Editing Open Problems (section)

=== Problem overview & questions ===

What is the power of CRNs augmented with stacks, without the [[initial context]] assumption?  (i.e., the concentrations of stack subunit species, $\bot_i$, are given as part of the input.)  The inspiration of this question can be found in the ability for DNA strand displacement systems to form polymers (complexes consisting of multiple DNA strands bound via hybridization of complementary domains).  While polymers can simulate stacks, they are more general and open up new possibilities for programming CRNs.

Interesting thought experiments in this model may include:

* Beginning from a population of input $X$, can stacks be used to output $\log(|X|)$ copies of an output species $Y$?
* Could consensus of two input species $X$ and $Y$ be computed exactly, and efficiently, with the use of a constant number of stack types?

* What additional computational power, if any, arises when the model is broadened to included:
** double ended stacks / queues
** stacks/queues/polymers that can be efficiently ''concatenated'' with another, at their ends
** polymers that can be efficiently "merged" with another, at multiple locations along the polymer