It has some attributes of both. With ring signatures it is mathematically proven (see white paper) that it is impossible to determine which of the possible spenders (outputs) is the real one. It is like mixing in the sense that each input has a finite set (generally 2-5 in practice though higher mix counts are allowed) of possible outputs and this ends up having a mixing-like effect on the entire blockchain.
It is unlike mixing in that the coins being "mixed" don't need to be spent at the same time (like coinjoin) nor do their owners need to take any active steps to mix (like most mixing systems), nor is there any coordinating central coordinating system that can either steal your coins (like centralized mixers) or serve as a point for tracking and logging (like most if not all coinjoin-based implementations).
Thanks for explaining! So what advantages will RingCT bring and is there a rough idea on when that will be implemented?
See:
https://lab.getmonero.org/pubs/MRL-0005.pdf
&
http://weuse.cash/2016/01/09/tying-up-loose-ends-with-ringct/
There is no specific ETA, somewhere this year is my guess.
That sounds awesome! And you can still monitor the total money supply unlike zcash?
Have any of the devs given an update on the progress for ringct lately?