MobX and the unique symbiosis of predictability and speed

"Oh, this is Meteor / Knockout / Angular all over again..." is the initial reaction many developers have when glancing over the reactivity model of MobX. Many developer still bear the scars of trying to fix problems that result from events firing at random, and derivation computing all over the place at undesirable moments.

But what if the basic idea was correct, but all the implementations we have seen so far were just sub-optimal? What if we could address the root issues of the predictability and performance of this reactivity model?

In this talk we will dive deep into the design principles and implementation details of MobX. In what ways is it fundamentally different from its predecessors? What makes it so insanely fast? (Making React manually as fast is, a few years later, still a complicated process!).

Prepare for a talk that dives into the nitty-gritty details of the implementation and design decisions of MobX. If MobX always looked like magic to you: after this talk no longer. And most importantly, we'll discover that predictability and speed are not competing concerns but rather catalysts of each other.