-diagramsbot- active/new-semantics 4b85725 Brent Yorgey: start working on FARM
              slides
<bergey> pr: Last thought for now: I'm planning to work on Paths in 3D this
         coming week.  Once we can define a square, and translate & rotate it
         in 3D, projecting the points back into 2D, maybe using the Camera
         type, would make sense.