I noticed that there are two different approaches on developing the general stokes theorem. In particular, Rudin did this on Chains (PMA) and Munkres defined the integration on manifolds by using partition of unity (Analysis on Manifolds)
As a beginner on this subject, I don't really appreciate the difference between those two approaches.
- So what are some advantages and disadvantages between those two methods?
- Which one is more easier?
- Which one is more general?