Conformal geometry is, in a sense, the study of geometry when you can measure *angles*, but not lengths. Though this viewpoint may seem a bit abstract, it plays an surprisingly interesting and important role in both smooth and discrete differential geometry. For one thing, it provides a setting for working with surfaces that is both very simple and very *regular*â€”recall for instance that we typically like to work with *regular* curves, because we can be confident that subsequent quantities will be well-defined (tangents, normals, curvatures, etc.); if we assume curves have an *arc length parameterization*, then life becomes particularly simple because we don’t have to worry about accounting for “stretching” as we go from the domain to the image of the curve. Likewise, with surfaces, we tend to work with *immersions*, which provide a useful notion of regularity. What’s the analogue of an arc-length parameterization for surfaces? In general it’s impossible to find a parameterization that has no stretching whatsoever, but we can always find one that at least preserves anglesâ€”a so-called *conformal parameterization*. Akin to arc-length parameterized curves, the amount of extra information about “stretching” that we need to carry around is now minimal. Moreover, the condition of angle preservation automatically gives us even more regularity than even a plain immersion: it automatically guarantees that our map is *smooth*. Note that this whole story applies equally well in both the smooth and discrete case: just as we had notions of discrete regularity for polygonal curves and discrete immersions for simplicial surfaces, we will also have a notion of discrete conformal equivalence for triangle meshes. Beyond these analytical properties, discrete conformal geometry leads to a huge number of fast, useful, and beautiful algorithms, which we will study and implement in the next few weeks.

(**Note**: these slides will also help you understand the homework, so please take a look!)